Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A health care operations research analysis of elderly fallers' emergency department services utilization… Woolcott, John Clifford 2011

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
24-ubc_2012_spring_woolcott_john.pdf [ 1.38MB ]
Metadata
JSON: 24-1.0072466.json
JSON-LD: 24-1.0072466-ld.json
RDF/XML (Pretty): 24-1.0072466-rdf.xml
RDF/JSON: 24-1.0072466-rdf.json
Turtle: 24-1.0072466-turtle.txt
N-Triples: 24-1.0072466-rdf-ntriples.txt
Original Record: 24-1.0072466-source.json
Full Text
24-1.0072466-fulltext.txt
Citation
24-1.0072466.ris

Full Text

A HEALTH CARE OPERATIONS RESEARCH ANALYSIS OF ELDERLY FALLERS’ EMERGENCY DEPARTMENT SERVICES UTILIZATION AND COST  by John Clifford Woolcott  Bachelor of Arts, Honours Economics, Wilfrid Laurier University, 1998 Masters of Arts, Economics, Carleton University, 2000   A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Pharmaceutical Sciences)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) December 2011  © John Clifford Woolcott, 2011  ii Abstract  Introduction: Falls in the elderly are a significant cause of morbidity.  Prescription medication use has been identified as an independent risk factor for falls.  Among all Emergency Department (ED) presentations by elderly persons, 14-40% are due to falls, placing considerable strain on ED resources.  Aims: In my thesis I aimed to 1) Provide updated estimates of the association between the use of specific medications and falling, 2) Determine whether the care provided to elderly fallers while patients in the ED follows published recommendations and was provided in a timely fashion, 3) Estimate the cost per fall resulting in an ED presentation, 4) Design a discrete event simulation (DES) model simulating care and then simulating other approaches to care including hypothetical changes.  Methods: 1) A Bayesian meta-analysis of studies assessing the association between specific classes of medication use and risk of a fall. 2) A cohort study of elderly fallers presenting to the ED.  3) DES of the ED care received by elderly fallers   iii Results: Use of anti-hypertensives, diuretics, sedatives and hypnotics, neuroleptics and anti-psychotics, antidepressants, benzodiazepines, and non-steroidal anti- inflammatory drugs are associated with an increased risk of falling. 1) In a sample of 101 ED fall presentations, 38% of elderly fallers leave the ED without a geriatric assessment and 14% are assessed by a physiotherapist. Less than 8% of fallers received care which met the wait time benchmarks. The estimated cost per fall causing an ED presentation is $11,408 with the cost per fall-related hospitalization estimated to be $29,363. 2) Providing care in a timely fashion could significantly reduce the time an elderly faller spends in the ED and the opportunity costs associated with waiting to be seen by physician or admission to hospital.  Summary: Many commonly used medications are associated with falls. The care provided by the elderly faller in the ED does not currently meet the recommendations of published guidelines, nor is it provided in a timely fashion. The economic burden of falls is significant. By not providing ED care that meets recommended wait time benchmarks significant opportunity costs are incurred by the ED.     iv Preface Sections of this thesis have been published as multi-authored manuscripts in peer- reviewed journals and are indicated with* beside the publication below. Details of the authors’ contributions are provided below. For Chapters 3-5 in which data were prospectively collected from elderly fallers, ethics approval was granted from the University of British Columbia Clinical Research Ethics Review Board (approval number H06-03142).  A version of Chapter 2* has been published. Woolcott JC, Richardson KJ, Wiens MO, Patel B, Marin J, Khan KM, Marra CA. (2009) Meta-analysis of the impact of 9 medication classes on falls in elderly persons. Arch Inter Med 2009 Nov 23. 169(21): 1952-1960.  Authors’ contributions: I was responsible for the original ideas behind the manuscript, data collection, analysis and writing the original publication. Carlo Marra guided all aspects of the research. Kathryn Richardson guided the statistical analysis of this manuscript.  Matthew Wiens, Bhavini Patel and Judith Marin were involved in data collection and analyses. Carlo Marra, Karim Khan, Kathryn Richardson, and Matthew Wiens participated in the design of the study, stimulated discussion and provided editorial assistance.  Chapter 3 is based on research carried out in the Vancouver General Hospital Emergency Department in elderly persons who have suffered a fall. I was responsible for the development of data collection strategy, undertaking/supervising data collection from elderly fallers in the ED, analysis of data and writing of the study  v results. Karim Khan and Carlo Marra participated in the design of the study and guided all aspects of the research. Suzana Mitrovic was involved in the data collection during the trial. Karim Khan and Carlo Marra stimulated discussion and provided editorial assistance.  A version of Chapter 4* has been published. Woolcott JC, Khan LM, Mitrovic S, Anis AH, Marra CA, The cost of fall related presentations to the ED: A prospective, in- person, patient-tracking analysis of health resource utilization. Osteoporos Int. 2011 Sep 3 (Epub ahead of print).  Author’s contributions: John Woolcott was responsible for the original ideas behind the manuscript, analysis and writing the original publication. Carlo Marra and Karim Khan guided all aspects of the research. Suzana Mitrovic participated in the data collection for this manuscript. Carlo Marra, Karim Khan, Suzana Mitrovic and Aslam Anis participated in the design of the study, stimulated discussion and provided editorial assistance.  Chapter 5 is based on research carried out in the Vancouver General Hospital Emergency Department in elderly persons who have suffered a fall. I was responsible for the development of data collection strategy, undertaking/supervision of data collection from elderly fallers in the ED, analysis of data, development of a discrete event simulation model and writing of the study results. Karim Khan and Carlo Marra participated in the design of the study and guided all aspects of the research while stimulating discussion and providing editorial assistance.   vi Table of Contents Abstract ................................................................................................................................................................ ii Preface .................................................................................................................................................................iv Table of Contents ...............................................................................................................................................vi List of Tables ..................................................................................................................................................... xiv List of Figures .................................................................................................................................................. xvii List of Abbreviations ...................................................................................................................................... xviii Acknowledgements .......................................................................................................................................... xix Chapter  1: Introduction .......................................................................................................................... 1 1.1 Definition of terms ........................................................................................................................ 4 1.2 Literature review ........................................................................................................................... 5 1.2.1 The epidemiology of falls: Definition and incidence ........................................................... 5 1.2.1.1 Definition of an injurious fall ......................................................................................... 7 1.2.1.2 Estimates of the incidence of falls ............................................................................... 8 1.2.1.2.1 International estimates of the incidence of falls among elderly persons .......... 8 1.2.1.2.2 Canadian estimates of the incidence of falls among elderly persons ............ 11 1.2.2 Risk factors for falls ............................................................................................................... 12 1.2.2.1 Environmental risk factors .......................................................................................... 13 1.2.2.1.1 Home hazards ........................................................................................................ 13 1.2.2.1.2 Weather conditions ................................................................................................ 14 1.2.2.2 Socioeconomic risk factors......................................................................................... 15 1.2.2.3 Biological risk factors .................................................................................................. 16 1.2.2.4 Behavioural risk factors .............................................................................................. 17 1.2.2.5 Medication use and fall risk ........................................................................................ 18 1.2.2.5.1 Risks associated with poly-pharmacy ................................................................. 18 1.2.2.5.2 Studies of specific medications ............................................................................ 20  vii 1.2.2.5.3 Studies of medication withdrawal ........................................................................ 26 1.2.3 Recommended care for elderly fallers ............................................................................... 26 1.2.3.1 UCLA Emergency Department Guidelines .............................................................. 26 1.2.3.2 American Geriatric Society/British Geriatrics Society/American Academy of Orthopaedic Surgeons Guidelines for the Prevention of Falls in Older Persons .................. 28 1.2.3.3 The PROFET Guidelines ............................................................................................ 31 1.2.4 Care of the elderly person presenting to the Emergency Department .......................... 33 1.2.4.1 The Emergency Department as a care provider for elderly persons ................... 33 1.2.4.2 Emergency Department presentations by elderly persons due to a fall .............. 36 1.2.4.2.1 International assessments of the elderly faller presenting to the Emergency Department ................................................................................................................................. 36 1.2.4.2.2 Canadian assessments of the elderly faller presenting to the Emergency Department ................................................................................................................................. 38 1.2.4.3 International Emergency Department interventions for the care of the elderly faller……… ...................................................................................................................................... 40 1.2.4.4 Canadian Emergency Department interventions for the care of the elderly faller……… ...................................................................................................................................... 41 1.2.5 The economic burden of falls .............................................................................................. 41 1.2.5.1 Estimation of costs for inclusion in a cost of illness study ..................................... 42 1.2.5.2 Categories of cost ........................................................................................................ 42 1.2.5.2.1 Direct medical costs ............................................................................................... 43 1.2.5.2.2 Direct non-medical costs ....................................................................................... 44 1.2.5.2.3 Indirect costs ........................................................................................................... 44 1.2.5.2.4 Intangible costs ....................................................................................................... 45 1.2.5.3 Perspectives and costs to be included in cost of illness studies .......................... 46 1.2.5.4 Estimates of the costs of falls .................................................................................... 47  viii 1.2.5.4.1 International estimates of the costs of falls ........................................................ 47 1.2.5.4.2 Canadian estimates of the costs of falls ............................................................. 49 1.2.5.4.3 Costs of falls in British Columbia ......................................................................... 51 1.2.6 Operations research and discrete event simulation ......................................................... 52 1.2.6.1 Operations research and discrete event simulation in the Emergency Department ...................................................................................................................................... 52 1.2.6.2 Methods of discrete event simulation ....................................................................... 53 1.3 Research studies: rationale, objectives, and potential contributions.................................. 54 1.3.1 Study 1 (Chapter 2): Meta-analysis of the impact of 9 medication classes on falls in elderly persons ..................................................................................................................................... 54 1.3.1.1 Rationale ....................................................................................................................... 54 1.3.1.2 Objective ....................................................................................................................... 55 1.3.1.3 Potential contribution ................................................................................................... 55 1.3.2 Study 2 (Chapter 3): The elderly faller’s Emergency Department management: a direct observational study of care delivery and wait times in an urban Canadian university hospital ………………………………………………………………………………………………...56 1.3.2.1 Rationale ....................................................................................................................... 56 1.3.2.2 Objective ....................................................................................................................... 57 1.3.2.3 Potential contribution ................................................................................................... 58 1.3.3 Study 3 (Chapter 4): the cost of fall-related presentations to the Emergency Department: A prospective, in-person, patient-tracking analysis of health resource utilization………………………………………………………………………………………………58 1.3.3.1 Rationale ....................................................................................................................... 58 1.3.3.2 Objective ....................................................................................................................... 59 1.3.3.3 Potential contribution ................................................................................................... 59  ix 1.3.4 Study 4 (Chapter 5): An operations research analysis of the Emergency Department care of the elderly faller: Simulation of current care and the impact of providing care that meets wait-time and falls prevention guidelines ............................................................................. 60 1.3.4.1 Rationale ....................................................................................................................... 60 1.3.4.2 Objective ....................................................................................................................... 61 1.3.4.3 Potential contribution ................................................................................................... 61 Chapter  2: Meta-Analysis of the Impact of 9 medication classes on falls in elderly people ................ 84 2.1 Introduction ................................................................................................................................. 84 2.2 Methods ....................................................................................................................................... 86 2.2.1 Data sources and searches ................................................................................................. 86 2.2.2 Study selection ...................................................................................................................... 87 2.2.3 Data extraction and quality assessment ............................................................................ 87 2.2.4 Data synthesis and analysis ................................................................................................ 88 2.3 Results ......................................................................................................................................... 91 2.3.1 Results of the meta-analyses .............................................................................................. 92 2.4 Discussion ................................................................................................................................... 95 Chapter  3: The elderly faller's Emergency Department management: A direct observational study of care delivery and wait-times in an urban Canadian university hospital. ............................................. 106 3.1 Introduction ............................................................................................................................... 106 3.2 Methods ..................................................................................................................................... 109 3.2.1 Study design and setting .................................................................................................... 109 3.2.2 Definition of a fall ................................................................................................................. 110 3.2.3 Patient recruitment and ethical approval ......................................................................... 110  x 3.2.4 Data collection ..................................................................................................................... 111 3.2.5 Assessment of care ............................................................................................................ 112 3.2.5.1 Guidelines for fall prevention ................................................................................... 112 3.2.5.2 Canadian Association of Emergency Physicians Emergency Department wait time guidelines .............................................................................................................................. 112 3.2.6 Outcomes ............................................................................................................................. 113 3.2.6.1 Concordance with falls guidelines ........................................................................... 113 3.2.6.2 Concordance with wait time guidelines .................................................................. 114 3.2.7 Statistical analysis ............................................................................................................... 114 3.3 Results ....................................................................................................................................... 115 3.3.1 Participant characteristics .................................................................................................. 115 3.3.2 Care provided in the Emergency Department ................................................................ 117 3.3.3 Concordance with falls guidelines .................................................................................... 118 3.3.4 Emergency Department wait time ..................................................................................... 119 3.4 Discussion ................................................................................................................................. 120 Chapter  4: The cost of fall related presentations to the Emergency Department: A prospective, in- person, patient-tracking analysis of health resource utilization .......................................................... 133 4.1 Introduction ............................................................................................................................... 133 4.2 Methods ..................................................................................................................................... 135 4.2.1 Study design and setting .................................................................................................... 135 4.2.2 Health resource utilization and cost data ......................................................................... 135 4.2.3 Statistical analysis ............................................................................................................... 136  xi 4.2.4 Sensitivity analysis .............................................................................................................. 136 4.3 Results ....................................................................................................................................... 137 4.3.1 Participant demographics and characteristics of care ................................................... 137 4.3.2 Cost of care .......................................................................................................................... 138 4.3.3 Sub-group analysis of fallers suffering a hip fracture ..................................................... 138 4.3.4 Sensitivity analysis .............................................................................................................. 139 4.4 Discussion ................................................................................................................................. 139 Chapter  5: An operations research analysis of the Emergency Department care of the elderly faller: Simulation of current care and the impact of providing care that meets wait time and falls prevention guidelines ............................................................................................................................................ 146 5.1 Introduction ............................................................................................................................... 146 5.2 Methods ..................................................................................................................................... 147 5.2.1 Hospital setting and Vancouver General Hospital Emergency Department ............... 147 5.2.2 Study approach ................................................................................................................... 148 5.2.3 The care path of the elderly faller ..................................................................................... 149 5.2.4 Development of the discrete event simulation model .................................................... 151 5.2.5 Model simulation ................................................................................................................. 152 5.2.6 Scenario analysis ................................................................................................................ 152 5.3 Results ....................................................................................................................................... 154 5.3.1 Model performance ............................................................................................................. 154 5.3.2 Scenario analyses ............................................................................................................... 154  xii 5.3.2.1 Scenario 1: For all elderly fallers presenting to the VGH ED, wait time from presentation to being seen by an ED physician should follow CAEP wait time benchmarks (Table 5-1) (as defined by the patient’s CTAS). ....................................................................... 154 5.3.2.2 Scenario 2: For all elderly fallers who are to be admitted to hospital, wait time from the decision to admit until discharge from the VGH ED should not exceed 120 minutes….. ..................................................................................................................................... 155 5.3.2.3 Scenario 3: All elderly fallers should receive care from a geriatric triage nurse and physiotherapist. ..................................................................................................................... 155 5.3.2.4 Scenario 4: This scenario simulated both the CAEP wait time benchmarks being met and post fall prevention guidelines followed for every faller presenting to VGH ED……….. ..................................................................................................................................... 156 5.4 Discussion ................................................................................................................................. 156 Chapter  6: Integrated discussion ....................................................................................................... 165 6.1 Overview of my findings .......................................................................................................... 165 6.2 Unique contributions, impact and implications .................................................................... 167 6.2.1 Medication use in the elderly: Methods advancement and new knowledge .............. 168 6.2.2 Care gap in the Emergency Department: Implications for health authorities and professional bodies ........................................................................................................................... 169 6.2.3 The first Canadian estimates of the cost of falls in the Emergency Department ....... 171 6.2.4 A discrete event simulation of the care provided to elderly fallers in the Emergency Department ......................................................................................................................................... 171 6.3 Limitations ................................................................................................................................. 172 6.4 Recommendations for future research ................................................................................. 174 References........................................................................................................................................................ 178 Appendices ....................................................................................................................................................... 198  xiii Appendix A Health care operations analysis to reduce attending times for seniors presenting to the Emergency Department with a fall: subject information and research project consent form ............... 198 Appendix B Health care operations analysis to reduce attending times for seniors presenting to the Emergency Department with a fall: incident and fall history questionnaire (researcher administered)  ............................................................................................................................................................................ 205 Appendix C Health care operations analysis to reduce attending times for seniors presenting to the Emergency Department with a fall: resource questionnaire (researcher administered) ....................... 208    xiv List of Tables  Table 1-1 Fall definitions as taken from Zecevic et al .................................................... 62 Table 1-2 Definitions of injurious falls ............................................................................... 63 Table 1-3 Incidence of falls reported in international and Canadian studies .............. 64 Table 1-4 Studies published between 1996 and 2007 measuring the association between poly-pharmacy and falls ...................................................................................... 66 Table 1-5 Results of Leipzig et al’s meta-analysis .......................................................... 67 Table 1-6 Studies published between 1996 and 2007 measuring the association between antidepressant use and falls .............................................................................. 68 Table 1-7 Studies published between 1996 and 2007 measuring the association between sedative/hypnotic use and falls .......................................................................... 70 Table 1-8 Studies published between 1996 and 2007 measuring the association between benzodiazepine (BZD) use and falls ................................................................. 71 Table 1-9 Studies published between 1996 and 2007 measuring the association between diuretic use and falls ............................................................................................ 72 Table 1-10 Studies published between 1996 and 2007 measuring the association between anti-psychotic/neuroleptic use and falls ............................................................ 73 Table 1-11 Studies published between 1996 and 2007 measuring the association between psychotropic use and falls .................................................................................. 74 Table 1-12 Studies published between 1996 and 2007 measuring the association between narcotic use and falls ........................................................................................... 75 Table 1-13 Studies published between 1996 and 2007 measuring the association between non-steroidal anti-inflammatory drug (NSAID) use and falls ......................... 76  xv Table 1-14 Studies published between 1996 and 2007 measuring the association between beta-blocker use and falls ................................................................................... 77 Table 1-15 Studies published between 1996 and 2007 measuring the association between calcium channel blocker use and falls .............................................................. 78 Table 1-16 Studies published between 1996 and 2007 measuring the association between angiotensin-converting enzyme inhibitor (ACEI) use and falls ..................... 79 Table 1-17 Studies published between 1996 and 2007 measuring the association between digoxin use and falls ............................................................................................ 80 Table 1-18 Studies published between 1996 and 2007 measuring the association between anti-hypertensive use and falls .......................................................................... 81 Table 1-19 UCLA Practice Guideline: ED management of falls in patients aged 65 years and older ..................................................................................................................... 82 Table 2-1  Drugs and studies included in meta-analysis ............................................. 102 Table 2-2 Characteristics of studies included in meta-analysis .................................. 103 Table 2-3 Pooled Bayesian odds ratios and sub-group sensitivity analysis ............. 105 Table 3-1 Components of guideline care for the elderly faller .................................... 126 Table 3-2 Canadian Association of Emergency Physicians (CAEP) wait-time benchmarks. ........................................................................................................................ 127 Table 3-3 Demographics of fallers .................................................................................. 128 Table 3-4 Recruited sample physiological risk factors (n = 99) .................................. 129 Table 3-5 Medication related risk factors (n = 99) ........................................................ 130 Table 3-6 Participant duration of wait times (n=101) .................................................... 131 Table 4-1 Unit costs of health resource utilizations ...................................................... 143  xvi Table 4-2 Costs of Emergency Department care .......................................................... 144 Table 5-1 Canadian Triage Acuity Scale Levels ........................................................... 160 Table 5-2 Emergency Department wait time distributions and sources of data ....... 162 Table 5-3 Probability of Emergency Department events ............................................. 163 Table 5-4 Simulation and scenario analysis results ..................................................... 164   xvii List of Figures  Figure 1-1 Fall risk factors  ................................................................................................. 65 Figure 1-2 Prevention of falls in older persons living in the community (American Geriatric Society/British Geriatric Society/American Academy of Orthopedic Surgeons Guideline) ............................................................................................................ 83 Figure 2-1 Flow diagram of study selection process .................................................... 101 Figure 2-2 Meta-analysis results ...................................................................................... 104 Figure 3-1 CAEP Guideline adherence for time spent waiting for Emergency Department physician and total time in Emergency Department ............................... 132 Figure 4-1 Cost of an Emergency Department fall (mean and 95%CI) ..................... 145 Figure 5-1 Vancouver General Hospital Emergency Department process map of elderly faller ......................................................................................................................... 161   xviii List of Abbreviations ACEI   Angiotensin-Converting Enzyme Inhibitor AGS/BGS/AAOS American Geriatric Society/British Geriatrics Society/American  Academy of Orthopedic Surgeons BZD   Benzodiazepine CAEP   Canadian Association of Emergency Physicians CI   Confidence Interval CIHI   Canadian Institute for Health Information CINAHL   Cumulative Index for Nursing and Allied Health Literature CrI   Credible Interval CTAS   Canadian Triage and Acuity Scale DES   Discrete Event Simulation EBM   Evidence-Based Medicine ED   Emergency Department EHS   Emergency Health Services EMBASE  Excerpta Medica Database HR   Hazard Ratio ICD   International Classification of Diseases LOS   Length of Stay MeSH   Medical Subject Headings NSAID   Non-Steroidal Anti-Inflammatory Drug OR   Odds Ratio ProFaNE  Prevention of Falls Network of Europe PROFET  Prevention of Falls in the Elderly Trial RR   Relative Risk SARI   Serotonin Antagonist and Re-uptake Inhibitor SES   Socioeconomic Status SD   Standard Deviation SSRI   Serotonin Re-uptake Inhibitor TCA   Tri-cyclic Antidepressants VGH   Vancouver General Hospital WHO   World Health Organization   xix Acknowledgements This thesis is the culmination of many people’s efforts throughout the past number of years. I am indebted to my two supervisors-Drs Carlo Marra and Karim Khan. Carlo, I could not have completed this work without your support, mentorship and friendship. You provided me with a rich learning environment which truly enhanced the Ph.D. experience. You have been a true inspiration and I am forever in your debt.  Karim, you took on a student who you knew very little about and challenged me to work in a completely new field of research. You exposed me to areas of research which were completely foreign to me and let me find my way while never letting me stray too far off course.  To my committee members- Drs. Larry Lynd, Joel Singer and Aslam Anis- I have worked with you all in some capacity or another since 2000 and your support over the past decade plus has been invaluable. Larry, you have been a constant source of support in so many of my academic and non-academic endeavours. Joel, you have always been available to discuss and provide me with words of encouragement to forge onward. Aslam, I will always be grateful that you took the risk of hiring me many years ago, and I am eternally grateful that you did, opening doors to new possibilities that I had no idea existed.  This research would not have been completed without the participants of my investigation into their ED experiences. I extend my gratitude and thanks to all of them. Thank you to all of the nurses in the VGH ED, your help in alerting me to potential participants and enlightening me on the practice of emergency medicine and all of its idiosyncrasies. I also thank Callista Haggis and Suzana Mitrovic who both were great helps in the data collection for this project.  Thank you to my great friends, John Morettie, Ken Harkness, Kevin Fowlie, Chad Robinson, Brett Carels and Ryan Picklyk for reminding me to never take any issue, problem or myself so seriously that it interfered with enjoying life.  Thank you to my parents, Donald and Ferne, you have been supportive throughout not just this Ph.D. journey but all of life’s challenges. To my sisters, brothers-in-law and nieces, you have been a constant source of support and I will never forget that this involved sacrifices from you as well.  Finally to my wife Romy, you are the rock on which this work has been built. You never questioned me throughout this process and made sacrifices so it could be completed. For that I give my thanks and love. To my daughter Anna, you were a constant inspiration and motivator, even if it was because you wanted me to watch Dora with you.  Our next chapter as a family will be even better than the last, I love you both.  1 Chapter  1: Introduction Falls are a significant cause of morbidity and mortality among the elderly.(1, 2) Approximately one third of all adults ≥65 years of age experience at least one fall each year and half of those suffer recurrent falls.(1, 3, 4) Falls have been associated with substantial health care burden including increased hospitalizations, Emergency Department (ED) visits, physician visits, admission to long term care facilities, diagnostic/laboratory investigations, and surgical procedures compared to non- fallers.(2, 5, 6)  As well, falls and fall-related complications have been identified as the 5th leading cause of death in the developed world with 40% of all injury deaths being attributed to falls.(7-11) The fall fatality rate for Canadians ≥65 years of age has been reported as 9.4 per 10,000.(12)  In addition to increased mortality, falls are responsible for 85% of all injury-related hospitalizations for persons ≥65 years of age.(13) The reported duration of hospital stay due to falls ranges from 4 to 15 days.(14) The most common causes of fall- related admissions to hospital are hip fractures, traumatic brain injuries, and upper limb injuries.(14-17) Over 90% of all hip fractures are as result of a fall.(18, 19)  Falls by the elderly also place a substantial burden on the ED and account for 10- 40% of all ED presentations by persons ≥65 years of age.(20-23)  Among fall-related ED visits by the elderly, 21-34% of fallers are admitted to hospital with the remainder discharged back into the community.(16, 17)   2 Many intrinsic and extrinsic risk factors for falls have been identified.(10, 24) Prescription medication use has been identified as an independent risk factor after adjustment for a number of potential confounders.(25, 26) Similarly, the care received by the elderly person who has experienced a fall, and specifically the care received in the ED, has been shown to significantly impact the risk an elderly person has of experiencing subsequent falls.(16, 27)  The impact of falls by elderly persons on Canadians and the Canadian health care system has been recognized by the federal, provincial, and territorial governments.(12, 28) However, there is little known about the costs of ED falls in Canada and the care a faller receives in the ED. As well, there still remains confusion with respect to the association between specific medication classes and the risk of experiencing a fall by elderly persons. To address these aforementioned research gaps, I investigated the: epidemiology, process, costs, and quality of care provided after an elderly person suffers a fall.  A comprehensive literature review (Section 1.2) was conducted to describe 1) the epidemiology of falls by the elderly (including the identified risks associated with experiencing a fall), 2) current guidelines and recommendations for care after an elderly person has suffered a fall, 3) the role of the ED in providing care to the elderly faller, 4) the economic costs of falls by the elderly, and 5) the use of operations research and the potential to model the care of the elderly faller who has  3 presented to the ED. I conclude this chapter with the rationale, objectives, and hypotheses for the four research studies that make up my thesis.  My initial study, described in Chapter 2, is a meta-analysis of the association between falling and the use of various medications classes by elderly persons. Using Bayesian methodologies I updated previously completed meta-analyses (25,26) to provide Odds Ratio (OR) estimates and 95% credible intervals (95% CrI) estimates for 9 specific medication classes.  As noted above, falls by the elderly are a common cause of ED presentations. As well, the care provided in the ED to the elderly faller has been shown to significantly impact an individual’s fall risk and likelihood of subsequent falls and ED admissions.(16, 27, 29, 30) To further investigate the ED care of elderly fallers, the burden which falls by the elderly places on the ED, and the impact of potential changes to the care delivered in the ED, I recruited and prospectively collected data on a sample of elderly fallers who had presented to the Vancouver General Hospital (VGH) ED. These data were used in Chapters 3-5.  In Chapter 3, I investigate the characteristics and care of a sample of elderly fallers with respect to the guidelines and recommendations for post fall care and wait times in the ED.(11, 16, 31-33)  Chapter 4 describes the results of my study estimating the cost of a fall that requires care in the ED.   4 In Chapter 5, using the methods of operations research, I modelled the care provided to the elderly faller. Operations research applies mathematical tools to develop system models and discrete event simulations (DES).(34) Applying the methods of operations research, a DES model was developed to simulate the care path of the elderly faller in the ED. Using the DES model I was able to assess potential changes to the ED care path and their impact on the time elderly fallers spent in the ED and the estimated costs/potential savings of making these changes. Finally, in Chapter 6, I provide an integrated discussion of the studies completed and suggest directions for future research.  1.1 Definition of terms While there is no single definition for the term ‘elderly person’, in the context of this thesis ‘elderly person’ refers to an individual ≥70 years of age.  In academic literature the term ‘fall’ has been given several definitions, which are reviewed in detail in Section 1.2.  For the purpose of this thesis, the term ‘fall’ is defined as “an unexpected event in which the participants (sic) come to rest on the ground, floor, or lower level.”(35)  By extension, the term ‘fall risk’ refers to the risk (i.e., probability) of this event occurring, either at a population level or an individual level. The term ‘fall risk factor’ is used as shorthand for the phrase ‘risk factor for falls’.  The term ‘faller’ is used to describe a person who has experienced at least one fall; the term ‘non- faller’ refers to a person who has never experienced a fall. The term ‘elderly faller’ refers to an elderly person who has experienced at least one fall.  When reporting on the living arrangements of an elderly person in the context of this thesis, “community  5 dwelling” refers to individuals who are not at the time of their fall living in a long term care facility; the term “non-community dwelling” refers to individuals currently living and receiving care in a long term care facility.  Given the relative lack of agreement on the definition of each term, throughout the thesis, when I discuss a study in which the authors have used the term ‘fall’, ‘elderly person’, or a related term (e.g., ‘senior citizen’) I note the definition used in that particular study.  This lack of standardized terminology highlights a significant issue in the fall literature, the potential compromise of study comparability.  1.2 Literature review 1.2.1 The epidemiology of falls: Definition and incidence Before investigating the incidence of falls, it is important to define what constitutes a fall. In 2006, Zecevic and colleagues published a study comparing the definitions of what constitutes a fall in academic research to the perceptions by community dwellers and health care providers of what constitutes a fall.(36) Data were collected from two groups: 477 community dwellers ≥55 years of age; and 31 health care providers. Zecevic showed that community dwellers and health care providers more often focused on the causes and consequences of a fall, while researchers tended to be more descriptive of the fall event itself. Zecevic reported that intuitively we understand the term ‘fall’, yet we struggle to articulate exactly what comprises a fall.(36)  Without a universal definition of the term ‘fall’ (hereafter ‘fall definition’), it is  6 difficult to compare findings across studies designed to assess the incidence, risk, prevention, and burden of falls.(14, 36-38)  In a systematic review of fall definitions used in randomized, controlled fall prevention trials, Hauer et al. noted that among 90 identified trials, only 46 provided a specific fall definition. The definitions used in these trials varied substantially.(38) This literature review revealed that in many of the trials, researchers made subjective decisions that influenced the methods of reporting and defining falls.(38) As noted by Zecevic et al. in their review, the differences in fall definitions (as seen in Table 1-1)(36) can be substantial and may lead to large differences in fall rates reported across studies.(35)  As noted in Table 1-1, the Kellogg fall definition was one of the first standardized definitions and was the product of an international working group for the prevention of falls by elderly persons. This fall definition was designed for use in research. It described a fall as: “…an event which results in a person coming to rest inadvertently on the ground or other lower level and other than as a consequence of the following: sustaining a violent blow, loss of consciousness, sudden onset of paralysis as in an stroke, an epileptic seizure.”(39)  Subsequent variations of this fall definition are similar, highlighting  “…a sudden unintentional or unexpected change in position resulting in an individual coming to rest with some part of their body on a lower level.“(36)   7 In 2006 the Prevention of Falls Network of Europe (ProFaNE) published a consensus statement that recommended an updated fall definition for use in falls research. ProFaNE defines a fall as “an unexpected event in which the participants (sic) come to rest on the ground, floor, or lower level.”(35)  The ProFaNE group further recommends the following question to identify if study participants have experienced a fall: “have you had any fall including a slip or trip in which you lost your balance and landed on the floor or ground or lower level?”(35). The most notable difference between the Kellogg and ProFaNE definitions is the absence of any mention by the ProFaNE definition of excluding falls due to external health issues i.e. stroke or loss of consciousness.  As noted by the ProFaNE group, using a standardized fall definition when investigating whether a fall has occurred allows for increased comparability across studies. As such, where data were prospectively collected as part of my thesis studies, I used the ProFaNE fall definition and advice on proper questioning to ascertain whether a fall had occurred.  1.2.1.1 Definition of an injurious fall The term ‘injurious fall’ is widely used in the literature, yet there is no standard definition of what constitutes an injurious fall.  As with inconsistent fall definitions, the lack of a standard definition of an injurious fall results in different incidence estimates and risk factor identification, while restricting comparisons across studies.(36-38)   8 A number of studies have used proxies to define an injurious fall, such as a fracture,(40-43) a faller’s need for medical attention(44-51), or the recording of an ICD-8,(52) ICD-9,(53-55) or ICD-10(56) code indicating a fall-related Emergency Department (ED) visit or hospitalization. Table 1-2 shows a sample of injurious fall definitions taken from English language studies identified using the PubMed search terms “injurious”, “falls”, and “elderly,” published between 1988 and January 2011.  1.2.1.2 Estimates of the incidence of falls A number of studies using different fall definitions, different methods of fall data collection and completed in unique populations have been published to assess the incidence of falls.(7-9, 57-60)  In the following sections, I describe two Canadian studies that investigated the incidence of falls among elderly persons (≥65 years of age).(9, 60) I also discuss five pivotal prospective cohort studies, completed in the United States, New Zealand, Australia, and Sweden, that documented the incidence of falls among community and non-community-dwelling elderly persons.(7, 8, 57-59)  1.2.1.2.1 International estimates of the incidence of falls among elderly persons A number of prospective cohort studies have documented the incidence of falls among elderly persons across international settings. In this section I highlight five studies regarded as pivotal in the assessment of the incidence of falls in the elderly using prospective methods of data collection.(7, 8, 57-59)   9 In 1988, Tinetti and colleagues prospectively monitored 336 community-dwelling Americans ≥75 years of age to estimate the incidence of falls using the Kellogg definition.(8, 39)  Participants completed monthly falls diaries (the preferred method for falls ascertainment due to its potential to reduce recall bias (35)) and bi-monthly telephone interviews over a 12-month period. It was found that 32% of study participants fell at least once and 9% fell at least twice. Although no definition of injurious fall was employed, it was reported that 24% of those who fell sustained a serious injury and 6% suffered a fracture.(8)  A year later, Campbell and colleagues published their investigation of a cohort of 761 New Zealand residents who were ≥70 years of age.(61) Using a fall definition of “…any unintended contact with the ground,” participants were asked to complete a monthly falls diary. They were also contacted monthly for a telephone interview to document any fall that had occurred since the previous follow-up. Five hundred and seven falls occurred during 8,914 months of follow-up. Among all participants, 35.2% (n=268) suffered at least one fall, with 39.6% of females and 28.4% of males reporting at least one fall.(61)  Lord et al. completed a 12-month study of 341 community-dwelling elderly persons ≥65 years of age using the Kellogg definition to define a fall.(39, 57) Falls data were collected bi-monthly using mailed questionnaires, which participants completed and returned to the investigator. During the 12 months of follow-up, 287 falls were  10 reported by 134 (39.3%) participants. Among fallers, 63 (47.0%) fell once, 45 (33.6%) fell twice, and 26 (19.4%) fell ≥3 times.(57)  In an assessment of persons >85 years of age completed in Sweden, von Heideken Wagert et al., followed 220 participants for six months.(58)  Using the ProFaNE fall definition, data on the incidence of falls and fall-related injuries were collected using diaries and bi-weekly telephone calls. It was found that 91 (41.1%) participants suffered a fall during the follow-up period.(58)  Most recently, in 2010, Delbaere and colleagues conducted a prospective cohort study of 500 community-dwelling people between 70 and 90 years of age in Australia.(59) Delbaere et al. used the ProFaNE definition and data were collected using monthly diaries and telephone interviews over a 12-month period. Falls were categorized as being injurious or non-injurious, where injurious falls were those causing an injury such as a bruise, laceration, or fracture.  In this study, 214 of participants (43.6%) suffered one or more falls and 166 (28.5%) suffered at least one injurious fall.  Among participants who reported a single fall during the follow-up period, 60% reported their fall to be injurious. Among recurrent fallers (n=94), 69 (73.4%) reported that at least one of their falls caused an injury such as a bruise, laceration or fracture.(59)   11 1.2.1.2.2 Canadian estimates of the incidence of falls among elderly persons In 1993 O’Loughlin et al. published the first study that prospectively assessed the incidence of falls among community-dwelling Canadians ≥65 years of age.(9) O’Loughlin et al. used the Kellogg definition (39) to identify fallers from non-fallers. She defined injurious falls as …”falls in which the subject reported sustaining one or more injuries that resulted from the fall.”(9)  Four hundred and nine community- dwelling individuals ≥65 years of age were recruited for their study. O’Loughlin et al. utilized falls diaries and asked participants to place a sticker on a calendar on all dates on which they suffered a fall. She conducted phone interviews every 4 weeks for 48 weeks to collect data on any fall that may have occurred since the last contact. O’Loughlin found that 118 (28.8%) of participants fell at least once during the 48 weeks of study and 47 (11.5%) fell at least twice, with 73 (17.9%) of all fallers reporting a fall with injury.  The incidence of falls increased with age, with 36.4% of females and 40.9% of males 80 to 92 years of age reporting a fall during the study period. Among fallers who were female, 26% reported an injurious fall.(9)  In 2009 Leclerc et al. completed a study on the risk of recurrent falling (≥2 falls) in a cohort of community-dwelling-persons ≥65 years receiving publicly funded, home- based nursing care.(60) Similar to O’Loughlin et al., participants used falls diaries to record the dates on which they suffered a fall. Over a six-month follow up period, Leclerc et al. conducted monthly telephone interviews with participants to determine fall status.(60) For this study the ProFaNE definition (35) was used to identify fall status. Among the 868 participants, 250 (28.8%) suffered a fall within 6 months of  12 study entry and 99 (11.4%) had ≥2 falls.(60)  The authors did not attempt to differentiate falls as injurious or non-injurious.  The two studies completed in Canada used different fall definitions in similar populations of individuals ≥65 years of age. While the length of follow-up was slightly longer in O’Loughlin et al’s study there was little difference (<1%) in the estimates of rates of falling or for multiple falling episodes.  However, as shown in Table 1-3, the Canadian estimates of the incidence of falls are lower than those reported in the international pivotal trials. This may be reflective of differing risk factors and potentially underreporting of falls in the Canadian studies.  1.2.2 Risk factors for falls A number of different factors are associated with an increased fall risk. The World Health Organization (WHO), as illustrated in Figure 1-1, categorizes fall risk factors into four dimensions: environment, socioeconomic, biological, and behavioural.(14) It is recognized that there are interactions between the four dimensions of fall risk such that the impact of multiple risk factors is not simply the sum of its parts. Rather, having multiple risk factors across dimensions often increases the intensity of the risk factor and thus the likelihood of suffering a fall.(8, 10, 14, 57) In the following section, I discuss the fall risk factors in each dimension and describe the unique impact of each on the occurrence of falls.   13 1.2.2.1 Environmental risk factors 1.2.2.1.1 Home hazards Environmental fall risk factors include the features of the physical surroundings both in and outside the home.(10) Within the home, potential fall risk factors include uneven or slippery floors, carpets or rugs, inappropriate furniture and/or obstacles, poor lighting, stairs that were poorly designed/maintained without handrails and the absence of proper safety devices.(8, 62, 63)  It is hypothesized that many home hazards may be minimized if not eliminated through simple interventions including home assessments and modifications.(64, 65) In 2006, Lord and colleagues completed a systematic review in which they assessed the available evidence regarding home hazards and the potential for home modifications to reduce falls.(66) Among the five identified randomized controlled trials only two showed a significant reduction in the incidence of falls after a home assessment. Nikolaus and Bach report that among persons ≥65 years of age, those who received a home assessment by an occupational therapist with advice on home modifications alongside training in using devices had a reduction in fall risk compared to those receiving usual care (incidence risk ratio: 0.69, 95%CI 0.51- 0.97).(65) Cumming et al., in a study of elderly (≥65 years of age) community- dwellers, compared the fall incidence between those who received usual care to those who received a home assessment and supervised home modifications.(67) He reported that the interventions were effective in reducing falls in those who had reported previous falls (risk ratio: 0.64, 95%CI 0.50-0.83). However, Cumming et al.  14 reported that among those who had not previously fallen, compared to those receiving usual care, the intervention had no impact on the likelihood of experiencing a fall during the 12 months of the study (relative risk: 1.03, 95%CI 0.75-1.41).(67) Following Cumming et al’s findings, Lord contended that exposure to home hazards alone are not sufficient to cause a fall in persons with no history of falls.(66, 67)  As such, Lord states that home hazard reduction should be targeted to the elderly with a history of falls and mobility limitations.(66)  1.2.2.1.2 Weather conditions There has been some investigation regarding the impact outdoor temperature has on falls. Specifically it has been put forward that inclement weather, including snow, ice, and rain, increases the propensity of outdoor falls.  In a population of Finnish seniors ≥70 years of age, Luukinen et al. report when the temperature was below minus 20 degrees Celsius there was a higher incidence rate (774 falls/1000 person years, 95%CI 227-1614) of outdoor falls than when the temperature was above 9 degrees Celsius (173 falls/1000 person years, 95%CI 144-206).(68) However, there was no statistically significant difference in the incidence rate of indoor falls (573 falls/1000 person year, 95%CI 158-1412) when the temperature was below minus 20 degrees Celsius or above 9 degrees Celsius (237 fall/1000 person years, 95%CI 144-206).  Luukinen et al. suggest that indoor falls are not impacted by weather changes since indoor environmental hazards do not change with the weather, conversely changes in weather significantly alter the environment and likelihood of experiencing a fall.(68) Similarly, Campbell, in a cohort of elderly women ≥70 years  15 in New Zealand saw a statistically significant increase in the risk of fall as temperature fell below 1 degree Celsius (OR 1.53, 95%CI 1.21-1.84) when compared to the probability of suffering a fall when the temperature was above 1 degree Celsius.(69)  As the two identified studies did not use the same temperatures in assessing different weather conditions, comparison between the two studies cannot be made. The relative dearth of studies assessing the impact of environmental factors highlights an area of falls research requiring further study.  1.2.2.2 Socioeconomic risk factors Low socioeconomic status (SES) is generally associated with impaired health.(70, 71) Although the WHO suggested that low SES is associated with increased falls there is surprisingly little known about the association between falls and SES.(14) As noted by Todd and colleagues, SES can be measured in a number of ways.(72) Identified measures of SES include aspects/categories based upon income, residency ownership or tenancy, occupation classification, highest level of education attained, geographical location of home, and proximity to healthcare, amongst other variables.(72)  Research on the relationship between SES and falls has yielded mixed results as no completed study has been powered sufficiently to assess the impact of any measures of SES on falling.(72)  In their meta-analysis of potential risk factors of falls for community-dwelling older people (≥65 years of age),  Deandrea et al. found that the most commonly used SES measure in the falls literature was education.(73) Seven completed studies were  16 identified that, among the risk factors for falling assessed, included education as a potential risk factor (low education/SES vs. intermediate/high education/SES).  The results of their meta-analysis showed that when compared to those with high SES, low SES was not associated with an increased likelihood of experiencing a fall (OR 1.01, 95%CI 0.88-1.16) or with experiencing multiple falls (OR 0.81, 95%CI 0.62- 1.05).(73)  1.2.2.3 Biological risk factors Included in biological risk factors for falls are non-modifiable factors such as age and gender, as well as potentially modifiable factors such as physiological and medical conditions.(14)  Advanced age has been shown to be a significant risk factor for falls, with both the incidence of falls and the severity of fall-related injuries increasing with age.(11, 74, 75) Deandrea completed a meta-analysis which identified that among community dwelling older people each 5 year increase in age was associated with statistically significant increase in falls (OR 1.12, 95%CI 1.07- 1.17).(73)  As well, compared to males, the risk of experiencing a fall or being a recurrent faller is higher among females, with OR estimates of 1.30 (95% CI 1.18- 1.42) and 1.34 (95% CI 1.12-1.60) respectively.(7, 8, 73)  A number of studies have shown a strong association between specific chronic conditions and falling.  In their meta-analysis, Deandrea et al. found several chronic conditions to be associated with an increased likelihood of experiencing a fall, including cognitive impairment, physical disability (undefined), depression, stroke,  17 urinary incontinence, rheumatic disease (undefined), dizziness and vertigo, hypotension, diabetes, pain, Parkinson’s disease, gait problems, vision impairment and hearing impairment.(73) As well, asthma, and chronic obstructive pulmonary disease have been observed to increase the risk of falls.(76, 77)  Chronic conditions are common in the elderly population,(78, 79) with 79% of Canadian older adults (≥65 years of age) reporting that they suffer from at least one chronic condition and 26% reporting at least three chronic conditions.(80, 81) Assessment of an elderly person’s chronic conditions should be rigorously conducted when determining the proper care to be provided. Health care providers should include education on falls and fall prevention strategies when providing care for a person with a chronic condition identified as increasing fall risk.  1.2.2.4 Behavioural risk factors Behavioural risk factors include perceptions, activities and lifestyle choices.  These risk factors are often seen as modifiable through either alerting the elderly person to the impact of these behaviours on their individual fall risk and/or through changes in care provided.(14)  Identified behavioural risk factors include medication use,(25, 26) physical inactivity,(11) fear of falling,(82) alcohol use,(31) and inappropriate footware.(14, 31)  The most studied behavioural fall risk factor is medication use, including poly-pharmacy (often referred to as >4 medications) (11, 14, 25, 26, 83)) and specific medication classes. The following sections describe the current knowledge with respect to medication use and fall risk.  18 1.2.2.5 Medication use and fall risk A number of studies have measured the association between medication use and fall risk.  Given that many elderly persons report a number of chronic medical conditions,(80, 81) it is not surprising that elderly persons are prescribed a number of medications. (84)  Ramage-Morin analysed Canadian population-based administrative data reporting that 97% of institutional residing seniors (≥65 years of age) and 76% of community dwelling seniors were currently taking at least one medication.(85) In addition, they report that 53% of all seniors living in an institution and 13% of community-dwelling seniors take more than five medications.(85) While, as described above, numerous medical conditions are associated with falls, the number and type of prescription medications are independently associated with experiencing a fall.(86)  1.2.2.5.1 Risks associated with poly-pharmacy As the number of medications taken by an elderly person rises, so too does their likelihood of experiencing a fall. (2, 59, 87-95)  In their review of the association between medication and falls, Boyle et al. highlight that the risk of concurrent use of multiple prescription medications, regardless of the class, is an independent risk factor for falls.(86)  Leipzig et al. reported in their systematic review that, compared to those taking three or fewer medications, the use of four or more medications was associated with an increased risk of single and recurrent (≥2) falls with OR estimates ranging from 0.81  19 (95%CI 0.55-1.39) to 2.9 (95%CI 1.27-6.65) for experiencing a single fall and 1.28 (95%CI 0.56- 2.93) to 2.91 (95%CI 1.26-6.53) for experiencing recurrent falls.(26) Due to heterogeneity between studies, specifically with respect to what was classified or identified as a medication, Leipzig et al. did not complete a meta- analysis to estimate the association between poly-pharmacy and falling. (26) Deandrea et al., in their meta-analysis of community-dwelling elderly persons (≥65 years of age) reported that each single drug increase was associated with increased likelihood of suffering a fall with an OR estimate of 1.06 (95%CI 1.04-1.08).(73)  Among studies published since Leipzig et al’s meta-analysis (Table 1-4), three reported on the association between poly-pharmacy and falling.(91, 96, 97) Comparing those using ≥4 medications to those using <4 medications, two studies showed using ≥ 4 medications to be associated with an increased likelihood of experiencing a fall.(91, 97)  Similarly, Neutel et al. compared users of 5 to 9 medications and ≥10 medications to those using ≤4 medications.(96) They found that elderly persons taking 5 to 9 medications had an increased likelihood of suffering a fall when compared to those taking ≤4 medications (OR 4.3, 95%CI 1.8-10.1). As well, for those taking ≥10 medications, their probability of suffering a fall was significantly higher (OR 6.1, 95%CI 1.9-15.9) than those taking ≤4 medications. Neutel et al’s findings remained unchanged after adjustment for a number of potential confounders including age, gender, duration of time spent in long term care facility, and dementia.(96)  20 Given the results of the previous meta-analyses and research studies it is recommended that potential benefits and risks of prescribing additional/multiple medications be continuously reassessed to minimize the potential for adverse events such as falls. Specifically, the potential harms of prescribing multiple medications should be considered in those with pre-existing conditions that have been associated with falls.(86)  1.2.2.5.2 Studies of specific medications In 1999, Leipzig et al. published a pair of meta-analyses that assessed the association between 22 different medication classes and fall risk among elderly persons ≥60 years of age.(25, 26)  These pooled analyses of English-language articles published between 1966 and March 1996 assessed the association between falls and the use of psychotropics, antidepressants, neuroleptics, tri-cyclic antidepressants, neuroleptics, sedative/hypnotics, benzodiazepines, long acting and short acting benzodiazepines, any diuretic, thiazide diuretics, loop diuretics,  beta- blockers, centrally acting anti-hypertensives, calcium channel blockers, angiotensin converting enzyme inhibitors, nitrates, type 1A antiarrhythmics, digoxin, narcotics, non-steroidal anti-inflammatory drugs, aspirin, and non-narcotic analgesics. Psychotropics, neuroleptics, sedative/hypnotics, antidepressants, both long and short acting benzodiazepines, diuretics, and type 1A antiarrythmics were associated with a statistically significant increase in fall risk (Table 1-5).(25, 26).   21 Hartikainen et al. completed a systematic review of studies published between 1996 and 2004 that assessed medication use and fall risk among persons ≥60 years of age.(98) The authors concluded, based on the OR estimates reported in the 22 studies, that psychotropics, benzodiazepines, antidepressants, and anti-psychotics were associated with increased risk of falls. However, they did not complete any pooling or meta-analysis of the available data. In addition to their comments on the association between the aforementioned medication classes and falling, Hartikainen et al. identified a number of factors which, if addressed, could improve the quality of observational studies investigating medication use as a risk factor for falling. Included in their 19 recommendations were comments on study populations, standardized outcome measures and data collection, classification and descriptions of medications, potential confounders to be included, and the potential for clinical implementation of results.(98)  While Leipzig et al’s meta-analyses only included literature published prior to 1996,(25, 26) a number of new studies have been published since 1996 that investigate the association between medication use and falls. Tables 1-6 to 1-18 report the characteristics and results of studies completed between 1996 and August 2007 following the inclusion criteria used by Leipzig et al.(25, 26)  The most commonly studied class of medications assessed was antidepressants for which 21 studies have been published since 1996 that assessed the association between the use of antidepressants and falling by elderly persons (Table 1-6).(55,  22 87, 96, 97, 99-114) Among these 21 studies 17 reported on the class of medications identified as antidepressants, 6 reported on the selective serotonin re-uptake inhibitor (SSRI) class of antidepressants,(99, 103, 108, 109, 112-114) 5 reported on tri-cyclic antidepressants (TCAs),(87, 99, 103, 114, 115) 2 studies reported on serotonin antagonist and reuptake inhibitors (SARI),(103, 109) and 1 study was completed on non-SSRIs.(113) The reported association between use of antidepressants and falling were quite variable with unadjusted OR estimates ranging from 0.8 (95%CI 0.1-4.4)(115) to 2.38 (95%CI 1.89-3.00)(103), when comparing antidepressant users to non-users.  After antidepressants, the most commonly studied class of medications was sedative/hypnotics (n=15) (Table 1-7).(55, 87, 91, 97, 104, 106, 108-111, 115-119) Similar to Leipzig et al’s findings from the work published prior to 1996, sedative hypnotics use was often shown to be associated with falling when comparing users to non-users.(25)  There were also a number of studies (n=14) which assessed the impact of benzodiazepines on falls (Table 1-8).(87, 89, 96, 100, 102, 103, 107-109, 112, 120- 123) Similar to Leipzig et al.’s findings that elderly person’s using this class of medication had a higher likelihood of experiencing a fall than non-users,(25) all of the studies that looked at benzodiazepines reported an OR greater than 1.0. As well, long-acting and short-acting benzodiazepine users each had an increased probability of being a faller when compared to non-users.  23 Fourteen studies were identified assessing the association between diuretic use and falls (Table 1-9).(89, 92, 95, 96, 100, 104, 108, 111, 116, 121, 124, 125) Similar, to the antidepressants, OR estimates assessing the use of diuretics and its association on falls varied, however only one study reported that diuretic use (specifically loop diuretics and thiazide diuretics(112)) was associated with a statistically significant increase in the likelihood of  falling.  There were 12 studies identified which assessed the use of anti- psychotics/neuroleptics on falls (Table 1-10).(55, 87, 100, 102, 106-109, 111, 119, 126) These studies all showed increased likelihood of suffering a fall when using anti-psychotics/neuroleptics when compared to non-users. This reiterated Leipzig et al’s finding that use of this class of medications was associated with falling.(25)  Eight studies assessed whether psychotropic users had an increased likelihood of suffering a fall compared to non-users (Table 1-11).(45, 99, 104, 109, 119, 122, 124, 125) All eight of the studies’ reported results that were similar to Leipzig et al’s meta- analysis findings that psychotropic users are at an increased probability of suffering a fall when compared to non-users.(25)  Similarly,eight studies measured the association between narcotic analgesics and falling (Table 1-12).(55, 87, 100-103, 108, 111) Only one study showed that narcotic analgesic use was associated with an increased risk of falling when compared to  24 non-users (adjusted OR 1.68, 95%CI 1.39-2.03).(55) All of the other seven studies had inconclusive results (p>0.05).(87, 101-103, 108, 111)  Use of non-steroidal anti-inflammatory drugs (NSAIDs) and its’ association with falling was investigated in six studies with large variation in the OR estimates (Table 1-13).(45, 89, 100, 104, 108, 111, 124) Using data reported by Gluck et al., when compared to those who were not using NSAIDs, NSAID use was associated with reduced likelihood of falling (OR 0.53, no 95% CI reported).(104) Conversely Walker et al., reported that NSAID users when compared to non-users had an increased likelihood of suffering a fall with an OR estimate of 10.02 (95%CI 2.6-38.58). However, this study included only 62 hospitalized patients.(100)  Identified studies which looked at beta-blocker use (n=6) (Table 1-14),(95, 101, 108, 111, 116, 124) calcium channel blocker use (n=5) (Table 1-15),(95, 97, 108, 111, 124) and angiotensin-converting enzyme inhibitors (n=4) (Table 1-16)(95, 97, 108, 124) all showed results similar to Leipzig et al.(25, 26) Comparing users to non- users for each of these three medication classes, it was estimated that there were no statistical differences between the two groups in their likelihood of experiencing a fall (p>0.05).(26)  Among the four studies identified that assessed digoxin use and its association with falls, all reported results similar to Leipzig et al.,(26) that compared to non-users, digoxin use was associated with increased falls (Table 1-17).(100, 108, 112, 125)  25 However, contrary to Leipzig’s findings,(26) one study reported that the increased likelihood of falling when comparing users of digoxin to non-users to be statistically significant (OR 3.3, 95%CI 2.1-5.1).(112)  The medication classes for which fewer than three studies were found that had been completed since 1996 were aspirin (n=1),(124) nitrates (n=2),(108, 124), and antiarrythmics (n=2)(91, 100). For each of these three medication classes the results of the identified studies correlated with the findings of Leipzig et al’s meta-analyses (Table1-5).(25, 26)  I also identified that nine studies had been completed assessing the impact of anti- hypertensives on the risk of experiencing a fall.(55, 91, 92, 95, 100, 105, 116, 118, 121)  This class of medication was not assessed in Leipzig et al’s meta-analysis. The results of these studies were largely inconclusive with no study reporting a statistically significant difference between users and non-users of anti-hypertensives probability of experiencing a fall.  Due to the significant research that has been published since 1996, an updated meta-analysis incorporating both Leipzig et al’s findings and new research completed since 1996 would allow for a greater understanding of the association between fall risk and the use of specific medication classes in the elderly population.   26 1.2.2.5.3 Studies of medication withdrawal While a number of medication classes have been associated with falls, only one randomized controlled trial has assessed the effectiveness of medication withdrawal on reducing falls.(127) In a double-blinded, randomized controlled trial including elderly persons (≥65 years of age) who were taking a psychotropic drug, Campbell et al. observed that the gradual withdrawal of psychotropics was effective in reducing falls.  Compared to patients who remained on their current dose of psychotropics, those who had their psychotropic dose reduced to zero over a 14- week period had a reduction in their overall rate of falls (0.52 vs. 1.16 falls per person year).(127)  1.2.3 Recommended care for elderly fallers A number of practice guidelines have been designed to improve post-fall care and to reduce the risk of future falls among elderly fallers. In the following section, I describe three of these guidelines, highlighting both their unique factors and their commonalities.(11, 16, 31, 83)  1.2.3.1 UCLA Emergency Department Guidelines In 1997, Baraff and colleagues designed a set of practice guidelines commonly referred to as the UCLA ED Guidelines.(31) These guidelines were designed to support ED evaluation, treatment, and outpatient referral for a target population of elderly fallers, specifically “community-dwelling persons older than 65 years who present to the ED after a fall.”(31)  Using a modified Delphi technique, practice  27 guidelines were developed from the available literature and presented to an expert committee of ED physicians and geriatricians. The expert committee’s goal was to reach a consensus on the form and content of the guidelines, while ensuring that undue demands were not made on physicians and nurses in the ED.(31)  The UCLA ED guidelines recommend that post-fall assessments in the target population include an essential history, physical examination, diagnostic testing, intervention determination, and a set of selected health prevention methods.  The essential history includes an assessment of medications, location and cause of falls, patient functional status, medical problems and vision testing. The physical examination requires assessments of vital signs, nutritional status, mental status, and the injury that caused ED presentation; cardiopulmonary examination; and the ‘get-up-and-go’ test, which includes an assessment of the individual’s ability to rise from a chair, walk 10 meters and return to their chair. The diagnostic testing for each elderly faller is to be dictated by the complaints and injuries presented; however, laboratory and/or radiological testing should be completed to confirm or rule out acute or sub-acute conditions and injuries. Interventions to be considered for elderly fallers include medication changes/referral for change to primary care providers, alcohol abuse interventions, geriatric assessments, social service referral, home health referral, physical therapy referral, optometry/ophthalmology referral, footwear/podiatry referral, and/or exercise program planning. Suggested preventative measures include recommendations for immunizations, specifically those for tetanus, pneumonia and influenza, and for calcium and Vitamin D use.  To  28 improve post-fall care, a questionnaire for use in the ED among patients ≥65 years was designed to identify the appropriate ED health care professional to administer assessments/care, potential additional assessments, and specific interventions (Table 1-19).(31)  1.2.3.2 American Geriatric Society/British Geriatrics Society/American Academy of Orthopaedic Surgeons Guidelines for the Prevention of Falls in Older Persons The American Geriatric Society/British Geriatrics Society/American Academy of Orthopaedic Surgeons (AGS/BGS/AAOS) Panel on Falls prevention first published their Guidelines for the Prevention of Falls in Older Persons in 2001 (11) and subsequently updated them in 2011.(83)  The goal of these recommendations is to “…assist health care professionals in their assessment of older patients who are at risk of falling and those who have fallen.”(11)  Similar to the methodology used by Baraff et al. in the development of the ED-specific guidelines,(31) the AGS/BGS/AAOS guidelines were developed following a literature review and recommendations made by the AGS/BGS/AAOS Panel on Falls Prevention.(11)  The AGS/BGS/AAOS guidelines provide a number of specific recommendations for assessment and interventions to prevent falls in older persons. Recommended assessments differ depending on whether the individual is receiving routine care (i.e., is not presenting after a fall); or has suffered ≥1 fall and/or has abnormalities of gait and/or balance. The guidelines recommend that in the context of routine care,  29 assessments should include an annual discussion on falls and a ‘get-up-and-go’ test to detect difficulty or unsteadiness.  For individuals who are unsteady or have difficulty with the get-up-and-go test, further gait and balance assessment is recommended. For individuals presenting with a history of falling as well as those with abnormalities of gait and/or balance, the recommended assessment includes a clinician evaluation of balance and gait (to be performed by a specialist/geriatrician, if deemed necessary by the initial assessor) and a full fall evaluation.(11) The AGS/BGS/AAOS full fall evaluation (11) includes an assessment of 1) history of fall circumstances 2) medications 3) acute or chronic medical problems and mobility levels 4) vision, gait and balance, and lower extremity joint function 5) basic neurological function included mental status, muscle strength; lower extremity peripheral nerves, proprioception, reflexes, tests of cortical, extrapyramidal, and cerebellar function 6) basic cardiovascular status including heart rate and rhythm, postural pulse and blood pressure and where appropriate, heart rate and blood pressure responses to carotid sinus simulation. Following the results of the recommended AGS/BGS/AAOS assessment, there are a number of AGS/BGS/AAOS interventions dependent on whether the individual is community-dwelling; a resident of a long term care or assisted living facility; or a patient in an acute hospital setting.  However, at the time of the guideline development, the evidence regarding fall prevention in the acute hospital setting was  30 not sufficient to make specific recommendations. Interventions recommended for both community-dwelling and long-term care or assisted living facility residents include: education programs on falls; gait training alongside advice on assistance devices; and review and modification of medication, especially for those using four or more medications and/or psychotropics.  In addition, community-dwelling individuals are recommended to receive or undergo a patient-specific exercise program, assessment of the home environment for potential fall risks and modification, treatment of postural hypotension, and treatment of any cardiovascular disorder.(11)  Other interventions that the group did not feel there was sufficient evidence to recommend for fall prevention were bone-strengthening medications, treatment of vision impairment, footwear assessments, and restraints.(11)  The 2011 update to the AGS/BGS/AAOS fall prevention guidelines include amendments to the recommendations regarding assessments and care for the elderly person at risk for falls.(83) With respect to assessments, the elderly person who reports a single fall without observed or reported gait or balance problems does not require a complete falls risk assessment. As well, prevention guidelines now require more intensive questioning surrounding the elderly person’s fall history and circumstances, symptoms at time of the fall, and injuries. Finally there is the expectation that an examination of the feet and footwear, an assessment of activity of daily living skills, an individual’s perceived functional ability and fear of falling, as well as an environmental assessment/adaptation be completed.(83)   31 Updates to the recommended interventions for community-dwelling elderly persons include new exercises, such as balance, gait and strength training. These activities are identified as effective fall prevention measures. It is also recommended that postural hypotension assessment and treatment be added to multi-factorial interventions,(83)as opposed to a single intervention.(11) The updated AGS/BGS/AAOS fall prevention guidelines also note that medication reduction or withdrawal be stressed for all individuals, not just those using four or more medications.(83)  Similarly, the updated AGS/BGS/AAOS fall prevention guidelines report that, among older adults, Vitamin D deficiency is identified as a common problem that can impair muscle strength and increase the risk of falling.(83)  Vitamin D supplementation has been shown to be beneficial in fall prevention, and as such should be considered for all older adults.(83, 128, 129)  Figure 1-2 shows the algorithm of care for the elderly faller as per the 2011 AGS/BGS/AAOS Guideline for the prevention of falls in older persons.(83)  1.2.3.3 The PROFET Guidelines In 1999 Close and colleagues published the results of their “prevention of falls in the elderly” (PROFET) trial.(16) PROFET was a UK-based randomized controlled clinical trial assessing the benefit of a structured inter-disciplinary assessment of elderly (≥65 years of age) fallers in preventing future falls.  Participants were identified after presentation to the Accidents and Emergency Department with a  32 primary diagnosis of a fall. Three hundred and ninety-seven participants were randomized to receive either a structured bi-disciplinary assessment or usual care (no assessment).  During the 12-month follow-up period, participants who received usual care (n=213) reported 510 falls by 111 (52%) individuals.  Compared to those receiving usual care, participants who received the intervention (n=184) had significantly fewer falls (183 falls by 59 individuals (32%), p=0.0002). Similarly, compared to those who received usual care, those who received the intervention had a lower risk of falling in the 12-month follow-up period (adjusted OR 0.39, 95%CI 0.23-0.6).(16)  The intervention proposed by the PROFET trial included both a medical assessment and occupational therapy assessment from which a participant-specific intervention was designed.  The medical assessment is a comprehensive general examination including investigation into visual acuity, balance, cognition, affect, and current medication use.  From the medical assessment, the primary cause of the fall and potential risk factors for future falls are identified. The occupational therapy assessment includes a visit to the participant’s home after the medical assessment in which functional (as measured by the Barthel Index (130) and other functional assessments) and environmental hazards are identified.  After completion of the two assessments, and in collaboration with the participant’s general practitioner, a participant-specific intervention for the prevention of falls is designed. The interventions include education on the risks of falling for all participants alongside  33 appropriate referrals to the relevant services based on the medical and occupational therapy assessments.(16)  1.2.4 Care of the elderly person presenting to the Emergency Department In the following section I outline the role of the Emergency Department (ED) in providing care for elderly persons. Specifically, I highlight the ED’s role in providing post-fall care and preventing future falls.  1.2.4.1 The Emergency Department as a care provider for elderly persons It is estimated that elderly persons account for almost one quarter of all ED visits.(131, 132)  The traditional role of the ED has been to provide “..care for patients with severe or life-threatening conditions that require immediate medical attention.”(133)  However, for the elderly, EDs provide a much wider breadth of services.  In assessments of the reasons for seeking treatment in the ED, it is reported that the elderly not only access the ED for treatment of emergencies, but as an entry point for acute and long-term health care, round-the-clock provision of primary medical care, and as a safety net when various components of the health care system are unable to adequately provide care.(132, 134, 135)  Compared to rest of the adult population, elderly ED patients are more likely to arrive via ambulance, require more urgent care, and to have higher rates of hospitalizations, an increased risk of adverse health outcomes, longer duration of ED visits and elevated risk of readmission.(131, 132)  Twenty-four percent of elderly  34 persons discharged from the ED to the community return to the ED within three months, with six-month rates of return estimated as high as 44%.(136)  Although there have been assessments of the impact of ED interventions designed for the care of the elderly, there is little agreement as to what type of intervention should be implemented.(136, 137)  Two review articles on the impact of interventions in geriatric populations on ED visits have been published.(136, 137) Among those interventions that were designed for elderly ED patients to reduce the likelihood of future ED visits and hospitalizations, many reported statistically insignificant results.(136, 137) Studies that assessed readmission to the ED by comparing an intervention group receiving some sort of geriatric assessment and follow-up to “usual care” had conflicting results.(30, 138-140)  McCusker et al. found that when compared to the usual care group, those that received a geriatric assessment and subsequent referral had a statistically significant increase (adjusted OR 1.6, 95%CI 1.0-2.6) in 30-day ED readmissions.(138) Conversely, Caplan and colleagues reported that receiving a geriatric assessment during their ED visit reduced an elderly person’s likelihood of being admitted to hospital in the 30 days post ED visit, when compared to those receiving usual care (16.5% vs. 22.2%, p=0.48). Also, the time until next ED visit was found to be longer in those in the intervention group (382 days), than for those who received usual care (348 days) (p=0.01).(30)  Mion et al., comparing those who  35 received a comprehensive ED-based geriatric assessment and subsequent referral to usual care found that there was no difference in the likelihood of being readmitted to the ED in the next 30 (OR 1.42, 95%CI 0.95-2.14) or 180 days (OR 0.90, 95%CI 0.66-1.24).(139)   Guttmann et al. assessed 8-day ED readmission rates and reported a relative risk reduction of 27% (95%CI 0-44%) in unscheduled ED visits among those who received an ED based nurse discharge plan (8.5%, 95%CI 6.6- 10.4%), when compared to those who received standard discharge care (11.6%, 95%CI 9.5-16.7%). However, this risk reduction was not statistically significant (p>0.05).(140)  Other outcomes reported in identified ED-based studies(136, 137) were the number of ED visits over a month/30 days(138, 141), 3-month(142), 24-week,(143) 10- month(144), or 1-year(145) time horizon; cost of ED visits;(143, 145) patient satisfaction;(138-140, 144) medication adherence;(140) and quality of life/health status.(30, 138-140, 142-144) No studies using total ED visits or cost reported statistically significant differences between study groups.(137) For those studies which reported patient satisfaction (n=4), two reported no change in patient satisfaction between the intervention group and control group,(138, 144) while the other two studies showed significant increases in patient satisfaction among participants in the intervention group compared to the control.(139, 140)  In 2009, Koehler et al. published results of a randomized controlled trial assessing the impact of a medication assessment by a clinical pharmacist, discharge planning  36 by a care coordinator and telephone follow-up compared to usual care.(146) The study reported the 30-day ED readmission rate was significantly lower in the intervention group (10.0% of participants readmitted to the ED in the 30 days post index ED visit) compared to participants who received usual care (38.1%) (p=0.04).(146)  1.2.4.2 Emergency Department presentations by elderly persons due to a fall The ED is often the first point of care after an elderly person has suffered a fall. Falls are recognized as the primary cause of ED admissions for elderly patients accounting for 10-40% of presentations with ≥50% of these being recurrent fallers.(16, 20, 21, 131)  The elderly faller presenting to the ED is likely to require radiological investigation and laboratory testing with a higher rate of presenting with a fracture and requiring admission to hospital than a non-faller.(21, 147)  1.2.4.2.1 International assessments of the elderly faller presenting to the Emergency Department One of the earliest studies on elderly fallers that presented to the ED was completed in 1996 by Davies and Kenny who assessed 188 fallers who had presented to an Accidents and Emergency Department in the United Kingdom.(20) Thirty percent of patients were admitted to hospital, with 24.4% of all presenting fallers diagnosed with a fracture.  Within this sample, many presented with identified fall risk factors including: fall-related medications (39%), gait abnormalities (35%), depression (31%), and taking ≥4 medications (27%).(20)  37 In 1999, Bell et al. evaluated the characteristics of elderly fallers presenting to an ED in Australia.(21) They reported that elderly fallers presenting to the ED showed high rates of fracture (36.7%), admission (57.2%) and prolonged hospital stay (34.8% of persons hospitalized had stays ≥10 days). Bell also reported that 30.1% of fallers had fallen before.(21)  The same year that Bell et al’s study was published, Close et al’s UK study PROFET (described previously in Section 1.2.3.3) showed that among 1031 fallers presenting to the Accidents and Emergency Department, 31% were admitted to hospital.(16)  Among study participants (n=397), Close et al. observed that, at presentation, 65.1% reported previous falls, and 28.4% were recurrent fallers (≥2 falls).(16)  In 2005, Paniagua and colleagues described the characteristics of the elderly faller presenting to a tertiary care hospital ED in the United States.(148) This retrospective study of persons who had an ED discharge ICD-9 code associated with falling showed 58.5% of fallers required admission to hospital. Among the identified risk factors, 45% were ≥80 years of age, 12% of fallers reported a previous fall, 43% reported taking 4 or more medications, 20% reported alcohol use, and 33% had ≥1 co-morbid condition that was associated with falling.(148)  Similarly, Davison and colleagues investigated a United Kingdom cohort of 314 elderly recurrent ED fallers, noting that at baseline 19.4% of participants had been hospitalized due to their index fall and 26.7% had suffered a fracture.(149) Within Davison’s intervention population (n=159), 93.1% had a balance impairment, 52.7%  38 were on a medication identified as associated with falling, and 47.9% had a visual impairment.(149)  A year later, in 2006, Kalula and colleagues identified a random sample of 100 elderly fallers presenting to an Accident and Emergency Department after a fall.(147) Kalula et al. reported that among their sample population, 89% received some sort of radiology or laboratory testing and that 52% required a referral to an orthopaedic surgeon. Fifty-five percent suffered a fracture, the most common of which was fracture of the femur (27% of all fractures).  Unfortunately, data collection on fall history and risk factors was completed in less than 20% of fallers, restricting further assessment on the generalizability of the sample population.(147)  Most recently, in 2010, Hill et al. conducted a randomized controlled trial to assess interventions such as education and home assessments/adaptations to combat fear of falling.(150)  Among 712 fallers ≥60 years of age who had presented to the ED, Hill found that 365 (51.3%) were multiple fallers, taking on average 5 medications (95%CI 4.7-5.4), with 424 (59.5%) reporting a fear of falling.(150)  1.2.4.2.2 Canadian assessments of the elderly faller presenting to the Emergency Department Based on a literature review using MEDLINE, CINAHL and EMBASE, three Canadian based studies that reported on the elderly faller who has presented to the ED were identified. Donaldson et al., in 2005, assessed the out-patient care of  39 elderly (≥70 years of age) fallers who had presented to the Vancouver General Hospital (VGH) ED.(151)  Among 226 female participants who presented to the VGH ED with a fall, 107 (47.3%) had suffered a fracture and 45 (19.9%) had suffered a hip fracture. In their assessment of the post fall care of 63 participants 18-months post-initial ED fall, they found that 32% of fallers had been referred to their family physician for post fall care, and 24% had been referred to a physiotherapist.(151)  Also at the VGH ED, Salter et al. observed that among 54 elderly fallers who were discharged to the community, only 2 (3.7%) received care that followed the AGS/BGS/AAOS fall prevention guidelines in the 6-months post VGH ED visit.(152) As well, they reported that, in the 6-month’s post VGH ED visit, the mean fall risk score, as measured by Lord’s Physiological Profile Assessment,(153) had significantly increased from baseline (1.6 vs. 2.24, p=0.0005).  Salter and colleagues also reported that the median time the elderly faller spent in the VGH ED prior to discharge to community was 223 minutes (Inter quartile range:135.8)(152)  Lee et al., in 2007, published the results of a randomized controlled trial assessing the impact of personal emergency response systems on anxiety and health care use on elderly (≥70 years of age) ED patients.(114) Among the 86 community-dwelling participants, 52 (60%) lived alone, with a mean Mini-Mental State Examination score of 27.9 (Standard Deviation (SD): 2.81).  Among all participants (n=86), 9 (10.5%) suffered a fall within 30 days of their index fall/ED visit. Among these 9 fallers, 8 (88.9%) suffered an injury (fracture, contusion, or laceration).(114)  40 1.2.4.3 International Emergency Department interventions for the care of the elderly faller The ED has been identified as a key access point for investigating an elderly person’s fall risk factors and providing referrals to fall prevention programs.(16, 148, 150-152, 154) Systematic reviews and meta-analyses of multi-factorial and single factor interventions for fall prevention (155, 156) have had conflicting results. However, significant heterogeneity between studies, interventions, sample populations and outcomes of interest could be biasing the results of meta-analyses towards the null.(155, 156) As a result, multi-factorial and single interventions continue to be recommended for the elderly faller who seeks any medical attention as result of a fall.(83)  As discussed in Section 1.2.3 there are a number of recommendations on the care that should be provided to an individual who has presented to the ED.(11, 16, 31, 83)   However, as shown in previous assessments, this care is rarely provided after discharge from the ED.(151, 152)   41 1.2.4.4 Canadian Emergency Department interventions for the care of the elderly faller In Canada, Lee et al’s 2007 randomized controlled trial is the only study to date to assess an ED based intervention for fallers discharged from the ED to the community. However, their intervention of providing the treatment group with personal emergency response systems was not designed to assess its’ impact on falls or fall risk.(114)  1.2.5 The economic burden of falls An economic burden of illness or cost of illness study provides integral information for the development of health care policy, priority setting, public health management and areas in need of research funding.(157) Through quantifying the effects of an event, condition, or disease into dollar terms, decision and policy makers are able to understand and compare the relative impact of events, conditions, and disease.  As noted by Rice, the estimates of a cost of illness study can be used to (157) 1) define the magnitude of a disease or injury in dollar terms, 2) justify intervention programs, 3) assist in the allocation of research dollars on specific diseases, 4) provide a basis for policy and planning relative to prevention and control initiatives, 5) provide an economic framework for program evaluation.   42 Given that a decision maker is always limited by budgetary and resource constraints, a transparent and clear estimate of the cost of an illness can better equip policy makers when faced with difficult decisions on what services or programmes to fund/provide.(157)  1.2.5.1 Estimation of costs for inclusion in a cost of illness study When completing a cost of illness study, it is necessary to quantify all relevant outcomes in monetary terms through an estimation of their cost. Cost estimation can be split into three unique tasks: identification, measurement and valuation. Cost identification refers to the determination of the items to be included in the cost estimates and their unit costs. Measurement refers to determination of the quantity of each resource that is used. Valuation refers to the estimation of the total cost of resources used, where the identified unit costs are multiplied by the measured unit of resource used.(158)  1.2.5.2 Categories of cost Cost categorization is an important and necessary aspect of the costing exercise since some costs may or may not be relevant depending on the perspective of the evaluation. Separating and reporting costs in distinct categories explicitly states what type of cost is being reported, and provides the decision maker with information on the costs included in the evaluation. The decision maker is then able to understand and determine whether the costs included are relevant given the perspective of the evaluation. The main categories of costs described here and of relevance to falls by  43 elderly persons are: direct medical costs, direct non-medical costs, indirect costs, and intangible costs.(158-160)  1.2.5.2.1 Direct medical costs Direct medical costs refer to costs of medically related resources used in providing treatment. Direct costs can include, among others, drug cost (including costs of drug acquisition, preparation and adverse events), scans and other diagnostic procedures, hospitalizations, physician visits and disease and/or treatment related complications.(158, 159, 161)  In estimates of the costs due to an elderly person suffering a fall, direct medical costs could include the costs of 1) physician visits for diagnosis and post fall care, 2) post fall rehabilitation including physiotherapy and/or occupation therapy, 3) medications prescribed due to injuries suffered from the fall, 4) modifications to current medications, 5) laboratory and radiology testing, 6) hospitalizations due to injuries suffered as result of the fall, 7) transfers to and from hospitals through/via emergency health services/ambulances, 8) nursing home admissions, 9) surgical interventions, 10) time spent receiving care as a patient in urgent care clinics and/or EDs.  44 1.2.5.2.2 Direct non-medical costs Direct non-medical costs refer to the costs accrued by patients and their families that are directly related to the disease, but are not medical in nature.(158, 160) Examples of direct non-medical costs include the costs associated with travelling to the hospital or physician’s office to receive care, the cost of child care while receiving treatment, as well as food and lodging expenses in cases where treatment requires that patients travel from their home for extended periods of time.(158, 160) With respect to estimating the direct non-medical costs due to an elderly person experiencing a fall, potential resource utilizations/costs to consider for inclusion are 1) health improvement activities (including exercise programs, strength and stability training), 2) changes to living conditions/addition of assistive devices, 3) domestic chores, including homemaking, cooking, 4) transportation costs, 5) other out-of pocket expenses.  1.2.5.2.3 Indirect costs Indirect costs refer to the value of changes in productivity due to illness, treatments, or mortality.(158-161) Often an acute illness, injury, or chronic condition will negatively impact an individual’s ability to work in either paid or unpaid labour. Alternatively, treatments which improve an individual’s health can increase their productivity/ability to work in paid or unpaid labour. As well, productivity costs/benefits can be incurred by other family members and caregivers and are not  45 restricted to the individual who has the illness, injury, chronic condition, is receiving treatment or has died.  The estimation of productivity costs is most commonly completed using either the human capital approach or the friction cost approach.(158, 162) The human capital approach will estimate the cost of being unable to work using the earnings of the individual whose productivity/ability to work has changed. With respect to unpaid labour such as homemaking, often a wage/cost will be imputed based on the cost of having an external/paid individual perform the duties previously completed without pay.  Alternatively the friction cost method estimates cost of lost production from the perspective of the employer. As such, included in the cost will be costs associated with training replacement workers and reduced productivity during the period of worker replacement.(158, 163) As shown by Goeree et al. in their comparison of the estimated productivity costs due to schizophrenic mortality, the method chosen can significantly impact the estimated productivity cost, with friction cost estimates often lower than estimates made using the human capital approach.(164)  1.2.5.2.4 Intangible costs Intangible costs are the costs/benefits which may not be associated with a resource utilization(s) for which a unit cost(s) can be applied. This category of costs includes such entities as pain and suffering to a patient because of a condition, or treatment of a condition.(158, 159, 161, 165) Placing a monetary value on intangible costs/benefits has been thought to be quite difficult, but can be done through the use  46 of “willingness to pay” studies to determine a person’s value of identified intangible costs.(159)  However, willingness to pay has been identified as controversial since intangible costs are being valued in monetary terms when there is no real market existing.(165) As such these intangible costs are often not reported in economic evaluations or costing studies.(158, 159, 161)  However, Meltzer in his 2001 Lancet article: “Introduction to health economics for physicians,” remarks that intangible costs and benefits may become crucial in any public debate or decision making process regarding the adoption of an intervention designed to prevent, treat or control a disease.(159)  It has been stated that intangible costs/benefits are not costs/benefits since they do not result in resources being denied/available for another use. Furthermore, since some entities identified as intangible costs such as pain, suffering, anxiety, and fatigue, can be measured using willingness to pay studies and/or utility/quality of life instruments, these items are not strictly intangible.(158)  1.2.5.3 Perspectives and costs to be included in cost of illness studies Perspective refers to the viewpoint from which the evaluation is completed. The perspective used in the evaluation should be determined and stated in the study question, and should be appropriate for the target audience.(166) The costs and outcomes included in the evaluation are determined by who the target audience is; as such, the results of an evaluation can change greatly depending on the perspective. Perspectives often used in estimating costs include societal, ministry of  47 health, other government ministries, the patient/individual, the employer, hospital or care provider, and the agency responsible for providing treatment or programmes.(158, 161)  Many guidelines for completing costing evaluations recommend using the societal perspective. The societal perspective takes into account the perspective of society as a whole including direct medical costs, direct non-medical costs, indirect costs, and intangible costs for which a monetary value has been estimated, related to an intervention and its associated outcomes.(162, 167) The societal perspective is often the preferred perspective to be taken for a burden of illness evaluation, since it takes a global view of all costs regardless of the payer.(158, 161)  1.2.5.4 Estimates of the costs of falls With respect to falls, a number of completed studies have attempted to estimate the cost of falls from a number of different perspectives. In the following sections, I outline some of both international and Canadian attempts to estimate the cost of falls. Given the substantial differences in healthcare systems and costs between countries, the international costs of falls estimated are not directly comparable.  1.2.5.4.1 International estimates of the costs of falls In 2009, Davis et al. and Heinrich et al. completed systematic reviews on the economic burden of falls in the elderly.(5, 6) Both systematic reviews identified cost studies completed from a number of different perspectives, including societal, the  48 health system, care providers, payer, and individual. Davis’ and Heinrich’s systematic reviews reported on 33 studies completed in the United States, United Kingdom, Australia, Switzerland, Ireland, Jamaica, Finland, Sweden and the European Union.  No Canadian study was identified in either literature review as meeting the inclusion criteria.(5, 6)  Davis et al., who provided a comparison of the costs of falls among community- dwelling elderly (≥60 years of age) reported costs of $10,749 (2009 US$) per injurious fall, and $26,483 (2009 US$) per fall requiring hospitalization.(5) Heinrich and colleagues included all studies, regardless of whether they were completed in community or non-community dwelling elderly (≥60 years of age). Heinrich et al. reported costs per fall ranging from $1,059 to $10,913 (2009 US$) and costs for fall related hospitalizations ranging from $5,654 to $42,840 (2009 US$).(6)  However, among the 33 estimates identified, only one- completed in Australia -used data on both falls and costs that were prospectively collected. The most common source of information for fall identification and costing was administrative data, which easily provides information on large cohorts of individuals.(5, 6) However, there are identified limitations to their use such as misclassification and underreporting of events.(168, 169) Similarly, charges for care and procedures were used as a proxy for cost data, even though charges and costs rarely are equal, as charges are not an accurate measure of economic burden.(170)   49 Both systematic reviews identified a number of reasons for the substantial range in costs. These included differences in the definition of a fall, type of costing (prevalent vs. incident fall cases), perspective, cost items collected and measured outcomes. These issues may restrict potential comparisons of results across the estimates.(5, 6)  Specifically, Heinrich et al. note the need for more detailed information on the cost estimates that are specific to the type of fall, post-fall diagnoses, and specific settings in which care was provided.(6)  Davis et al. note that the lack of specific cost-per-fall estimates are a significant gap in the literature as a proper understanding of the burden of falls is necessary to help determine how to best allocate resources for preventive efforts.(5)  1.2.5.4.2 Canadian estimates of the costs of falls A 2011 literature search, employing the search strategies of Davis et al.(5) and Heinrich et al.,(6) did not identify any estimates of cost per fall completed in a Canadian setting using data prospectively collected data from an elderly persons who have suffered a fall.  Markle-Reid and colleagues completed a cross-sectional study examining prevalence, correlates and 6-month costs of health services associated with falls among 109 participants ≥75 years of age receiving home care and at risk of falling. Risk of falling was defined as a previous history of falling, fear of falling, or unsteadiness on their feet. Participants were divided into two groups based on whether or not they had reported at least one fall during the preceding six  50 months.(74) For this study the definition of fall mirrored the ProFaNE definition.(35, 74)  Using a societal perspective, it was found that there was no statistically significant difference between the costs of fallers and non-fallers in total per-person health care costs. However, the mean cost of fallers was $5,749 higher than non- fallers (p=0.39), with higher mean costs of 911/emergency services (difference $11) ambulance services (difference $108), and acute hospitalizations (difference $5,534). Fallers did have lower six-month use and costs associated with family physician visits (difference -$39), endocrinologist visits (difference -$9),(171) and general surgeon visits (difference -$42).(74) However, this study was not powered to truly assess the association between healthcare costs and falls and did not measure costs from the time of the fall, but rather measured costs in the six months prior to data collection. The cross-sectional nature of the study restricts its generalizability. Also, these estimates did not address the substantial potential for biases associated with recall, confounding by indication (including baseline chronic conditions that could predispose an individual to suffer a fall as well as increase the healthcare needs of an individual), and lack of investigation of the temporal relationship between falls and additional healthcare costs.  Using data from the Canadian Institutes for Health Information (CIHI), the SmartRisk group have estimated the burden of falls in Canada from a societal perspective.(171) The burden of illness associated with falls by Canadians ≥65years of age was estimated at $2.03 Billion (2004 Canadian$). However, fallers were identified by the ICD-10 code associated with an ED presentation or acute hospital stay. As a result,  51 falls were classified in a very broad manner that did not differentiate the type of fall, and as such, included falls from skates, skis, boards, blades, furniture, stairs, ladders/scaffolding, and diving.  As such it is difficult to accurately comment on the cost of low-trauma falls in this population.  Similarly, studies assessing the impact of fractures on health resource utilization (172) have not investigated the impact of fall- related fractures specifically.  This is significant as fall-related fractures may have characteristics unique from fractures caused by other, non-fall events. This highlights the need for future cost of illness studies from the Canadian healthcare setting to obtain an accurate estimate of the cost of injurious and other falls in older adults.  1.2.5.4.3 Costs of falls in British Columbia As part of a pilot study to understand the burden of falls on an urban centred tertiary care teaching hospital, I collected data on a population of fallers who had presented to the Vancouver General Hospital (VGH) ED.  Using the VGH ED census, which collects data on every individual who presents to VGH ED, I tracked patients ≥70 years of age who had presented as result of fall. From hospital reports and VGH ED census database, I collected information on patient demographics, diagnosis, admission status, Canadian Triage Acuity Scale (CTAS measured on an I-V scale where I is the most acute), and hospital length of stay. Using a third party payer perspective of the VGH hospital, I estimated the total hospital costs resulting from an elderly faller, including costs of time in the ED and hospitalizations from the fully allocated Vancouver General Hospital Cost Model valued in 2006. I identified 390  52 falls by 281 individuals between December 1, 2006 and March 31, 2007. The majority of fallers were women ≥80 years of age. The total cost of the 390 falls was estimated at $2,520,514. Among those hospitalized (34%) the mean length of stay was 31.6 days with a mean cost of $18,375.(17)  1.2.6 Operations research and discrete event simulation Operations research applies mathematical tools to develop system models and simulations.(34)   In applying these modeling techniques it is possible to gain an understanding of the current status of a system as well as gain insight into the consequences of making alterations to the system without actually altering it.(173) Discrete event simulation (DES), an operations research technique, models the activities of a system as a network of interdependent discrete events.(174) DES models reflect the actual events using the data elements that reflect the entities and activities of the system. These simulations are built with the goal of modeling the behavior of the system and estimating system performance under various scenarios. This method of research has been applied in health care to simulate and model health care delivery in a number of different settings, including surgical scheduling, patient scheduling of diagnostic services, and portering of patients.(34, 175, 176)  1.2.6.1 Operations research and discrete event simulation in the Emergency Department As noted by Connelly and Bair, operations research allows for a quantitative analysis of ED patient flow, wait times, treatment times, and assessments of factors which  53 influence each.(174)  One of the earliest simulations of the ED was completed by Saunders et al. in 1989 who attempted to create a simulation model that could be applied to any ED. Saunders contended that their model realistically reflected the complexity of the ED and allowed for estimation of output data including patient throughput times, queues, and resource utilizations.(173) Subsequent to Saunders, a number of assessments have been completed by different groups in attempts to assess wait times, identify potential patient care bottlenecks and determine staffing needs.(173, 174, 177)  While these assessments have been completed on the systems of the ED, very few assessments have been completed in the ED for specific sub-types of ED patients. As such, there are limitations to the use of these ED system models for an assessment of the impact of a system level change on a specific sub-type of patient.  1.2.6.2 Methods of discrete event simulation When completing a DES, the first step is determination of the questions or problem for which the model will be designed to answer. Subsequent to this, the scope of the study must be defined. Next, an assessment of the current system as it exists and the development of a process/flow diagram is to be completed. These process or flow diagrams provide the backbone of the model as well as outline the type of data required to build the simulation.  Data is then collected on the relevant inputs to build the simulation model. After creation of the model, sensitivity and scenario analyses can be completed to assess the impact of system/model alterations.  54 As noted above, no operations research methods have been utilized to simulate the ED care provided to the elderly or more specifically, to the elderly individual presenting to the ED as result of a fall. As such, there is an opportunity to apply these methods to assess the experiences of elderly fallers presenting to the ED and identify areas of congestion that impact the time, resource utilizations and costs of care.  1.3 Research studies: rationale, objectives, and potential contributions 1.3.1 Study 1 (Chapter 2): Meta-analysis of the impact of 9 medication classes on falls in elderly persons 1.3.1.1 Rationale There currently exists an appreciation that medication use is associated with increased risk of falling.  Using articles published between 1966 and 1996 Leipzig et al. published two meta-analyses measuring the association between suffering a fall and the use of various medication classes in elderly persons.(25, 26) Subsequently, Hartikainen et al. completed a systematic review that reported the results of articles published between 1996 and 2004 but made no formal statistical attempts to pool the results.(98)  Although there have been a number of studies published since 1996 many of these studies did not have the requisite sample size to properly assess the association between use of a specific medication and suffering falls. As such, there is not a definitive understanding on the impact that the use of various medications have on the risks of suffering falls.   55 1.3.1.2 Objective The primary objective of this analysis was to complete a Bayesian meta-analysis incorporating results of Leipzig et al’s(25, 26) work with new study data completed between 1996 and 2007 on medication classes previously assessed. I also aimed to complete meta-analyses on additional drug classes that were not originally assessed by Leipzig et al.  It was my expectation that the results of the previously completed meta-analyses by Leipzig et al.(25, 26) would be shown to be correct in their identification of medications associated with increased risk of falling. Similarly, the use of Bayesian methodologies would increase understanding of the potential risk associated with specific medication classes through estimation of Odds Ratio and 95% Credible Intervals.  1.3.1.3 Potential contribution This was the first meta-analysis to update the previous findings of Leipzig et al. This study highlighted the increased risk of suffering a fall associated with antidepressant, benzodiazepine, and sedative/hypnotic use and the need for caution when prescribing these medications in the elderly.  As well, this study is one of the first Bayesian meta-analyses (178) that incorporated information from a previously completed meta-analysis with new information, representing a significant methodological contribution.   56 For studies 2, 3, 4 (Chapters 3, 4, and 5) I note that my role was to collect, compile and analyse data on care and costs of care for a cohort of elderly fallers (age ≥70 years) who presented to the VGH ED.  Fallers were identified using the ProFaNE falls definition of “…an unexpected event in which the participants (sic) come to rest on the ground, floor or lower level” which was designed specifically for research studies investigating falls and fall prevention.(35)  Dr Karim Khan is the principal investigator of this study which was funded by the Michael Smith Foundation for Health Research.  1.3.2 Study 2 (Chapter 3): The elderly faller’s Emergency Department management: a direct observational study of care delivery and wait times in an urban Canadian university hospital 1.3.2.1 Rationale The Emergency Department is an important contact point between seniors and the medical system.(16, 31, 151, 152) Previously, it has been estimated that between 14-40% of all presentations to the ED are for falls by elderly persons.(20, 21) Previous research undertaken in the VGH ED reported that, annually, ≥1400 elderly persons present to the VGH ED as result of a fall.(151, 152)  As well, among elderly fallers who are discharged to the community-representing 70% of all elderly fallers presenting to the ED-it was observed that in the six months following their ED presentation no patient received care which corresponded to the American Geriatrics Society guidelines for post-fall care.(152)   While there exist three distinct guidelines for the care of the elderly faller who has presented to the ED,(11, 16, 31,  57 83), no study has assessed the care currently being provided in the ED to the elderly faller.  The duration of time spent in the ED is a growing concern and has been identified as a substantial health care issue.(33) Elderly fallers presenting to the VGH ED waited on average over 4 hours to be discharged to the community(152) and more than 10 hours to be admitted to hospital.(17)  The Canadian Triage Acuity Scale (CTAS) was developed to define a patient’s need for timely care based on the perceived urgency of their presenting complaint using a five point scale.(179) The CTAS has also been used to benchmark acceptable waiting times for discharge from the ED, assessment by a ED nurse, and assessment by an ED physician by the Canadian Association of Emergency Physicians (CAEP).(32, 33, 179)  No study has assessed whether these benchmarks are being met when providing ED care to elderly fallers.  1.3.2.2 Objective To help understand the current level of care and ability to provide care in an acceptable period of time, I prospectively collected data on a cohort of elderly fallers who had presented to the VGH ED as result of a fall. For my recruited population, I compared the care received, timing of care received relative to their presentation time, and total time spent in the ED to the current guidelines for post fall care(11, 16, 31, 83) and wait times in the ED.(32, 33, 179)   58 It was my expectation that elderly fallers presenting to the VGH ED are not receiving care which meets the current recommendations provided by the published fall prevention guidelines. Similarly, the care received in the VGH ED does not meet the current standards for acceptable wait times.  1.3.2.3 Potential contribution This study is the first observational study to prospectively collect data on the care given to the elderly faller presenting to the ED. My results suggest that neither the care provided nor the timing of the care provided was within the recognized standards. These data allowed for identification of current care gaps as well as identifying specific areas of concern with respect to the duration of wait times experienced by the elderly faller.  1.3.3 Study 3 (Chapter 4): the cost of fall-related presentations to the Emergency Department: A prospective, in-person, patient-tracking analysis of health resource utilization 1.3.3.1 Rationale Two recent systematic reviews which included studies completed in the United States, Europe, New Zealand reported that the total cost per fall ranged from $10,749-$26,676 (2009 United States Dollars).(5, 6) However, both systematic reviews noted that current studies were limited by a lack of prospectively collected data on health resource utilizations.(5, 6) The authors also noted that better estimates of the costs of falls requires more detailed information on the costs  59 specific to type of fall, post-fall diagnoses, and care setting while using a uniform definition of what constitutes a fall.(6)  As well, neither of the systematic reviews of the peer-reviewed literature identified a study which estimates the cost per fall in the Canadian healthcare setting.(5, 6) This lack of a Canadian specific cost per fall estimate is a significant gap in the literature. A more in-depth proper understanding of the burden of falls is necessary to help determine how to better allocate resources for preventive efforts.(5)  1.3.3.2 Objective I attempt to fill some of the identified gaps by estimating the cost per fall by an elderly person resulting in a presentation to an ED using prospectively collected data on health resource utilizations and their costs.  Similarly, using data on discharge location and diagnosis, I estimate the cost per fall which did not require hospitalization, cost per fall which did require admission to hospital, cost per fall- related fracture, and cost per fall-related hip fracture.  1.3.3.3 Potential contribution Given the growing population of elderly persons in Canada(180) and the high incidence of falls, an understanding of the cost of a fall will allow decision makers to better understand the need for fall prevention programmes relative to other demands for healthcare funding and support. This is the first Canadian based study to use prospectively collected data to estimate the cost per fall, filling an identified gap in  60 the area of falls in the elderly. These data on the cost per fall and cost per sub-type of fall will provide decision makers with a greater understanding of the burden that a single fall places on the healthcare system. They will also complement the currently available global estimates provided in larger administrative database studies.  1.3.4 Study 4 (Chapter 5): An operations research analysis of the Emergency Department care of the elderly faller: Simulation of current care and the impact of providing care that meets wait-time and falls prevention guidelines 1.3.4.1 Rationale Operations research utilizes mathematical methods to model a system.(34)  This method of research has been applied in health care to simulate and model health care delivery in a number of different settings, including surgical scheduling,(34) patient scheduling of diagnostic services, and portering of patients.(175, 176) Discrete event simulation (DES), a type of operations research model, has been applied to emergency medicine to assess wait times and to identify potential patient care bottlenecks.(174, 177, 181)  These DES models can provide insights on the current activities in the ED, as well as assess the impact of hypothetical changes to the system. However, no operations research methods have been utilized to simulate the ED care provided to older people. I propose to assess the experiences of elderly fallers presenting to the ED and identify areas of congestion that impact the time, resource utilizations and costs of care.   61 1.3.4.2 Objective Using prospectively collected data and data taken from the VGH ED census on the care received by the elderly faller in the VGH ED, I aimed to build a DES model which simulates the care received by an elderly person who has presented to the VGH ED as result of a fall. I then used this DES model to estimate the impact of changing the timing and types of care delivered in the VGH ED. Specifically, I will be measuring the impact of following the CAEP benchmarks for wait times(32, 33) and the impact of providing guideline post fall care as defined by the three published guidelines described in Section 1.2.1-1.2.3.(11, 16, 31, 83)  Primary outcomes of interest will be the duration of time an elderly faller spends in the ED, and the costs or potential costs averted from these changes.  1.3.4.3 Potential contribution This research will provide the first simulation of the care that elderly fallers receive in an ED.  Through the completion of this research, I will also be able to report on the potential impact changes to provision of care in the ED could have on the time an elderly faller spends in the ED.  As well, I will be able to report on the costs/opportunity costs of not implementing the changes to the timing and delivery of care.      62 Table 1-1 Fall definitions as taken from Zecevic et al.(36) Author Year Selected Fall Definitions Kellogg group 1987 “A fall is an event which results in a person coming to rest inadvertently on the ground or other lower level and other than as a consequence of the following: Sustaining a violent blow, Loss of consciousness, Sudden onset of paralysis, as in a stroke, An epileptic seizure.” (p.4) Lach et al. 1991 “…an unexpected loss of balance resulting in coming to rest on the floor, the ground, or an object below knee level.” (p. 198) Buchner et al. 1993 “Unintentionally coming to rest on ground, floor, or other lower level; excludes coming to rest against furniture, wall, or other structure.” (p. 301) Means et al. 1996 “…any involuntarily change from a position of bipedal support (standing, walking, bending, reaching, etc.) to a position of no longer being supported by both feet, accompanied, by (partial or full) contact with the ground or floor.” (p. 1032) Berg, Alessio, Mills, & Tong 1997 “…losing your balance such that your hands, arms, knees, buttocks or body touch or hit the ground or floor.” (p. 262) Canadian Institute for Health Information 2002 “…an unintentional change in position where the elder ends up on the floor or ground.” Carter et al. 2002 “…inadvertently coming to rest on the ground or other lower level with or without loss of consciousness and other than as the consequence of sudden onset of paralysis, epileptic seizure, excess alcohol intake or overwhelming external force.” (p. 999) Cesari et al.  2002 “…a sudden loss of gait causing the hit of any part of the body to the floor…” (p. M723) Tideiksaar 2002 “…any event in which a person inadvertently or intentionally comes to rest on the ground or another lower level such as a chair, toilet or bed.” (p. 15) Note: The most diverse fall definitions are presented in the table. Those not shown (Campbell et al., 1999; Campbell et al., 1997; Covinsky et al., 2001; Cummings et al., 1988; Feder et al., 2000; Florida Hospital Association, 2001; Gillespie et al., 2002; Kron et al., 2003; Lajoie & Gallagher, 2004; Lamb et al., 2003; Lamb et al., 2005; Li et al., 2005; Lord et al., 2001; McMurdo et al., 2000; Nevitt et al., 1991; Nurse Assist, 2001; Province et al., 1995; Salva et al., 2004; Tinetti et al., 1988; Tinetti et al., 1997; Tinetti & Speechley, 1989) were slight variations and their repetition was considered redundant.    63 Table 1-2 Definitions of injurious falls Authors Year Outcome Outcome Definition Reported Incidence Nevitt et al.(15)  1991 major injury “..fracture, joint dislocation or laceration requiring sutures.” 6% of all falls Nevitt et al.(15)  1991 minor injury “…lacerations without sutures, bruises, abrasions, sprains, and other minor soft tissue injury 77% of all falls Koski et al.(182)  1998 major injurious fall fractures, joint dislocations, lacerations, and intracranial injuries 32% of all fallers Rizzo et al.(183)  1998 serious fall injuries ... all fractures and joint dislocations, head injuries resulting in loss of consciousness and hospitalization, internal injuries resulting in hospitalization, and joint injuries resulting in hospitalization or in decreased mobility or activity for at least 3 days after the fall 51% of all falls Robertson et al.* (184) 2002 serious injury serious injury if the fall resulted in a fracture or admission to hospital with an injury or required stitches 7% of all falls Robertson et al.* (184) 2002 moderate injury moderate injury if bruising, sprains, cuts, abrasions, or reduction in physical function for at least 3 days resulted or if the participant sought medical help 35% of all falls Fitzharris et al.(185) 2010 injurious falls .. where the following self-reported outcome occurred: a cut, scrape, gash, bruise or fracture was sustained; a head injury resulted or where the fall resulted in hospitalisation 56% of all falls Gill et al.(186)  2008 injurious falls as falls that resulted in contusions (bruises), abrasions (scrapes), lacerations (cuts), sprains or strains, back pain, fractures, head injuries and other unspecified injuries 65% of fallers O'Loughlin et al.(9)  1993 injurious falls …falls in which the subject reported sustaining one or more injuries that resulted from the fall 18% of all fallers Duh et al.(187)  2008 injurious falls an injurious event claim, within 30 days after a fall claim, for fractures of the hip/pelvis/femur, vertebrae/ribs, humerus or lower limbs; Colles' fracture; or head injuries/haematomas.” 75% of all falls *Derivations of this definition have been used in a number of different studies investigating injurious falls. (188-190)  64 Table 1-3 Incidence of falls reported in international and Canadian studies Study Study Setting Population Fall Definition Data Collection Method Sample Size Study Duration Fall Incidence Rate Tinetti et al. 1988 (8) United States ≥75 yrs community dwelling Kellogg Bi-monthly phone interview/monthly falls diaries 336 12 months 32.0% Campbell et al. 1989 (61) New Zealand ≥70 yrs community dwelling “..any unintended contact with the ground” Monthly interviews/falls diaries 761 12 months 35.2% Lord et al. 1994 (57) Australia ≥65 yrs community dwelling Kellogg Bi-monthly mailed falls diaries 341 12 months 39.3% von Heidekien Wagert et al. 2009 (58) Sweden ≥85 yrs community dwelling ProFaNE Bi-weekly phone interviews/falls diaries 220 6 months 40.0% Delbaere et al. 2010 (59) Australia 70-90 yrs community dwelling ProFaNE Monthly interviews/falls diaries 500 12 months 43.6% O’Loughlin et al.1993 (9) Canada ≥65 yrs community dwelling Kellogg 4-week phone interviews/falls diaries 409 48 weeks 29.0% Leclerc et al. 2009 (60) Canada ≥65 yrs community dwelling ProFaNE Monthly interviews/falls diaries 868 6 months 28.8%    65 Figure 1-1 Fall risk factors (14)    66 Table 1-4 Studies published between 1996 and 2007 measuring the association between poly- pharmacy and falls Study Authors (Year) Type of Study Number of Medications Measure of Association Unadjusted estimated measure of association (95%CI) Adjusted estimated measure of association (95%CI) Tromp et al. (1998) (91) Cohort 4 or more Odds ratio 1.4 (1.1-2.1) Gerdhem  et al. (2005) (92) Cohort 4 or more Odds ratio 1.30 (0.95-1.78) Mauer et al. (2005)  (97) Cohort 4 or more Hazard ratio 2.64 (1.04-6.66) Neutel et al. (2002)  (96)   Case crossover 5 to 9 Odds ratio 4.3 (1.8-10.1) 4.0 (1.6-9.9) Neutel et al. (2002) (96) Case crossover >10 Odds ratio 6.1(2.6-14.5) 5.5 (1.9-15.9)   67 Table 1-5 Results of Leipzig et al’s meta-analysis(25, 26) Medication Class Number of studies included in meta-analysis Pooled Odds Ratio estimate (95% Confidence Interval) Psychotropics 19 1.73 (1.52 – 1.97)* Antidepressants 27 1.66 (1.41 – 1.95)* Neuroleptics 22 1.50 (1.25 – 1.79)* Tri-cyclic antidepressants 8 1.40 (0.96 – 2.02) Sedative hypnotics 22 1.54 (1.40 – 1.70)* Benzodiazepines (Both short and long-acting) 13 1.48 (1.23 – 1.77)* Short acting benzodiazepines 9 1.44 (1.09 – 1.90)* Long acting benzodiazepines 9 1.32 (0.98 – 1.77) Diuretics (Any) 26 1.08 (1.02 – 1.16)* Thiazide diuretics 12 1.06 (0.97 – 1.16) Loop diuretics 11 0.90 (0.73 – 1.12) Beta-blockers 18 0.93 (0.77 – 1.11) Centrally acting anti-hypertensives 11 1.16 (0.87 – 1.55) Calcium channel blockers 13 0.94 (0.77 – 1.14) Angiotensin-converting enzymes 10 1.20 (0.92 – 1.58) Nitrates 14 1.13 (0.95 – 1.35) Type 1A antiarrhythmics 10 1.59 (1.02 – 2.48)* Digoxin 17 1.22 (1.05 – 1.42)* Narcotic analgesics 13 0.97 (0.78 – 1.20) Non-steroidal anti-inflammatory drugs 13 1.16 (0.97 – 1.38) Aspirin 9 1.12 (0.80 – 1.57) Non-narcotic analgesics 9 1.09 (0.88 – 1.34) *Exposure associated with statistically significant increased likelihood of suffering a fall vs. non-exposure     68 Table 1-6 Studies published between 1996 and 2007 measuring the association between antidepressant use and falls Study Authors (Year) Type of Study Drug Evaluated Measure of Association Unadjusted estimated measure of association (95%CI) Adjusted estimated measure of association (95%CI) Thapa et al. (1998) (99) Cohort Amitriptyline Rate ratio per 100 person yr 2.2 (2.0, 2.5) 1.9 (1.7, 2.1) Walker et al. (2005) (100) Case-control Antidepressants Odds ratio 1.43 (0.49, 4.14) Desmet et al. (2004) (101) Cohort Antidepressants Odds ratio 6.13* 5.26 (2.26-12.20) Ebly et al. (1997) (102) Cross-sectional Antidepressants Odds ratio 2.23* Ensrud et al. (2002) (103) Cohort Antidepressants Odds ratio 2.38 (1.89, 3.00) 2.40 (1.90, 3.02) Ensrud et al. (2002) (103) Cohort Antidepressants Odds ratio  1.22 (0.97, 1.53) Gluck et al. (1996) (104) Case-control Antidepressants Odds ratio 1.87* Heitterachi et al. (2002) (105) Cohort Antidepressants Relative risk 0.97 (0.46, 2.02) Hien et al.  (2005) (106) Cohort Antidepressants Hazard ratio 1.56 (1.19, 2.04) 1.45 (1.09, 1.93) Hien et al. (2005) (106) Cohort Antidepressants Relative risk 1.96 (1.45, 2.64) Kallin et al. (2002) (107) Cohort Antidepressants Odds ratio  4.66 (1.23, 17.59) Kallin et al. (2004) (108) Cross-sectional Antidepressants Odds ratio 1.51 (1.19, 1.91) 1.33 (1.02, 1.75) Kelly et al. (2003) (55) Case-control Antidepressants Odds ratio  1.46 (1.21, 1.78) Landi et al. (2005) (109) Case-control Antidepressants Odds ratio 1.01 (0.74, 1.36) 0.92 (0.67, 1.26) Lawlor et al. (2003) (110) Cross-sectional Antidepressants Odds ratio 2.02 (1.58, 2.59) 1.53 (1.15, 2.02) Lord et al. (2003) (119) Cohort Antidepressants Odds ratio 1.34 (1.05, 1.72) Maurer et al. (2005 )(97) Cohort Antidepressants Hazard ratio 0.35 (0.13, 0.89) Mustard and Mayer (1997) (111)) Case-control Antidepressants Odds ratio  0.92 (0.75, 1.12) Neutel et al. (2002) (96) Case-crossover Antidepressants Odds ratio 1.7 (0.9, 3.1) 2.0 (0.5, 5.2) Thapa et al. (1998) (99) Cohort Doxepine Rate ratio per 100 person yr 2.4 (2.1, 2.8) 2.0 (1.7, 2.3) Thapa et al. (1998) (99) Cohort Fluoxetine Rate ratio per 100 person yr 2.4 (2.1, 2.8) 1.8 (1.6, 2.1) Thapa et al. (1998) (99) Cohort Imipramin Rate ratio per 100 person yr 2.6 (2.2, 3.1) 2.2 (1.8, 2.6) Souchet et al. (2005 (112)) Case-control Imipraminic antidepressant Odds ratio 3.4 (2.4, 4.8) 3.6 (2.5, 5.1) Arfken et al. (2001) (113) Case-control Non-SSRI antidepressants Odds ratio  1.40 (0.65, 3.03) Thapa et al. (1998) (99) Cohort Nortriptyline Rate ratio per 100 person yr 2.3 (2.0, 2.5) 2.0 (1.8, 2.3) Thapa et al. (1998) (99) Cohort Paroxetine Rate ratio per 100 person yr 2.3 (2.1, 2.6) 1.7 (1.5, 1.9) Ensrud et al. (2002) (103) Cohort SARI Odds ratio  1.11 (0.55, 2.23) Landi et al. (2005) (109) Case-control SARI Odds ratio  0.62 (0.26, 1.76) Thapa et al. (1998) (99) Cohort Sertraline Rate ratio per 100 person yr 2.6 (2.3, 3.0) 1.8 (1.5, 2.1) Nygaard et al. (1998) (115) Cohort Antidepressants Odds ratio 0.8 (0.1, 4.4)  69 Study Authors (Year) Type of Study Drug Evaluated Measure of Association Unadjusted estimated measure of association (95%CI) Adjusted estimated measure of association (95%CI) Kallin et al. (2004) (108) Cross-sectional SNRIs Odds ratio 0.97 (0.60, 1.57) Souchet et al. (2005) (112) Case-control SRI Odds ratio 2.6 (1.8, 3.6) 2.2 (1.5, 3.1) Arfken et al. (2001) (113) Case-control SSRI Odds ratio  2.01 (1.23-3.28) Ensrud et al. (2002) (103) Cohort SSRI Odds ratio  2.61(1.51, 4.50) Landi et al. (2005) (109) Case-control SSRI Odds ratio  0.99 (0.69, 1.41) Lee et al. (2003) (114) Cross-sectional SSRI Odds ratio 1.02* Thapa et al. (1998) (99) Cohort SSRI Rate ratio per 100 person yr 2.4 (2.2, 2.6) 1.8 (1.6, 2.0) Thapa et al. (1998) (99) Cohort SSRI < 20 mg/d Rate ratio per 100 person yr  1.5 (1.3, 1.7) Thapa et al. (1998) (99) Cohort SSRI ≥20 mg/d Rate ratio per 100 person yr  1.9 (1.7, 2.2) Kallin et al. (2004) (108) Cross-sectional SSRIs Odds ratio 1.67 (1.31, 2.13) Weiner et al. (1998) (87) Cohort TCA Odds ratio  1.50 (0.69, 3.27) Thapa et al. (1998) (99) Cohort TCA ≤10 mg/d Rate ratio per 100 person yr  1.2 (1.0, 1.5) Thapa et al. (1998) (99) Cohort TCA >50 mg/d Rate ratio per 100 person yr  2.4 (2.1, 2.8) Thapa et al. (1998) (99) Cohort TCA 11-25 mg/d Rate ratio per 100 person yr  2.0 (1.8, 2.3) Thapa et al. (1998) (99) Cohort TCA 26-50 mg/d Rate ratio per 100 person yr  2.1 (1.8, 2.3) Ensrud et al. (2002) (103) Cohort TCA high anticholinergic activity Odds ratio 1.32 (0.90, 1.94) Ensrud et al. (2002) (103) Cohort TCA low anticholinergic activity Odds ratio 1.22 (0.53, 2.77) Kallin et al. (2004) (108) Cross-sectional TCA Odds ratio 1.44 (0.71, 2.92) Thapa et al. (1998) (99) Cohort Trazodone Rate ratio per 100 person yr 1.9  (1.7, 2.1) 1.2 (1.0, 1.4) Thapa et al. (1998) (99) Cohort Trazodone < 50 mg/d Rate ratio per 100 person yr  1.5 (1.2, 1.8) Thapa et al. (1998) (99) Cohort Trazodone ≥50 mg/d Rate ratio per 100 person yr  1.1 (1.0, 1.3) Ensrud et al. (2002) (103) Cohort TCA Odds ratio  1.06 (0.81, 1.40) Landi et al. (2005) (109) Case-control TCA Odds ratio  0.85 (0.32, 2.20) Lee et al. (2003) (114) Cross-sectional TCA Odds ratio 2.04* Thapa et al. (1998) (99) Cohort TCA Rate ratio per 100 person yr 2.4 (2.1, 2.6) 2.0 (1.8, 2.2) Nygaard et al. (1998) (115) Cohort ≥2 antidepressants Odds ratio  1.0 (0.2, 4.9) *Calculated from available data   70 Table 1-7 Studies published between 1996 and 2007 measuring the association between sedative/hypnotic use and falls Study Authors (Year) Type of Study Drug Evaluated Measure of Association Unadjusted estimated measure of association (95%CI) Adjusted estimated measure of association (95%CI) Mustard and Mayer (1997) (111) Case-control Anxiolytics/sedatives/ hypnotics Odds ratio  1.35 (1.09, 1.68) Rozenfeld et al. (2003) (116) Cross-sectional Anxiolytics/sedatives Odds ratio 1.14 (0.77, 1.69) Kallin et al. (2004) (108) Cross-sectional Flunitrazepam Odds ratio 1.28 (0.90, 1.82) Lawlor et al. (2003) (110) Cross-sectional Hypnotic/anxiolytics Odds ratio 1.94 (1.41, 2.68) 1.41(1.00, 1.98) Avidan et al. (2005) (117) Cohot Hypnotics Odds ratio 1.29 (1.13-1.48) 1.13 (0.98-1.30) Weiner et al. (1998) (87) Cohort Other sedatives Odds ratio  1.05 (0.52, 2.12) Hien et al. (2005) (106) Cohort Sedative/anxiolytic Hazard ratio 1.37 (1.10, 1.72) 1.19 (0.94, 1.50) Hien et al. (2005) (106) Cohort Sedative/anxioytic Relative risk 1.90 (1.42, 2.55) Chu et al. (2005) (118) Cohort Sedative/hypnotics  Odds ratio 0.87* Chu et al. (2005) (118) Cohort Sedative/hypnotics  Odds ratio 0.78* Landi et al. (2005) (109) Case-control Sedative/hypnotics  Odds ratio 1.25 (0.98, 1.61) 1.08 (0.83, 1.41) Maurer et al. (2005) (97) Cohort Sedative/hypnotics  Hazard ratio 1.10 (0.50, 2.4) Gluck et al. (1996) (104) Case-control Sedatives Odds ratio 1.00* Kelly et al. (2003) (55) Case-control Sedatives Odds ratio  1.15 (0.97, 1.36) Lord et al. (2003) (119) Cohort Sedatives Odds ratio 1.27 (1.01, 1.60) Neutel et al. (2002) (96) Case-crossover Sedatives Odds ratio 1.6 (0.8, 3.2) 1.3 (0.5, 3.1) Tromp et al. (1998) (91) Cohort Sedatives Odds ratio 1.2 (0.9, 1.7) Nygaard et al. (1998) (115) Cohort Single anxiolytic/Hypnotic Odds ratio 0.6  (0.1, 2.8) Nygaard et al. (1998) (115) Cohort Two or more Anxiolytics/Hypnotics Odds ratio 0.6 (0.1, 2.3) Kallin et al. (2004) (108) Cross-sectional Zopiclone Odds ratio 1.31 (0.81, 2.12) *Calculated from available data   71 Table 1-8 Studies published between 1996 and 2007 measuring the association between benzodiazepine (BZD) use and falls Study Authors (Year) Type of Study Drug Evaluated Measure of Association Unadjusted estimated measure of association (95%CI) Adjusted estimated measure of association (95%CI) de Rekeneire et al. (2003) (120) Cross-sectional Benzodiazapine Odds ratio  1.6 (1.0, 2.6) Ebly et al. (1997) (102) Cross-sectional Benzodiazapine Odds ratio 1.34* Ensrud et al. (2002) (103) Cohort Benzodiazapine Relative risk  1.34 (1.09, 1.63) Frels et al. (2002) (121) Case-control Benzodiazapine Odds ratio  2.3 (1.4, 3.7) Hanlon et al. (2002) (89) Cohort Benzodiazapine Odds ratio  0.99 (0.71, 1.39) Kallin et al. (2002) (107) Cohort Benzodiazapine Odds ratio 1.12* Kallin et al. (2004) (108) Cross-sectional Benzodiazapine Odds ratio 1.32 (1.03, 1.68) Landi et al. (2005) (109) Case-control Benzodiazapine Odds ratio  1.54 (1.23, 1.91) 1.36 (1.08, 1.71) Neutel et al. (2002) (96) Case-crossover Benzodiazapine Odds ratio 1.7 (1.0, 2.9) 1.3 (0.7, 2.4) Passaro et al. (2000) (122) Cohort Benzodiazapine Odds ratio 1.7 (1.2, 2.3) Souchet et al. (2005) (112) Case-control Benzodiazapine Odds ratio 5.1 (4.0, 6.3) 4.7 (3.7, 5.9) Walker et al. (2005) (100) Case-control Benzodiazapine Odds ratio 2.80 (0.84, 9.31) Weiner et al. (1998) (87) Cohort Benzodiazapine Odds ratio  1.68 (0.72, 3.90) Neutel et al. (2002) (96) Case-crossover BZD and anti- psychotic Odds ratio 11.4 (1.5, 89.0) Ray et al. (2000) (123) Cohort BZD current user Odds ratio 1.44 (1.33, 1.56) Souchet et al. (2005) (112) Case-control BZD long half- life Odds ratio 3.1 (2.1, 4.5) 3.0 (2.1, 4.4) Ray et al. (2000) (123) Cohort BZD recent user Odds ratio 1.23 (1.07, 1.42) Souchet et al. (2005) (112) Case-control BZD short half- life Odds ratio 5.1 (4.0, 6.7) 4.9 (3.7, 6.3) Passaro et al. (2000) (122) Cohort Combinaiton BZD Odds ratio 2.0 (1.1, 3.9) 1.6 (0.8, 3.3) Ray et al. (2000) (123) Cohort Intermediate- acting BZD Odds ratio 1.45 (1.33, 1.59) Landi et al. (2005) (109) Case-control Long half-life BZD Odds ratio  1.45 (1.00, 2.19) Passaro et al. (2000) (122) Cohort Long half-life BZD Odds ratio 1.02 (0.5, 2.2) 0.8 (0.4, 1.8) Ensrud et al. (2002) (103) Cohort Long-acting BZD Relative risk  1.61 (1.17, 2.20) Ray et al. (2000) (123) Cohort Long-acting BZD Odds ratio 1.73 (1.40, 2.14) Passaro et al. (2000) (122) Cohort New prescription BZD Odds ratio 1.6 (1.1, 2.4) Landi et al. (2005) (109) Case-control Short half-life BZD Odds ratio   1.32 (1.02-1.72) Passaro et al. (2000) (122) Cohort Short half-life BZD Odds ratio 1.7 (1.2, 2.6) 1.8 (1.2, 2.8) Ensrud et al. (2002) (103) Cohort Short-acting BZD Relative risk  1.19 (0.92, 1.54) Ray et al. (2000) (123) Cohort Short-acting BZD Odds ratio 1.15 (0.94, 1.40) Passaro et al. (2000) (122) Cohort Very short BZD  Odds ratio 2.1 (1.2, 4.06) 1.9 (1.03, 3.3) *Calculated from available data   72 Table 1-9 Studies published between 1996 and 2007 measuring the association between diuretic use and falls Study Authors (Year) Type of Study Drug Evaluated Measure of Association Unadjusted estimated measure of association (95%CI) Adjusted estimated measure of association (95%CI) Hanlon et al. (2002) (89) Cohort Diuretics Odds ratio  1.07 (0.80, 1.42) Fisher et al. (2003) (95) Case-control Diuretics Odds ratio 0.6 (0.3, 1.4) Frels et al. (2002) (121) Case-control Diuretics Odds ratio 0.67 (0.39, 1.0) Gerdhem et al. (2005) (92) Cohort Diuretics Odds ratio 1.18 (0.82, 1.68) Gluck et al. (1996) (104) Case-control Diuretics Odds ratio 1.61* Kallin et al. (2004) (108) Cross-sectional Diuretics Odds ratio 0.88 (0.69, 1.11) Lee et al. (2003) (114) Cross-sectional Diuretics Odds ratio  0.99 (0.76-1.27) Mustard and Mayer (1997) (111) Case-control Diuretics Odds ratio  0.97 (0.82, 1.15) Neutel et al. (2002) (96) Case-crossover Diuretics Odds ratio 1.4 (0.8, 2.4) 1.0 (0.5, 1.9) Sieri and Berreta (2004) (125) Cross-sectional Diuretics Odds ratio 0.20* Walker et al. (2005) (100) Case-control Diuretics Odds ratio 1.20 (0.39, 3.6) Rozenfeld et al. (2003) (116) Cross-sectional Diuretics  Odds ratio 1.39 (0.94, 2.06) Souchet et al. (2005) (112) Case-control Loop diuretics Odds ratio 2.5 (1.8, 3.6) 1.1 (0.7, 1.5) Maurer et al. (2005) (97) Cohort Loop diuretics Hazard ratio 1.72 (0.96, 3.07) Fisher et al. (2003) (95) Case-control Thiazide Odds ratio 0.4 (0.1, 1.5) Souchet et al. (2005) (112) Case-control Thiazide diuretics Odds ratio 3.4 (1.5, 7.5) 1.9 (0.8, 4.3) *Calculated from available data  73 Table 1-10 Studies published between 1996 and 2007 measuring the association between anti- psychotic/neuroleptic use and falls Study Authors (Year) Type of Study Drug Evaluated Measure of Association Unadjusted estimated measure of association (95%CI) Adjusted estimated measure of association (95%CI) Ebly et al. (1997) (102) Cross-sectional Anti-psychotic Odds ratio 1.11* Mustard and Mayer(1997) (111) Case-control Anti-psychotics Odds ratio  1.31 (1.06, 1.61) Kelly et al. (2003) (55) Case-control Anti-psychotic Odds ratio  1.35 (0.90, 2.02) Lord et al. (2003) (119) Cohort Anti-psychotics Odds ratio  1.27 (0.92, 1.75) Landi et al. (2005) (109) Case-control Anti-psychotic Odds ratio 1.78 (1.33, 2.37) 1.48 (1.09, 2.02) Walker et al. (2005) (100) Case-control Anti-psychotics Odds ratio 0.92 (0.61, 16.17) Landi et al. (2005) (109) Case-control Atypical anti- psychotic Odds ratio  1.45 (1.00, 2.11) Kallin et al. (2004) (108) Cross-sectional Haloperidol Odds ratio 1.19 (0.68,2.09) Kallin et al. (2004) (108) Cross-sectional Olanzapine Odds ratio 1.89 (0.99, 3.62) Hien et al. (2005) (106) Cohort Olanzapine Relative risk 2.50 (1.10, 5.68) Hien et al. (2005) (106) Cohort Olanzapine Hazard ratio 2.35 (1.43, 3.87) 1.74 (1.04, 2.90) Rozenfeld et al. (2003) (116) Cross-sectional Other psychoactives agents Odds ratio 2.04 (1.05, 3.99) Weiner et al. (1998) (87) Cohort Neuroleptics Odds ratio  1.20 (0.20, 7.26) Kallin et al. (2002) (107) Cohort Neuroleptics Odds ratio 1.25* Kallin et al. (2004) (108) Cross-sectional Neuroleptics Odds ratio 1.53 (1.30, 1.96) 1.38 (1.04, 1.82) Horikawa et al. (2005) (126) Cohort Neuroleptics Odds ratio  3.47 (1.15, 10.48) Hien (2005) (106) Cohort Risperidone + psychotropric Relative risk 3.12 (1.50, 6.50) Hien (2005) (106) Cohort Risperidone    Relative risk 1.00 (0.15, 6.67) Landi et al. (2005) (109) Case-control Typical anti- psychotic Odds ratio  1.49 (1.10, 2.51) Hien et al. (2005) (106) Cohort Typical anti- psychotic Hazard ratio 1.48 (0.96, 2.26) 1.35 (0.87, 2.09) Hien et al. (2005) (106) Cohort Typical anti- psychotic + psychotropic Relative risk 2.57 (1.62, 4.08) *Calculated from available data  74 Table 1-11 Studies published between 1996 and 2007 measuring the association between psychotropic use and falls Study Authors (Year) Type of Study Drug Evaluated Measure of Association Unadjusted estimated measure of association (95%CI) Adjusted estimated measure of association (95%CI) Gluck et al. (1996) (104) Case-control Psychotropic Odds ratio 0.32* Lee et al. (2003) (124) Cross-sectional Psychotropic Odds ratio 1.38 (0.73-2.63) Sieri & Berreta (2004) (125) Cross-sectional Psychotropic Odds ratio 1.68* Landi et al. (2005) (109) Case-control Psychotropic Odds ratio 1.47 (1.24, 1.74) Schwartz et al. (1999) (45) Cohort Psychotropic Rate ratio per 1000 person years 2.10 (1.12, 3.95)  Thapa et al. (1996) (99) Cohort Psychotropic Odds ratio 2.49 (1.43, 4.33) Passaro et al. (2000) (122) Cohort Other psychotropic drugs Odds ratio 3.2 (2.3, 4.4) 2.3 (1.6, 3.2) Lord et al. (2003) (119) Cohort Any psychotropics Odds ratio 1.47 (1.20,1.81) Lord et al. (2003) (119) Cohort ≥ 2 psychotropics Odds ratio 1.30 (1.00, 1.69)  *Calculated from available data  75 Table 1-12 Studies published between 1996 and 2007 measuring the association between narcotic use and falls Study Authors (Year) Type of Study Drug Evaluated Measure of Association Unadjusted estimated measure of association (95%CI) Adjusted estimated measure of association (95%CI) Kelly et al. (2003) (55) Case-control Narcotic pain killers Odds ratio 1.68 (1.39, 2.03) Ebly et al. (1997) (102) Cross- sectional Narcotics Odds ratio 1.44* Ensrud et al. (2002) (103) Cohort Narcotics Relative risk 1.02 (0.79, 1.31) Mustard and Mayer (1997) (111) Case-control Narcotics/opioids Odds ratio 1.18 (0.96, 1.45) Desmet et al. (2004) (101) Cohort Opiate derivatives Odds ratio 3.67* Kallin et al. (2004) (108) Cross- sectional Opiates Odds ratio 1.18 (0.89, 1.54)  Walker et al. (2005) (100) Case-control Opioids  Odds ratio 0.33 (0.11, 0.96)  Weiner et al. (1998) (87) Cohort Opioids analgesics Odds ratio 2.43 (0.80, 7.40) *Calculated from available data  76 Table 1-13 Studies published between 1996 and 2007 measuring the association between non- steroidal anti-inflammatory drug (NSAID) use and falls Study Authors (Year) Type of Study Drug Evaluated Measure of Association Unadjusted estimated measure of association (95%CI) Adjusted estimated measure of association (95%CI) Gluck et al. (1996) (104) Case-control NSAIDs  Odds ratio 0.53* Hanlon et al. (2002) (89) Cohort NSAIDs Odds ratio 1.02 (0.78, 1.32) Kallin et al. (2004) (108) Cross-sectional NSAIDs Odds ratio 1.58 (1.03, 2.42)  Lee et al. (2003) (114) Cross-sectional NSAIDs Odds ratio . 1.42 (0.98, 2.05) Mustard and Mayer (1997) (111) Case-control NSAIDs Odds ratio 1.11 (0.92, 1.33) Schwartz et al. (1999) (45) Cohort NSAIDs Rate ratio per 1000 person years 1.70 (1.12, 2.58)  Walker et al. (2005) (100) Case-control NSAIDs  Odds ratio 10.02 (2.6, 38.58)  *Calculated from available data  77 Table 1-14 Studies published between 1996 and 2007 measuring the association between beta- blocker use and falls Study Authors (Year) Type of Study Drug Evaluated Measure of Association Unadjusted estimated measure of association (95%CI) Adjusted estimated measure of association (95%CI) Desmet et al. (2004) (101) Cohort Beta-blocker Odds ratio 0.34* Fisher et al. (2003) (95) Case-control Beta-blocker Odds ratio 1.2 (0.5, 3.4) Kallin et al. (2004) (108) Cross-sectional Beta-blocker Odds ratio 0.96 (0.71, 1.30)  Lee et al. (2003) (114) Cross-sectional Beta-blocker Odds ratio 1.14 (0.92-1.40) Mustard and Mayer (1997) (111) Case-control Beta-blocker Odds ratio 1.04 (0.64, 1.63) Rozenfeld et al. (2003) (116) Cross-sectional Beta-blocker Odds ratio 1.54 (0.92, 2.58)  *Calculated from available data  78 Table 1-15 Studies published between 1996 and 2007 measuring the association between calcium channel blocker use and falls Study Authors (Year) Type of Study Drug Evaluated Measure of Association Unadjusted estimated measure of association (95%CI) Adjusted estimated measure of association (95%CI) Fisher et al. (2003) (95) Case-control Calcium Channel Blockers Odds ratio 1.4 (0.6, 3.2)  Kallin et al. (2004) (108) Cross-sectional Calcium Channel Blockers Odds ratio 0.95 (0.60, 1.49)  Lee et al. (2003) (114) Cross-sectional Calcium Channel Blockers Odds ratio 1.00 (0.82, 1.22) Maurer et al. (2005) (97) Cohort Calcium Channel Blockers Hazard ratio 2.18 (0.98, 4.85)  Mustard and Mayer (1997) (111) Case-control Calcium Channel Blockers Odds ratio 1.02 (0.68, 1.51)  79 Table 1-16 Studies published between 1996 and 2007 measuring the association between angiotensin-converting enzyme inhibitor (ACEI) use and falls Study Authors (Year) Type of Study Drug Evaluated Measure of Association Unadjusted estimated measure of association (95%CI) Adjusted estimated measure of association (95%CI) Fisher et al. (2003) (95) Case-control ACEI Odds ratio 1.0 (0.5, 2.3) Kallin et al. (2004) (108) Cross-sectional ACEI Odds ratio 0.81(0.54, 1.21) Lee et al. (2003) (114) Cross-sectional ACEI  Odds ratio 0.98 (0.76, 1.27) Maurer et al. (2005) (97) Cohort ACEI Hazard ratio 2.08 (1.18, 3.68)   80 Table 1-17 Studies published between 1996 and 2007 measuring the association between digoxin use and falls Study Authors (Year) Type of Study Drug Evaluated Measure of Association Unadjusted estimated measure of association (95%CI) Adjusted estimated measure of association (95%CI) Kallin et al. (2004) (108) Cross-sectional Digoxin Odds ratio 1.02 (0.72, 1.46)  Sieri & Beretta (2004) (125) Cross-sectional Digoxin Odds ratio 1.11* Souchet et al. (2005) (112) Case-control Digoxin Odds ratio 3.3 (2.1, 5.1) 1.1 (0.7, 1.7) Walker et al. (2005) (100) Case-control Digoxin Odds ratio 1.94 (0.31, 12.16)  *Calculated from available data  81 Table 1-18 Studies published between 1996 and 2007 measuring the association between anti- hypertensive use and falls Study Authors (Year) Type of Study Drug Evaluated Measure of Association Unadjusted estimated measure of association (95%CI) Adjusted estimated measure of association (95%CI) Chu et al. (2005) (118) Cohort Anti-hypertensive Odds ratio 1.54* Chu et al. (2005) (118) Cohort Anti-hypertensive Odds ratio 1.27* Fisher et al. (2003) (95) Case-control Anti-hypertensive Odds ratio 0.8 (0.4, 1.6)  Frels et al. (2002) (121) Case-control Anti-hypertensive Odds ratio 1.19* Gerdhem et al. (2005) (92) Cohort Anti-hypertensive Odds ratio 1.10 (0.81, 1.49)  Heitterachi et al. (2002) (105) Cohort Anti-hypertensive Relative risk 0.95 (0.60-1.51)  Kelly et al. (2003) (55) Case-control Anti-hypertensive Odds ratio 0.98 (0.84, 1.15) Rozenfeld et al. (2003) (116) Cross- sectional Anti-hypertensive Odds ratio 1.11 (0.78, 1.58)  Tromp et al. (1998) (91) Cohort Anti-hypertensive Odds ratio 1.4 (1.0, 1.9) 0.91 (0.68, 1.26) Walker et al. (2005) (100) Case-control Anti-hypertensive Odds ratio 1.23 (0.45, 3.39)  *Calculated from available data  82 Table 1-19 UCLA Practice Guideline: ED management of falls in patients aged 65 years and older(31) Please circle response to each question and make intervention, when appropriate Assessment   Intervention Triage nurse >4 medications Y N PMD/GA Dalmane, Valium, Elavil Y N ED MD/PMD Ethanol Y N ED MD/PMD Seizure/loss of consciousness Y N ED evaluation Primary nurse Vital and orthostatic vital signs Y N ED evaluation >1 fall in preceding 3 months Y N PMD/GA Long lie (>5 minutes) or assistance required to get up Y N PMD/GA Recent change in mental status Y N ED evaluation Environmental hazards (e.g., poor lighting) Y N HH Vision examination in preceding year Y N Optometry Complex social problems Y N SS Elder abuse Y N SS Inability to live independently Y N SS Physician Abnormal orientation, including month and year Y N ED evaluation Abnormal hydration Y N ED evaluation Nutritional deficiency Y N SS Abnormal gait or balance: Get-Up-and-Go Test Y N PMD/PT Foot problems Y N Podiatry Physician assessment of cause of fall (may be more than one) Sports or occupation Y N Treat injury Weakness/poor balance Y N  PMD/PT Syncope/near-syncope Y N ED evaluation Delirium/dementia Y N ED evaluation Seizure Y N ED evaluation New stroke Y N ED evaluation Complex medical problems Y N PMD/GA Preventive measures for older patients Exercise program Y N Information Calcium and vitamin D supplements Y N Information Pneumovax Y N ED/PMD Influenza vaccine Y N ED/PMD PMD, primary physician; GA, geriatric assessment; ED MD, emergency physician; HH, home health; SS, social services; PT, physical therapy.  83  Figure 1-2 Prevention of falls in older persons living in the community (American Geriatric Society/British Geriatric Society/American Academy of Orthopedic Surgeons Guideline)(11, 83)                    Any indication for additional intervention? Initiate multifactorial/multicomponent intervention to address identified risk(s) and prevent falls: 1. Minimize medications 2. Provide individually tailored exercise program 3. Treat vision impairment (including cataract) 4. Manage postural hypotension 5. Manage heart rate and rhythm abnormalities 6. Supplement vitamin D 7. Manage foot and footware problems 8. Modify the home environment 9. Provide education and information Reassess periodically Yes No No Evaluate gait and balance [E] 1. Obtain relevant medical history, physical examination, cognitive and functional assessment 2. Determine multifactorial fall risk: a. History of falls b. Medications c. Gait, balance, and mobility d. Visual acuity e. Other neurological impairments f. Muscle strength g. Heart rate and rhythm h. Postural hypotension i. Feet and footware j. Environmental hazards [F] Are abnormalities in gait or unsteadiness identified? Yes 5 6 7 8 9 10 Older person encounters healthcare provider [A] Screen for fall(s) or risk for falling (See questions in sidebar) [B] Answers positive to any of the screening questions? (See sidebar) [C] Does the person report a single fall in the past 12 months? [D] Yes Yes No No 1 2 3 4 Sidebar: Screening for Fall(s) Questions: 1. Two or more falls in prior 12 months? 2. Presents with acute fall? 3. Difficulty with walking or balance?  84 Chapter  2: Meta-Analysis of the Impact of 9 medication classes on falls in elderly people1 2.1  Introduction Falling in elderly people is a major, yet under recognized public health concern. Falls and fall related complications are the 5th leading cause of death in the developed world and more than 30 percent of those older than 65 will fall at least once annually.(7-9, 11, 57)  Furthermore, falls are the primary reason for 85% of all injury- related hospitalizations for those ≥65 years of age and >40% of nursing home admissions.(13, 151)  The annual costs associated with falls and fall related complications have been estimated to be in the billions of dollars worldwide.(5, 6, 171, 191) As a result, research examining the contributions of different risk factors on falls is urgently needed.  Fall risk is multi-factorial with many intrinsic and extrinsic risk factors.(24) Prescribed medications have been identified as an important contributor to falls and the risk of falling in seniors.(11, 14, 25, 26, 98)  Poly-pharmacy, the concurrent use of 4 or more medications, has been identified as a risk factor for falling.(11, 26, 192) As well, a number of guidelines for the care of elderly fallers recommend that an assessment of medications be completed to ensure that individuals’ medications do not unnecessarily increase their risk of falling.(11, 16, 31, 83)  1 A version of Chapter 2 has been published. Woolcott JC, Richardson KJ, Wiens WO, Patel B, Marin J, Khan KM, Marra CA. (2009) Meta-analysis of the impact of 9 medication classes on falls in elderly persons. Arch Inter Med 2009 Nov 23. 169(21): 1952-1960.   85  Several commonly used medications have been associated with both the probability of falling as well as sustaining fractures after a fall.(25, 26, 153, 193)  However, determining which medications contribute to falls and which do not remains a clinical challenge. Although a number of studies have assessed the association between specific medication and medication classes on the probability of experiencing one or more falls during the timeframe being studied, differences in study methodologies, setting, power, and fall definitions have made it difficult to conclusively state the impact of various medications on falling. As well, although there is evidence that certain medications are associated with falls, the prevalence of prescribing of medications to seniors has increased substantially over the past decade.(84)  Using papers published between 1966 and 1996, Leipzig et al. published two meta- analyses which assessed the association between falling and the use of various medications in seniors.(25, 26)  Subsequent to Leipzig’s meta-analyses, Hartikainen et al. completed a systematic review describing studies published after 1996 that examined the impact of medication use on falls but conducted no formal statistical analyses to pool these data.(98)  This study provides a quantitative update to the previous meta-analyses of Leipzig et al. I completed a Bayesian meta-analysis incorporating the results of Leipzig et al’s work with newly (post 1996) published data for medications that were previously assessed. Furthermore, I sought to complete meta-analyses on additional drugs classes not originally assessed by Leipzig et al.(25, 26)   86 2.2 Methods 2.2.1 Data sources and searches Along with three research assistants (Mathew Wiens, Judith Marin, and Bhavini Patel), I conducted a computerized EBM, CINAHL, EMBASE and MEDLINE search of literature published between April 1996 and August 2007 to identify all potentially eligible studies. The MeSH term ‘therapeutic uses’, encompassing all indexed classes of drugs and individual agents, was combined with the MeSH terms ‘accidental fall’ or ‘home accident’. All MeSH terms were expanded to include all sub-headings. Furthermore, the MeSH terms ‘epidemiology’ or ‘pharmacoepidemiology’ were combined with ‘accidental fall’ or ‘home accident’ to capture studies where exposure to drugs was not the primary objective, but may have been a secondary objective. A similar algorithm was applied in EMBASE. The MeSH terms ‘analgesic, anti-inflammatory, antirheumatic and antigout agents’ or ‘central nervous system agents’ or ‘agents interacting with transmitter, hormone or drug receptors’ or ‘cardiovascular agents’ were combined with the MeSH terms ‘accident’ or ‘falling or ‘home accident’. All terms were expanded to capture all relevant articles. All potentially eligible studies were considered regardless of publication type. All references of retrieved articles were also searched for potentially eligible studies. Furthermore, leading investigators in the area of falls in elderly people were contacted to obtain studies that may have not been captured with the planned search strategy.   87 2.2.2 Study selection Studies were considered eligible for inclusion if they presented original data of randomized controlled trial, case-control, cohort or cross-sectional designs assessing the association between medication usage and falls in people 60 years of age or older. These criteria for study inclusion mirror the criteria used by Leipzig et al.(25, 26)  2.2.3 Data extraction and quality assessment Studies were assessed independently by me and/or MW, JM, BP for methodological quality using a published checklist by Downs and Black,(194) and any disagreements were resolved by a third co-investigator (Carlo Marra).  In addition to completing the quality assessment checklist for each study, I also looked at the methods of fall and medication use ascertainment. Using the criteria used previously by Leipzig et al.,(25) a study which ascertained medications at the time of the fall and documented fall occurrence prospectively or from fall reports were identified as having “good” medication/falls ascertainment with all others identified as “poor”.  Many studies evaluated outcomes of several classes of drugs, thus one study could provide the risks for several exposure types. All exposures were required to be presented as the odds ratios (OR) associated with exposure or non exposure, or as 2 x 2 tables of reporting falls by a given exposure relative to non exposure, along with 95% confidence intervals (CI). Where these results were not published, authors were contacted and asked to provide the necessary data to calculate ORs and 95%  88 CI. If the available data did not allow for the calculation of ORs and 95% CIs, the exposure was excluded from final analysis. Additional information collected from the included studies was study type, study setting (hospital/long term care facility vs. community), the mean age of participants, time of medication ascertainment, and method of fall ascertainment. If provided, adjusted ORs and 95% CIs and the covariates adjusted for were also extracted.  For each study included in the meta-analysis, the fall definition used was compared to the Prevention of Falls Network Europe (ProFaNE) fall definition. The ProFaNE fall definition of “an unexpected event in which participants (sic) come to rest on the ground, floor, or lower level” (35) is the current gold standard for fall definition and is recommended for use in fall injury prevention trials.   I also mirrored the methodology of Leipzig et al’s meta-analysis,(25) comparing each study’s fall definition to the Kellogg Working Group’s fall definition.(39) The Kellogg fall definition is different from the ProFaNE fall definition by excluding those falls which are “…as a consequence of sustaining a violent blow, loss of consciousness, sudden onset of paralysis as in a stroke or an epileptic seizure.”(39)  2.2.4 Data synthesis and analysis The primary method for the assessment of medication risk was through pooled OR estimates updated using Bayesian meta-analysis methodology. Using Bayesian random effects models allowed the integration of prior information with newly available information to provide a posterior OR estimate with a 95% credible interval  89 (the Bayesian equivalent to the Frequentist CI).  These methods have been identified as having a number of advantages over Frequentist methods of meta- analysis including the ability to adjust for greater uncertainty and complexity of issues while incorporating pertinent/prior known information regarding the association being assessed.(178) The Bayesian results also allow us to make probabilistic statements about the effect size, i.e. I am able to answer the question, "given the observed data, what is the probability that medication use increases the chances of falling"?  Using the fixed-effects pooled results from the previous meta-analyses completed by Leipzig et al.(25, 26) as prior unadjusted ORs, I calculated updated Bayesian pooled estimates of the ORs for the impact of medication use on the likelihood of having a fall during the study period. The reported ORs for each study were assumed to follow a lognormal distribution and the between-study precision was modelled using a vague prior with a gamma distribution. The prior is suitably vague as it has a large variance to represent a lack of information about the possible heterogeneity between studies and is a commonly chosen prior for the between-study precision.(195) For medication classes not assessed by Leipzig et al., a ‘non-informative’ prior with a lognormal distribution centred at zero with a wide variance (1000) to reflect the lack of previous evidence was used.  Although Leipzig et al. reported fixed-effects meta-analyses, they found evidence of heterogeneity in their pooled estimates of the medication classes  90 sedatives/hypnotics and neuroleptics/anti-psychotics.(25) Ideally a random-effects meta-analysis would have been performed for these classes to allow for the between-study and within-study variability and would have the effect of reducing the relative weighting given to the more precise studies.(178) Thus, as I was skeptical about the fixed-effects pooled estimates for sedatives/hypnotics and neuroleptics/anti-psychotics, I also performed a sensitivity analysis where I inflated the variance of the fixed-effects estimate to five times its original variance as a prior estimate, thus giving it less weight in the Bayesian analysis.  To provide a contrast to the Bayesian results and a comparison to Leipzig et al’s previous findings, I also estimated pooled Frequentist ORs and 95% CIs for the newly identified studies using random-effects models by weighting each study by the inverse of its variance.  Meta-analyses were completed on those medication classes with four or more published studies completed in the period 1996-2007. In addition, pooled results were estimated by subgroups of studies defined by residential type (long term care, community, or other), falling frequency (greater or less than 35%), age of participants (mean age greater or less than 75 years), ascertainment of medications and falls (good or poor), and study design (cohort, case-control, or cross-sectional). Where there were only one or two new studies, due to the instability in the Bayesian models, Frequentist random effects inverse-variance models were used to provide a pooled OR only. Bayesian posterior adjusted ORs were also estimated for those  91 medication classes with four or more studies providing adjusted ORs and 95% CIs. WinBUGS version 1.4 was used to perform the Bayesian analyses with two separate chains with 10,000 Markov Chain Monte Carlo iterations completed for each medication class.  2.3 Results As shown in Figure 2-1, the search strategy identified 11,118 articles of which 22 were used in the meta-analyses. Of the 22 studies that met the inclusion criteria, none were randomized controlled trials. Table 2-1 contains a summary of the medications assessed in each included study, while Table 2-2 displays the specific settings, size, and characteristics of populations, as well as the temporal relationships between medication ascertainment, falls ascertainment, and index fall. Of the observational studies assessed, ten were cohort studies,(87, 89-92, 103, 113, 118, 122)  five were case control studies(95, 96, 100, 104, 121) and seven were cross sectional studies.(102, 108-110, 116, 120, 124)  Of the cohort studies, eight(87, 89-92, 103, 118, 122) were prospective studies with follow-up ranging from 6 months(87) to 37.8 months,(91) with all other prospective studies using a one year follow up interval. The definition of a fall used in three(95, 116, 118)  of the analyses was that of the Kellogg working group,(39) while nine(90, 91, 96, 102, 108, 109, 120- 122, 124)  of the analyses used a fall definition similar to the ProFaNE group.(35) Six studies were considered to have 'good' medication/falls ascertainment.(96, 100, 104, 108, 121, 122)   92 2.3.1 Results of the meta-analyses The meta-analyses were completed on nine unique drug classes including 79,081 participants. For each drug class assessed, Figure 2-2 shows the OR and 95% CI for each independent study alongside the associated Frequentist random-effects pooled ORs with 95% CIs and the Bayesian pooled ORs with 95% credible intervals (95% CrI) updated from the prior information. Among all results, when compared to those who were not using an antidepressant, using an antidepressant had the strongest association with experiencing a fall with an updated Bayesian OR of 1.68 (95%CrI 1.47-1.91).  The lowest OR point estimate was for the narcotics class, where when compared to those not using a narcotic, those who reported using a narcotic had an estimated pooled OR of 0.96 (95%CrI 0.78-1.18).  In many cases the updated Bayesian estimates were similar to the prior OR estimates by Leipzig et al.(25, 26)  However, one notable exception was the beta- blocker drug class. When compared to those who were not using a beta blocker, Leipzig et al. reported that exposure to a beta-blocker had an OR estimate of 0.93 (95%CI 0.77-1.11).(26)  Using the studies completed between 1996-2007 I estimated a Frequentist OR of 1.14 (95%CI 0.97-1.33) and an updated Bayesian OR of 1.01 (95%CrI 0.86-1.17). The difference between the prior OR and the new Frequentist OR was close to statistical significance (p=0.05). In many cases the prior OR differed from the new Frequentist OR estimates, although the differences were not statistically significant. Anti-hypertensives were not included in the previous meta-analysis by Leipzig et al.(25, 26) Compared to those who were not using an  93 anti-hypertensive, the Bayesian meta-analysis using a 'non-informative' prior resulted in a positive association between anti-hypertensive use and falling (OR 1.24, 95%CrI 1.01-1.50). The Bayesian OR estimate for comparing those who did not use anti-hypertensives to those who did was similar to the Frequentist random- effects OR estimate (OR 1.26, 95%CI 1.08-1.46).  Use of sedatives/hypnotics and neuroleptics/anti-psychotics was positively associated with falling (OR 1.47 (95%CrI 1.35–1.62) and OR 1.59 (95%CrI 1.37– 1.83) respectively) when using the fixed-effects estimate from Leipzig et al. as a prior. The conclusions did not differ when the skeptical priors (variance inflated by a factor of five) of 1.54 (95%CI 1.24–1.91) and 1.50 (95%CI 1.00–2.24) were used instead, as they resulted in Bayesian pooled estimates of 1.38 (95%CrI 1.22–1.60) and 1.68 (95%CrI 1.36–2.07) respectively.  The between study variance estimated by the Bayesian models for the new studies was 0.29 for narcotics, 0.03 for antidepressants, 0.01 for non-steroid anti- inflammatory drugs (NSAIDs) and less than 0.006 for the remaining six drug classes. The evidence of considerable heterogeneity between the new narcotic studies was mainly due to the study by Walker (100) reporting a protective association with falling.  Stratification of the studies by participants residing in a long term care facility or community/other type of residence, percentage of fallers >35%, mean age of  94 participants >75 years of age, ascertainment of medications and falls (retrospective or prospective), or study design had little effect on the Bayesian OR estimates (Table 2-3). A few differences in the stratified ORs were observed, particularly across population (in the beta-blockers and neuroleptics/anti-psychotics classes) and study type (in the sedatives/hypnotics, benzodiazepines and narcotics classes). An increased likelihood of falling (i.e. the entire credible interval >1) was associated with the use of sedatives/hypnotics, neuroleptics/anti-psychotics, antidepressants, benzodiazepines, and NSAIDs, based on those studies considered to have 'good' medication and falls ascertainment.  A subset of studies provided adjusted ORs for the drug classes included in this meta-analysis (Table 2-1). Most of these studies adjusted for age, sex and co- morbidities, whilst disability, cognition, previous falls, and other medications were also commonly adjusted for (Table 2-2). The updated Bayesian posterior OR, compared to those who did not use, for use of diuretics, neuroleptics/anti-psychotics, antidepressants, and benzodiazepines using a prior estimate of 1.85 (95%CI 1.20- 2.85) for antidepressants from the Leipzig  et al.(25, 26) and a non-informative prior for the rest were 0.99 (95%CrI 0.78–1.25), 1.39 (95%CrI 0.94–2.00), 1.36 (95%CrI 1.13–1.76) and 1.41 (95%CrI 1.20–1.71) respectively.(87, 89, 90, 96, 100, 103, 108- 110, 120, 121, 124)     95 2.4 Discussion Using Bayesian methodology, I have completed one of the first meta-analyses to use informed priors in the OR calculations. By incorporating the results of previous meta-analyses by Leipzig et al.,(25, 26) the OR estimates provide a needed update of the association between falls and nine different medication classes.  I also estimated Frequentist pooled ORs to measure the association between falls and medication classes using research completed since 1996 for comparison to the Bayesian estimates. My results extend the current knowledge on specific medication classes’ impact on the risk of falls while complementing previous meta-analyses and systematic reviews that incorporated research completed before 1996 and 2004 respectively.(25, 26, 98)  Use of diuretics was associated with an increased fall risk in the unadjusted meta- analysis, but not when adjusted for covariates. For the other cardiac drug classes, anti-hypertensives were associated with falling and although the OR point estimate for beta blockers was greater than one, the posterior probability that their use increased the risk of falling was 55%. Also of note was the difference between the prior and new information for the beta blocker class leading to a combined updated Bayesian estimate of no association with falling.  It was also noticed that the post- 1996 studies had a higher prevalence of beta blocker use than studies reported by Leipzig et al.(26)   96 This meta-analysis also showed that psychotropic drugs were associated with increased falls.  The overall pooled Bayesian OR estimate and the sensitivity analyses undertaken on the sedative/hypnotics, antidepressants, and benzodiazepine classes revealed that their use substantially increased the likelihood of falls. Although neuroleptics/anti-psychotics were associated with falling in the main unadjusted results of the meta-analysis, after adjusting for potential confounders the association statistical significance was attenuated.  In contrast to Leipzig et al., who suggested that hospitalized patients who used neuroleptics/anti- psychotics would have fewer falls,(25, 26) I observed no such association between the use of neuroleptic/anti-psychotic use and falls in that setting.  An important consideration when estimating the level of association between specific medication use and falling is the impact of and adjustment for confounding, specifically confounding by indication. In a study of association between medication use and falling, confounding by indication can occur when the medication class assessed is a marker for a clinical diagnosis which in itself changes the risk of experiencing a fall and also requires treatment with the medication being assessed.(196)  Adjustment for confounding by indication is usually completed by the use of propensity score methods. Propensity scores can estimate the probability of exposure and can allow for the matching of individuals across groups with similar propensity scores or probabilities of exposure and is thought to be most appropriate when the treatment is frequent with rare outcomes.(197, 198)  Multivariable modelling incorporates potential confounders into regression analysis to estimate  97 adjusted measures of association. However the results are dependent on the potential confounders that are included in the regression models.(197)  None of the studies included in the meta-analysis utilized propensity score matching to control for confounding by indication, yet many included multivariable modelling and reported adjusted ORs.  However, recent evidence suggests that often similar results are achieved using conventional multivariable models as compared to propensity score methods.(198) Reassuringly, I found that the pooled adjusted ORs were similar to the unadjusted ORs leading me to conclude that the role of confounding was quite small in this regard.  During the search for relevant articles, I identified six studies reporting an unadjusted hazard ratio (HR) or relative risk (RR)(45, 97, 105, 106, 119) and four studies reporting an adjusted HR or RR(106, 123, 129, 199) for a specific medication class included in this meta-analysis and their association with falling.  The high incidence of falling in elderly persons did not allow us to compare the reported RRs and HRs to the OR estimates of the other identified studies and could not be included in the main meta-analysis. In addition, there were too few of these studies reporting on the same medication class to pool their results.  However, the individual studies’ adjusted and unadjusted hazard ratio estimates and my OR estimates resulted in similar conclusions.  It should be noted that the number of participants included in the studies completed after 1996 was generally greater than the numbers included in Leipzig et al’s initial  98 meta-analyses of psychotropics.(25)  This increase in study sample sizes could be partly attributed to either a larger population available for assessment due to increased prevalence of diseases requiring psychotropic treatment or more frequent prescribing of these types of drugs or a combination of both. I observed a slight increase in the proportion of participants taking each psychotropic in the studies completed after 1996 when compared to the studies completed prior to 1996 and included in Leipzig et al’s meta-analysis.(25)  The one exception to this trend of increased proportion of psychotropic use in studies completed after 1996 was in the sedatives/hypnotics drug class. However, if the larger sample sizes are a result of increased prevalence of conditions for which psychotropics are indicated it is possible that the results are confounded by indication, as many of these conditions themselves are associated with an increased likelihood of falling.  A primary strength of this study is the use of Bayesian meta-analyses which allowed for information from the previous meta-analysis to be integrated with the more recently completed studies to evaluate the level of association between use of specific drug classes and experiencing a fall.  Applying the Bayesian methodology also allows me to make statements about the probability that the ORs are greater than one in cases when the associated 95% CrI includes one.  Although using a Frequentist random-effects model to pool all of the new studies and either the old studies, or the Leipzig et al. pooled estimate, would give similar estimates to the Bayesian methods; the Bayesian methodology allows for the between-study variance to differ from the Leipzig et al. studies and the new studies, taking into  99 account the fact that different types of evidence are being synthesized.(178) The Bayesian methodology also allows for greater uncertainty than the Frequentist approach as both the overall population effect and the between study precision in the random effects meta-analyses are estimated by the data.(200)  A limitation of this meta-analysis is that relatively few studies met the inclusion criteria of using falls as an outcome. Although the number of new studies included was small for every drug class assessed besides diuretics, the total number of additional participants included in the meta-analysis was greater than that in the previous meta-analysis by Leipzig et al.(25, 26)  A second limitation is the method of falls and medication ascertainment in many of the studies. Using the previously mentioned methodology of Leipzig et al.,(25, 26) 16 of the studies were noted to be of poor quality when considering the timing and reporting method of the falls and the medications used by study participants at the time of the fall.  Medications use is identified as a modifiable risk factor for falling, yet only one randomized controlled trial has looked at withdrawing medications from a population of users and the impact on falls.(127)  Although this study showed that the discontinuation of psychotropics can reduce the probability that an individual would fall, no other randomized controlled trials have assessed the introduction or withdrawal of specific medication classes on falls in elderly people. Given the divergent results of some observational assessments within specific medication  100 classes, the results of this meta-analysis reiterate the need for caution when prescribing these medications to elderly persons.  In addition to exercising caution when prescribing potential fall-inducing medications to those at risk, regular assessments of their medications should be performed. This assessment is particularly important for those who have previously experienced a fall and is in line with the three most commonly used guidelines for care to prevent falls.(11, 16, 31) However, similar to other recommendations to prevent falls in the elderly, regular medication assessment is rarely done.(151, 152)  It is hoped that future research in this area can be completed with larger sample sizes in both community and long term care facility settings and thus improve the quality of information about fall risks available to physicians and pharmacists when they are deciding which types of pharmacotherapy to provide.    101 Figure 2-1 Flow diagram of study selection process         References identified from initial search (n = 11,118)   251 EBM  2511 CINAHL  4261 EMBASE  4087 MEDLINE  8 Cochrane Database Studies excluded (n = 11,065)  Article did not look at association between medication use and risk of falling  Article with data before 1996  Studied subjects less than 60 years of age Studies considered and assessed for meta-analysis (n = 53) Studies excluded (n = 31)  Insufficient data results  Inappropriate study design Studies included in the meta-analysis (n = 22)  102 Table 2-1  Drugs and studies included in meta-analysis Author Anti- hypertensives Diuretics Beta Blockers Sedative/ Hypnotics Neuroleptics/ Anti- Psychotics Antidepressants Benzodiazepines Narcotic Analgesics NSAIDs Arfken et al. 2001 (113)      X Avidan et al.  2005  (117)    X A Chu  et al. 2005 (118) X   X deRekeneire et al. 2003 (120)       X A Ebly et al. 1997 (102)     X X X X Ensrud et al. 2002 (103)      X A X A X A Fisher et al. 2003 (95) X X X Frels et al. 2002 (121) X X A     X A Gerdhem et al. 2005 (92) X X Gluck et al. 1996 (104)  X  X  X   X Hanlon et al. 2002 (89)  A     A  A Kallin et al. 2004 (108)  X X  X A X A X X X Landi et al. 2005 (109)    X A X A X A X A Lawlor  et al. 2003 (110)    X A  X A Lee et al. 2006 (124)  X A X A      X A Neutel et al. 2002 (96)  X A    X A X A Passaro et al. 2000 (122)       X Rozenfeld et al. 2003 (116) X X X X Tromp et al. 1998 (91) X   X Tromp et al. 2001 (90)       X A Walker et al. 2005 (100)  X A   X A X A X A X A X A Weiner et al. 1998 (87)     A  A A X indicates crude OR or data to calculate crude OR provided A indicates adjusted OR provided (see Table 2-2 for list of confounders adjusted for)      103 Table 2-2 Characteristics of studies included in meta-analysis Author Setting Years & Duration of Data Collection N Age (mean) Drugs Time of Medication Ascertainment Method & Recall Time of Fall Ascertainment Study Design Confounders adjusted for Arfken et al. 2001 (113) Long term care facility 1995 (mean of 3 months) 368 81 Ad baseline incident report cohort N/A Avidan et al. 2005 (117) Long term care facility 2001 (150-210 days) 34163 84 Se baseline recall (180 days) cohort A,G, Adl, Cp, Rug, Boi, V, B, M, Ed, Altc Chu et al. 2005 (118) Community 1998-1999 (12 months) 1516 73 Hy, Se baseline recall (2 months) cohort N/A deRekeneire et al. 2003 (120) Community 1997-1998 (12 months) 3050 70-79 B interview recall (12 months) cross sectional A, R, Ss, Bmi, Ui, Wt, B, C, Lm Ebly et al. 1997 (102) Community 1991-1992  (not stated) 2035 80 Ad, B, N, Na interview recall (not stated) cross sectional N/A Ensrud et al. 2002 (103) Community 1992-1994 (12 months) 8127 77 Ad, B, Na interview recall (4 months) cohort A, G, Sr, Mc, Fp, Di, F, De, Wc, C, Oe* Fisher et al. 2003 (95) Long term care facility not stated  (12 months) 119 87 Be, D, Hy baseline incident report case control N/A Frels et al. 2002 (121) Hospital, Acute Medical not stated (4 months) 362 73 B, D, Hy at fall incident report case control A, G, S, Hc, F, Dis, Ma Gerdhem et al. 2005 (92) Community 1995-1999 (12 months) 978 75 D, Hy baseline recall (12 months) cohort N/A Gluck et al 1996 (104) Hospital, Acute Medical not stated 100 84 Ad, D, NSAID, Se at fall incident report case control Matched on A and G Hanlon et al. 2002 (89) Community 1989-1990  (12 months) 2996 72 B, D,  interview recall (12 months) cohort A, G, R, E, In, U, O, Alc, Sm, De, Nag, Art, Di,  Fr, Sl, Inc, Sr, Red, Opsy, Kallin et al. 2004 (108) Long term care facility 2000 (1 week) 3604 83 Ad, B, Be, D, N, Na, NSAID at fall incident report cross sectional A, G, F, C, W, P, Cp Landi et al. 2005 (109) Community 2000-2002 (90 days) 2854 77 Ad, B, N, Se interview recall (90 days) cross sectional A, G, M, Mc, De, Adl, Cp, Fp, Wa, Ga, Fof Lawlor et al. 2003 (110) Community not stated (12 months) 4050 69 Ad, Se interview recall (12 months) cross sectional A, Bmi, Hg, Cd, Di, Td, As, De, V, Art, Alc, Soc Lee et al. 2006(124) Community 2001-2003  (12 months) 4000 72 Be, D, NSAID interview recall (12 months) cross sectional A, G, V, He, Lmp, Di, S, Sr, Sln, M Neutel et al. 2002 (96) Long term care facility 1995-1996 (12 months) 227 80-90 Ad, B, D at fall incident report case control A, G, LOS, Pr, Cp, M Passaro et al. 2000 (122) Hospital, Acute Medical 1991-1993 (8 months) 7908 65-80 B at fall incident report cohort N/A Rozenfeld et al. 2003 (116) Community 1996 (12 months) 631 69 Be, D, Hy, Se interview recall (12 months) cross sectional N/A Tromp et al. 1998 (91) Community 1992-1995 (38 months) 1370 73 Hy, Se baseline recall (12 months) cohort N/A Tromp et al. 2001 (90) Community 1995-1996 (12 months) 1285 75 B baseline recall (daily)* cohort U, F, V Walker et al. 2005 (100) Hospital, Acute Medical 2002 (12 months) 124 74 Ad, B, D,N, Na, NSAID at fall incident report case control A, G, Aci, Art, Chf, Cp, Fp, Hip, F, V, Ga, Peri, S, Sur, Ui Weiner et al. 1998 (87) Community not stated  (6 months) 305 74.4 B, N, Na baseline recall (daily)** cohort A, Cp, De, Mo * Tromp used daily completed falls calendars collected every 3 mths for fall ascertainment / **Weiner used daily completed falls calendars collected every 30 days for fall ascertainment  / ***Ensrud used oral estrogen use only for adjOR for antidepressants DRUG ABBREVIATIONS: Ad – antidepressants; B – Benzodiazepines; Be – Beta blockers; D – Diuretics; Hy – Anti-hypertensives; N – Neuroleptics; Na – Narcotic Analgesics; NSAID – Non-steroidal anti-inflammatory drugs; S – Sedative Hypnotics CONFOUNDERS: A – Age; G – Gender; LOS – LTC length of stay; M – Number of medications/presence of poly-pharmacy; Adl – Activities of daily living impairment; Cp – Cognitive performance or impairment; Rug – Resource utilization group; Boi – Burden of illness; V – Vision problems; B – Balance test score; Ed – Emergency department visit; E – Education; U – underweight; O – Overweight; Fr – broken bones; Sl – Sleeping problems;  Altc – Admission to LTC; R – Race; Ss – Study site; Bmii – Body Mass Index; UI – Urinary Continence Issues; Wt – 6m walk time; C – Difficulty/ability to rise from chair; Lm – Leg muscle strength; Mc – Medical conditions; Ga – Gait speed or problems; Oe – Oral estrogen use; F – Previous falls; S – Previous stroke; Hc – Additional health conditions; Dis – Disoriented; Ma – Requiring max. assistance while in hospital; W – Walks with helper; P – Reports pain; De – Depression; Wa – Wandering; Fof – Reported fear of falling; Fp – Foot problems; Hg – Hemoglobin concentration; Cd – Circulatory disease; Sm – Smoking status; Red – Reduced activities; Di – Diabetes; Td – Thyroid disease; As – Asthma or bronchitis; Art – Arthritis; Alc – Heavy alcohol consumption; Soc – Adult social class; He – Heart disease; Sr – Self-rated health; Sln – Stride length; Pr – Programs/intensity of care; Opsy – Other psychotropic drug use; Wc – Weight change; In – Income; Aci – Acute illness or infection; Chf – Congestive heart failure; Hip – Hip fracture; Peri – Peripheral neuropathy; Sur – Surgery or anesthesia; Mo – Mobility issues; Nag – Nagi disabilities; Lmp – Lower muscular pain; Inc - Incontinence  104 Figure 2-2 Meta-analysis results Odds ratios and 95% Credible Intervals or 95% Confidence Intervals on a logarithmic scale for individual or pooled study data for each class of medication. Outcome is occurrence of at least one fall.    Anti-hypertensives Antidepressants Neuroleptics/anti-psychotics Diuretics  Beta-blockers Benzodiazepines Sedatives/hypnotics Narcotics NSAIDs  105 Table 2-3 Pooled Bayesian odds ratios and sub-group sensitivity analysis  Anti-hypertensives Diuretics Beta-blockers Sedatives/hypnotics Anti-psychotics Antidepressants Benzodiazepines Narcotics NSAIDs Study Characteristic n OR 95% CI n OR 95% CI n OR 95% CI n OR 95% CI n OR 95% CI n OR 95% CI n OR 95% CI n OR 95% CI n OR 95% CI No. of subjects 4,976   10,145   8,354   44,684   8,617   20,469   32,684   12,811   7,828 No. taking drug 1,482   2,374   1,432   1,737   1,303   3,021   5,050   1,314   429 All studies 6 1.24 1.01 – 1.50 9 1.07 1.01 – 1.14 4 1.01 0.86 – 1.17 7 1.47 1.35 – 1.62 4 1.59 1.37 – 1.83 9 1.68 1.47 - 1.91 11 1.57 1.43 – 1.72 4 0.96 0.78 - 1.18 4 1.21 1.01 – 1.44 Population  Community 4 1.29 1.00 - 1.65 3 1.08 1.01 - 1.16 2 0.98* 0.79 - 1.18 5 1.50 1.36 - 1.67 2 1.76* 1.41 - 2.18 4 1.68 1.39 - 2.03 6 1.55 1.36 - 1.73 2 1.22 0.79 - 1.66 1 1.12 0.91 - 1.36  <35% fallers 3 1.34 0.93 - 1.91 2 1.09 1.00 - 1.17 1 0.94 0.75 - 1.16 3 1.62* 1.44 - 1.84 1 1.75 1.36 - 2.24 3 1.62 1.22 - 2.03 5 1.53 1.31 - 1.75 2 1.32 0.87 - 1.80 1 1.08 0.86 - 1.34  ≥35% fallers 1 1.11^ 0.78 - 1.58 1 1.04 0.87 - 1.24 1 1.13 0.72 - 1.70 2 1.22* 1.00 - 1.48 1 1.78^ 1.33 - 2.38 1 2.30 1.41 - 3.69 1 1.58 1.05 - 2.45 0 0.71 0.27 - 1.82 0 1.34 0.83 - 2.15   Long-term care 1 0.80^ 0.40 - 1.60 3 1.02 0.84 - 1.25 2 1.18* 0.83 - 1.88 1 1.38 1.14 - 1.74 1 1.55 1.24 - 1.95 3 1.63 1.35 - 1.98 2 1.43 1.11 - 1.85 1 1.03 0.61 - 1.54 1 1.41 1.05 - 1.89  Other 1 1.19^ 0.77 - 1.83 3 1.06 0.82 - 1.38 0 0.43* 0.19 - 0.97 1 1.56 1.19 - 2.05 1 1.02* 0.65 - 1.61 2 1.70 1.17 - 2.48 3 1.91 1.43 - 2.55 1 0.94 0.72 - 1.24 2 1.62^0.21 - 12.60  ≤75 yrs 4 1.33 1.03 - 1.68 4 1.11* 1.03 - 1.20 2 0.93 0.71 - 1.18 4 1.54 1.39 - 1.72 1 1.72 1.32 - 2.27 2 1.66 1.33 - 2.04 5 1.55 1.26 - 1.83 1 0.96 0.73 - 1.27 2 1.26 0.92 - 1.70  >75 yrs 2 1.04^ 0.79 - 1.38 5 0.96* 0.84 - 1.08 2 1.02 0.81 - 1.28 3 1.37 1.19 - 1.60 3 1.63 1.27 - 2.03 7 1.79 1.47 - 2.16 6 1.56 1.39 - 1.77 3 1.24 0.75 - 1.65 2 1.17 0.95 - 1.43  Good 1 1.19^ 0.77 - 1.83 5 1.04 0.89 - 1.23 1 0.87 0.69 - 1.07 1 1.66 1.25 - 2.22 2 1.34* 1.05 - 1.68 4 1.68 1.39 - 2.06 5 1.65 1.39 - 1.98 2 0.81 0.55 - 1.20 3 1.59* 1.11 - 2.24  Poor 5 1.24 0.97 - 1.54 4 1.09 1.02 - 1.16 3 1.02 0.84 - 1.21 6 1.43 1.30 - 1.58 2 1.79* 1.46 - 2.17 5 1.64 1.37 - 1.94 6 1.54 1.36 - 1.72 2 1.11 0.85 - 1.42 1 1.12* 0.93 - 1.35 Study type  Case-control 2 1.09^ 0.80 - 1.50 5 1.11 0.94 - 1.32 1 0.87 0.55 - 1.37 1 1.62* 1.31 - 2.00 1 1.23 0.92 - 1.63 3 1.83 1.42 - 2.35 3 2.18* 1.57 - 3.12 1 0.21*^0.10 - 0.45 2 1.62^0.21 - 12.60  Cohort 3 1.34 0.93 – 1.91 1 1.05 0.97 - 1.15 0 1.00 0.78 - 1.30 3 1.24* 1.05 - 1.45 0 1.90 1.35 - 2.67 2 1.67 1.36 - 2.02 3 1.51* 1.29 - 1.75 1 1.49*^1.22 - 1.83 0 NA  Cross-sectional 1 1.11^ 0.78 - 1.58 3 1.11 1.00 - 1.24 3 1.02 0.79 - 1.24 3 1.56* 1.39 - 1.76 3 1.67 1.38 - 2.00 4 1.57 1.25 - 1.96 5 1.49* 1.24 - 1.73 2 1.18*^0.90 - 1.53 2 1.44^ 0.49 - 4.03   * >95% posterior probability that the difference between ORs is greater than zero    ^ Attained by random effects inverse-variance model (Frequentist) due to unstable Bayesian model    n columns refer to new studies only  106 Chapter  3: The elderly faller's Emergency Department management: A direct observational study of care delivery and wait-times in an urban Canadian university hospital. 3.1 Introduction Care provided and time spent in the Emergency Department (ED) have been shown to have a significant impact on many future outcomes including duration of hospital stay, success in rehabilitation from surgery, and potential for subsequent ED visits.(16, 201, 202) Of considerable concern is the care provided to elderly ED patients as they are at high risk for repeat ED visits and increased wait times when compared to younger ED patients.(132, 136, 203, 204)  Approximately 30% of the elderly experience at least one fall each year and half of these will suffer multiple falls.(7, 8, 57)  When an elderly person falls, and an injury is suspected, the ED is often their first point of contact for receiving care.  As such, 14- 40% of all ED presentations by elderly persons are due to falls with over 50% of these being recurrent fallers.(16, 20, 21, 154)  As well, among the elderly, falls are responsible for more than 85% of all injury related hospitalizations and cause 90% of all hip fractures.(13, 205)  Health care delivery performance can be measured in various ways. Two ways relevant to elderly fallers in the ED are (i) ‘wait times’ or time spent in the ED waiting for different types of care and discharge, and,  (ii) delivery of recommended ‘guideline care’ to the elderly fallers themselves.  107 Emergency Department length of stay and wait times provide benchmarks to which patient flow through the ED can be compared. Many hospitals and health authorities use these data to complete assessments of care and design interventions in an attempt to minimize wait times and reduce overcrowding in the ED. Within Canada, the Canadian Association of Emergency Physicians (CAEP) has identified ED length of stay recommendations and benchmarks on acceptable waiting times to see a physician, time spent in the ED, and time spent waiting to be admitted to hospital after the decision to admit has been made.(32, 33)  Three unique guidelines or recommendations for post fall care, each of which can be delivered in the ED, have been developed.(11, 16, 31, 83)  These guidelines assist in identifying a faller’s fall risk factors, as well as help determine interventions to reduce their likelihood of experiencing a fall.(10, 83, 127, 206) Given the volume of elderly persons presenting to the ED as result of a fall, the ED is a key point for the assessment of fall risk factors and implementation of interventions designed to prevent falls.(16, 20, 21, 27, 154, 177)  Two Canadian studies have assessed the post-ED care of an elderly person who presented to the ED as result of a fall. Donaldson et al.(151) reported that in a sample of 63 female elderly fallers who had presented to the Vancouver General Hospital (VGH) ED, only 32% received post fall care from their family physician, and 24% had been referred to a physiotherapist in the 18 months after their ED visit. Subsequent to Donaldson et al’s research, Salter et al.,(152) in the same research  108 setting, reported on the care received in the 6-months after discharge from the VGH ED for 54 elderly fallers. Salter et al. observed that only 2 elderly fallers (3.7%) received care that followed the guidelines for the prevention of falls as defined by the American Geriatrics Society/British Geriatrics Society/American Academy of Orthopaedic Surgeons (AGS/BGS/AOS) Guideline for the Prevention of Falls in Older Person within six months of their visit to the VGH ED.(11)  These post-ED health service delivery studies provided important evidence of a ‘care gap’-a failure of clinical care to meet guideline recommendations. Whether the ED visit itself-the time from presentation to discharge from the ED-corresponds with published fall guidelines has not previously been investigated.  To fill this research gap, I undertook a study assessing specific ED care received by the elderly faller who presented in that ED setting.   In this prospective study, I identified elderly persons (≥70 years of age) who presented to the VGH ED due to a fall between October 31, 2007 and November 1, 2008. Data were collected on the care they received and I assessed its concordance with each of the individual guidelines for post fall care and acceptable ED wait times. I also aimed to identify demographic and ED-visit characteristics which were potential predictors of receiving care which partially or fully followed published guidelines for post fall care or CAEP wait time benchmarks.   109 3.2 Methods 3.2.1 Study design and setting My study was a prospective cohort study of elderly fallers. I identified and recruited elderly persons presenting to the VGH ED as a result of a fall between October 31, 2007 and November 1, 2008 who met the study inclusion criteria.  The VGH ED is open 24 hours daily every day of the year and is staffed with ED physicians, ED nurses, social workers, pharmacists, geriatric nurses and physiotherapists, as well as access to on-call specialists.  Among the services available to an elderly person in the VGH ED are geriatric triage nurses who are available seven days a week from 7:00 to 19:00 hours. They provide care to elderly patients and assist in referring patients to existing services for post-ED care.  I had insufficient resources to recruit and monitor patients continuously for purposes of this study. After an assessment of arrival times taken from previously completed administrative database studies completed on elderly fallers who had presented to the VGH ED(151, 152), recruitment efforts were spread out such that recruitment would capture day and evening shifts in proportion to the rate at which falls had presented in prior years. Thus, as 53% of elderly fall presentations occurred between 8:00 am and 4:00 pm in the year leading up to study, I allocated 53% of data collection efforts to day shifts.   110 3.2.2 Definition of a fall Identifying falls and fallers is best done prospectively as the proximity of the researcher to the patient gives greater confidence that the definition of a “fall” is consistently applied.(5, 6, 35)  I used the ProFaNE fall definition which is: “an unexpected event in which the participants come to rest on the ground, floor or lower level.”(35) This definition was designed specifically for use in research studies.(35)  Participants were 70 years of age or older, were able to speak and read English, and had presented to the VGH ED as result of a fall. I excluded fallers who had reported a physician diagnosis of vision impairment which precluded their ability to read the consent form. I also excluded those who were cognitively impaired, which was defined as any record of the patient having a diagnosis of dementia or Alzheimer’s disease, or a reported Mini Mental State Exam Score of less than 24.  Patients who had received care for their fall prior to presenting to the VGH ED, including patients who had been referred to the ED by a general practitioner or family physician, were also excluded from the study.  3.2.3 Patient recruitment and ethical approval Upon presentation to the ED, demographic information and injury type are routinely reported into the VGH ED census database. The VGH ED census was monitored to prospectively identify potential participants and, upon identification of a faller, the ED chart was inspected for exclusion criteria by me or a research assistant (Suzana Mitrovic). Potential participants who met the inclusion criteria were subsequently  111 approached for informed consent. This study received ethics approval from the University of British Columbia Clinical Research Ethics Review Board (approval number H06-03142).  3.2.4 Data collection Upon recruitment, patients’ data were collected including details of the fall that resulted in the ED presentation, current health and chronic conditions, falls history, living arrangements and medications. These data were collected from patient interviews, medical records, and the VGH ED record. A key source of information was the geriatric nurse assessment, which includes information on the individual’s initial complaint/injury, medical history, current medications, mobility aids, living situation, ability to complete activities of daily living, falls history, mobility, degree of continence, cognition, current social and health issues alongside the geriatric triage nurse’s recommendation for future care.  Participants were tracked until discharged from the ED (via discharge to their home or admission to hospital). While tracking patients, I or a research assistant (SM), recorded the time of care, duration, type of care and resources used in all interactions with nurses, ED physicians, specialists, and other hospital healthcare professionals.  As well, data were collected on any laboratory and radiology investigations (including x-ray, ultrasound, and computerized tomography) which were ordered while the patient was in the care of the ED. As well, for those patients admitted to hospital, data were collected from patient medical records on surgical  112 procedures and duration of hospital stay related to their fall.  ED length of stay, time to see an ED physician, time to discharge from the ED, and time from decision to admit until admission to the hospital (where applicable)  were collected using the study case report forms, and VGH ED Census data.  3.2.5 Assessment of care 3.2.5.1 Guidelines for fall prevention When comparing observed care to guideline care the guidelines of care used were the UCLA ED guidelines designed by Baraff et al. in 1996;(31) AGS/BGS/AAOS Guidelines for fall prevention published in 2001 and updated in 2011; (11, 83)  and Close et al’s PROFET care guidelines.(16) These guidelines are the product of randomized controlled trials, clinical investigation, and consensus among researchers.(11, 16, 31, 83)  Specific information on the recommendations of each guideline is presented in Table 3-1.   Each of the three guidelines recommend similar assessments of an elderly person’s health after a fall yet they do differ on the health professional recommended to perform the investigations/interventions.  3.2.5.2 Canadian Association of Emergency Physicians Emergency Department wait time guidelines When presenting to the ED, triaging determines the priority in which the patient is to receive care relative to other patients in the ED. To aid in prioritizing patients through triage, a number of patient classification instruments have been designed.  The Canadian Association of Emergency Physicians (CAEP) designed the Canadian  113 Emergency Department Triage and Acuity Scale (CTAS) to categorize patients. Upon presentation in the ED, each patient is classified into one of five CTAS acuity levels: Resuscitation (Level I), Emergent (Level II), Urgent (Level III), Less Urgent (Level IV), and Non Urgent (Level V).(32, 33, 179)  The CAEP Wait Time Guidelines are based on the individual patients’ CTAS Level and sets standards on acceptable wait times to be seen by an ED nurse, an ED physician, the total time spent in the ED, and, when the patient’s injuries require hospitalization, the acceptable wait for transfer to an in-hospital bed.(32, 33, 179) Details on the CAEP Wait Time Guidelines are shown in Table 3-2.  3.2.6 Outcomes 3.2.6.1 Concordance with falls guidelines My primary goal was to provide a description of the care received by elderly fallers presenting to the VGH ED, comparing the care provided to the recommendations from the UCLA ED guidelines(31),  AGS/BGS/AAOS guidelines(11, 83), and PROFET care.(16)  Following a previously used methodology, (151, 152) participants were separated into three unique groups: “No Guideline Care”, “Partial Guideline Care”, and “Guideline Care”. “Guideline Care” referred to full completion of all guideline recommendations. Since the geriatric triage nurse assessment incorporated many of the aspects in the AGS/BGS/AAOS, UCLA ED, and PROFET care guidelines, “Partial Guideline Care” referred to being seen by a geriatric triage  114 nurse, and “No Guideline Care” referred to receiving care only from an ED physician and/or nurse.  3.2.6.2 Concordance with wait time guidelines In assessing whether wait-time benchmarks were met, I compared the recorded durations of waiting time until the participant was1) seen by an ED physician, 2) seen by an ED nurse, 3) discharged from the ED, and 4) where appropriate, admitted to hospital after the decision to admit, to the relevant CAEP wait time standard.  3.2.7 Statistical analysis To estimate whether the sample of recruited participants was representative of all elderly fallers who presented to the VGH ED during the year of recruitment (October 31, 2007-November 1, 2008), I compared my sample of patients with data available from the VGH ED census for all elderly fallers. Key variables that were compared between study participants and the full cohort were age, sex, arrival method, discharge location, diagnosis, and the duration of time in the ED. For continuous variables (age, duration of time in ED) t-tests were used to test the hypothesis that the two samples did not differ, an a Chi-squared test for all categorical variables (proportion female, arrival mechanism, diagnosis, and discharge location).  To determine the characteristics which predicted an elderly faller to receive “Partial Guideline Care” or “No Guideline Care” I used multi-variable, logistic regression  115 analysis using the type of care received (“Partial Guideline” vs. “No Guideline Care”) as the dependent variable. I estimated unadjusted and adjusted Odds Ratio (OR) estimates using the independent variables age (≥80 years of age, <80 years of age), gender, time of presentation (7:00 -7:00 pm; 7:01 pm-6:59am), and CTAS.  Similar to the assessment of characteristics associated with receiving guideline care, I used multi-variable, logistic regression to identify characteristics associated with wait times that exceeded the recommended CAEP benchmarks. I estimated adjusted and unadjusted ORs between the independent variables of participant characteristics (including age, gender, time of presentation, and CTAS) and the dependent variable of whether the recommended CAEP benchmarks were met (after adjustment for individual CTAS score where appropriate). Data were analyzed using Microsoft Access 2003 and SAS 9.1 software.  3.3 Results 3.3.1 Participant characteristics During the period of recruitment there were 70,251 visits to the VGH ED, of which 7,764 (11%) were by individuals ≥70 years of age with 1484 (19%) of these as a result of a fall. Five hundred and ninety-four (40.0%) of the fall presentations during this period were ineligible due to patients’ cognitive impairment (n=309), inability to read or speak English (n=256), or visual impairment (n=63). Similarly, due to multiple fallers receiving treatment in the ED at the same time, 87 individuals could not be recruited due to the fallers either receiving care before they could be  116 approached to participate in the study or because the patients were in different areas of the ED not allowing data to be accurately collected from more than one patient at a time.  One hundred and eighty-eight individuals who met the eligibility criteria were invited to participate in the study with 99 participants consenting, representing 101 unique falling episodes.  Study participants’ average age was 82.5 (SD:6.2) years of age and 74% were female.  Table 3-3 outlines study participant demographics and the characteristics of the care received in the ED. Table 3-3 also provides descriptive characteristics of all elderly fallers who presented to the VGH ED during the period of recruitment. There were no differences (p>0.05) between my sample and the registry of patients who presented over the entire year in terms of age, total time in the ED, total time spent in the ED for those being discharged to community, and total time spent in the ED for those being admitted to hospital.  All study participants had at least one identified fall risk factor, the most common of which were the reporting of taking a prescription medication associated with increased fall risk  (78/99, 79%),(25, 26, 207) taking ≥4 prescription medications (67/99, 68%) or previously having suffered a fall (43/99, 43%). Among study participants, 41% lived alone, while the remainder lived with a spouse or other family.  Ninety-three (94%) participants reported having a chronic condition with the most common being hypertension (62/99=63%) followed by osteoarthritis (52/99=53%) and hearing impairment (26/101=27%). Ninety-one (92%) participants  117 reported that they were taking at least one prescription medication. Table 3-4 and 3- 5 report participants’ recorded physiological and behavioural fall risk factors.  Fifty-seven of the 101 (58%) falls occurred indoors. The most commonly reported injuries were fractures (33/101=33%) and lacerations (11/101=11%). More than one- third of the participants required admission to hospital (37/99=38%).  Twenty of the participants (20/101=20%) presented to the ED between 7:00 pm and 7:00 am. Further information on the participants’ and their falls are found in Tables 3-3, 3-4, and 3-5.  3.3.2 Care provided in the Emergency Department During their time in the ED, participants received care from a number of health care providers. For all participants (n=99), during every ED fall presentation (n=101) they were assessed by an ED physician, 62 (62%) were seen by a geriatric triage nurse, 14 (14%) were seen by a physiotherapist, and 48 (48%) required a specialist consultation regarding their injuries. Ninety-four (94%) of the participants required a laboratory or radiology assessment while in the ED.   118 3.3.3 Concordance with falls guidelines Although the assessment of the geriatric triage nurse assessment closely mirrored the recommended guideline care, some aspects of the full “Guideline Care” were not provided. The most notable omission was that no participant received an assessment/discussion of the possible association between poly- pharmacy/medication use and the falling episode requiring presentation to the ED, which is a recommendation of all three of the guidelines for care of the elderly faller.(11, 16, 31, 83) As well, while the guidelines recommend the inclusion of an ED physician, triage nurse, primary nurse, geriatric specialist, physiotherapist, and pharmacist as necessary in the provision and assessment of care, no study participant received care from all of these health professionals while in the ED.  Contrary to the recommendations contained in the UCLA ED guidelines,(31) none of the patients had data collected or a discussion on current environmental hazards, sedative use, or changes in mental status. As well, there was no provision or discussion of identified preventative mechanisms such as calcium/vitamin D supplements,(31, 128) pneuomoccoccal vaccination, and influenza vaccination.(31) Similarly, for only 2 (2%) participants was UCLA ED guideline recommendation of a discussion of the faller’s nutritional status/deficiency followed.(31) With respect to the recommendations in the AGS/BGS/AAOS guidelines(11, 83), there were no assessments of lower extremity peripheral nerves, proprioception, vision, or reflexes. Similarly, when looking at the care provided to fallers and its concordance to the  119 PROFET care,(16) no participant had an assessment of their use of social services, their perceived ability to go out, smoking status, or Barthel Score.  Of those discharged to the community (n=63), 38 (60%) were seen by a geriatric triage nurse and as such received “Partial Guideline Care.” Nine (14%) of the fallers discharged to the community received care, as recommended by the UCLA guidelines,(31) from a physiotherapist. The most common post fall care recommendation provided was to follow-up with the patients’ family physician (n=19, 20%).  Among the 101 ED fall presentations, “Partial Guideline Care” was received by 62 participants (n=61%).  In the multi-variable logistic regression analysis, the only variable which significantly influenced the likelihood of a geriatric assessment was the time of presentation after adjustment for age, gender, total time in the ED, severity of injury, and need for hospitalization. Patients who presented to the ED between 7:01 pm -6:59am were less likely to receive a geriatric triage nurse assessment than those presenting outside this time (adjusted OR 3.4, 95%CI 1.2- 9.9).  3.3.4 Emergency Department wait time For all falls (n=101), the mean time spent in the ED was 479 (SD: 408) minutes. On average, fallers who were discharged to the community waited 88 (SD: 57) minutes to see an ED physician and spent 283 (SD:151) minutes in the ED. For those  120 admitted to hospital, the mean time spent in the ED was 808 minutes (SD: 493), of which 60 (SD: 56) minutes were spent waiting for an ED physician assessment. Table 3-6 shows the average time spent waiting for an ED Physician and the total time in the ED by CTAS. Figure 3-1 shows the proportions of patients who met the CAEP Wait Time Guidelines to see an ED physician and total time in the ED. Among those patients admitted to hospital (n=38), only 10 (26.3%) were admitted to an inpatient bed within 2 hours of the decision to admit.  Logistic regression analysis on fallers discharged to the community showed that female fallers were more likely to spend time in the ED which exceeded the CAEP benchmarks (adjusted OR 8.30, 95%CI 1.64-41.95). As well, when assessing the OR estimates for meeting the CAEP wait time guidelines for seeing an ED physician, when compared to those with a CTAS level of IV,  a CTAS level of III was associated with a wait time exceeding the CAEP benchmarks (OR 5.22, 95%CI 1.12-22.45). For all other CTAS levels, the results were not statistically significant.  3.4 Discussion My study provided the first Canadian assessment of the care received by an elderly faller in an ED using prospectively collected data. I observed that no study participant received full “Guideline Care” as recommended by any of the three clinical guidelines for post fall care.(11, 16, 31, 83)  The 62% of fallers who were seen by a geriatric triage nurse received “Partial Guideline Care.”  Similarly, the total time spent in the ED by elderly fallers greatly exceeded identified wait time  121 benchmarks. Compared to men, women were at a higher risk of having a wait time that exceeded CAEP benchmarks.(32, 33)  Before discussing the implications of these findings; it is important to address the question ‘Are these 101 falls representative of elderly fallers in a major urban center or might there be a substantial sampling bias?’ I attempted to ‘sample’ fallers over the year by targeting data collection in the appropriate day/night shift proportion that accurately reflected when fallers attended the ED the year before. The recruitment rate when one of the research assistant (SM) or I were in attendance (‘open for recruitment’) was 54% which is comparable with other studies in this setting.(16, 208)  Importantly, I compared my sample of 101 patients with other samples of ‘senior emergency department fallers’ in 3 ways. I compared the sample against (a) all elderly fallers who presented to the VGH ED in 2008, (b) a sample of 58 consecutive fallers recruited at the VGH ED in 2003,(152) and (c) a highly-cited large randomized controlled trial that recruited senior fallers in the UK.(16) There was no statistically significant difference between the recruited sample and the global population of elderly fallers to the VGH ED in 2008/09. When I compare the available variables of Salter et al.(152) and Close et al.(16) to the recruited sample, there were no differences in age, gender, arrival method, admission rates, and time spent in the ED.   122 Of note, 38% of my sample of elderly fallers and 39% of those discharged to the community were not seen by a geriatric triage nurse. Eighty-six percent did not receive any assessment of their gait or balance by a physiotherapist. As well, many of the participants reported medication use, previous falls, and chronic conditions that have been associated with increased fall risks yet they did not receive any specific counseling regarding these medications or how to minimize their risks of subsequent falls.  This finding is similar to that of Snooks et al., who observed that while “Emergency Calls” for fall-related injuries were common and an opportunity to provide preventative care, this opportunity was often missed.(154)  Given that all patients in the recruited sample reported at least one environmental, physiological, or behavioural risk factor, the lack of full “Guideline Care” provision represents a significant care gap and a missed opportunity. It is recognized that the risk of falling is multi-factorial and that multiple risk factors increase the intensity of each independent risk factor for an individual.(8, 10, 14, 57)  The missed opportunity of not properly assessing or referring the elderly ED faller is highlighted by reports that tailored interventions in this population can result in fewer future falls.(11, 16, 83, 149, 155)  As well, since during the recruitment period almost 20% of all ED presentations in persons ≥70 years of age were fall related, it is clear that falls place a substantial burden on the resources of the VGH ED.  The time at which the elderly faller presented to the VGH ED had a significant impact on the likelihood of receiving aspects of full “Guideline Care.”  Fallers  123 presenting at times when the geriatric triage nurse was not on duty (19:01-6:59) significantly reduced the likelihood of receiving an assessment from a geriatric triage nurse. As well, when the ED has many elderly patients presenting for care at the same time, it may be difficult for an elderly faller to have access to the geriatric nurse for assessment sometime during their ED stay.  However, since previous research reported that geriatric triage nurse assessments and tailored interventions reduced the risk of future ED visits,(29, 30) my findings highlight that this missed opportunity may have an impact on all-cause ED visits as well as potential future falls.  A primary limitation in administering care in the ED is the limited time that an ED physician can spend with a patient.(209)  Previous analyses have found that an ED physician spends 5.2 minutes completing a history and physical examination of a patient.(209)  Given the number of factors that must be assessed in elderly fallers, it is unlikely that full geriatric assessments of falls risk can be completed during that time.  However, as noted in each of the guidelines, there are other health professionals which should provide care to an ED faller. It is noted in all of the guidelines that assessments and interventions from geriatric specialists, physiotherapists, and pharmacists should be included in the care of the elderly faller.(11, 16, 31, 83) I observed that, in my sample of 101 falls, only physiotherapists were incorporated into their ED care, and only in the care of 9 elderly fallers.   124 Fewer than 8% of study participants were treated and discharged from the ED within the established wait time durations. Most notable is the inability of the hospital to accept patients to an in-patient bed within two hours of the decision to admit. This is a concern as long waits for in-patient beds and subsequent wait for surgery are associated with an increase of patient complications.(201, 202)  Similarly there is an associated cost with time spent in the ED, which given the intensity of care provided is higher than that of other areas of the hospital.(210)  A limitation of my study was the exclusion of any elderly person with an identified cognitive impairment. However, little research has been completed on falls prevention among this population and fall prevention guidelines are aimed at cognitively intact individuals.(11, 31, 83) As well, the research assessing the impact of ED interventions on fall risk and falls incorporated into the PROFET care guidelines was completed in cognitively intact elderly persons.(16) Additionally, the care of non-English speaking elderly fallers was not assessed; while this population has been shown in other analyses to be underrepresented in healthcare research, the logistics of recruitment were too complex given the resources of this study.  The care provided to an elderly person presenting to the ED does not conform to published guidelines that could reduce the risk of future falls. Similarly, elderly fallers who presented to the ED are not receiving care within the prescribed wait time benchmarks; women were more likely to have waits that exceeded CAEP wait time benchmarks. This lack of compliance with recommend guideline care indicates that  125 a key opportunity for identifying fall risks and/or implementing fall prevention interventions is being missed.  This present study showed that 10 years after the landmark ED study in the UK by Close et al.(16) and after the development of 3 sets of fall prevention guidelines,(11, 16, 31, 83) patients are not receiving ‘standard of care’ in a major Canadian ED.  This raises important questions as to whether the clinical pathways in the guidelines are realistic. Future research should assess the impact of the ED care provided and future falls in the Canadian healthcare setting.  126 Table 3-1 Components of guideline care for the elderly faller Guideline Designed for To be administered by Specific assessments of UCLA ED Guidelines (31) Elderly person present Triage nurse, primary nurse, physician, geriatric specialist, physiotherapist, pharmacist Medications, location and cause of fall, functional status, environmental factors, medical problems, visual testing, vital signs, nutritional status, mental status, injury, cardiopulmonary examination, get up and go test AGS/BGS/AAOS Fall Guidelines (11, 83) Older persons requiring medical attention due to a fall Clinician Fall history, medications, acute or chronic medical conditions, mobility levels, vision, gait and balance, lower extremity joint function, basic neurological function including mental status, muscle strength, lower extremity peripheral nerves, proprioceptions, reflexes, tests of cortical, extra pyramidal and cerebella function; and assessment of basic cardiovascular status including heart rate and rhythm, postural pulse and blood pressure, and if appropriate, heart rate and blood pressure responses to carotid stimulation PROFET Care Guidelines (16) Person presenting to ED as a result of a fall Geriatrician Medical history, previous falls, location and cause of fall, ability to get up after fall, time spent on floor, injury, medications, living arrangements, use of social services, perceived ability to go out, mobility, smoking status, alcohol intake, AMT, Barthel score           127 Table 3-2 Canadian Association of Emergency Physicians (CAEP) wait-time benchmarks.(32, 33, 179) CTAS Level Nursing Response Time Physician Response Time Acceptable ED Time Level I Immediate Immediate 360 minutes Level II Immediate <15 minutes 360 minutes Level III <30 minutes <30 minutes 360 minutes Level IV <60 minutes <60 minutes 240 minutes Level V <120 minutes <120 minutes 240 minutes                 128 Table 3-3 Demographics of fallers Characteristic Study participants (n=99) Elderly fallers presenting to VGH ED (Oct 31/07-Nov 1/08) (n=1484) Salter  et al. (n=58)(152) PROFET trial (n=397)(16) Mean age in years (Standard Deviation) 82.5 (6.2) 83.7 (7.1)* 78.5 (5.7) 78.2 (7.5) Female n (%) 73 (74%) 1062 (71%)* 34 (63%) 269 (68%) Arriving via ambulance n (%) 66 (66%) 1021 (69%)* Not available Not available Admitted to Hospital n (%) 37 (37%) 578 (39%)* Not available Not available Diagnosed with a fracture n (%) 33 (33%) 464 (31%)* Not available Not available Mean total time spent in the ED in minutes (SD) 479 (404.8) 453 (428.5) Not available Not available Mean total time spent in the ED in minutes (SD) for hospitalized patients 808 (489.5) 773 (511.3)* Not available Not available Mean total time spent in the ED in minutes (SD) for patients discharge to community 283 (151.1) 304 (203.1)* 223 (138.5) Not available * p>0.05         129 Table 3-4 Recruited sample physiological risk factors (n = 99) Risk Factors Total (%) > 80 years of age 64 (65%) Previous fall 43 (43%) Reported Any Chronic Condition 93 (94%) Hypertension 62 (63%) Arthritis 52 (53%) Hearing Impairment 26 (26%) Diabetes 15 (15%) Parkinson's 1 (1%) Stroke 4 (4%) TIAs 5 (5%) Incontinence 12 (12%) Cancer 10 (10%) COPD/Asthma 6 (6%) Vertigo 3 (3%)            130 Table 3-5 Medication related risk factors (n = 99) Risk Factor Total (%) Medication associated with a risk of falling 78 (79%) ≥4 medications 67 (68%) Anti-hypertensives 59 (59%) Ace inhibitors 24 (24%) Type 1A antiarrhythmics 11 (11%) Diuretics 34 (34%) Beta-blockers 17 (17%) Sedative/hypnotics 18 (18%) Neuroleptics/Anti-psychotics 7 (7%) Antidepressants 19 (19%) Benzodiazepines 13 (13%) Narcotics 16 (16%) NSAIDs 17 (17%)            131 Table 3-6 Participant duration of wait times (n=101)  CTAS Level II (Emergent) (n=6) CTAS Level III (Urgent) (n=57) CTAS Level IV (Less Urgent) (n=38) Average time spent waiting for ED physician, in minutes (SD) 34.20 (25.07) 79.55 (61.53) 89.7 (60.24) Average total time spent in ED, in minutes (SD) 582.6 (452.05) 566.28 (471.58) 335.86 (223.64) Percent seen by ED Physician within CAEP wait time benchmark 0% 23% 42% Percent seen by nurse within specified time CAEP wait time benchmark 67% 51% 47% Percent who left ED within specified time CAEP wait time benchmark 50% 39% 23%         132 Figure 3-1 CAEP Guideline adherence for time spent waiting for Emergency Department physician and total time in Emergency Department  Group 1: Met guidelines for ED physician wait time and total ED wait times Group 2: Exceeded ED physician wait time guideline, met ED wait time Group 3: Met ED physician wait time guideline, exceeded ED wait time Group 4: Exceeded ED physician and ED wait time guidelines   Group 1 8% Group 2 20% Group 3 31% Group 4 41%  133 Chapter  4: The cost of fall related presentations to the Emergency Department: A prospective, in-person, patient-tracking analysis of health resource utilization1 4.1 Introduction Falls and specifically fall-related fractures in the elderly are a major source of mortality, morbidity, and disability.(1, 2)  Among persons ≥65 years of age, over 85% of all injury-related hospitalizations are due to falls.(13) Two recent systematic reviews which included studies completed in the United States, Europe, New Zealand and Australia reported that the total cost per fall ranged from $10,749- $26,676 (2009 United States Dollars).(5, 6) The increasing population of seniors (180) means total healthcare costs will increase dramatically over the coming decades and there is a need for accurate estimates of this burden for planning purposes and to help determine how to allocate resources for preventive efforts.  Fall-related costs have traditionally been estimated from administrative data. Thirty- two of 33 studies reviewed recently relied on administrative data to estimate fall- related costs.(5, 6) Although administrative datasets can provide information on large cohorts of individuals, their limitations include estimation errors from coding mistakes, and misclassification of events.(168, 169) Also, many studies using administrative data use charge data as a proxy for cost. However, charges for care  1 A version of Chapter 4 has been published. Woolcott JC, Khan LM, Mitrovic S, Anis AH, Marra CA, The cost fall related presentations to the ED: A prospective, in-person, patient-tracking analysis of health resource utilization. Osteoporos Int. 2011 Sep 3 (Epub ahead of print)   134 and procedures often differ greatly from the true cost and are not an accurate measure of economic burden.(170)  Falls by elderly individuals place a substantial burden on the Emergency Department (ED) with falls responsible for 10-40% of all ED presentations by elderly persons.(16, 20, 21, 27) Similarly, as described in Chapter 3, falls by persons ≥70 years of age are responsible for almost 20% of all presentations by elderly persons to the Vancouver General Hospital  (VGH) ED.  As well, the duration of time spent in the ED was shown to exceed recommended wait times. However, no Canadian study has prospectively assessed the costs of a fall by an elderly person.  Prospective data capture is the preferred method for identifying falls as the proximity of researchers to the patients in this setting provides greater confidence that the definition of a “fall” is consistently applied. (5, 6, 35)  The recommended fall definition is: “an unexpected event in which the participants (sic) come to rest on the ground, floor or lower level.”(35)  Thus, better estimates of the cost of falls requires more detailed information on costs which are specific to the type of fall, post-fall diagnoses and care provided in specific settings.(6) These detailed estimates of the cost per fall would complement global estimates provided by larger administrative database studies.   135 In this prospective study, I identified elderly persons who presented to the VGH ED due to a fall between October 31, 2007 and November 1, 2008. Data were prospectively collected on their healthcare resource utilizations to estimate the total cost of a fall requiring an ED presentation in Vancouver, British Columbia, Canada.  4.2 Methods 4.2.1 Study design and setting As described in Chapter 3 I prospectively collected data on a sample of elderly fallers presenting to the VGH ED between October 31, 2007 and November 1, 2008 who met the inclusion criteria and consented to participate.  This study received ethics approval from the University of British Columbia Clinical Research Ethics Review Board (approval number H06-03142).  4.2.2 Health resource utilization and cost data The data collection methodology is described in Chapter 3. Health resource utilization data included collected information on all laboratory and radiology investigations (including x-ray, ultrasound, and computerized tomography), as well as reports from specialist consultations, surgical procedures and duration of hospital stay. In addition, I recorded the ED length of stay (LOS) as measured in minutes recorded on case report forms and VGH ED census data. Hospital and rehabilitation facility length of stay, measured in days, was collected from the patient’s medical records.   136 A unit cost was assigned for each component of health resource utilization.  In estimating the component costs of healthcare resource utilizations I followed the methods accepted by the Canadian Agency for Drugs and Technologies in Health Document for the Costing Process.(162)   Unit cost estimates are shown in Table 4- 1.(210-214)  4.2.3 Statistical analysis Following the recommendations published by Baladi et al.,(162) cost-per-fall estimates were calculated using the formula    k i iijC 1   for j=1 to 101, where jC  is the cost of the fall, i  refers to the total utilization of each health resource and i  refers to the unit cost of each health resource utilization.  The Canadian Consumer Price Index for Health and Personal Care was used to adjust costs to reflect 2009 Canadian dollars.(215) Given the high incidence of fractures, particularly hip/pelvic fractures, I also estimated cost-per-fracture by modifying the above formula.  Data were analyzed using Microsoft Access 2003 and SAS 9.1 software.  4.2.4 Sensitivity analysis To assess variability in cost-per-fall between Canadian provinces, I repeated the analysis above using comparable unit cost data from the Alberta Health Care Insurance Plan.(216-218)  Like British Columbia’s, Alberta’s health care system is  137 publically funded; however, there are differences between the two systems in unit costs for procedures, interventions, and hospital care.  4.3 Results 4.3.1 Participant demographics and characteristics of care As described in Chapter 3, I recruited 99 participants, representing 101 falls. Among study participants the average age was 82.5 (SD:6.2) years of age and 74% were female. Ninety-four percent of study participants reported having at least one chronic condition, and reported taking an average of 5 (SD:2.89) prescription medications.  The most common injuries were fractures (33/101=33%) and lacerations (11/101=11%). The most common fractures were fractures of the hip/pelvis (18/33=55%), upper body (12/33=36%) and face (2/33=6%).  Twenty- seven participants (27%) required a surgical procedure during their ED visit prior to discharge home (n=6) or during their in-hospital stay (n=21).  More than a third of falls resulted in an admission to hospital (38/100=38%). Among participants who were hospitalized, the average length-of-stay in-hospital was 19 days (SD: 17 range 2-69).   Of hospitalized participants, 23/38 or 61% were discharged to a rehabilitation facility, where the mean length-of-stay was 39 days (SD: 20).   138 4.3.2 Cost of care The total cost of care for the 101 falls was $1,152,252. The mean cost of a fall resulting in an ED presentation was $11,408 (SD: $19,655). For fallers admitted to hospital, costs incurred during their in-hospital stay were responsible for 70% of the total cost. Among those admitted to hospital, the average total cost was $29,363 (SD: $22,661). An extrapolation to the total number of fall-related ED visits at VGH ED (n=1,484) yields an estimated total cost of ED care for falls of $16.9 million for the 12 months from October 31, 2007 through to and including November 1, 2008.  Table 4-2 provides costs specific to health care resources consumed in the ED.  The total cost of ED care for the 101 falls sustained by participants was $98,213, (mean cost of ED care per-fall of $972, SD: $591, range: $56-$2519). Among falls that did not require admission to hospital (63/101=62%) the mean cost of ED care was $674 (SD: $429).  Among the 33 participants whose falls caused a fracture, their total costs of care were $777,356 with an average cost of care of $23,556 (SD: $25,955).  4.3.3 Sub-group analysis of fallers suffering a hip fracture As noted above, hip fractures were identified as the most common fracture. As well, hip fracture was the most common reason for hospitalization (18/38=47%). Among fallers who had suffered a hip fracture, their length of stay in hospital was on average 24 days (SD: 17). All those with a hip fracture required surgery and 10 (56%) required a stay in a rehabilitation hospital.  The mean cost for a fall-related hip  139 fractures were $39,507 (SD: $17,932). Figure 4-1 shows the component costs for the base case model and sub groups.  4.3.4 Sensitivity analysis Following a similar costing methodology using unit cost data from Alberta, I estimated the mean cost of fall resulting in an ED visit to be $11,959.  For fallers hospitalized as result of their fall the mean cost of in-hospital procedures, consultation, x-rays and laboratory tests was $31,856.  The estimated mean cost of a fracture was $25,403, while the cost of a hip fracture was $42,231.  4.4 Discussion Over a 12 month period I prospectively collected data on 101 falls resulting in an ED visit.  Using the costing methodologies recommended by the Canadian Agency for Drug and Technology in Health(162), I found the mean cost-per-fall to exceed $11,000. Similar to previously completed cohort studies in a population of fallers presenting to the ED, almost 40% of fallers were hospitalized due to the injuries suffered as result of a fall.(16, 17) My findings are the first Canadian estimates of the costs of a fall-related ED visit.(5, 6)   I also note that the average costs in the Canadian province in which the study was completed, British Columbia, were comparable to those reported in the neighboring province, Alberta, which has a similar health delivery system.   140 Although this study focused only on the direct medical costs of falling, the non- medical costs of falls- including costs due to long-term care admissions, household changes, modifications to lifestyle and activities for both elderly persons who have fallen and their caregivers- are likely substantial and would significantly increase the total costs of a falls.(53)  For those falls that resulted in a hip fracture (17.8%), their associated direct medical cost was $39,507. Using administrative data on hip fractures among seniors >85 years of age in Ontario, Wiktorowicz et al. estimated the one-year cost of a hip fracture.(172) Their estimate, which included both direct and non-direct medical costs, was $28,977.(172)  My estimate included only the costs of in-hospital and rehabilitation care and did not account for any indirect costs or costs of care after discharge from hospital or rehabilitation facility.  By only estimating the cost of in- hospital care, I believe this to be a conservative estimate of the cost of fall-related hip fractures. These hip fracture data are independent of any potential ‘sampling bias’ and have important implications for governments anticipating the economic burden of hip fractures. I appreciate that costs of medical conditions vary in different countries so these data will not generalize outside of Canada. Nevertheless, the finding that the prospectively collected, validated costs were 36% greater than the widely quoted ‘cost’ in Canada may be of interest to health economists and policy makers in other countries where hip fracture is a major cost driver.   141 My study has several strengths. The major strength is in the collaboration among health economists, clinical researchers, and ED staff. This allowed me or a research assistant (SM) to personally track patients through the ED to accurately capture costs prospectively and in real-time for each participant. This meant I captured the individual care received by each elderly faller including investigations, procedures, consultations, and length of stay. As a result, rather than extrapolating the care and costs of care from Case Mix Group data or Resource Intensity Weight data, I was able to micro-cost according to the care received by each participant. This costing approach used the VGH Fully Allocated Cost Model which is designed specifically for use in this type of detailed costing estimates (210) alongside published costs of for consultations, procedures, and medications. (211-214)  In addition to these strengths on the economics element, my clinical partnership meant I used the accepted fall definition (35) -again in real time so I was not relying on coding/chart review. This improves my confidence in case finding compared with studies that used diagnoses as coded in administrative datasets.  There are some limitations to this study which would result in conservative estimates of cost to the healthcare system.  Participants were followed until discharge from the ED, hospital, or rehabilitation facility and no subsequent costs incurred were included in my estimates. If post-hospital costs had been included, as they were in the study by Tiedemann et al.,(53) my cost-per-fall estimates would have been larger. As well, I did not recruit non-English speaking or cognitively impaired  142 individuals who may have different fall-related health resource utilizations and costs from the study sample.  The incidence of falls is high among elderly persons yet there has been little micro- costing of these events. Using a bottom-up rather than top-down costing approach, I have identified key cost areas and have provided new data as to the costs of falls in the elderly between the time of Emergency Department presentation and their first discharge from clinical care. This study highlights the substantial burden falls puts on the Canadian healthcare system and the need to increase efforts designed to prevent falls in elderly persons.   143 Table 4-1 Unit costs of health resource utilizations Item Cost Emergency Department care(210) $47.15 per hour In hospital bed cost (210) $619.03 per day Rehabilitation hospital cost (210) $473.74 per day Physiotherapy (214) $60.59 per consultation Occupational therapist (214) $60.59 per consultation Orthopedic consultation (212) $156.78 per consultation Neurological consultation (212) $169.21 per consultation Trauma consultation (212) $161.09 per consultation Geriatrician consultation (212) $170.46 per consultation Family practice consultation (212) $161.09 per consultation Hospitalist consultation (212) $89.58 per consultation Psychiatry consultation (212) $288.73 per consultation Plastics consultation (212) $72.33 per consultation Internal medicine consultation (212) $161.09 per consultation Cardiology consultation (212) $166.15 per consultation Ophthalmology consultation (212) $89.58 per consultation Rheumatology consultation (212) $163.24 per consultation         144 Table 4-2 Costs of Emergency Department care Variable Mean Range Standard deviation Total cost of x-rays in ED (n=101)  $83 $0-498 $68 Total cost of blood work in ED (n=101)  $45 $0-148 $42 Cost of specialist consultations (n=49)  $156 $72-170 $18 Total cost of ED care (n=101)  $ 972 $56-2519 $591 Total cost of ED for discharge to community (n=63)  $673 $56-1,719 $429 Total cost of ED care for hospitalized (n=38)  $1,490 $461-2,519 $462                145 Figure 4-1 Cost of an Emergency Department fall (mean and 95%CI)     146 Chapter  5: An operations research analysis of the Emergency Department care of the elderly faller: Simulation of current care and the impact of providing care that meets wait time and falls prevention guidelines 5.1 Introduction Among all ED presentations by elderly persons, 14-40% are due to falls with over 50% of these being recurrent fallers.(16, 20, 21, 154)  As described in Chapter 3, on average, elderly fallers presenting to the VGH ED do not receive care that follows recommendations on acceptable wait times or care provision. To address how and where changes could be made to improve elderly faller’s ED care to meet the recognized benchmarks requires a model of how care is provided to the elderly faller in the ED at present.  Operations research applies mathematical tools to develop system models and discrete event simulations (DES).(34)  DES can be used to provide insights on the current activities of a system as well as to assess the impact of hypothetical changes to the system.(34, 173)  DES has been used to model a number of health care systems including surgical scheduling, portering, diagnostic services, and the ED.(34, 175, 176)  Within the ED, these DES have been used to model current care and potential interventions to reduce patient wait times, improve the provision of care, and identify staffing needs.(173, 174, 177, 181)  No previous study has applied DES to model the care of the elderly faller in the ED. I aimed to build a DES model which mirrored the current care received by the elderly  147 faller presenting to the VGH ED (‘base case’ case). I also modelled the impact of ensuring elderly fallers, while a patient of the VGH ED, received care within the Canadian Association of Emergency Physicians (CAEP) identified wait times and/or care which followed the post-fall guidelines on time and costs of providing ED care.  5.2 Methods 5.2.1 Hospital setting and Vancouver General Hospital Emergency Department For my study, I prospectively collected data on a cohort of elderly persons who had presented to the VGH ED as a result of a fall between October 31, 2007 and November 1, 2008 who met the study inclusion criteria. As noted previously in Chapters 3 and 4, the VGH ED operates 24 hours a day, 7 days a week and 365 days a year, handling more than 73,000 ED presentations each year, of which more than 1,400 are as a result of falls by persons ≥70 years of age.  As outlined in Chapter 3, at the time of the study, the VGH ED was staffed with ED physicians, ED nurses, geriatric triage nurses, social workers, porters, nurse practitioners, physiotherapists, clinical pharmacists as well as having access to specialist consultants on call.2 Similarly, the staff in the VGH ED gave usual care including the ordering of laboratory and/or radiographic testing (including x-rays,  2 Physician specialties available/on-call for ED consultations at VGH ED include: orthopaedics, neurology, plastics, family practice, hospitalists, cardiology and internal medicine.  148 computerized tomography, ultrasound, and echocardiography) according to the severity of the fall.  5.2.2 Study approach The study began with myself, and other co-investigators (Dr Karim Khan, Dr Riyad Abu-Laban) holding a number of meetings and discussions with staff working in the VGH ED. We met with members of the various teams in the VGH ED including nursing staff, geriatric triage nurses, and ED physicians to describe the aims of the research and receive feedback on the VGH ED operations and the care provided to the elderly faller.  I completed a pilot assessment of the available data using the VGH ED Census database. Data were collected on elderly fallers (≥70 years of age) who had presented between December 1, 2006 and March 30, 2007.  From the VGH ED census database I collected data on the duration of wait times (using the time of presentation as baseline) to be seen by an ED physician, an ED nurse, and the total time in the ED before discharge to the community or admission to hospital.  For elderly fallers admitted to hospital, I collected data on the duration of time between the request for an in-hospital bed and the patient leaving the ED. Other data available from the VGH ED census included the method of arrival (via ambulance or self/walk-in), radiology investigations while in the ED and discharge disposition (to community or admitted to hospital). However, care provided to the elderly faller beyond the elements listed above was not recorded in the VGH ED census  149 database. Therefore, to augment the VGH ED census data and obtain further elements required for creating a DES model, I recruited 99 participants (representing 101 falls) to record both the care that they received and the timing of that care while in the VGH ED. Study inclusion/exclusion criteria, methods of patient recruitment and data collection are described in further detail in Chapter 3.  By prospectively recording data from time of presentation to discharge, I gained an understanding of all aspects of patient care provided to elderly fallers while in the ED, examples of which include a more detailed understanding of the type of care provided based on the presenting Canadian Triage Acuity Scale (CTAS) level (179) and the points in the ED care path where the patient was waiting for care. I used these data from the 99 study participants alongside data collected from the VGH ED census on every elderly (≥70 years of age) individual faller (n=1484) who presented to the VGH ED between October 31, 2007 and November 1, 2008 to build a process map of the care received by the elderly faller in the ED as well as inform the DES model with respect to patient characteristics, care path and wait times.  5.2.3 The care path of the elderly faller Upon presentation to the VGH ED, specific characteristics of every patient are collected including the arrival time, age, gender, address, and reason for presentation. These data are entered into the VGH ED census database for every presentation to the VGH ED.  Initially, patients are seen in the registration/triage area where the triage nurse records, on the patient’s ED chart, vital signs (including  150 temperature, blood pressure, heart rate), co-morbidities, and presenting condition/injuries and assigns a CTAS level reflecting the severity of the presenting injuries. CTAS levels range from I (most severe injury requiring immediate care) to V (least severe injury for which care should be provided by an ED physician within 2 hours of presentation).(32, 33, 179)  Description of specific CTAS levels and the CAEP wait time benchmarks by CTAS level are presented in Table 5-1.  The triage nurse also assigns a bed and/or area where patients wait for subsequent care by the VGH ED nursing staff and physicians. Within the VGH ED, patients receive care in either the “acute” (a-side, b-side, or triage/hallway area) or “treatment” areas. If assigned to one of the “acute” care areas patients are assisted by an ED porter who helps move them to the assigned care area. Patients assigned to the “treatment” area usually have less serious injuries and often are able to move to the area without assistance.  In some cases, the triage nurse determines the need for immediate diagnostics and requests patients undergo laboratory, echocardiography, or radiology testing.  For those patients who have laboratory or echocardiography tests ordered by a triage nurse, the relevant technician performs these tests at bedside (i.e. the patient is not moved). For those patients for whom radiology tests were ordered, porters escort patients to and from the Radiology Department. After the procedure a nurse and/or ED physician will then assess the patients. A nurse may provide an initial assessment and treatment plan that may include providing pain medication,  151 requesting any relevant tests not previously requested, and recording current vital signs on the patients’ ED medical record. Similarly, an ED physician will provide a complete assessment of the patients including ordering relevant tests and recording their initial perceptions and diagnosis. For patients who have mild injuries an ED physician may provide immediate care and make a diagnosis, after which the patient will be discharged from the ED to the community. Alternatively an ED physician may request laboratory, echocardiography, or radiology tests or request a consultation from on-call specialists.  Specialists will assess/provide care to the elderly faller as well as assist in determining whether the individual injuries require admission to hospital. Often patients are seen repeatedly by nurses and physicians as patients are monitored for any change in condition or need for pain management as they wait for additional care, discharge, or admission to hospital. Figure 5-1 provides a process map of the care path of the elderly faller in the VGH ED.  5.2.4 Development of the discrete event simulation model The DES model was developed through an iterative process in which the system was divided into a number of different activities that elderly fallers could experience while in the ED. The likelihood of each event was estimated from the data collected from either the VGH ED census or my prospective cohort of elderly fallers presenting to the VGH ED.  The summary of model parameters and variables are reported in Tables 5-2 and 5-3. Construction of the DES model and the running of base case and all scenario analyses was done in Arena 12.0 software on a standard PC (Intel Core 2 Duo CPU, 1.20 GHZ, 2GB of Ram).  152 5.2.5 Model simulation Reflecting that the VGH ED is open 24 hours a day, 365 days a year, I designed the model to simulate a 365 day period.  Outcomes identified for comparison between the collected and simulated data were the number of fallers presenting in a 365 day period, discharge locations of elderly fallers, total time spent in the ED, time spent waiting to be seen by the ED physician, and time spent waiting for admission to hospital. I also estimated the cost of time spent in the ED using the VGH fully allocated cost model which estimates that ED care has a cost of $47.15 per hour.(210) Following established protocol for system analysis the model incorporated a warm-up period (in this case a 30 day period) to ensure the model was in a steady state before gathering any results for analysis.(219)  Mean estimates of outcomes of interest were taken from the results of 500 simulations of the DES model for the base case and all scenario analyses.  5.2.6 Scenario analysis Hypothetical systemic changes which I implemented in the care model were 1) meeting Canadian Association Emergency Physician (CAEP) guidelines for wait time to see the ED physician,(32, 33, 179) 2) meeting CAEP recommendation of admission to hospital from the ED within 2 hours of the decision to admit,(32, 33, 179) 3) provision of care which followed the current guidelines for post fall care,(11, 16, 31, 83) and 4) all three scenarios. These scenarios were chosen based on previous analyses (Chapters 3 and 4) reporting that among 101 falls by elderly persons resulting in a presentation to VGH ED less than 8% were both receiving  153 care and discharged within the CAEP guidelines and that no patients were receiving care which fully met any of the published guidelines for post fall care. I compared baseline results to the results of each scenario.  Outcomes of interest were time spent in the ED, as well as costs of ED care. In the base case DES model and when simulating changes in care delivery, I did not incorporate the costs of staffing changes or changes to the care provided to other non-elderly fallers in the VGH ED.  To build scenario analyses on wait time changes, distributions of wait times were created such that 95% of individuals would be seen by a physician within the wait time benchmark (as defined by the patient CTAS) and/or that 95% of patients to be admitted to hospital were discharged from the ED to the hospital within 120 minutes of the decision to admit/request for bed.  Similarly, I simulated the provision of guideline care (as defined as being seen by a geriatric nurse and physiotherapist) as per UCLA ED guidelines;(31) the American Geriatric Society/British Geriatric Society and the American Academy of Orthopedic Surgeons Panel on Fall Prevention (AGS/BGS/AAOS) Guidelines for fall prevention;(11, 83)  and Close et al’s PROFET care.(16) As noted in Chapter 3, during the period of data collection no patient was observed receiving full guideline care as defined by the identified guidelines even if they had been seen by a geriatric triage nurse and physiotherapist. As such, I incorporated a 20% increase in the estimated time spent with each of the geriatric triage nurse and physiotherapist to account for the additional time needed to provide full guideline care.  154 5.3 Results 5.3.1 Model performance My simulation of 500 replications of a 365 day period estimated that, annually, 1520 elderly persons would experience a fall resulting in a VGH ED presentation.  Among these, 15 (0.98%) would be classified as CTAS I, 163 (10.72%) as CTAS II, 868 (57.10%) as CTAS III, 457 (30.06%) as CTAS IV and 17(1.12%) as CTAS V. The model estimates that 967 (63.62%) would receive care from a geriatric triage nurse and 198 (13.03%) would receive care from a physiotherapist.  The total cost of ED care was estimated to be $560,204 (range: $387,257-$660,540). Table 5-4 shows the DES model results for the base case and all of the scenario analyses.  5.3.2 Scenario analyses 5.3.2.1 Scenario 1: For all elderly fallers presenting to the VGH ED, wait time from presentation to being seen by an ED physician should follow CAEP wait time benchmarks (Table 5-1) (32, 33, 179) (as defined by the patient’s CTAS). Compared to the base case estimates, for fallers who were hospitalized (n=612, range: 547-700), the impact of this change was a reduction in the mean time spent in the VGH ED by an average of 44 minutes (range: 68-31 minutes). Similarly, compared to base case results, for those fallers discharged (n=925, range: 828- 1017) it was estimated that the total time spent in the ED was on average 41 minutes (range: 53-27 minutes) shorter. By meeting these benchmarks, the estimated total time spent in the VGH ED by elderly fallers could be reduced by  155 1071 hours over the course of a single year. The opportunity cost to the ED of not meeting this benchmark is estimated to be $50,483.  5.3.2.2 Scenario 2: For all elderly fallers who are to be admitted to hospital, wait time from the decision to admit until discharge from the VGH ED should not exceed 120 minutes.(32, 33, 179) In this scenario, for patients admitted to hospital (n=615, range: 529-708), the estimated average time an elderly faller spent in the VGH ED was 390 minutes (range: 370-440 minutes). Compared to the base case results, this reduced wait time reflects an estimated decrease of 396 minutes (range: 346- 416 minutes) per patient. In meeting these benchmarks, the total time spent in the VGH ED by elderly fallers could be reduced by 4059 hours over the course of a single year resulting reduction in the opportunity costs to the ED of $191,381.  5.3.2.3 Scenario 3: All elderly fallers should receive care from a geriatric triage nurse and physiotherapist. When compared to the time spent in the ED in the base case analysis, ensuring that each patient received care from the geriatric triage nurse and physiotherapist resulted in an increase in the average total time spent in the ED by 24 minutes (range:12-53 minutes). For those admitted to hospital the estimated increase in time spent in the VGH ED was 26 minutes (range: 9-65 minutes), and for those discharged to community the increase in time spent in the VGH ED was 25 minutes  156 (range: 8-48 minutes). The estimated additional opportunity costs to the ED associated with providing this care is estimated to be $28,648.  5.3.2.4 Scenario 4: This scenario simulated both the CAEP wait time benchmarks(32, 33, 179) being met and post fall prevention guidelines(11, 16, 31, 83) followed for every faller presenting to VGH ED. Under this best case scenario, when compared to the base case results, the total time spent in the VGH ED would change. The estimated average time a community faller (n=917, range: 827-1030) would spend in the VGH ED fell by 16 minutes (range: 5-28 minutes). For the elderly faller who requires hospital admission (n=613, range: 522-708), it was estimated that their average time in the VGH ED would be reduced, on average, by 412 minutes (range: 394-424). This reduction in time spent in the ED represents a potential reduction in the annual estimated VGH ED costs associated with elderly fallers of $210,379.  5.4 Discussion My model provides the first DES assessment simulating the care of elderly fallers while they are patients in the ED. Using data prospectively collected from elderly fallers and from hospital databases I was able to model the ‘as-is’ or current care provision,  as well as, assess the impact of incorporating changes to the timing and type of care received.  I found that when compared to the base case results, through the provision of care within the identified CAEP benchmarks on acceptable wait times, the estimated time an elderly faller spent in the VGH ED could be reduced  157 significantly. Not surprisingly, simulations including the provision of addition care to all fallers following the published guidelines for post fall care resulted in an increase in the time an elderly faller would spend in the VGH ED compared to the estimated average time spent in the ED in the base case.  My findings highlight that waiting for ED physician care and beds impacts the total time an elderly faller spends in the ED. In simulations in which care was provided to the elderly faller that adhered to the CAEP benchmarks for acceptable wait times, there were significant reductions in the total time an elderly faller spent in the VGH ED. This finding also highlights a potentially inefficient use of resources as the intensity and cost of a VGH ED bed ($1131 per day) is much greater than the cost of an acute hospital bed ($619 per day).(210)  In contrast to scenario assessments of the impact of meeting CAEP wait time benchmarks, when simulating the potential impact of the provision of post-fall guideline care, the time spent in the VGH ED increased. When compared to base case results, the annual VGH ED cost of providing care to elderly fallers was estimated to increase by $28,648. However, Close et al. found that compared to fallers who did not receive PROFET care, elderly fallers receiving PROFET care in the ED had a significant reduction of fall risk (OR 0.39, 95%CI 0.23-0.66) and lower likelihood of being hospitalized (OR 0.61, 95%CI 0.35-1.05) in the 12-months post ED fall.(16)  As well, I found that the cost of a fall requiring presentation to the VGH ED had an estimated average cost of $11,408 (SD: $19,655) and a fall requiring  158 hospitalization had an estimated average cost of $29,363 (SD: $22,661). Therefore, further assessment should be completed to understand the total impact on costs and effectiveness associated with providing guideline care in the ED.  Finally, in my best case scenario simulations where patients were meeting the CAEP wait time benchmarks and receiving guideline care, I observed that the estimated time an elderly faller spends in the VGH ED was less than that in the base case even though the quality of care was improved. The decrease in time spent in the VGH ED was significantly lower among those admitted to hospital (estimated reduction of 411 minutes, range: 394-424). This again highlights the substantial bottleneck in the system which exists for patients waiting for admission to a hospital bed at VGH.  The limitations of this study are primarily related to the extrapolation of results from the single patient sub-set of elderly fallers presenting to the VGH ED to other patient sub-groups and/or EDs.  Data were collected on the care received by elderly fallers presenting to the VGH ED and did not explicitly incorporate data related to the ED physician staffing, acute in-hospital beds, laboratory/radiology capacity, and other factors. As such no conclusion can be made on the VGH staffing levels and/or ED patient care for patients who were not elderly fallers.  As well, before any inter patient or facility comparisons can be made beyond the population of elderly fallers presenting to the VGH ED, an assessment of the  159 characteristics of patient populations, facility processes, and capacity should be completed to ensure generalizability. Finally, it should also be noted that the recruited sample (n=101) was comprised of cognitively intact, English speaking persons over 70 years of age.  In summary, my study highlighted the potential savings in both time spent in the VGH ED and VGH ED resources/dollars which could be directed towards the care of other ED patients if the current guidelines of care were met.(32, 33, 179)  Similarly, it was shown that providing care that met the current post fall guidelines(11, 16, 31, 83)  would increase the time an individual spent in the VGH ED. Further ED research and cost effectiveness analyses should be completed to assess whether the increased costs associated with providing guideline care to an elderly faller are offset by a reduction of future fall presentations to the VGH ED.    160 Table 5-1 Canadian Triage Acuity Scale Levels(32, 33, 179) CTAS Level Level of Illness/Acuity Physician Response Time Sentinel Diagnosis Level I Resuscitation Immediate Cardiac arrest Level II Emergent <15 minutes Chest pain Level III Urgent <30 minutes Moderate asthma Level IV Less Urgent <60 minutes Minor trauma Level V Non Urgent <120 minutes Common cold     161 Figure 5-1 Vancouver General Hospital Emergency Department process map of elderly faller P h y s i c i a n G e r i a t r i c  N u r s e R e g i s t e r e d  N u r s e P a t i e n t T r i a g e  N u r s e Arrive to ED by Ambulance Walk into ED Triage Asses Acuity Level (2-4) Assign Area A Assign ED Bed Based on Acuity Level Assign Area B Assign Treatment Area Level 1 & 5 are rare Testing Required? Make Diagnosis Lab work ordered No Yes Testing Required? Order TestsYes No Order TestsYes Order TestsYes Testing Required? This process is started based on availability. Starting this process can be delayed. This process is started based on availability. Starting this process can be delayed. Queuing can occur Diagnosis made Leave ED Admit to Hospital Discharge to Home Consultatio n Required? Request Consult Wait for Conult Ortho Consult Finalize Treatment Plan Yes No Patient Requires Admission No Register with ED Request Lab Work / Recommen d X-Ray Send Patient for Testing Perform Assessment Perform Assessment Perform Assessment Send Patient for Testing Send Patient for Testing Review Results Prepare Patient for Admission Yes Wait for Transfer to Hospital Treat Patient in ED Receive Blood Testing / X- Ray Move to ED Bed Arrange for Transportation Home Discharge Patient To Yes Assess Perform Road Test Hallway/ temporary holding Yes   162 Table 5-2 Emergency Department wait time distributions and sources of data Parameter Distribution* Time Distribution Source of Data Wait for Triage Beta -0.5+31*Beta(0.193,1.03) VGH ED census database data collected on  all VGH ED elderly fall presentations (n=1484) Time spent in Triage Normal Normal (mean=4, SD=6) Recruited Population (n=101) Wait for move to area for treatment Weibull -0.5+Weibull(6.02,0.732) Recruited Population (n=101) Wait for ED Physician care Gamma 2+Gamm(59,1.1) VGH ED census database data collected on  all VGH ED elderly fall presentations (n=1484) Wait for testing Beta -0.001+137*Beta(0.498,0.471) Recruited Population (n=101) Wait for discharge to community post initial EP care Beta -0.001 + 588 *Beta(1.55, 7.14) Recruited Population (n=101) Wait for admission to hospital Log Normal -0.001 + LogNormal (486, 1.27e+003) VGH ED census database data collected on  all VGH ED elderly fall presentations (n=1484) Wait for Triage Beta -0.5+31*Beta(0.193,1.03) VGH ED census database data collected on  all VGH ED elderly fall presentations (n=1484) Time spent in Triage Normal Normal (mean=4, SD=6) Recruited Population (n=101) Wait for move to area for treatment Weibull -0.5+Weibull(6.02,0.732) Recruited Population (n=101) Wait for ED Physician care Gamma 2+Gamm(59,1.1) VGH ED census database data collected on  all VGH ED elderly fall presentations (n=1484) Wait for testing Beta -0.001+137*Beta(0.498,0.471) Recruited Population (n=101) * Distributions were fitted using Arena software 10.2 Input Analyzer for identifying distributions of best fit. All times were measured in minutes      163 Table 5-3 Probability of Emergency Department events Parameter Event Probability Probability of EHS arrival 69% Probability of arrival via walk-in 31% Probability of assignment of CTAS I 1% Probability of assignment of CTAS II 11% Probability of assignment of CTAS III 57% Probability of assignment of CTAS IV 30% Probability of assignment of CTAS V 1% Probability of requiring radiology testing 69% Probability of requiring laboratory testing 49% Probability of receiving an echocardiogram 31% Probability of receiving a specialist consultation 54% Probability of being seen by a geriatric triage nurse 60% Probability of being seen by a physiotherapist 14%   164 Table 5-4 Simulation and scenario analysis results Outcome of interest Base Case: (taken from 500 iterations of model) Scenario 1: Patients wait-time to be seen by ED physician within CAEP benchmarks(32, 33, 179) Scenario 2: Patients wait-time to be admitted to hospital within CAEP benchmarks(32, 33, 179) Scenario 3: All patients to receive post fall guideline care(11, 16, 31, 83) Scenario 4: All patients to receive post fall guideline care(11, 16, 31, 83) and to be seen within CAEP wait time benchmarks(32, 33, 179) Total number of elderly fallers during a 365-day period 1520 (SD:48.38) 1537 1541 1514 1528 Average time spent in ED, in minutes 469 (range:315-553) 451 (range:394-527) 335 (range: 329-382) 493 (range: 481-515) 318 (range: 299-347) Average time spent in ED for those admitted to hospital, in minutes 786 (range: 672-1006) 742 (range: 617-935) 390 (range: 370-440) 812 (range:663-972) 374 (range: 362-388) Average time spent in ED for those discharge to community, in minutes 299 (range:275-318) 258 (range: 241-271) 301 (range: 270-322) 324 (range: 284-347) 283 (range:272-294) Number of patients admitted to hospital n(%) 606 (39.87%) 611 (39.75%) 615 (39.90%) 599 (39.56%) 613 (40.12%) Time spent waiting to see an ED physician 86 (range:74-95) 26 (range: 23-29) 82 (range: 68-93) 88 (range: 77-101) 25 (range: 21-29) Time spent waiting for admission to hospital after a bed request 485 (range: 364-805) 484 (range: 366-773) 92 (range: 79-115) 480 (range: 370-791) 91 (range: 82-117)  165 Chapter  6: Integrated discussion In this chapter I summarize the findings from Chapters 2-5, discuss the significance of my research and how it extends the body of knowledge in the area of falls in the elderly. I also outline limitations of my research findings, and provide suggestions on future research.  6.1 Overview of my findings I began my thesis with a meta-analysis which examined the association between an older person using one of 9 specific medication classes and the likelihood of experiencing a fall.  This meta-analysis updated Leipzig et al’s meta-analysis (25, 26) using data published before 1996 and extended it by including 22 studies published between 1996 and 2007.  My estimates of unadjusted OR and 95% CrI indicated that, when compared to non-use, use of each of anti-hypertensives, diuretics, sedatives and hypnotics, neuroleptics and anti-psychotics, antidepressants, benzodiazepines, and non-steroidal anti-inflammatory medications were associated with increased falls in elderly. I found that neither beta-blocker use nor narcotic use by elderly persons were associated with an increased likelihood of suffering a fall. After adjustment for a number of confounders, I estimated adjusted OR and 95% CrI for the use of diuretics, neuroleptics and anti-psychotics, antidepressants, and benzodiazepines. After adjustment I observed that compared to non-use, antidepressant and benzodiazepine use are each associated with falling by elderly persons.   166 My second study complemented studies by Donaldson et al. (2005)(151) and Salter et al. (2006),(152) who reported on post discharge care of elderly fallers  who had presented to the Vancouver General Hospital (VGH) Emergency Department (ED). Neither of those studies assessed the care received during the patient’s time in the ED itself. I answered the question: “What specific care is being provided in the ED and how long does it take to arrive?” My primary outcomes were 1) concordance between the care received by the elderly faller to the current clinical guidelines for post fall care(11, 16, 31, 83) and 2) whether care was provided within the CAEP benchmarks for these patients.(32, 33, 179) I found that none of 99 participants (101 falls) received full guideline care, and 62 received partial guideline care. Fewer than 8% of my sample were treated by an ED physician and were discharged from the ED within the CAEP benchmarks.  Included in the data prospectively collected from the 101 fall-related ED presentations were the direct health resource utilizations of an elderly person who presented to the ED with a fall. I applied unit costs to each component of resource utilization by the elderly faller from the time of ED presentation to discharge from either: the VGH ED, hospital or rehabilitation hospital. I followed the methods recommended by the Canadian Agency for Drugs and Technologies in Health for completing an economic evaluation of health technologies.(162) I estimated that the mean cost per fall by an elderly person was $11,408. Falls which resulted in a hospitalization were estimated to cost, on average, $29,363. For falls which caused  167 a fracture, the estimated average cost was $23,556 and falls which caused a hip fracture had an estimated average cost of $39,507.  From my data on the care and resource utilization of the elderly faller within the ED (Chapters 3 and 4), supplemented with data taken from the VGH ED census, I built a discrete event simulation model of the care path of the elderly faller while a patient of the VGH ED.  By simulating potential changes in the type and timing of care provided in the VGH ED I estimated that through decreasing the wait times to be seen by an ED physician, elderly fallers could be discharged from the ED approximately 40 minutes sooner. This increased wait time reflects an opportunity cost associated with providing ED care to elderly fallers of $50,483. Similarly, ensuring that patients to be admitted to hospital are discharged from the ED within 2 hours of the decision to admit would reduce the annual cost of providing ED care by $191,381. Simulating the provision of current guideline care(11, 16, 31, 83) to elderly fallers led to an estimated annual increase in costs of $28,648.  6.2 Unique contributions, impact and implications My thesis contributes significantly to the current body of knowledge in four main ways. These include both methodological and content specific advances which I outline below.   168 6.2.1 Medication use in the elderly: Methods advancement and new knowledge Poly-pharmacy in the elderly is a major issue in internal medicine and my study was the first to address the relation between medication and falls in the elderly since Leipzig’s 1999 publications.(25, 26)  That paper contained no data published after 1996. Since that time new drugs have been marketed for the elderly and new classes of drugs have been developed. Bayesian meta-analysis was an appropriate method to both integrate Leipzig et al’s findings with studies published since 1996. My meta-analysis is a significant methodological contribution as it is one of the first meta-analyses to use fully informed priors in the calculation in OR estimates.(178) As well, the use of Bayesian methodology allowed me to make statements on the probability that the ORs are greater than 1 when the 95% CrI included 1, a key difference in the outcome and potential interpretation of Frequentist OR estimates.(178, 200)  Importantly, these methodological advances allowed me to provide novel estimates of association between 9 medication classes and the risk of falling. The clinical implication of my research is that sedatives/hypnotics, neuroleptics/anti-psychotics, antidepressants, and benzodiazepines, have particularly high association with falls. Given the frequency of use of these drugs in the elderly, physicians should screen carefully for other fall risk factors and consider fall prevention strategies (e.g. strength and balance training) when prescribing these drugs. The importance of the research was reflected in its publication in the general medical journal ‘Archives of  169 Internal Medicine’ and the subsequent medical press and general public news coverage the paper received.  6.2.2 Care gap in the Emergency Department: Implications for health authorities and professional bodies By directly observing and noting the care of elderly fallers during their time in the VGH ED, I was able to report on the specifics of care provided. While labour intensive, this data was rich in detail and provided key insights into how care is provided. The care provided to elderly fallers did not meet the post-fall guidelines. There were significant gaps in care and importantly, a missed opportunity to reduce elderly fallers’ subsequent fall risk.(16, 154)  I also found that although geriatric triage nurses are present in the ED from 7:00 am – 7:00 pm, ostensibly to provide care to patients such as elderly fallers, 38% of participants in my sample did not receive an assessment from the geriatric triage nurse. This highlights a health system failure to translate evidence into practice. Geriatric triage nurses’ role was tested in the DEED II trial which showed that care from these professionals can to reduce ED and hospital admissions.(30) As shown in Chapter 3, the goal of the geriatric triage nurse, to provide improved/specialized care to elderly persons while patients of the VGH ED, are not being met for the sub-group of elderly fallers.  I also observed that the timing of the care provided and total duration spent in the ED by elderly fallers exceeded the current wait time benchmarks.(32, 33, 179)  My study was the first to prospectively collect data to assess the care of the elderly faller  170 with respect to identified guidelines for care and wait times and as such, key learning on care gaps in the VGH ED  A strength of this study was that by distributing recruitment efforts to ensure data were collected on elderly fallers in proportion to the rate at which falls had presented to the ED reduced some of the potential systematic bias of collecting data on only those fallers who presented during a specific time/period of the day. As well, prospectively collected data allowed me to apply the accepted fall definition for use in fall research studies(35), rather than rely on coding/chart review to identify fallers. As such, my confidence in the case finding is much greater compared to previously completed studies that used diagnostic codes from administrative databases.  My study assessing the care of the elderly faller in the ED showed that although there are geriatric triage nurses dedicated to providing care to elderly patients presenting to the VGH ED, many elderly fallers did not receive an assessment from the geriatric triage nurse. As well, the care provided to elderly fallers did not follow the post-fall guidelines for fall prevention, highlighting a significant gap in care and is a missed opportunity to potentially reduce an elderly faller’s future fall risk.(16, 154) My work highlights the gap between knowledge and execution – the ‘implementation and dissemination’/knowledge translation challenge. This highlights an important avenue for further research (see below).   171 6.2.3 The first Canadian estimates of the cost of falls in the Emergency Department As highlighted by Heinrich et al.(6) and Davis et al.,(5) there are gaps in knowledge with respect to the estimates of the costs per fall (as defined by standard criteria) using prospectively collected data.  In Chapter 4, following these authors’ recommendations(5, 6) and the current Canadian guidelines for estimating costs,(162) I have provided the first Canadian estimate of the average cost of a falls by elderly people who have presented to the ED.  International comparisons are of limited value due to the structure of the different health care systems as well as economic differences. However, my estimates of cost per fall are in the same order of magnitude as the only previous, comparable studies in Australia,(220) the United States,(221, 222) and Europe.(223)  I also used prospectively collected data to provide estimates of: cost per fall related hospitalization, cost per fall related fracture, and cost per fall related hip fracture.  This study further highlighted the substantial burden that elderly falls place on the Canadian health care system.  6.2.4 A discrete event simulation of the care provided to elderly fallers in the Emergency Department Although operations research has been used in a number of different health care settings, my discrete event simulation (DES) model was the first to simulate the care path of an elderly faller who has presented to the ED. As well, I was able to estimate the impact of changing the timing and type of care received in the ED by the elderly faller providing estimates of potential time and cost savings/increases.  172 When simulating wait times to be seen by an ED physician or admission to hospital that met the current CAEP benchmarks,(32, 33, 179) the average total time spent in the ED by elderly fallers’ decreased compared to the base case.  When simulating care that met the guidelines for post-fall care,(11, 16, 31, 83) it was observed that the estimated average time spent in the ED increased. However, in the best case scenarios in which wait time benchmarks for time to see an ED physician and total wait time to be admitted to hospital was met alongside the provision of guideline care, the average total time elderly fallers spent in the ED declined when compared to my base case results.  6.3 Limitations As with any study, my thesis includes limitations, which have been detailed within each chapter (Chapter 2-5).  My meta-analysis determined updated pooled OR estimates for those medication classes for which ≥4 published studies had been completed that met our inclusion criteria. As a result, I was able to calculate a Bayesian OR estimate for only 8 of the 22 medication classes that Leipzig et al. initially assessed.(25, 26) However, the total number of subjects included in studies completed since 1996 was much greater than the total subjects included in Leipzig et al.’s meta-analysis.(25, 26)  As well, both Leipzig et al.’s meta-analysis and my updated meta-analysis report that, in many studies, the methods used to determine whether a subject had experienced a fall and the types of medications being used during the study’s analytical horizon were not standardized.(25, 26)  Sixteen of the  173 22 studies included in my meta-analysis were deemed to be of poor quality based on the methods used for reporting falls and the medications used at the time of fall.  An important limitation to my findings from Chapters 3-5 is that I prospectively collected data only from cognitively intact elderly fallers in the ED, who were competent in English. Therefore my results cannot be extrapolated beyond this population of elderly fallers.  Also my research setting was a tertiary care ED.  As such, generalizability to other types of institutions is limited given that patient characteristics, facility processes, and capacity differ in different settings.  For my assessment of the provision of care (Chapter 3), I only used data collected while the participant was a patient was in the ED. As such I am not able to assess the post ED-care and do not know to what extent participants received guideline care for the prevention of falls after discharge from the ED.  Data collection on health resource utilization ceased upon discharge from ED, hospital care or rehabilitation hospital. This truncated my estimates of cost per fall resulting in ED presentation. It is likely that had post-hospital costs been included my estimates of the cost of a fall by an elderly person presenting to the ED would have been much higher. It is likely that my estimate of the cost per fall is quite conservative as previous research has shown that post-hospital health resource utilization (including indirect costs) as result of an elderly person experiencing a fall are significant and exceed the direct medical costs.(171)  174 6.4 Recommendations for future research The articles chosen for inclusion in our meta- analysis reported very divergent results. For only 1 medication class- benzodiazepines- were the OR estimates consistent in the association between medication use and falling, with all of the identified studies reporting OR estimates greater than 1. In all of the other 8 medication classes for which there were sufficient studies to complete an updated meta-analysis OR estimate, the results varied from being greater than 1 to less than 1. In the 9 medication classes for which there were sufficient studies to complete an updated meta-analysis, the studies often were underpowered with very wide confidence intervals for the unadjusted and adjusted OR estimates. This highlights a need for larger studies, powered to detect differences between populations and thus improve the quality of information about the risks associated with the use of specific medication classes. As shown in our sub-group analyses the risk of falling associated with the use of many medication classes differed if the study was completed in a population of elderly long-term care residents or elderly community- dwellers. As such, future research should be completed independently in both long- term care facilities and in the community setting to better inform physicians and pharmacists when deciding on the proper pharmacotherapy to provide.  As well, future research should assess the impact of medication withdrawal on the probability of experiencing a fall.  Robertson et al. showed in their randomized controlled trial that the withdrawal of psychotropics can be an effective intervention to reduce an elderly person’s probability of experiencing a fall.(184)  Future research  175 should be completed to assess the potential impact of withdrawal of other medications that are known to increase falls. Specific medications which should be assessed to see if their withdrawal will reduce falls compared to those who continue to use: antihypertensives, diuretics, sedative and hypnotics, sedatives and hypnotics, neuroleptics and antipsychotics, antidepressants, and NSAIDs.  The results of my second research study (Chapter 3) showed a significant care gap where many patients presenting to the ED were not receiving guideline care (UCLA ED, the AGS/BGS/AAOS, or PROFET).(11, 16, 31, 83)  This fact raises a number of questions regarding the feasibility of delivering this care, therefore it is my recommendation that future research be undertaken to assess whether this care can be provided to elderly fallers while they are in the ED.  Given the finding that wait times experienced by elderly fallers exceeded the current benchmarks, further investigation as to why patients are not being seen within the recommended times should be undertaken. Specifically, future studies should assess the reasons for excess wait times and potential solutions. These assessments should be undertaken for all ED patients, but particular focus should be paid to the elderly female faller, a sub-group shown to be at increased risk of having an excessive wait time when compared to men and adjusted for age, time of presentation, and Canadian Triage Acuity Scale level.(32, 33, 179)   176 Additional research should be completed to understand the total economic burden of falls in Canada. There are many other types of falls and fall-related outcomes beyond the ones for which I have provided an estimate of cost.  Similar to the work completed by Tiedemann et al. in the Australian healthcare setting,(53) I recommend that future studies investigating the cost per fall use a longer analytical horizon and include post discharge resource utilizations. The development of a registry collecting data on all resource utilization and indirect costs in a 12 month period post ED-visits would allow for a more complete understanding of the total economic burden and long term outcomes associated with suffering a fall.  As described in Chapter 5, when developing the simulation model, I restricted my analysis to assess the ED care provided to elderly fallers.  As such, we were not able to provide assessments on the staffing of the ED, care provided to other patients in the ED, or impact of any changes made to the care of elderly fallers on other patients in the ED. Future studies should be of a larger scope to better assess these issues.  It was also observed that the impact of improving the care provided to meet the current fall prevention guidelines was associated with increased time spent in the ED and as a result, increased cost. However, given the expectation that these interventions could reduce falls, further investigation should be completed to determine whether the cost of preventative care could be offset by reduced future fall-related ED presentations. Through completing a full cost effectiveness analysis  177 comparing these strategies, estimating incremental cost effectiveness ratios for the relevant outcomes (cost per fall averted, cost per fracture averted, cost per hospitalization averted, cost per life year gained and cost per quality adjusted life year), would provide decision makers with important information to help determine whether these interventions are an efficient use of their limited budgets.  In summary, with respect to falls in the elderly, this thesis addressed questions on the association of medication classes on the risk of experiencing a fall, the care provided to an elderly faller in the ED, the costs of a fall resulting in an ED presentation, and potential impact of changing the timing and type of care provided in the ED.  In each chapter, I followed the recommendations of previously completed studies to update the current base of knowledge and identified potential areas of future improvement. As I am keenly aware that publication and presentation of results is not enough to prompt change in behaviour, in the future I will continue to actively promote increased knowledge translation of my findings as well as future investigation into the issue of falls in the elderly.   178 References  1. Tinetti M, Williams C. Falls, injuries due to falls, and the risk of admission to a nursing home. - N Engl J Med.1997 Oct 30;337(18):1279-84.(0028-4793 (Print)). 2. Tinetti M, Kumar C. The patient who falls: "it's always a trade-off". - JAMA.2010 Jan 20;303(3):258-66.(1538-3598 (Electronic); 0098-7484 (Linking)). 3. Nevitt M, Cummings S, Kidd S, Black D. Risk factors for recurrent nonsyncopal falls. A prospective study. - JAMA.1989 May 12;261(18):2663-8.(0098-7484 (Print)). 4. Self-reported falls and fall-related injuries among persons aged > or =65 years-- united states, 2006. - MMWR Morb Mortal Wkly Rep.2008 Mar 7;57(9):225-9.(1545- 861X (Electronic); 0149-2195 (Linking)). 5. Davis JC, Robertson MC, Ashe MC, Liu-Ambrose T, Khan KM, Marra CA. International comparison of cost of falls in older adults living in the community: A systematic review. - Osteoporos Int.2010 Feb 27.(1433-2965 (Electronic); 0937- 941X (Linking)). 6. Heinrich S, Rapp K, Rissmann U, Becker C, Konig H. Cost of falls in old age: A systematic review. - Osteoporos Int.2010 Jun;21(6):891-902.Epub 2009 Nov 19.(1433-2965 (Electronic); 0937-941X (Linking)). 7. Campbell A, Borrie M, Spears G. Risk factors for falls in a community-based prospective study of people 70 years and older. - J Gerontol.1989 Jul;44(4):M112- 7.(0022-1422 (Print)). 8. Tinetti M, Speechley M, Ginter S. Risk factors for falls among elderly persons living in the community. - N Engl J Med.1988 Dec 29;319(26):1701-7.(0028-4793 (Print)). 9. O'Loughlin J, Robitaille Y, Boivin J, Suissa S. Incidence of and risk factors for falls and injurious falls among the community-dwelling elderly. - Am J Epidemiol.1993 Feb 1;137(3):342-54.(0002-9262 (Print)). 10. Lord SR, Sherrington C, Menz HB. Falls in older people: Risk factors and strategies for prevention. New York: Cambridge University Press; 2007. 11. Guideline for the prevention of falls in older persons. American Geriatrics Society, British Geriatrics Society, and American Academy of Orthopaedic Surgeons panel on falls prevention. - J Am Geriatr Soc.2001 May;49(5):664-72.(0002-8614 (Print)).  179 12. Division of Aging and Seniors, PHAC Canada. Report on senior's fall in Canada. Ontario, division of aging and seniors. public Health agency of Canada. 13. Weir E, Culmer L. Fall prevention in the elderly population. CMAJ. 2004 September 28, 2004;171(7):724. 14. WHO global report on falls prevention in older age. France: World Health Organization; 2007. 15. Nevitt M, Cummings S, Hudes E. Risk factors for injurious falls: A prospective study. - J Gerontol.1991 Sep;46(5):M164-70.(0022-1422 (Print); 0022-1422 (Linking)). 16. Close J, Ellis M, Hooper R, Glucksman E, Jackson S, Swift C. Prevention of falls in the elderly trial (PROFET): A randomised controlled trial. - Lancet.1999 Jan 9;353(9147):93-7.(0140-6736 (Print)). 17. Woolcott JC, Davis JC, Buchanan J, Abu-Laban RB, Khan KM, Marra CA. The direct costs of injurious falls in seniors. International Society for Pharmacoeconomics and Outcomes Research Thirteenth Annual International Meeting May 3-7, 2008. 18. Parkkari J, Kannus P, Palvanen M, Natri A, Vainio J, Aho H, et al. Majority of hip fractures occur as a result of a fall and impact on the greater trochanter of the femur: A prospective controlled hip fracture study with 206 consecutive patients. - Calcif Tissue Int.1999 Sep;65(3):183-7.(0171-967X (Print); 0171-967X (Linking)). 19. Nguyen N, Pongchaiyakul C, Center J, Eisman J, Nguyen T. Identification of high-risk individuals for hip fracture: A 14-year prospective study. - J Bone Miner Res.2005 Nov;20(11):1921-8.Epub 2005 May 31.(0884-0431 (Print); 0884-0431 (Linking)). 20. Davies A, Kenny R. Falls presenting to the accident and emergency department: Types of presentation and risk factor profile. - Age Ageing.1996 Sep;25(5):362- 6.(0002-0729 (Print); 0002-0729 (Linking)). 21. Bell A, Talbot-Stern J, Hennessy A. Characteristics and outcomes of older patients presenting to the emergency department after a fall: A retrospective analysis. - Med J Aust.2000 Aug 21;173(4):179-82.(0025-729X (Print); 0025-729X (Linking)). 22. Sattin R, Lambert Huber D, DeVito C, Rodriguez J, Ros A, Bacchelli S, et al. The incidence of fall injury events among the elderly in a defined population. - Am J Epidemiol.1990 Jun;131(6):1028-37.(0002-9262 (Print); 0002-9262 (Linking)). 23. National Center for Injury Prevention and Control:. Statistics and activities. int J trauma nurs 1998;4(1):18-22. .  180 24. Graafmans W, Ooms M, Hofstee H, Bezemer P, Bouter L, Lips P. Falls in the elderly: A prospective study of risk factors and risk profiles. - Am J Epidemiol.1996 Jun 1;143(11):1129-36.(0002-9262 (Print); 0002-9262 (Linking)). 25. Leipzig R, Cumming R, Tinetti M. Drugs and falls in older people: A systematic review and meta-analysis: I. psychotropic drugs. - J Am Geriatr Soc.1999 Jan;47(1):30-9.(0002-8614 (Print); 0002-8614 (Linking)). 26. Leipzig R, Cumming R, Tinetti M. Drugs and falls in older people: A systematic review and meta-analysis: II. cardiac and analgesic drugs. - J Am Geriatr Soc.1999 Jan;47(1):40-50.(0002-8614 (Print); 0002-8614 (Linking)). 27. Close J, Hooper R, Glucksman E, Jackson S, Swift C. Predictors of falls in a high risk population: Results from the prevention of falls in the elderly trial (PROFET). - Emerg Med J.2003 Sep;20(5):421-5.(1472-0213 (Electronic)). 28. Scott V, Wagar L, Elliott S. Falls & related injuries among older Canadians: Fall‐related hospitalizations & intervention initiatives. Victoria BC: Prepared on behalf of the Public Health Agency of Canada, Division of Aging and Seniors.; 2010. 29. Caplan G, Brown A, Croker W, Doolan J. Risk of admission within 4 weeks of discharge of elderly patients from the emergency department--the DEED study. discharge of elderly from emergency department. - Age Ageing.1998 Nov;27(6):697- 702.(0002-0729 (Print); 0002-0729 (Linking)). 30. Caplan G, Williams A, Daly B, Abraham K. A randomized, controlled trial of comprehensive geriatric assessment and multidisciplinary intervention after discharge of elderly from the emergency department--the DEED II study. - J Am Geriatr Soc.2004 Sep;52(9):1417-23.(0002-8614 (Print)). 31. Baraff L, Della Penna R, Williams N, Sanders A. Practice guideline for the ED management of falls in community-dwelling elderly persons. Kaiser Permanente medical group. - Ann Emerg Med.1997 Oct;30(4):480-92.(0196-0644 (Print)). 32. Canadian Association of Emergency Physicians. Position statement on emergency department overcrowding. Canadian Association of Emergency Physicians; 2007. 33. Canadian Association of Emergency Physicians. Taking action on the issue of overcrowding in Canada's emergency department. 2005. 34. Patrick J, Puterman M. Reducing wait times through operations research: Optimizing the use of surge capacity. - Healthc Q.2008;11(3):77-83.(1710-2774 (Print); 1710-2774 (Linking)).  181 35. Lamb S, Jorstad-Stein E, Hauer K, Becker C. Development of a common outcome data set for fall injury prevention trials: The prevention of falls network Europe consensus. - J Am Geriatr Soc.2005 Sep;53(9):1618-22.(0002-8614 (Print)). 36. Zecevic A, Salmoni A, Speechley M, Vandervoort A. Defining a fall and reasons for falling: Comparisons among the views of seniors, health care providers, and the research literature. - Gerontologist.2006 Jun;46(3):367-76.(0016-9013 (Print); 0016- 9013 (Linking)). 37. Zecevic AA, Salmoni AW, Lewko JH, Vandervoort AA, Speechley M. Utilization of the seniors falls investigation methodology to identify system-wide causes of falls in community-dwelling seniors. The Gerontologist. 2009 October 01;49(5):685-96. 38. Hauer K, Lamb S, Jorstad E, Todd C, Becker C. Systematic review of definitions and methods of measuring falls in randomised controlled fall prevention trials. - Age Ageing.2006 Jan;35(1):5-10.(0002-0729 (Print); 0002-0729 (Linking)). 39. The prevention of falls in later life. A report of the Kellogg international work group on the prevention of falls by the elderly. - Dan Med Bull.1987 Apr;34 Suppl 4:1-24.(0907-8916 (Print); 0907-8916 (Linking)). 40. Luukinen H, Koski K, Laippala P, Kivela S. Factors predicting fractures during falling impacts among home-dwelling older adults. - J Am Geriatr Soc.1997 Nov;45(11):1302-9.(0002-8614 (Print); 0002-8614 (Linking)). 41. Resnick B. Falls in a community of older adults: Putting research into practice. - Clin Nurs Res.1999 Aug;8(3):251-66.(1054-7738 (Print); 1054-7738 (Linking)). 42. van Doorn C, Gruber-Baldini A, Zimmerman S, Hebel J, Port C, Baumgarten M, et al. Dementia as a risk factor for falls and fall injuries among nursing home residents. - J Am Geriatr Soc.2003 Sep;51(9):1213-8.(0002-8614 (Print); 0002-8614 (Linking)). 43. Sadigh S, Reimers A, Andersson R, Laflamme L. Falls and fall-related injuries among the elderly: A survey of residential-care facilities in a Swedish municipality. - J Community Health.2004 Apr;29(2):129-40.(0094-5145 (Print); 0094-5145 (Linking)). 44. Yasumura S, Haga H, Niino N. Circumstances of injurious falls leading to medical care among elderly people living in a rural community. - Arch Gerontol Geriatr.1996 Sep-Oct;23(2):95-109.(0167-4943 (Print); 0167-4943 (Linking)). 45. Schwartz A, Villa M, Prill M, Kelsey J, Galinus J, Delay R, et al. Falls in older Mexican-American women. - J Am Geriatr Soc.1999 Nov;47(11):1371-8.(0002-8614 (Print); 0002-8614 (Linking)).  182 46. Richardson JK. Factors associated with falls in older patients with diffuse polyneuropathy. - J Am Geriatr Soc.2002 Nov;50(11):1767-73.(0002-8614 (Print); 0002-8614 (Linking)). 47. Lehtola S, Koistinen P, Luukinen H. Falls and injurious falls late in home-dwelling life. - Arch Gerontol Geriatr.2006 Mar-Apr;42(2):217-24.Epub 2005 Aug 25.(0167- 4943 (Print); 0167-4943 (Linking)). 48. Lin M, Hwang H, Wang Y, Chang S, Wolf S. Community-based tai chi and its effect on injurious falls, balance, gait, and fear of falling in older people. - Phys Ther.2006 Sep;86(9):1189-201.(0031-9023 (Print); 0031-9023 (Linking)). 49. Findorff M, Wyman J, Nyman J, Croghan C. Measuring the direct healthcare costs of a fall injury event. - Nurs Res.2007 Jul-Aug;56(4):283-7.(0029-6562 (Print)). 50. Iinattiniemi S, Jokelainen J, Luukinen H. Exercise and risk of injurious fall in home-dwelling elderly. - Int J Circumpolar Health.2008 Jun;67(2-3):235-44.(1239- 9736 (Print); 1239-9736 (Linking)). 51. Shumway-Cook, Anne, Ciol, MA, Hoffman,J, Dudgeon, BJ, Yorkston, K, Chan,L. Falls in the Medicare population: Incidence, associated factors, and impact on health care. Physical Therapy. 2009;89(4):324-32. 52. Malmivaara A, Heliovaara M, Knekt P, Reunanen A, Aromaa A. Risk factors for injurious falls leading to hospitalization or death in a cohort of 19,500 adults. - Am J Epidemiol.1993 Sep 15;138(6):384-94.(0002-9262 (Print); 0002-9262 (Linking)). 53. Tiedemann A, Murray S, Munro B, Lord S. Hospital and non-hospital costs for fall-related injury in community-dwelling older people. - N S W Public Health Bull.2008 Sep-Oct;19(9-10):161-5.(1034-7674 (Print); 1034-7674 (Linking)). 54. Jensen J, Lundin-Olsson L, Nyberg L, Gustafson Y. Falls among frail older people in residential care. - Scand J Public Health.2002;30(1):54-61.(1403-4948 (Print); 1403-4948 (Linking)). 55. Kelly K, Pickett W, Yiannakoulias N, Rowe B, Schopflocher D, Svenson L, et al. Medication use and falls in community-dwelling older persons. - Age Ageing.2003 Sep;32(5):503-9.(0002-0729 (Print); 0002-0729 (Linking)). 56. Saari P, Heikkinen E, Sakari-Rantala R, Rantanen T. Fall-related injuries among initially 75- and 80-year old people during a 10-year follow-up. - Arch Gerontol Geriatr.2007 Sep-Oct;45(2):207-15.Epub 2006 Dec 20.(0167-4943 (Print); 0167- 4943 (Linking)).  183 57. Lord S, Ward J, Williams P, Anstey K. Physiological factors associated with falls in older community-dwelling women. - J Am Geriatr Soc.1994 Oct;42(10):1110- 7.(0002-8614 (Print); 0002-8614 (Linking)). 58. von Heideken Wagert P, Gustafson Y, Kallin K, Jensen J, Lundin-Olsson L. Falls in very old people: The population-based umea 85+ study in sweden. - Arch Gerontol Geriatr.2009 Nov-Dec;49(3):390-6.Epub 2009 Jan 13.(1872-6976 (Electronic); 0167-4943 (Linking)). 59. Delbaere K, Close J, Heim J, Sachdev P, Brodaty H, Slavin M, et al. A multifactorial approach to understanding fall risk in older people. - J Am Geriatr Soc.2010 Sep;58(9):1679-85.doi: 10.1111/j.1532-5415.2010.03017.x.(1532-5415 (Electronic); 0002-8614 (Linking)). 60. Leclerc B, Begin C, Cadieux E, Goulet L, Allaire J, Meloche J, et al. A classification and regression tree for predicting recurrent falling among community- dwelling seniors using home-care services. - Can J Public Health.2009 Jul- Aug;100(4):263-7.(0008-4263 (Print); 0008-4263 (Linking)). 61. Campbell A, Borrie M, Spears G. Risk factors for falls in a community-based prospective study of people 70 years and older. - J Gerontol.1989 Jul;44(4):M112- 7.(0022-1422 (Print)). 62. Carter S, Campbell E, Sanson-Fisher R, Redman S, Gillespie W. Environmental hazards in the homes of older people. - Age Ageing.1997 May;26(3):195-202.(0002- 0729 (Print); 0002-0729 (Linking)). 63. Connell B, Wolf S. Environmental and behavioral circumstances associated with falls at home among healthy elderly individuals. Atlanta FICSIT group. - Arch Phys Med Rehabil.1997 Feb;78(2):179-86.(0003-9993 (Print); 0003-9993 (Linking)). 64. Josephson K, Fabacher D, Rubenstein L. Home safety and fall prevention. - Clin Geriatr Med.1991 Nov;7(4):707-31.(0749-0690 (Print); 0749-0690 (Linking)). 65. Nikolaus T, Bach M. Preventing falls in community-dwelling frail older people using a home intervention team (HIT): Results from the randomized falls-HIT trial. - J Am Geriatr Soc.2003 Mar;51(3):300-5.(0002-8614 (Print); 0002-8614 (Linking)). 66. Lord S, Menz H, Sherrington C. Home environment risk factors for falls in older people and the efficacy of home modifications. - Age Ageing.2006 Sep;35 Suppl 2:ii55-ii59.(0002-0729 (Print); 0002-0729 (Linking)). 67. Cumming RG, Thomas M, Szonyi G, Salkeld G, O'Neill E, Westbury C, et al. Home visits by an occupational therapist for assessment and modification of environmental hazards: A randomized trial of falls prevention. J Am Geriatr Soc. 1999 Dec;47(12):1397-402.  184 68. Luukinen H, Koski K, Kivela S. The relationship between outdoor temperature and the frequency of falls among the elderly in Finland. - J Epidemiol Community Health.1996 Feb;50(1):107.(0143-005X (Print); 0143-005X (Linking)). 69. Campbell A, Spears G, Borrie M, Fitzgerald J. Falls, elderly women and the cold. - Gerontology.1988;34(4):205-8.(0304-324X (Print); 0304-324X (Linking)). 70. Roos N, Burchill C, Carriere K. Who are the high hospital users? A Canadian case study. J Health Serv Res Policy. 2003 Jan;8(1):5-10. 71. Glazier RH, Badley EM, Gilbert JE, Rothman L. The nature of increased hospital use in poor neighbourhoods: Findings from a Canadian inner city. Can J Public Health. 2000 Jul-Aug;91(4):268-73. 72. Todd C, Ballinger C, Whitehead S. Reviews of socieo-demographic factors related to falls and environmental interventions to prevent falls amongst older people living in the community. WHO [Internet]:June 3, 2011. Available from: http://www.who.int/ageing/projects/3.Environmental and socioeconomic risk factors on falls.pdf. 73. Deandrea S, Lucenteforte E, Bravi F, Foschi R, La Vecchia C, Negri E. Risk factors for falls in community-dwelling older people: A systematic review and meta- analysis. - Epidemiology.2010 Sep;21(5):658-68.(1531-5487 (Electronic); 1044-3983 (Linking)). 74. Markle-Reid M, Browne G, Gafni A, Roberts J, Weir R, Thabane L, et al. A cross-sectional study of the prevalence, correlates, and costs of falls in older home care clients 'at risk' for falling. - Can J Aging.2010 Mar;29(1):119-37.(0714-9808 (Print); 0714-9808 (Linking)). 75. Campbell A, Borrie M, Spears G, Jackson S, Brown J, Fitzgerald J. Circumstances and consequences of falls experienced by a community population 70 years and over during a prospective study. - Age Ageing.1990 Mar;19(2):136- 41.(0002-0729 (Print); 0002-0729 (Linking)). 76. Roig M, Eng J, MacIntyre D, Road J, FitzGerald J, Burns J, et al. Falls in people with chronic obstructive pulmonary disease: An observational cohort study. - Respir Med.2011 Mar;105(3):461-9.(1532-3064 (Electronic); 0954-6111 (Linking)). 77. Ryynanen O, Kivela S, Honkanen R, Laippala P, Saano V. Medications and chronic diseases as risk factors for falling injuries in the elderly. - Scand J Soc Med.1993 Dec;21(4):264-71.(0300-8037 (Print); 0300-8037 (Linking)). 78. Griffith L, Raina P, Wu H, Zhu B, Stathokostas L. Population attributable risk for functional disability associated with chronic conditions in canadian older adults. - Age Ageing.2010 Sep 1.(1468-2834 (Electronic); 0002-0729 (Linking)).  185 79. Fried LP, Bandeen-Roche K, Kasper JD, Guralnik JM. Association of comorbidity with disability in older women: The Women’s health and aging study. J Clin Epidemiol. 1999 1;52(1):27-37. 80. Woolcott J, Ashe M, Miller W, Shi P, Marra C. Does physical activity reduce seniors' need for healthcare?: A study of 24 281 Canadians. - Br J Sports Med.2010 Sep;44(12):902-4.Epub 2009 Oct 23.(1473-0480 (Electronic); 0306-3674 (Linking)). 81. Sawatzky R, Liu-Ambrose T, Miller W, Marra C. Physical activity as a mediator of the impact of chronic conditions on quality of life in older adults. - Health Qual Life Outcomes.2007 Dec 19;5:68.(1477-7525 (Electronic); 1477-7525 (Linking)). 82. Tinetti M, Powell L. Fear of falling and low self-efficacy: A case of dependence in elderly persons. - J Gerontol.1993 Sep;48 Spec No:35-8.(0022-1422 (Print); 0022- 1422 (Linking)). 83. Summary of the updated American Geriatrics Society/British Geriatrics Society clinical practice guideline for prevention of falls in older persons. - J Am Geriatr Soc.2011 Jan;59(1):148-57.doi: 10.1111/j.1532-5415.2010.03234.x.(1532-5415 (Electronic); 0002-8614 (Linking)). 84. Morgan S, Raymond C, Mooney D, Martin D. The Canadian RX atlas 2nd edition (the Canadian prescription drug atlas). UBC centre for health services and policy research. 2008. 85. Ramage-Morin PL. Medication use among senior Canadians. - Health Rep.2009 Mar;20(1):37-44.(0840-6529 (Print); 0840-6529 (Linking)). 86. Boyle N, Naganathan V, Cumming RG. Medication and falls: Risk and optimization. Clin Geriatr Med. 2010 Nov;26(4):583-605. 87. Weiner D, Hanlon J, Studenski S. Effects of central nervous system polypharmacy on falls liability in community-dwelling elderly. - Gerontology.1998;44(4):217-21.(0304-324X (Print); 0304-324X (Linking)). 88. Baranzini F, Diurni M, Ceccon F, Poloni N, Cazzamalli S, Costantini C, et al. Fall-related injuries in a nursing home setting: Is polypharmacy a risk factor? - BMC Health Serv Res.2009 Dec 11;9:228.(1472-6963 (Electronic); 1472-6963 (Linking)). 89. Hanlon J, Landerman L, Fillenbaum G, Studenski S. Falls in African American and white community-dwelling elderly residents. - J Gerontol A Biol Sci Med Sci.2002 Jul;57(7):M473-8.(1079-5006 (Print); 1079-5006 (Linking)). 90. Tromp A, Pluijm S, Smit J, Deeg D, Bouter L, Lips P. Fall-risk screening test: A prospective study on predictors for falls in community-dwelling elderly. - J Clin Epidemiol.2001 Aug;54(8):837-44.(0895-4356 (Print); 0895-4356 (Linking)).  186 91. Tromp A, . Smit J, Deeg D, Bouter L, Lips P. Predictors for falls and fractures in the longitudinal aging study Amsterdam. - J Bone Miner Res.1998 Dec;13(12):1932- 9.(0884-0431 (Print); 0884-0431 (Linking)). 92. Gerdhem P, Ringsberg K, Akesson K, Obrant K. Clinical history and biologic age predicted falls better than objective functional tests. - J Clin Epidemiol.2005 Mar;58(3):226-32.(0895-4356 (Print); 0895-4356 (Linking)). 93. Clough-Gorr K, Erpen T, Gillmann G, von Renteln-Kruse W, Iliffe S, Beck J, et al. Preclinical disability as a risk factor for falls in community-dwelling older adults. - J Gerontol A Biol Sci Med Sci.2008 Mar;63(3):314-20.(1079-5006 (Print); 1079-5006 (Linking)). 94. Gassmann K, Rupprecht R, Freiberger E. Predictors for occasional and recurrent falls in community-dwelling older people. - Z Gerontol Geriatr.2009 Feb;42(1):3- 10.Epub 2009 Apr 10.(1435-1269 (Electronic); 0948-6704 (Linking)). 95. Fisher A, McLean A, Davis M, Le Couteur D. A multicenter, case-control study of the effects of antihypertensive therapy on orthostatic hypotension, postprandial hypotension, and falls in octo- and nonagenarians in residential care facilities. curr therap res. 2003; 64(3):206-14. . 96. Neutel C, Perry S, Maxwell C. Medication use and risk of falls. - Pharmacoepidemiol Drug Saf.2002 Mar;11(2):97-104.(1053-8569 (Print); 1053-8569 (Linking)). 97. Maurer M, Burcham J, Cheng H. Diabetes mellitus is associated with an increased risk of falls in elderly residents of a long-term care facility. - J Gerontol A Biol Sci Med Sci.2005 Sep;60(9):1157-62.(1079-5006 (Print); 1079-5006 (Linking)). 98. Hartikainen S, Lonnroos E, Louhivuori K. Medication as a risk factor for falls: Critical systematic review. - J Gerontol A Biol Sci Med Sci.2007 Oct;62(10):1172- 81.(1079-5006 (Print); 1079-5006 (Linking)). 99. Thapa PB, Gideon P, Brockman KG, Fought RL, Ray WA. Clinical and biomechanical measures of balance as fall predictors in ambulatory nursing home residents. J Gerontol A Biol Sci Med Sci. 1996 Sep;51(5):M239-46. 100. Walker P, Alrawi A, Mitchell J, Regal R, Khanderia U. Medication use as a risk factor for falls among hospitalized elderly patients. - Am J Health Syst Pharm.2005 Dec 1;62(23):2495-9.(1079-2082 (Print); 1079-2082 (Linking)). 101. Desmet C, Beguin C, Swine C, Jadoul M, Universite Catholique de Louvain Collaborative Group. Falls in hemodialysis patients: Prospective study of incidence, risk factors, and complications. Am J Kidney Dis. 2005 Jan;45(1):148-53.  187 102. Ebly E, Hogan D, Fung T. Potential adverse outcomes of psychotropic and narcotic drug use in canadian seniors. - J Clin Epidemiol.1997 Jul;50(7):857- 63.(0895-4356 (Print); 0895-4356 (Linking)). 103. Ensrud K, Blackwell T, Mangione C, Bowman P, Whooley M, Bauer D, et al. Central nervous system-active medications and risk for falls in older women. - J Am Geriatr Soc.2002 Oct;50(10):1629-37.(0002-8614 (Print); 0002-8614 (Linking)). 104. Gluck T, Wientjes H, Rai G. An evaluation of risk factors for in-patient falls in acute and rehabilitation elderly care wards. - Gerontology.1996;42(2):104-7.(0304- 324X (Print); 0304-324X (Linking)). 105. Heitterachi E, Lord S, Meyerkort P, McCloskey I, Fitzpatrick R. Blood pressure changes on upright tilting predict falls in older people. - Age Ageing.2002 May;31(3):181-6.(0002-0729 (Print); 0002-0729 (Linking)). 106. Hien le T, Cumming R, Cameron I, Chen J, Lord S, March L, et al. Atypical antipsychotic medications and risk of falls in residents of aged care facilities. - J Am Geriatr Soc.2005 Aug;53(8):1290-5.(0002-8614 (Print); 0002-8614 (Linking)). 107. Kallin K, Lundin-Olsson L, Jensen J, Nyberg L, Gustafson Y. Predisposing and precipitating factors for falls among older people in residential care. Public Health. 2002 Sep;116(5):263-71. 108. Kallin K, Gustafson Y, Sandman P, Karlsson S. Drugs and falls in older people in geriatric care settings. - Aging Clin Exp Res.2004 Aug;16(4):270-6.(1594-0667 (Print); 1594-0667 (Linking)). 109. Landi F, Onder G, Cesari M, Barillaro C, Russo A, Bernabei R. Psychotropic medications and risk for falls among community-dwelling frail older people: An observational study. - J Gerontol A Biol Sci Med Sci.2005 May;60(5):622-6.(1079- 5006 (Print); 1079-5006 (Linking)). 110. Lawlor D, Patel R, Ebrahim S. Association between falls in elderly women and chronic diseases and drug use: Cross sectional study. - BMJ.2003 Sep 27;327(7417):712-7.(1468-5833 (Electronic); 0959-535X (Linking)). 111. Mustard CA, Mayer T. Case-control study of exposure to medication and the risk of injurious falls requiring hospitalization among nursing home residents. Am J Epidemiol. 1997 Apr 15;145(8):738-45. 112. Souchet E, Lapeyre-Mestre M, Montastruc JL. Drug related falls: A study in the French pharmacovigilance database. Pharmacoepidemiol Drug Saf. 2005 Jan;14(1):11-6.  188 113. Arfken C, Wilson J, Aronson S. Retrospective review of selective serotonin reuptake inhibitors and falling in older nursing home residents. - Int Psychogeriatr.2001 Mar;13(1):85-91.(1041-6102 (Print); 1041-6102 (Linking)). 114. Lee JS, Hurley MJ, Carew D, Fisher R, Kiss A, Drummond N. A randomized clinical trial to assess the impact on an emergency response system on anxiety and health care use among older emergency patients after a fall. Acad Emerg Med. 2007 Apr;14(4):301-8. 115. Nygaard HA. Falls and psychotropic drug consumption in long-term care residents: Is there an obvious association? Gerontology. 1998;44(1):46-50. 116. Rozenfeld S, Camacho L, Veras P. Medication as a risk factor for falls in older women in brazil. - Rev Panam Salud Publica.2003 Jun;13(6):369-75.(1020-4989 (Print); 1020-4989 (Linking)). 117. Avidan A, Fries B, James M, Szafara K, Wright G, Chervin R. Insomnia and hypnotic use, recorded in the minimum data set, as predictors of falls and hip fractures in Michigan nursing homes. - J Am Geriatr Soc.2005 Jun;53(6):955- 62.(0002-8614 (Print); 0002-8614 (Linking)). 118. Chu L, Chi I, Chui A. Incidence and predictors of falls in the Chinese elderly. ann acad med singapore. 2005; 34(1): 60-72. . 119. Lord S, March L, Cameron I, Cumming R, Schwarz J, Zochling J, et al. Differing risk factors for falls in nursing home and intermediate-care residents who can and cannot stand unaided. - J Am Geriatr Soc.2003 Nov;51(11):1645-50.(0002-8614 (Print); 0002-8614 (Linking)). 120. de Rekeneire N, Visser M, Peila R, Nevitt M, Cauley J, Tylavsky F, et al. Is a fall just a fall: Correlates of falling in healthy older persons. the health, aging and body composition study. - J Am Geriatr Soc.2003 Jun;51(6):841-6.(0002-8614 (Print); 0002-8614 (Linking)). 121. Frels C, Williams P, Narayanan S, Gariballa S. Iatrogenic causes of falls in hospitalised elderly patients: A case-control study. - Postgrad Med J.2002 Aug;78(922):487-9.(0032-5473 (Print); 0032-5473 (Linking)). 122. Passaro A, Volpato S, Romagnoni F, Manzoli N, Zuliani G, Fellin R. Benzodiazepines with different half-life and falling in a hospitalized population: The GIFA study. gruppo italiano di farmacovigilanza nell'anziano. - J Clin Epidemiol.2000 Dec;53(12):1222-9.(0895-4356 (Print); 0895-4356 (Linking)). 123. Ray W, Thapa P, Gideon P. Benzodiazepines and the risk of falls in nursing home residents. - J Am Geriatr Soc.2000 Jun;48(6):682-5.(0002-8614 (Print); 0002- 8614 (Linking)).  189 124. Lee J, Kwok T, Leung P, Woo J. Medical illnesses are more important than medications as risk factors of falls in older community dwellers: A cross sectional study. age ageing. 2006 35(3):246-51. . 125. Sieri T, Beretta G. Fall risk assessment in very old males and females living in nursing homes. Disabil Rehabil. 2004 Jun 17;26(12):718-23. 126. Horikawa E, Matsui T, Arai H, Seki T, Iwasaki K, Sasaki H. Risk of falls in Alzheimer's Disease: A prospective study. Intern Med. 2005 Jul;44(7):717-21. 127. Campbell A, Robertson M, Gardner M, Norton R, Buchner D. Psychotropic medication withdrawal and a home-based exercise program to prevent falls: A randomized, controlled trial. - J Am Geriatr Soc.1999 Jul;47(7):850-3.(0002-8614 (Print); 0002-8614 (Linking)). 128. Murad MH, Elamin KB, Abu Elnour NO, Elamin MB, Alkatib AA, Fatourechi MM, et al. The effect of vitamin D on falls: A systematic review and meta-analysis. J Clin Endocrinol Metab. 2011 Jul 27. 129. Flicker L, Mead K, MacInnis R, Nowson C, Scherer S, Stein M, et al. Serum vitamin D and falls in older women in residential care in Australia. - J Am Geriatr Soc.2003 Nov;51(11):1533-8.(0002-8614 (Print); 0002-8614 (Linking)). 130. Mahoney F, Barthel D. Functional evaluation: The Barthel Index. - Md State Med J.1965 Feb;14:61-5.(0025-4363 (Print); 0025-4363 (Linking)). 131. Samaras N, Chevalley T, Samaras D, Gold G. Older patients in the emergency department: A review. - Ann Emerg Med.2010 Sep;56(3):261-9.(1097-6760 (Electronic); 0196-0644 (Linking)). 132. Aminzadeh F, Dalziel W. Older adults in the emergency department: A systematic review of patterns of use, adverse outcomes, and effectiveness of interventions. - Ann Emerg Med.2002 Mar;39(3):238-47.(0196-0644 (Print); 0196- 0644 (Linking)). 133. Royal College of Physicians and Surgeons of Canada. Objective of training and specialty training requirements: Emergency medicine. Ottawa. . 2003. 134. Lowenstein S, Crescenzi C, Kern D, Steel K. Care of the elderly in the emergency department. - Ann Emerg Med.1986 May;15(5):528-35.(0196-0644 (Print); 0196-0644 (Linking)). 135. Wofford J, Schwartz E, Byrum J. The role of emergency services in health care for the elderly: A review. - J Emerg Med.1993 May-Jun;11(3):317-26.(0736-4679 (Print); 0736-4679 (Linking)).  190 136. Hastings S, Heflin M. A systematic review of interventions to improve outcomes for elders discharged from the emergency department. - Acad Emerg Med.2005 Oct;12(10):978-86.(1553-2712 (Electronic); 1069-6563 (Linking)). 137. McCusker J, Verdon J. Do geriatric interventions reduce emergency department visits? A systematic review. - J Gerontol A Biol Sci Med Sci.2006 Jan;61(1):53-62.(1079-5006 (Print); 1079-5006 (Linking)). 138. McCusker J, Dendukuri N, Tousignant P, Verdon J, Poulin de Courval L, Belzile E. Rapid two-stage emergency department intervention for seniors: Impact on continuity of care. - Acad Emerg Med.2003 Mar;10(3):233-43.(1069-6563 (Print); 1069-6563 (Linking)). 139. Mion L, Palmer R, Meldon S, Bass D, Singer M, Payne S, et al. Case finding and referral model for emergency department elders: A randomized clinical trial. - Ann Emerg Med.2003 Jan;41(1):57-68.(0196-0644 (Print); 0196-0644 (Linking)). 140. Guttman A, Afilalo M, Guttman R, Colacone A, Robitaille C, Lang E, et al. An emergency department-based nurse discharge coordinator for elder patients: Does it make a difference? - Acad Emerg Med.2004 Dec;11(12):1318-27.(1069-6563 (Print); 1069-6563 (Linking)). 141. Dellasega C, Zerbe T. A multimethod study of advanced practice nurse postdischarge care. - Clin Excell Nurse Pract.2000 Sep;4(5):286-93.(1085-2360 (Print); 1085-2360 (Linking)). 142. Miller D, Lewis L, Nork M, Morley J. Controlled trial of a geriatric case-finding and liaison service in an emergency department. - J Am Geriatr Soc.1996 May;44(5):513-20.(0002-8614 (Print); 0002-8614 (Linking)). 143. Naylor M, Brooten D, Campbell R, Jacobsen B, Mezey M, Pauly M, et al. Comprehensive discharge planning and home follow-up of hospitalized elders: A randomized clinical trial. - JAMA.1999 Feb 17;281(7):613-20.(0098-7484 (Print); 0098-7484 (Linking)). 144. Gagnon A, Schein C, McVey L, Bergman H. Randomized controlled trial of nurse case management of frail older people. - J Am Geriatr Soc.1999 Sep;47(9):1118-24.(0002-8614 (Print); 0002-8614 (Linking)). 145. Brooks M, Ertl J. Social work home visits: Impact on recidivism and health care costs. - Continuum Soc Soc Work Leadersh Health Care.2000 Nov-Dec;20(6):3-9. 146. Koehler B, Richter K, Youngblood L, Cohen B, Prengler I, Cheng D, et al. Reduction of 30-day postdischarge hospital readmission or emergency department (ED) visit rates in high-risk elderly medical patients through delivery of a targeted  191 care bundle. - J Hosp Med.2009 Apr;4(4):211-8.(1553-5606 (Electronic); 1553-5592 (Linking)). 147. Kalula S, de Villiers L, Ross K, Ferreira M. Management of older patients presenting after a fall--an accident and emergency department audit. - S Afr Med J.2006 Aug;96(8):718-21.(0256-9574 (Print); 0256-9574 (Linking)). 148. Paniagua M, Malphurs J, Phelan E. Older patients presenting to a county hospital ED after a fall: Missed opportunities for prevention. - Am J Emerg Med.2006 Jul;24(4):413-7.(0735-6757 (Print); 0735-6757 (Linking)). 149. Davison J, Bond J, Dawson P, Steen I, Kenny R. Patients with recurrent falls attending accident & emergency benefit from multifactorial intervention--a randomised controlled trial. - Age Ageing.2005 Mar;34(2):162-8.(0002-0729 (Print); 0002-0729 (Linking)). 150. Hill K, Womer M, Russell M, Blackberry I, McGann A. Fear of falling in older fallers presenting at emergency departments. - J Adv Nurs.2010 Aug;66(8):1769- 79.Epub 2010 Jun 16.(1365-2648 (Electronic); 0309-2402 (Linking)). 151. Donaldson M, Khan K, Davis J, Salter A, Buchanan J, McKnight D, et al. Emergency department fall-related presentations do not trigger fall risk assessment: A gap in care of high-risk outpatient fallers. - Arch Gerontol Geriatr.2005 Nov- Dec;41(3):311-7.Epub 2005 Jun 27.(0167-4943 (Print); 0167-4943 (Linking)). 152. Salter A, Khan K, Donaldson M, Davis J, Buchanan J, Abu-Laban R, et al. Community-dwelling seniors who present to the emergency department with a fall do not receive guideline care and their fall risk profile worsens significantly: A 6-month prospective study. - Osteoporos Int.2006;17(5):672-83.Epub 2006 Feb 21.(0937- 941X (Print); 0937-941X (Linking)). 153. Lord S, Menz H, Tiedemann A. A physiological profile approach to falls risk assessment and prevention. - Phys Ther.2003 Mar;83(3):237-52.(0031-9023 (Print); 0031-9023 (Linking)). 154. Snooks H, Halter M, Close J, Cheung W, Moore F, Roberts S. Emergency care of older people who fall: A missed opportunity. - Qual Saf Health Care.2006 Dec;15(6):390-2.(1475-3901 (Electronic)). 155. Gates S, Fisher J, Cooke M, Carter Y, Lamb S. Multifactorial assessment and targeted intervention for preventing falls and injuries among older people in community and emergency care settings: Systematic review and meta-analysis. - BMJ.2008 Jan 19;336(7636):130-3.Epub 2007 Dec 18.(1468-5833 (Electronic); 0959-535X (Linking)).  192 156. Campbell A, Robertson M. Rethinking individual and community fall prevention strategies: A meta-regression comparing single and multifactorial interventions. - Age Ageing.2007 Nov;36(6):656-62.(1468-2834 (Electronic); 0002-0729 (Linking)). 157. Rice DP. Cost of illness studies: What is good about them? - Inj Prev.2000 Sep;6(3):177-9.(1353-8047 (Print); 1353-8047 (Linking)). 158. Drummond, MF, Sculpher, MJ, Torrance, GW, O'Brien, BJ, Stoddart,GL. Methods for the economic evaluation of health care programmes. third edition. Toronto: Oxford University Press; 2005. 159. Martin I M. Introduction to health economics for physicians. The Lancet. 2001 9/22;358(9286):993-8. 160. Goossens MEJB, Mölken MPMHR, Vlaeyen JWS, van der Linden SMJP. The cost diary: A method to measure direct and indirect costs in cost-effectiveness research. J Clin Epidemiol. 2000 7;53(7):688-95. 161. Rascati K. Essentials of pharmacoeconomics. Baltimore: Lippincott WIlliams & Wilkins; 2009. 162. Baladi J. A guidance document for the costing process. version 1.0. . 1996. 163. Koopmanschap MA, Rutten FF, van Ineveld BM, van Roijen L. The friction cost method for measuring indirect costs of disease. J Health Econ. 1995 Jun;14(2):171- 89. 164. Goeree R, O'Brien BJ, Blackhouse G, Agro K, Goering P. The valuation of productivity costs due to premature mortality: A comparison of the human-capital and friction-cost methods for schizophrenia. Can J Psychiatry. 1999 Jun;44(5):455- 63. 165. Xie F, Thumboo J, Fong KY, Lo NN, Yeo SJ, Yang KY, et al. A study on indirect and intangible costs for patients with knee osteoarthritis in Singapore. Value Health. 2008 Mar;11 Suppl 1:S84-90. 166. Philips, Z, Ginnelly, L, Sculpher, MJ, Claxton, K, Golder, S, Riemsma, R, Wollacott, N, Glanville,J. Review of guidelines for good practice in decision-analytic modelling in health technology assessment. Health Technol Assess. 2004;8(36):1,158; 1. 167. Canadian Coordinating Office for Health Technology Assessment. Guidelines for economic evaluation of pharmaceuticals: Canada. 2nd ed. ; 1997. 168. Iezzoni LI. Assessing quality using administrative data. - Ann Intern Med.1997 Oct 15;127(8 Pt 2):666-74.(0003-4819 (Print)).  193 169. Preen D, Holman C, Lawrence D, Baynham N, Semmens J. Hospital chart review provided more accurate comorbidity information than data from a general practitioner survey or an administrative database. - J Clin Epidemiol.2004 Dec;57(12):1295-304.(0895-4356 (Print)). 170. Finkler S. The distinction between costs and charges. Ann Intern Med. 1982;96(1):102,109; 102. 171. SMARTRISK. (2009). The Economic Burden of Injury in Canada. SMARTRISK: Toronto, ON. 172. Wiktorowicz M, Goeree R, Papaioannou A, Adachi J, Papadimitropoulos E. Economic implications of hip fracture: Health service use, institutional care and cost in Canada. - Osteoporos Int.2001;12(4):271-8.(0937-941X (Print)). 173. Saunders CE, Makens PK, Leblanc LJ. Modeling emergency department operations using advanced computer simulation systems. Ann Emerg Med. 1989 Feb;18(2):134-40. 174. Connelly L, Bair A. Discrete event simulation of emergency department activity: A platform for system-level operations research. - Acad Emerg Med.2004 Nov;11(11):1177-85.(1069-6563 (Print); 1069-6563 (Linking)). 175. Odegaard F, Chen L, Quee R, Puterman M. Improving the efficiency of hospital porter services, part 1: Study objectives and results. - J Healthc Qual.2007 Jan- Feb;29(1):4-11.(1062-2551 (Print); 1062-2551 (Linking)). 176. Odegaard F, Chen L, Quee R, Puterman M. Improving the efficiency of hospital porter services, part 2: Schedule optimization and simulation model. - J Healthc Qual.2007 Jan-Feb;29(1):12-8.(1062-2551 (Print); 1062-2551 (Linking)). 177. Coats T, Michalis S. Mathematical modelling of patients flow through an accident and emergency department. - Emerg Med J.2001 May;18(3):190-2.(1472- 0205 (Print); 1472-0205 (Linking)). 178. Sutton A, Abrams K. Bayesian methods in meta-analysis and evidence synthesis. - Stat Methods Med Res.2001 Aug;10(4):277-303.(0962-2802 (Print); 0962-2802 (Linking)). 179. Murray M, Bullard M, Grafstein E, CTAS National Working Group, CEDIS National Working Group. Revisions to the Canadian emergency department triage and acuity scale implementation guidelines. CJEM. 2004 Nov;6(6):421-7. 180. Belanger A, Martel L, Caron-Malenfant E. Catalogue no.91-52--XIE population projections for Canada, provinces and territories. [Internet]. 2005 December 2005:July 9, 2010.  194 181. Asplin B, Magid DJ, Rhodes K, Solberg L, Lurie N, Camargo C. A conceptual model of emergency department crowding. - Ann Emerg Med.2003 Aug;42(2):173- 80.(0196-0644 (Print); 0196-0644 (Linking)). 182. Koski K, Luukinen H, Laippala P, Kivela S. Risk factors for major injurious falls among the home-dwelling elderly by functional abilities. A prospective population- based study. - Gerontology.1998;44(4):232-8.(0304-324X (Print); 0304-324X (Linking)). 183. Rizzo J, Friedkin R, Williams C, Nabors J, Acampora D, Tinetti M. Health care utilization and costs in a Medicare population by fall status. - Med Care.1998 Aug;36(8):1174-88.(0025-7079 (Print); 0025-7079 (Linking)). 184. Robertson M, Campbell A, Gardner M, Devlin N. Preventing injuries in older people by preventing falls: A meta-analysis of individual-level data. - J Am Geriatr Soc.2002 May;50(5):905-11.(0002-8614 (Print); 0002-8614 (Linking)). 185. Fitzharris M, Day L, Lord S, Gordon I, Fildes B. The whitehorse NoFalls trial: Effects on fall rates and injurious fall rates. - Age Ageing.2010 Nov;39(6):728- 33.Epub 2010 Sep 4.(1468-2834 (Electronic); 0002-0729 (Linking)). 186. Gill D, Zou G, Jones G, Speechley M. Injurious falls are associated with lower household but higher recreational physical activities in community-dwelling older male veterans. - Gerontology.2008;54(2):106-15.Epub 2008 Feb 6.(1423-0003 (Electronic); 0304-324X (Linking)). 187. Duh M, Mody S, Lefebvre P, Woodman R, Buteau S, Piech C. Anaemia and the risk of injurious falls in a community-dwelling elderly population. - Drugs Aging.2008;25(4):325-34.(1170-229X (Print); 1170-229X (Linking)). 188. Szabo S, Janssen P, Khan K, Lord S, Potter M. Neovascular AMD: An overlooked risk factor for injurious falls. - Osteoporos Int.2010 May;21(5):855- 62.Epub 2009 Jul 23.(1433-2965 (Electronic); 0937-941X (Linking)). 189. Pariente A, Dartigues J, Benichou J, Letenneur L, Moore N, Fourrier-Reglat A. Benzodiazepines and injurious falls in community dwelling elders. - Drugs Aging.2008;25(1):61-70.(1170-229X (Print); 1170-229X (Linking)). 190. Nachreiner N, Findorff M, Wyman J, McCarthy T. Circumstances and consequences of falls in community-dwelling older women. - J Womens Health (Larchmt).2007 Dec;16(10):1437-46.(1540-9996 (Print); 1540-9996 (Linking)). 191. Scuffham P, Chaplin S, Legood R. Incidence and costs of unintentional falls in older people in the United Kingdom. - J Epidemiol Community Health.2003 Sep;57(9):740-4.(0143-005X (Print)).  195 192. Agostini JV, Han L, Tinetti ME. The relationship between number of medications and weight loss or impaired balance in older adults. - J Am Geriatr Soc.2004 Oct;52(10):1719-23.(0002-8614 (Print); 0002-8614 (Linking)). 193. Yip Y, Cumming R. The association between medications and falls in Australian nursing-home residents. - Med J Aust.1994 Jan 3;160(1):14-8.(0025-729X (Print); 0025-729X (Linking)). 194. Downs S, Black N. The feasibility of creating a checklist for the assessment of the methodological quality both of randomised and non-randomised studies of health care interventions. - J Epidemiol Community Health.1998 Jun;52(6):377-84.(0143- 005X (Print); 0143-005X (Linking)). 195. Spiegelhalter D, Thomas A, Best N, Gilks W. BUGS: Bayesian inference using Gibbs sampling version 0.50. MRC biostatistics unit: Cambridge, 1995. . 196. Psaty B, Koepsell T, Lin D, Weiss N, Siscovick D, Rosendaal F, et al. Assessment and control for confounding by indication in observational studies. - J Am Geriatr Soc.1999 Jun;47(6):749-54.(0002-8614 (Print); 0002-8614 (Linking)). 197. Sturmer T, Glynn R, Rothman K, Avorn J, Schneeweiss S. Adjustments for unmeasured confounders in pharmacoepidemiologic database studies using external information. - Med Care.2007 Oct;45(10 Supl 2):S158-65.(0025-7079 (Print); 0025-7079 (Linking)). 198. Sturmer T, Joshi M, Glynn R, Avorn J, Rothman K, Schneeweiss S. A review of the application of propensity score methods yielded increasing use, advantages in specific settings, but not substantially different estimates compared with conventional multivariable methods. - J Clin Epidemiol.2006 May;59(5):437-47.Epub 2005 Oct 13.(0895-4356 (Print); 0895-4356 (Linking)). 199. Vellas B, Wayne S, Garry P, Baumgartner R. A two-year longitudinal study of falls in 482 community-dwelling elderly adults. - J Gerontol A Biol Sci Med Sci.1998 Jul;53(4):M264-74.(1079-5006 (Print); 1079-5006 (Linking)). 200. Cooper N, Sutton A, Abrams K, Turner D, Wailoo A. Comprehensive decision analytical modelling in economic evaluation: A Bayesian approach. - Health Econ.2004 Mar;13(3):203-26.(1057-9230 (Print); 1057-9230 (Linking)). 201. Lefaivre K, Macadam S, Davidson D, Gandhi R, Chan H, Broekhuyse H. Length of stay, mortality, morbidity and delay to surgery in hip fractures. - J Bone Joint Surg Br.2009 Jul;91(7):922-7.(0301-620X (Print)). 202. Clague J, Craddock E, Andrew G, Horan M, Pendleton N. Predictors of outcome following hip fracture. admission time predicts length of stay and in-hospital mortality. - Injury.2002 Jan;33(1):1-6.(0020-1383 (Print); 0020-1383 (Linking)).  196 203. Li G, Lau J, McCarthy M, Schull M, Vermeulen M, Kelen G. Emergency department utilization in the United States and Ontario, Canada. - Acad Emerg Med.2007 Jun;14(6):582-4.Epub 2007 Apr 30.(1553-2712 (Electronic); 1069-6563 (Linking)). 204. Goodacre S, Webster A. Who waits longest in the emergency department and who leaves without being seen? - Emerg Med J.2005 Feb;22(2):93-6.(1472-0213 (Electronic); 1472-0205 (Linking)). 205. Gullberg B, Johnell O, Kanis J. World-wide projections for hip fracture. - Osteoporos Int.1997;7(5):407-13.(0937-941X (Print); 0937-941X (Linking)). 206. Davis J, Robertson M, Ashe M, Liu-Ambrose T, Khan K, Marra C. Does a home-based strength and balance programme in people aged > or =80 years provide the best value for money to prevent falls? A systematic review of economic evaluations of falls prevention interventions. - Br J Sports Med.2010 Feb;44(2):80- 9.(1473-0480 (Electronic); 0306-3674 (Linking)). 207. Woolcott J, Richardson K, Wiens M, Patel B, Marin J, Khan K, et al. Meta- analysis of the impact of 9 medication classes on falls in elderly persons. - Arch Intern Med.2009 Nov 23;169(21):1952-60.(1538-3679 (Electronic); 1538-3679 (Linking)). 208. Davis J, Guy P, Ashe M, Liu-Ambrose T, Khan K. HipWatch: Osteoporosis investigation and treatment after a hip fracture: A 6-month randomized controlled trial. - J Gerontol A Biol Sci Med Sci.2007 Aug;62(8):888-91.(1079-5006 (Print); 1079-5006 (Linking)). 209. Innes G, Stenstrom R, Grafstein E, Christenson J. Prospective time study derivation of emergency physician workload predictors. - CJEM.2005 Sep;7(5):299- 308.(1481-8035 (Print)). 210. Najafzadeh M, Marra C, Sadatsafavi M, Aaron S, Sullivan S, Vandemheen K, et al. Cost effectiveness of therapy with combinations of long acting bronchodilators and inhaled steroids for treatment of COPD. - Thorax.2008 Nov;63(11):962-7.Epub 2008 Jul 11.(1468-3296 (Electronic)). 211. Shalensky K, Hill S, editors. Formulary of Vancouver General Hospital , UBC Hospital, GF strong. ; 2009. 212. MSP-MSC payment schedule [Internet]. Victoria, British Columbia: British Columbia Ministry of Health Services; 2009 [updated April 1, 2009. Available from: http://www.health.gov.bc.ca/msp/infoprac/physbilling/payschedule/index.html.  197 213. Ambulance fee changes [Internet].; 2007 [updated September 12, 2007. Available from: http://www2.news.gov.bc.ca/news_releases_2005- 2009/2007HEALTH0101-001106-Attachment1.htm. 214. [Internet].; 2008 [updated march 31, 2008. Available from: http://www.hsabc.org/webuploads/files/member_services/collective_agreements/hsp /RatesOfPay.pdf. 215. Consumer price index, health and personal care, by province. [Internet].: Statistics Canada; 2010 [updated March 29, 2010. Available from: http://www40.statcan.gc.ca/l01/cst01/econ161a-eng.htm. 216. Alberta case cost report [Internet].; 2008 [updated December 2008. Available from: http://www.health.alberta.ca/documents/Case-Cost-Hospital-05-06.pdf. 217. Alberta wage and salary survey [Internet].; 2009. Available from: http://alis.alberta.ca/wageinfo/Content/RequestAction.asp?aspAction=GetWageDetai l&format=html&RegionID=20&NOC=3142. 218. Alberta health care insurance plan procedure list [Internet].; 2010 [updated January 1, 2010. Available from: http://www.health.alberta.ca/documents/SOMB- medical-procedures.pdf. 219. Kelton W, Sadowski R, Sturrock D. Simulation with arena, fourth edition. New York, New York: McGraw-Hill; 2007. 220. Hall S, Hendrie D. A prospective study of the costs of falls in older adults living in the community. - Aust N Z J Public Health.2003;27(3):343-51.(1326-0200 (Print)). 221. Finkelstein EA, Chen H, Miller TR, Corso PS, Stevens JA. A comparison of the case-control and case-crossover designs for estimating medical costs of nonfatal fall-related injuries among older Americans. Med Care. 2005 Nov;43(11):1087-91. 222. Englander F, Hodson TJ, Terregrossa RA. Economic dimensions of slip and fall injuries. J Forensic Sci. 1996 Sep;41(5):733-46. 223. Cotter PE, Timmons S, O'Connor M, Twomey C, O'Mahony D. The financial implications of falls in older people for an acute hospital. Ir J Med Sci. 2006 Apr- Jun;175(2):11-3.    198 Appendices Appendix A  Health care operations analysis to reduce attending times for seniors presenting to the Emergency Department with a fall: subject information and research project consent form  199      Health Care Operations Analysis to Reduce Attending Times for Seniors Presenting to the Emergency Department with a Fall Subject Information and Research Project Consent Form  Principal Investigator  Karim Khan, MD, PhD,    Department of Family Practice    University of British Columbia    (604) 827-4190  Co-Investigators: Bonnie Lillies-Vancouver Coastal Health Authority    Carlo Marra- Department of Pharmacy Practice, UBC    Boris Sobolev-Department of Healthcare and Epidemiology,    Dr Riyad Abu-Laban, Department of Medicine UBC  Sponsorship: Michael Smith Foundation for Health Services Research               200 Introduction  You have been invited to participate in this study because you are over 70 years of age and you presented to the emergency department as result of a fall. This study is looking at the treatment received by individuals over 70 years of age who come to the emergency department as result of a fall. We hope that our study will assist us in designing healthcare and improving the outcomes and treatment of those who fall and present to the emergency department. This study is being sponsored by the Michael Smith Foundation for Health Services Research.  If you agree to be involved, you may feel free to withdraw at any time. This will not affect your treatment.  Please take time to read the following information carefully before you decide.  Who is conducting the study?  The study is being conducted by Dr Karim Khan, Dr Carlo Marra, Dr Boris Sobolev, Dr Riyad Abu-laban, and Ms Bonnie Lillies of the University of British Columbia and Vancouver Coastal Health Authority. Funding has been received from the Michael Smith Foundation for Health Research to complete this study.  Purpose of the study:  The purpose of this study is to carefully record the current experiences of a senior (greater than or equal to 70 years of age) faller who has presented to Vancouver General Hospital Emergency Department. Through investigating subject information and subject experiences we will be able to understand if there exist potential areas for future improvement.  Who can participate in this study?  You can participate in this study if you are over 70 years of age and are presenting to the Emergency Department as result of a fall.     201 Who cannot participate in this study? You should not participate in this study if you do not speak English, or you have been diagnosed with impaired cognition or suffer from a cognitive disease such as Alzheimer disease or dementia.  What will happen during the study?  Should you decide to participate in this trial, our research assistant will provide you with several questionnaires to provide us with some initial information about you, your fall and your general health. You may choose not to answer any question for any reason.  Following your initial information collection, the research assistant will follow your Emergency Department (ED) stay until discharge or admission to hospital. The research assistant will record every event in your ED visit and the resources used. This information will not interfere with your care or the duration of your ED visit.  Length of study:  The study will be conducted only at the Vancouver General Hospital Emergency Department. Your participation in this study will only be while you are in the emergency department. You are not expected to be involved in any future follow-up information collection.  Risks and Discomforts:  Risks from this study are minimal. Throughout your time spent in the Emergency Department a researcher will be observing your experiences and recording the care you are receiving. The researcher will also be recording information about the Emergency Department. None of this information collection or reporting will impact your care or time spent in the Emergency Department. There is no part of this study that goes beyond or changes the current clinical practice you would receive if you chose not to be involved in this study.       202 What happens if I decide to withdraw my consent to participate?  Your participation in this research is entirely voluntary. You may withdraw at any time, and without providing any reasons for your decision.  If you decide to enter the study and to withdraw at any time in the future, there will be no penalty or loss of benefits to which you are otherwise entitled, and your future medical care will not be affected. If you choose to enter the study and then decide to withdraw at a later time, all data collected about you during your enrolment in the study will be retained for analysis.  By law, this data cannot be destroyed. Signing this consent form in no way limits your legal rights against sponsor, investigators or anyone else.  What happens after the study is finished? Once your participation in the study is concluded we will be analyzing the data received. The results will be anonymous and not individually identifiable and will not be provided to you.  What will the study cost me?  You will not incur any personal expenses as a result of participating in the study. You will not be paid for participating in this study.  Will my taking part in this study be kept confidential?  Your confidentiality will be respected.  No information that discloses your identity will be released or published without your specific consent to this disclosure.  However, research records and medical records identifying you may be inspected in the presence of the investigator or his or her designate by representatives of the UBC Research Ethics Board for the purpose of monitoring the research.  However, no records which identify you by name or initials will be allowed to leave the investigators’ offices.  Who do I contact if I have questions about the study during my participation?  If you have any questions or desire further information about this study before or during participation, you can contact Dr. Karim Khan at (604) 827-4190.   203 Who do I contact if I have any questions or concerns about my rights as a subject during the study?  If you have any concerns about your rights as a research subject and/or your experiences while participating in this study, contact the Research Subject Information Line in the University of British Columbia, Office of Research Services at 604-822-8598.   Summary information on your rights and welfare while in this study:  It is understood that you are free to withdraw from any or all parts of the study at any time without penalty. Your confidentiality will be respected. No information that discloses your identity will be released or published without your specific consent to the disclosure. However, research records and medical records identifying you may be inspected in the presence of the Investigator or his or her designate by representatives of Health Canada and the UBC Research Ethics Board for the purpose of monitoring the research. However, no records which identify you by name or initials will be allowed to leave the Investigators' offices.  Please be assured that you may ask questions at any time.  We will be glad to discuss your results with you when they have become available and we welcome your comments and suggestions. Should you have any concerns about this study or wish further information please contact Dr. Karim Khan (604 (604) 827-4190) at the University of British Columbia. The UBC phone number for research subjects to call should they have any concerns about their rights or experience as research subjects is 604-822-8598 and is called the 'Research Subject Information Line in the UBC Office of Research Services'. This information line is not intended to provide urgent service to subject with immediate needs for medical care for research-related injury.       204 Subject Consent:  I, (Please print your name) understand the purpose and procedures of this study as described and I voluntarily agree to participate.  I understand that at any time during the study I will be free to withdraw without jeopardizing any medical management, employment or educational opportunities.  I have received the consent form and understand the contents of these pages, the proposed procedures and possible risks.  I understand that I am not waiving any legal rights. I have had the opportunity to ask questions and have received satisfactory answers to all inquiries regarding this study.  By signing this form I agree that:  I have read or have had read to me the above pages concerning this study  I had a chance to ask questions regarding this study  I voluntarily agree to participate as a subject in the research study under the conditions described  I have been given a signed and dated copy of the consent form    _________________________ Signature of Subject ______________________ Date      ________________________  _______________________  ______________ Name of Witness (Please Print) Signature of Witness Date    __________________________  _____________________  _______________ Name of Investigator (Please Print) Signature of Investigator Date    205  Appendix B  Health care operations analysis to reduce attending times for seniors presenting to the Emergency Department with a fall: incident and fall history questionnaire (researcher administered)   206  ID NUMBER  __________   Health Care Operations Analysis to Reduce Attending Times for Seniors Presenting to the Emergency Department with a Fall: Incident and Fall History Questionnaire (Researcher Administered)                  207 Date ED  Age                          DOB MRN Date of Fall City Ethnicity Fall Mechanism After falling how long did the subject remain on the ground/ floor? (Circle one) None     < 5 minutes  5mins to 1 hour     > 1 hour Was the subject able to get up unaided? (Circle one)  Yes               No              Unknown  LOC      Yes        No        Unknown   Discharge Plan (please circle all which apply)  GP Referral Falls Clinic Referral Fall Risk Assessment Foot wear Assessment Home Hazard Assessment Medication Review Medication Change Mental State Assessment Occupational Assessment Orthopedics Referral Osteoporosis Assessment Physiotherapist Referral Vision Assessment Other ______________________________ No Plan Reported FALL HISTORY:  Has this subject presented with a fall at VGH ED in the past (Circle one)?     Yes    No ?   if so, when?  Please record the date, ED visit (if applicable for previous falls), and outcome (i.e. fracture, discharge, hospital admission)  Previous Fall 1  Date:  Previous Fall 2  Date Previous Fall 3  Date ED Date ED Date ED Date  Outcome  Outcome Outcome    208 Appendix C  Health care operations analysis to reduce attending times for seniors presenting to the Emergency Department with a fall: resource questionnaire (researcher administered)   209     ID NUMBER  __________      Health Care Operations Analysis to Reduce Attending Times for Seniors Presenting to the Emergency Department with a Fall : Resource Questionnaire (Researcher Administered)  210 Patient ID:    _______ __________________ Date of Presentation (Day/Month/Year) _____/_____/_____ Initial Complaint       _________________________  Method of arrival (Circle one):  Ambulance   EHS   Walk-in   Triage Priority of Subject (circle one):    1  2  3  4  5   ED beds Location   _________________________  No. available ED Beds  _________________________   Time of arrival (use 24 hour clock)  ____:____  Triage Time (use 24 hour clock)  ____:____  ED Bed time      ____:____  Seen by RN      ____:____  Seen by MD      ____:____  Seen by Geriatric Nurse    ____:____     211 1. Event 1:  _________________________ Resources used _________________________ Start Time of event 1      __:___ End Time of event 1      __:___  2. Event 2:  _________________________ Resources used _________________________ Start Time of event 2      __:___ End Time of event 2      __:___  3. Event 3:  _________________________ Resources used _________________________ Start Time of event 3      __:___ End Time of event 3      __:___  4. Event 4:  _________________________ Resources used _________________________ Start Time of event 4      __:___ End Time of event 4      __:___  5. Event 5:  _________________________ Resources used _________________________ Start Time of event 5      __:___ End Time of event 5      __:___  6. Event 6:  _________________________ Resources used _________________________ Start Time of event 6      __:___ End Time of event 6        __:___  7. Event 7:  _________________________ Resources used _________________________ Start Time of event 7      __:___ End Time of event 7       __:___  8. Event 8:  _________________________ Resources used _________________________ Start Time of event 8      __:___ End Time of event 8       __:___  Discharge Plan: Discharge time: Discharged by:  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0072466/manifest

Comment

Related Items