Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Nursing care and post-operative delirium in the cardiac surgery intensive care unit Taipale, Priscilla Gail 2010

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

Item Metadata

Download

Media
24-ubc_2010_fall_taipale_priscilla.pdf [ 2.07MB ]
Metadata
JSON: 24-1.0071390.json
JSON-LD: 24-1.0071390-ld.json
RDF/XML (Pretty): 24-1.0071390-rdf.xml
RDF/JSON: 24-1.0071390-rdf.json
Turtle: 24-1.0071390-turtle.txt
N-Triples: 24-1.0071390-rdf-ntriples.txt
Original Record: 24-1.0071390-source.json
Full Text
24-1.0071390-fulltext.txt
Citation
24-1.0071390.ris

Full Text

  NURSING CARE AND POST-OPERATIVE DELIRIUM IN THE CARDIAC SURGERY INTENSIVE CARE UNIT  by  Priscilla Gail Taipale  BScN., Laurentian University, 2001   A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF   MASTER OF SCIENCE IN NURSING   in   The Faculty of Graduate Studies       THE UNIVERSITY OF BRITISH COLUMBIA  (Vancouver)     October 2010     Priscilla Gail Taipale, 2010      ii Abstract  Post-operative delirium is a debilitating and costly adverse event that has detrimental effects on patients‟ recovery and complicates nursing care. Its numerous risk factors make the disorder seem unavoidable and unpreventable. Although pre-operative and intra-operative risk factors for delirium may not be controllable, the post-operative risk factors directly related to nursing practice are directly controllable.  Practices to control pain through analgesia and sedation administration given at nurses‟ prerogative may be associated with the onset of delirium in the immediate post-operative period. This study examined opioid and benzodiazepine administration given pro-re-nata (PRN) (“as needed”) by nurses to cardiac surgery patients to determine whether a relationship exists between delirium and nurses‟ drug administration. One hundred twenty-two patients were assessed during the first three days following cardiac surgery for delirium with the Confusion Assessment Method for the Intensive Care Unit (CAM-ICU). Data were collected regarding potential risk factors and opioid analgesia and benzodiazepine dosages given to the patients. A retrospective chart review was conducted to determine whether the patients had a physician‟s clinical assessment and diagnosis of delirium. Post-operative delirium occurred in 37.7% to 44.3% of the study sample, depending on how the cases that had positive CAM-ICU assessments and no clinical diagnoses of delirium were handled. The amount of opioid analgesia given to these patients varied widely; however, the total dosage over the 72-hour study period had no statistically significant relationship with the development of delirium (Median = 77.2 morphine equivalents (MEs) for group without delirium vs. 79.3 MEs for group with delirium; Mann-Whitney U = 1697, Z = -0.72, p = .47). The amount of Midazolam administered also varied widely.  There was a statistically significant and positive relationship between the dosage of Midazolam given and the development of post-     iii operative delirium (Median = 2.0 mg. for group without delirium vs. 4.0 mg. for group with delirium; Mann-Whitney U = 1393, Z = -2.31, p = .021). The results of this study indicate that better nursing education and changes in nurses‟ practice may be required to protect patients from experiencing drug-induced post-operative delirium.                        iv Preface  Ethical approval was obtained for this study from the Clinical Research Ethics Board (CREB) at the University of British Columbia on February 24, 2009. UBC CREB NUMBER: H09-00241  Ethical approval was also granted from Vancouver Coastal Health Authority Research Institute on March 3, 2009, #V09-0067.                           v Table of Contents Abstract ............................................................................................................................ ii  Preface ........................................................................................................................... iv  Table of Contents ............................................................................................................. v  List of Tables ................................................................................................................ viii  List of Figures .................................................................................................................. x  Acknowledgements ......................................................................................................... xi  Dedication. ..................................................................................................................... xii  Chapter One: Introduction ............................................................................................... 1  Delirium ....................................................................................................................... 1 Purpose of the Study ................................................................................................... 4  Chapter Two: Review of the Literature ............................................................................ 5  Background ................................................................................................................. 5 Delirium, the Acute Care Setting, and Assessment ..................................................... 7 Delirium: Risk Factors and Cardiac Surgery ............................................................. 10 Post-operative Pain, Analgesia, Sedation and Delirium in the ICU ........................... 16 Sedation Concerns and Delirium ............................................................................... 18     Pain and Delirium ...................................................................................................... 21 Pain Assessments in Cardiac Surgery ...................................................................... 23 Nursing Care:  A Modifiable Risk for Delirium? ......................................................... 25 Understanding Sedation Practices ............................................................................ 25 Patients’ Pain and Nurses’ Responses ..................................................................... 29 Summary ................................................................................................................... 33  Chapter Three: Methods ............................................................................................... 35  Research Question .................................................................................................... 35 Research Design ....................................................................................................... 36 Theory, Definitions and Measurement ....................................................................... 39 Measurement Tools ................................................................................................... 40 Mini mental state examination ............................................................................... 40     vi Confusion assessment method for the intensive care unit ..................................... 41 Research Protocol ..................................................................................................... 42 Recruitment and participation. ............................................................................... 42               Pre-admission patient recruitment .................................................................. 42               Hospital in-patient recruitment. ....................................................................... 42 Methods .................................................................................................................... 43 Measurement ............................................................................................................ 45 Measuring opioid analgesia. .................................................................................. 45 Measuring Midazolam. ........................................................................................... 45 Determining the presence of delirium. ................................................................... 46 Data Analysis ............................................................................................................ 47 Determining Sample Size .......................................................................................... 47 Ethics ........................................................................................................................ 47 Chapter Summary ..................................................................................................... 48  Chapter Four: Analysis and Results .............................................................................. 49  Sample Characteristics ............................................................................................. 49 Study Sample ............................................................................................................ 52     Pre-operative Risk Factors ....................................................................................... 52     Intra-operative Risk Factors ...................................................................................... 54     Post-operative Risk Factors ...................................................................................... 56 Prevalence of Post-operative Delirium ...................................................................... 60 Conservative and liberal estimates of the prevalence rate of delirium. .................. 62 Opioid Analgesia and Post-operative Delirium .......................................................... 67 Morphine equivalent doses. ................................................................................... 68 The relationship between opioid dosages and delirium. ........................................ 70 Midazolam and Post-operative Delirium .................................................................... 74 Summary of Chapter Results .................................................................................... 81  Chapter Five: Discussion .............................................................................................. 83  Delirium Risk Factors ................................................................................................ 83 Risk Factors .............................................................................................................. 84 Pre-operative Risk Factors ........................................................................................ 85 Intra-operative Risk Factors ...................................................................................... 88 Duration of surgery, anaesthesia, and cardiopulmonary bypass. .......................... 88 Hypothermia, pulmonary artery catheters, and bleeding. ...................................... 89     vii Post-operative Risk Factors and Adverse Events ..................................................... 90 Hypotension, inotropes, and low ejection fractions as risks. .................................. 91 Risks of atrial fibrillation and respiratory acidosis / alkalosis. ................................. 92 Overall risk factor profile. ....................................................................................... 93 Prevalence of Post-operative Delirium ...................................................................... 93 Administration of Opioid Analgesia to Cardiac Surgery Patients ............................... 95 Delirium and Opioid Exposure ................................................................................... 98 Midazolam Administration ......................................................................................... 99 Delirium and Midazolam exposure. ...................................................................... 100 Implications for Nursing Care .................................................................................. 101 Study Limitations ..................................................................................................... 105  Chapter Six: Conclusion ............................................................................................. 107  References ................................................................................................................. 109  Appendices ................................................................................................................. 116  Appendix A: Equianalgesic Dosing (Morphine Equivalents VCH) ............................... 116  Appendix B: VCHRI Ethics Approval Certificate. ........................................................ 117  Appendix C: UBC CREB Ethics Approval Certificate .................................................. 118  Appendix D: Participant Consent Form. ...................................................................... 119  Appendix E: Data Collection Tool ............................................................................... 122  Appendix F: Chart Review  Form. ............................................................................... 128             viii List of Tables Table 1: Risk Factors for Delirium ................................................................................. 11 Table 2: Inclusion and Exclusion Criteria for Participant Recruitment ........................... 37 Table 3: Recruitment of Participants ............................................................................. 50 Table 4: Patient Demographics and Health Status ....................................................... 52 Table 5: Pre-operative Medical Characteristics and Risk Factors ................................. 53 Table 6: Number of Pre-operative Risk Factors for Delirium Present in Study Sample .................................................................................................................. 54 Table 7: Intra-operative Risk Factors ............................................................................ 56 Table 8:Number of Intra-operative Risk Factors ........................................................... 56 Table 9: Post-operative Risk Factors ............................................................................ 58 Table 10: Number of Post-operative Risk Factors ........................................................ 58 Table 11: Sum of Risk Factors (all periods) .................................................................. 59 Table 12: Confusion Assessment Method for the ICU Delirium Assessment ................ 60 Table 13: Clinical Indication of Delirium as Determined by a Physician ........................ 60 Table 14: Agreement of Delirium Assessments ............................................................ 61 Table 15: Prevalence Rate of Delirium: Conservative Estimate .................................... 62 Table 16: Prevalence Rate of Delirium: Liberal Estimate .............................................. 62 Table 17: Frequency Distribution of Risk Factors Stratefied by Delirium ...................... 63 Table 18: Association between Intra-operative Risk Factors and Delirium ................... 64 Table 19: Participant Ages and Delirium ....................................................................... 65 Table 20: Distribution of Gender and Surgery Type in Delirium and No Delirium Groups ..................................................................................................... 65 Table 21: Comparison of Total Number of Risk Factors between Delirium Groups ...... 66 Table 22: Risk Factors of the Delirium and No Delirium Groups by the Phase of Surgery ................................................................................................... 66 Table 23: Total Morphine Equivalents Administered: Post-operative Period ................ 68 Table 24: Total Opioid Given in First 72 Hours Following Surgery ............................... 69 Table 25: Morphine Equivalents Given in Study Period: Delirium Group ...................... 70 Table 26: Morphine Equivalents Given in Study Period: No Delirium Group ................ 70     ix Table 27: The Relationship between Opioid Analgesia and Delirium Group (Conservative Classification) ................................................................................. 72 Table 28: The Relationship between Opioid Analgesia and Delirium Group (Liberal Classification) ........................................................................................... 73 Table 29: Midazolam Dosages Received by Study Participants ................................... 75 Table 30: Total Dosage pf Midazolam Received in CSICU ........................................... 75 Table 31: Total Amount of Midazolam Received by “No” Delirium Participants ............ 76 Table 32: Total Amount of Midazolam Received by Delirium Positive Participants ...... 76 Table 33: The Relationship between Midazolam and Delirium Group (Liberal Classification) ........................................................................................... 80 Table 34: The Relationship between Midazolam and Delirium Group (Conservative Classification) ................................................................................. 80                 x   List of Figures Figure 1: FLOWCHART of Participant Recruitment .......................................................51 Figure 2: Intra-operative Times ......................................................................................55 Figure 3: Post-operative Events .....................................................................................57 Figure 4: Morphine Equivalents Administered Postoperative .........................................67 Figure 5: Time of Onset of Delirium and Morphine Equivalents Administered ...............71 Figure 6: Benzodiazepine Administration .......................................................................74 Figure 7: Relationship Between Delirium & Midazolam: Conservative Analysis ............77 Figure 8: Relationship Between Delirium & Midazolam: Liberal Analysis ......................78 Figure 9: Midazolam Totals and Time to Delirium ..........................................................79  xi   Acknowledgements  The learning and success of this project is a direct result of the help, support, and partnerships of numerous individuals and groups whom I would like acknowledge with extreme gratitude. First, I wish to acknowledge and thank the patients who consented to participate in this study. Their willingness and desire to contribute to knowledge generation with the intent of possibly improving the experiences of others is admirable and greatly appreciated.  I would like to acknowledge and thank the health care team members in the cardiac services division of Vancouver General Hospital; the clerical staff, nurses, physicians, and surgeons of CSICU, CP10 and the pre-anaesthetic clinic who were gracious and accommodating by accepting us into their units. I would like to specifically thank Susan Taylor for her help during the study, as well as Dr. Rob Hewko who was a valuable resource in the planning stages.  To my research team; Connie, Deb and Jaime who sacrificed their own time for a 7 month period to help me meet my learning goals.  I could not have done it alone and your commitment to the project is truly appreciated. Your interest, and support helped to keep this project alive, I thank you all.  I would like to acknowledge the support of the Vancouver General Hospital Alumnae Association. I am thankful for your interest in developing research skills in nurses and am truly grateful for the help that was offered. Thank you to my thesis committee members at the University of British Columbia School of Nursing. Dr. Carol Jillings and Dr. Paul Galdas graciously dedicated time and provided guidance to help make this project a reality. I have felt nothing but an overwhelming sense of encouragement.  Finally, I‟d like to express my tremendous appreciation to my thesis chair, Dr. Pam Ratner who allowed, and encouraged me to pursue this endeavor.  She has provided continuous assistance, support and time over that past 2 years of this study. Her attention to detail, commitment to good research has taken my learning, appreciation, and love of research to a whole new level. It has truly been an honour to work with her.            xii   Dedication  To the nurses in the cardiac surgery intensive care unit at Vancouver General Hospital who have not only befriended me, and mentored me, but who also encouraged and supported me with their continuous interest.  You have helped to shape my nursing practice and career. I admire you all.  Finally, To my father who taught me the value of hard work.            1   Chapter One: Introduction   This thesis presents a study of cardiac surgery patients and the prevalence of post- operative delirium in a cardiac surgery intensive care unit (ICU). The study focused on pro-re- nata Midazolam and opioid administration practices of nurses and the potential relationship they may have with the development of post-operative delirium. Chapter 1 briefly describes delirium and the important role that nurses have in ensuring that evidenced-based practice guides the care they provide. The idea for this study, as well as the purpose is explained in Chapter 1 to underpin how the research questions, methods and data analysis strategies were selected. In Chapter 2, the literature regarding delirium, assessments, risk factors and the study variables, sedation and analgesia administration is critically reviewed to gain an understanding of what is known about current nursing sedation and analgesia practices as well as delirium in the acute care setting. Chapter 3 presents the research questions, rationale for the chosen design, and the definitions and terms used in the study. Chapter 3 also explains the measurement tools used, the process used for subject recruitment, the sample size determination, some ethical considerations, and the data analysis plan. Chapter 4 provides the results of the study obtained through parametric and non- parametric statistical analysis of the collected data. Finally, Chapter 5 provides an in-depth discussion of the study results and the implications for nursing care. Delirium Post-operative delirium experienced by cardiac surgery patients is a debilitating and costly significant adverse event that perplexes health care professionals. Numerous technological advances have not only improved patients‟ physical outcomes, but have also allowed for less stringent criteria regarding cardiac surgical candidacy, especially when considering age and the presence of co-morbidities (Sockalingham et al., 2005). In spite of this, or perhaps because of it,  2   post-operative delirium continues to affect about 30% of patients who undergo cardio-pulmonary bypass surgery (Laferlita, Klein, Choi, & Serfontaine, 2005); a complication that has significant implications for patients‟ health, well being and prognosis. Nurses caring for patients who undergo cardiac surgery play a pivotal role in managing patients‟ analgesia and sedation administration in the recovery phase. There is emerging evidence to suggest that these interventions may be associated with the onset of delirium and therefore highlight the importance of ensuring that nurses‟ actions are guided by knowledge of “best practices”. Researchers who have addressed the problem of delirium following cardiac surgery have associated this central nervous system dysfunction (Norkeine et. al., 2007) with a wide variety of pre-operative, intra-operative and post-operative risk factors (Miller & Wesley, 2006). In spite of research conducted within the previous 30 years, the incidence of post-operative delirium has remained constant (Van Der Mast & Friths, 1996) while intra-operative surgical techniques have improved and pre-operative risk factors have been identified. Although multiple conclusions have been drawn, the list of risk factors continues to lengthen and this adverse reaction to surgery continues to have detrimental effects on patients‟ lives lasting sometimes several months of the recovery period. Surprisingly, the post-operative period following cardiac surgery has received little research attention (Fong, Sands, & Leung, 2006) even though patients who experience delirium have longer ICU and hospital stays, increased morbidity and mortality, and possible long-term cognitive dysfunction (Gao et al., 2005). The use of benzodiazepines such as Midazolam, and opiates such as morphine (and its derivatives) in the elderly population has shown to increase confusion and delirium as well as lead to deterioration of memory and concentration (Rasmussen et al., 1999). Midazolam, a benzodiazepine frequently used in a variety of health-care settings and commonly used in ICUs has been shown to have a positive correlation with delirium in medical, surgical and trauma patients (Pandharipande et al., 2008). Opioid analgesics also have been shown to correlate with  3   the incidence of delirium (Miller & Ely, 2006). However, research related to post-operative pain levels of cardiac surgery patients has indicated that uncontrolled pain may increase the incidence of delirium (Fong, Sands, & Leung, 2006). These unclear results demonstrate the importance of studying these risk factors.  The idea for this study originated during an 18-month experience working in a cardiac surgery setting in the United Kingdom. During this time, it was noted that not only did there appear to be a lower incidence of post-operative delirium, but pain control and sedation practices differed dramatically from those experienced at a similar unit in Vancouver, Canada. This posed the initial question of why cardiac surgery patients situated in one hospital unit were experiencing higher rates of delirium and whether this could be related to nursing practices and specifically the way in which analgesia and sedation were administered. Currently, in the cardiac surgery ICU of interest, opioid analgesia and Midazolam are extensively used in the first 24 hours following surgery. These two medications may be increasing the incidence of delirium in the post-operative phase of cardiac surgery.      The literature pertaining to delirium in cardiac surgery patients currently lacks studies that systematically address the issue of risk factors immediately following surgery (Rothenhausler et al., 2005). The multifaceted nature of this problem is often under emphasized, and under treated, until patients‟ behaviour becomes problematic (Meagher, 2008). Because delirium following cardiac surgery has been shown to be one of the most frequent post-operative complications (Norkeine et al., 2007) it should not be an expected or accepted part of the recovery phase, especially for those with advanced age (Pandharipande et al., 2008). Research literature has shown that patients who experience delirium have increases in morbidity, mortality, and length of stay. However, the most significant adverse event of this phenomenon is impairment in cognitive function and quality of life (Rothenhausler et al., 2005).  4    Delirium has had a consistent prevalence in this group of patients and management and treatment have largely focused on behaviour control, rather than the identification of risk factors and the adjustment of treatment protocols to minimize the effects of those modifiable factors (Meagher, 2008). Critical thinking skills need to be utilized by nurses when administering narcotics and sedative agents; these agents may provide protection against the development of delirium, improve short-term behaviour problems, or add to the risk and worsen impairment (Meagher). Therefore, research to assist in the recognition, treatment and management of delirium, in consideration of modifiable risk factors, is essential. Purpose of the Study The purpose of this research was to explore the potentially modifiable risk factors of opioid and benzodiazepine administration that may be associated with the development of post- operative delirium in cardiac surgery patients. Answering this research question will help determine whether nurse-initiated treatment protocols for pain control and sedation increase patients‟ risk of developing delirium in the recovery period. The information garnered was anticipated to aid in the determination of the knowledge and critical thinking skills required by nurses who practice with substantial autonomy in this clinical area.  5   Chapter Two: Review of the Literature  This review of published literature examines the phenomenon of post-operative delirium experienced by cardiac surgery patients and research pertaining to the concepts of pain, analgesia and sedation administration in an ICU-setting. This review is limited to scholarly works published between the years of 1996 and 2010 with the objective of highlighting the complexity of delirium in relation to the risk factors of analgesia and sedation administration in the ICU. The literature search was conducted through the academic databases of PUBMED, the Cumulative Index of Nursing and Allied Health Literature (CINAHL) and Google Scholar. The terms delirium, cardiac surgery, nursing care, intensive care, sedation, analgesia, decision-making and pro re nata (PRN) were entered and resulted in 41 articles being selected for this review. Background  Organ failure following cardiac surgery is a devastating and life-threatening complication. Fortunately, clinicians have developed expert knowledge regarding specific clinical presentations and patient‟s responses, and have the ability to evaluate diagnostic test results and laboratory values to indicate the presence of organ failure of the lungs, heart, kidneys and liver. Health care providers also have access to empirical evidence of the risk factors as well as signs and symptoms of organ failure resulting in rapid diagnosis and intervention with proven treatment regimens. Although there is a large body of empirical evidence regarding organ dysfunction, confusion exists specifically regarding the central nervous system (CNS) dysfunction of post-operative delirium that affects patients recovering from cardiac surgery. These CNS effects of cardiac surgery remain obscure adverse events of which healthcare professionals lack complete understanding, leading to inconsistencies in standards for both assessment and treatment. Post-operative delirium is a perplexing CNS dysfunction plaguing  6   patients who undergo cardiac surgery with cardio-pulmonary bypass. Numerous research studies and reviews have attempted to explain, evaluate and determine relevant risk factors, appropriate assessment tools, and treatment options, yet no clear picture has emerged; the literature is replete with contradictory findings. Delirium is not unique to cardiac surgery patients; it is seen in all areas of acute care, including intensive care, medicine, surgery and geriatrics. Regardless of the environment, delirium presents as a problem of fluctuating confusion, agitation, inattention, disorganized thinking, and disorientation (American Psychiatric Association, 1994; Bourne, 2008; Rudolph & Marcantonio, 2008; Sockalingham et al., 2005; Ely et al., 2001;). This disorder affects the delivery of health care to affected individuals, and with little agreement about the appropriate response, mismanagement frequently occurs.  Post-operative delirium experienced by cardiac surgery patients is an important patient problem and nursing care can dramatically influence its recognition, diagnosis and treatment. Nurses have the critical role of providing care to a wide variety of patients with unique healthcare concerns, risk factors and experiences, and they are the practitioners with the most consistent contact with patients (Idemoto & Kresevic, 2007; McCaffrey et al., 2000). Basic fundamentals of care, such as diet, activity, and hygiene as well as complex decision-making, advanced practical skills, and medication administration are all aspects of nurses‟ work, which can become more difficult when delirium is present. Therefore it is crucial to understand the role nurses have to play in delirium prevention through recognition and identification of risk factors and adjustment of care. The study of the incidence of delirium in cardiac surgery patients and the potential relationship between the administration of benzodiazepines, opioid analgesics and post-operative delirium required an examination of the literature to gain an understanding of the complexity of delirium in the cardiac surgery and ICU settings. This literature review focuses on pain control, sedation and analgesia administration in the ICU, and the potential relationships between these  7   factors and delirium. This research did not explicitly study nurses‟ education levels, their work environments, or their decision-making skills, but it is a reflection of current nursing practice regarding the administration of PRN medications for pain and sedation. It was important that the contextual factors influencing nursing practice are explored. Therefore it was important to review the literature pertaining to delirium, assessment, pain control, sedation and opioid administration practices of ICU nurses as well as the concept of decision-making in an ICU to fully appreciate the relationship between nursing care and post-operative delirium. Delirium, the Acute Care Setting, and Assessment  Delirium is a disorder found in all areas of acute care and a wide variety of definitions are found in the academic literature. Confusion, agitation, disorientation, and aggressive behaviour are common descriptors documented by nurses in patients‟ records to describe the acute onset of altered consciousness and cognition (American Psychiatric Association, 1994). The literature provides several definitions of delirium to clarify this illness process, and also presents a variety of delirium assessment tools developed to detect and measure its frequency. Several studies have used these various tools in an attempt to predict delirium by incorporating the influences of risk factors. Although it can be argued that delirium may now be easier to detect, occurrence rates continue to range from 11% to 83% (Von Rompaey et al., 2008) demonstrating that inconsistencies still exist in its identification and measurement and in our true understanding of the disorder.  A review by Van Der Mast et al. (1996) found that the measurement and diagnosis of delirium in studies conducted prior to 1996 were inconsistent because several measurement tools and instruments had been used. They identified several concerns with the research they reviewed regarding the lack of homogeneity in samples, making comparisons across samples impossible, and that specific risk factors or strong predictors for delirium were not identified. In spite of the  8   14 years since Van Der Mast et al.‟s review, it is evident that delirium continues to be a problem throughout acute care with various patient populations, and the perplexity in delirium prediction and prevention still exists.  The consequences of delirium include increased length of hospital and ICU stays, increased morbidity and mortality, potential for post-traumatic stress disorder, and possible long- term cognitive impairment with an impact on quality of life (Rothenhausler et al., 2005; Ely et al., 2001). With more attention focused on the cost and consequences of delirium, assessment tools have become more user friendly for all practitioners and are now an integral part of delirium research. Although delirium assessment has become more common, standardized assessment tools are still not used in practice settings, adding to the lack of knowledge nurses and doctors have about the disorder and contributing to the confusion that still surrounds delirium. The inappropriate or lack of use of delirium assessment tools was shown in Devlin et al.‟s (2008) survey of ICU nurses about current practices regarding delirium assessments, which were compared with sedation assessments. The results showed that a majority of the critical care nurses studied routinely used validated tools to assess sedation; however, delirium screening protocols were described as “unclear”. Slightly less than one half (40%) of the respondents did not have unit-specific protocols for delirium assessment, and only 3% of the nurses considered delirium to be an important potential condition in need of assessment, in comparison with the need to assess patients‟ level of consciousness, presence of pain, and improper placement of invasive devices (Devlin et al., 2008). Nurses in this study were also unable to recognize the fluctuating signs and symptoms of delirium. Devlin et al. concluded that nurses who did not routinely assess for delirium did not know that it is under diagnosed, and did not know that a hypoactive form of delirium without the symptoms of agitation and inability to follow  9   commands existed. This study helps us to understand the important links between clinicians‟ education about and practice related to delirium and where some gaps in their knowledge exist.  The work conducted by Van Der Mast et al. (1996) and Devlin et al. (2008) highlights the continuing lack of knowledge about delirium as well as the deficiencies of delirium assessment in clinical practice and the common misconceptions that nurses may have about the signs and symptoms of the disorder. It is also evident that hypoactive delirium has received little attention even though it has been found to be most detrimental to long-term prognosis (Inouye et al., 2001). Although current research is continuing to increase practitioners‟ knowledge, the incidence rates of delirium in cardiac surgery patients and in ICUs are still alarmingly high. This may be a result of advances made in cardiology and cardiac surgery, as well as the changes that have occurred in relation to surgical candidates‟ age and prior health status (Sockalingham et al., 2005). These changes now allow patients in their eighth decade of life and those with complex chronic health conditions to be eligible for surgery (Tan et al., 2008; Santos et al., 2004; Segatore et al., 1998). Although this patient population may contribute to a higher prevalence rate of delirium, it is worrisome that the problems noted by Van Der Mast et al. (1996) of a lack of standard assessment still seems to be present as noted by Devlin et al.‟s (2008) more recent finding that many ICUs still do not have delirium protocols for assessment and treatment, and that nurses are still not aware of the importance of accurate and consistent delirium screening.  In reviewing the literature for delirium assessment tools, the Confusion Assessment Method (CAM) and the Confusion Assessment Method for the Intensive Care Unit (CAM-ICU) (Ely et al., 2001) are noted to be relatively more frequently used in practice. The CAM-ICU tool was developed to detect delirium in this environment where patients are sedated and are limited in their ability to speak because of intubation. The study conducted to validate this particular tool found a prevalence of delirium of 83.3%; however, there were differences in prevalence related to patients‟ age, gender, and race, although severity of illness was not associated with the  10   development of delirium. This may be an example of the effects of having a heterogeneous sample in a study. The study was not able to provide a clear picture of delirium risk factors. Therefore there may be deficiencies in not only assessing and using a tool to measure delirium, but also in the development of an accurate tool to measure delirium because of its complexity. Delirium: Risk Factors and Cardiac Surgery  Delirium has been reported as an adverse event following cardiac surgery for more than 40 years (Rudolph et al., 2009). Research over the last four decades has provided practitioners with an ever increasing and lengthy list of risk factors with hopes of enhancing the identification and prevention of this illness. A concern with the current literature is the lack of agreement of risk factors that potentially lead to delirium, as found in Ely et al.‟s (2001) study. This imprecision in risk factor identification may be due to study samples that are not representative of specific patient groups, which may lead to an underestimation of the effects of specific risk factors (Socklingham et al., 2005). Older age has consistently been thought to be a major risk factor for delirium (Burns et al., 2004); however, when considering the literature, it is important to recognize that age alone is not a discriminate predictor and younger patients also experience this episodic, fluctuating and debilitating condition.  Post-operative delirium experienced by cardiac surgery patients has been linked to an interaction of the many predisposing or precipitating risk factors present in the population (Von Rompaey et al., 2008). These risk factors can be explained in relation to the stage of care they occur in. Patients who require cardiac surgery are affected by, or exposed to, pre-operative, intra- operative and post-operative factors (Chang et al., 2008; Norkeine et al., 2007). Table 1 (page 11) is a list compiled from the research showing the various risks for delirium.    11   Table 1 Risk Factors for Delirium Pre-operative Risk Factors  Intra-operative Risk Factors Post-operative Risk Factors Prior stroke or TIAs  Psychiatric condition Age  Elevated Blood Urea Nitrogen (BUN)  Low ejection Fraction Hypertension  Smoking  Elevated creatinine  Co-morbidity Pain  Diabetes  Substance use  Low levels of education  Atrial fibrillation  Pre-existing dementia  Elevated body mass index  Male gender  Peripheral vascular disease Diazepam as pre-anaesthetic agent  Blood loss and transfusion  Cardiopulmonary bypass  Use of anticholinergics  Type of Surgery  Hypothermia  Anaesthetic agents used  Retrograde cardioplegia  Emergency surgery  Length of anaesthetic time  Serum albumin level  Acute renal failure  Pneumonia  Hypotension  Cardiogenic shock / low ejection fraction  Diuresis  Inadequate fluid balance  Mechanical ventilation Benzodiazepines  Narcotics  Pain  Hypoxemia / Hypercarbia  High dose inotropes  Hepatic dysfunction  Metabolic acidosis  Use of urinary catheter  Use of inotropes From: Rudolph et al. (2009), Tan et al. (2008), Van Rompaey et al. (2008), Chang et al. (2007),  and Santos et al. (2004).   The sheer number of risk factors presented in Table 1, may be a clue as to why delirium in cardiac surgery patients is often unrecognized or ignored; the misdiagnosis rate is estimated to be between 66% and 84% (Chang et al., 2008; Norkeine et al., 2007). Commonly, patients requiring cardiac surgery present with health histories inclusive of multiple pre-operative risk  12   factors that may explain why health care practitioners do not routinely consider an individual‟s probability or risk of developing delirium. Pre-operative and intra-operative risk factors are often uncontrollable, and their effects unanticipated, which may explain the under emphasis of using reliable and valid screening tools for delirium. Practitioners may consider intra-operative risk factors to be a result of pre-operative disease, or an adverse event resulting from iatrogenic causes, such as length of time on bypass and length of anesthesia (which are often a result of a patient‟s anatomy or pathophysiology of disease) and they may be overlooked as potential contributors to this disorder.  Risk factors associated with delirium have been studied at length and conflicting results are presented in various studies. Rothenhausler et al.‟s (2005) study found that the diagnosis of delirium was not associated with gender, age, marital status, pre-operative psychological testing scores, a psychiatric illness or length of hospital stay. This is not in agreement with various other studies that have found that age and history of a psychiatric illness put patients at risk for delirium, as demonstrated in Norkeine et al.‟s (2007) study that found a delirium prevalence rate of 3% with an increase risk in older male patients. Norkeine et al.‟s study also contradicted other research studies regarding pre-operative health conditions and their link to delirium risk; they found that delirium did not correlate with pre-operative hypertension, chronic obstructive pulmonary disease, an increased body mass index, or diabetes, but was affected by pre-operative hemodynamic status. Norkeine et al.‟s study found that patients with a low ejection fraction who required inotropic support, and those who had peripheral vascular disease had a higher incidence of delirium, as well as those who received blood products intra-operatively, those patients older than 65 years, and those requiring inotropic support in the post-operative period. Norkeine et al. concluded that of the 1,367 cardiac patients they studied, the incidence of delirium was strongly associated with factors already present when the patient was admitted to hospital, which demonstrates the importance of considering pre-operative risk for delirium prior to surgery.  13    Although some literature indicates that the risk of delirium may be related to health status, not all studies have supported the relationship. Tan et al. (2008) studied 53 cardiac patients and found a relationship between specific pre-morbid conditions and the development of delirium. Participants with a history of cerebral vascular disease, a high co-morbidity, an increased pre-operative creatinine, and an increased level of pre-operative pain had a significantly higher rate of delirium leading these researchers to believe that these factors were possible predictors of delirium. They also found that subsyndromal delirium (having some core diagnostic symptoms that do not meet the criteria for diagnostic threshold) and delirium were associated with a history of left ventricular dysfunction and diabetes. The results of Tan et al.‟s study also led to the speculation that pre-operative stressors such as hypo-perfusion, atrial fibrillation, pneumonia and metabolic derangements that reduce the availability of acetylcholine in the brains of stroke patients further precipitated delirium. Finally ischemic time (aortal cross clamping time) was associated with delirium, but Tan et al. did not find an association between other previously studied risk factors thought to significantly predict delirium, including a positive history of alcohol abuse, depression, visual and hearing impairments, or the 3-day opioid morphine equivalent dosage given in the post-operative period. Again this research shows some agreement as well as disagreement with other work, which may be a result of varying sample characteristics.   Predicting delirium on the basis of specific risk factors has proven to be inconsistent. As mentioned, Tan et al.‟s (2008) study found that diabetes and left ventricular dysfunction predicted delirium, whereas Rudolph et al.‟s (2009) study to develop a prediction rule for delirium in a cardiac surgery cohort found that four variables were independently associated with delirium: prior stroke or TIA, low Mini-Mental State Examination scores, abnormal serum albumin levels, and geriatric depression scale scores. Their work presents a differing view with greater emphasis placed on the pre-operative cognitive and neurological status of patients, which  14   seems limiting when considering all of the risk factors possibly linked to delirium. These groups of researchers did not agree about the pre-operative conditions that would allow practitioners to accurately predict delirium. Since Tan et al.‟s (2008) study had only 53 participants, their sample size may have been a factor, and in Rudolph et al.‟s study, patients who had surgery with cardiopulmonary bypass  (CPB) considered „on-pump‟ and those who did not have CPB, considered „off-pump‟ were included. Their results may be questionable because CPB has long been thought to contribute to micro emboli in the central nervous system (Socklingham et al., 2005) thus possibly being a significant risk. Rudolph et al. also found that alcohol consumption was associated with a reduced risk for delirium, body mass index was significantly lower in patients with delirium, and the urgency or type of surgery was not associated with delirium. These are also conflicting findings; alcohol use has been thought to increase the risk for delirium (Marcantonio et al., 1994), emergency surgery has been cited as being a risk factor for delirium (Chang et al., 2008), and patients undergoing valve surgery have been found to be at increased risk for delirium (Herrmann et al., 1999). Therefore, in spite of the discrepancies among studies, the question of whether any list of risk factors or pre-operative conditions can predict delirium may be the better question. Knowing that delirium has been linked to numerous risk factors and research continues to add to this extensive list, it appears that there is no definitive answer or specific patient profile for predicting delirium. The study by Rudolph et al. may lead practitioners to believe that delirium can be predicted based on these few risk factors, without acknowledging or appreciating the complexity of all risk factors and their contribution to this disorder.  The most comprehensive study to explain risk factors, included in this literature review, was Chang et al.‟s (2008) study of 288 cardiac surgery patients where 52 patient-related risk factors for delirium were reviewed. Chang et al. found that patients with and without delirium differed on 29 pre-operative, intra-operative and post-operative risk factors. The pre-operative  15   risk factors were: older age, less education, single marital status, psychological disorders, diabetes, history of stroke, history of renal disease, left ventricular ejection fraction of less than 30%, atrial fibrillation, emergency surgery and cardiogenic shock. Intra-operative risk factors were: complex procedures, circulatory arrest greater than 30 minutes, blood transfusion, and body temperature less than 25 degrees. Post-operative risk factors were left ventricular ejection fraction of less than 30%, atrial fibrillation, blood transfusion, cardiogenic shock, hypoalbuminemia, infection, dehydration, renal insufficiency, hepatic dysfunction, hypercarbia, and treatment with anticholinergics. Chang et al. concluded that hematocrit less than 30%, cardiogenic shock, hypoalbuminemia and acute infection were significant independent risk factors for post-operative delirium. This study draws attention to risk factors and variables that may not have been previously considered, and shows that patients present with a wide array of conditions before, during, and after surgery that potentially increase their risk of developing delirium. This study was specific to cardiac surgery patients and enhances practitioners‟ understanding of the influence and relationship of co-morbidities, and the surgery and care provided to patients. It also helps practitioners to recognize that many factors increase the risk for, and lead to, the development of delirium, and brings awareness to the complexity of the disorder and the relationships among factors, whereas other studies have attempted to identify delirium risk factors with parameters that may be too limiting. It is anticipated that post- operative nurses and physicians can work to prevent many of these risk factors because they are evidently linked to the care provided to patients and possibly to the development of delirium.  When considering the incidence of post-operative delirium experienced by cardiac surgery patients in relation to the list of pre-operative, intra-operative, and post-operative risk factors it may seem that delirium is unpreventable. Non-modifiable risk factors such as age and pre-operative health conditions, as well as some of the iatrogenic, intra-operative factors such as bypass pump time, anaesthetic time, and blood loss may not be considered by health care  16   practitioners as possible risks for delirium and instead are viewed as common patient characteristics and occurrences due to surgery. Although the literature may highlight older age, practitioners should not consider this to be the single, discriminate predictor for delirium. Emphasis on patients‟ age and its relationship with delirium may be the reason that delirium is under recognized, often ignored (Norkeine et al., 2007) and potentially why this syndrome has become accepted as a harmless, common occurrence among elderly patients (Von Rompaey et al., 2008). The research examined for this literature review demonstrates the conflicting evidence that exists in delirium studies regarding the predictive power of risk factors for delirium. The complex interplay of risk factors in all operative time periods for delirium possibly creates an improbability of deriving homogenous study samples and makes comparisons of patient groups across studies difficult if not impossible. Although pre-operative and intra-operative risk factors are not likely to be modifiable by physicians and nurses, all of these risk factors should be considered for every patient to determine the overall risk and potential of delirium. Because the delirium literature does not present a consensus of the relevant risk factors, this assessment of risk is necessary and may be a key factor that influences decisions made related to the practice and care that contributes to post-operative delirium. In particular, the post-operative risk factors of pain, and the use of opioids and benzodiazepines, which are given “as needed” by nursing staff, are directly related to the decisions nurses make regarding care and are undoubtedly modifiable. Post-operative Pain, Analgesia, Sedation and Delirium in the ICU Opioids and sedative agents are two drug classes that have been under constant suspicion of contributing to the development of delirium. The administration of sedatives and analgesia in the ICU brings forth unique contextual differences that have been widely studied and have resulted in dramatic changes to the sedation and analgesia practices of physicians and nurses  17   during the last decade. Clinical practice guidelines from the Society of Critical Care Medicine (1995) suggest that an easily arousable and calm patient is most appropriate, and should be the desired level of sedation while in intensive care (Jacobi et al., 2002). Allowing for an increase in patients‟ levels of consciousness can expedite time to extubation, and possibly prevent numerous adverse events from prolonged sedation, mechanical ventilation and hospitalization (Jacobi et al.; Mehta et al., 2009), including delirium. Opioids and sedation present as two of the largest groups of medications that are risk factors for the development of delirium. Bourne (2008) observed that the primary use of sedation and analgesia in the ICU is to aid in mechanical ventilation. The balance between level of sedation, patient comfort, and ease of ventilation is difficult to achieve, and detrimental effects are seen often with over-sedation including: prolonged ventilation, ventilator assisted pneumonia, and cardiovascular instability, whereas under-sedation may cause high levels of distress, discomfort, agitation, anxiety and pain (Guttormson et al., 2010; Jacobi et al., 2002, Ramsay, 2000). The effects of under- and over-sedation, as well as the use of opioids and sedative agents in the ICU, are all important risk factors for delirium and require further examination to gain an understanding of when, how and why sedation is used in the ICU, particularly with respect to nurses‟ clinical decision-making. Current research shows that there are various explanations for sedation practices, such as nursing knowledge, unclear physicians‟ orders, sedation protocols not individualized to patients, a lack of assessment, and inadequate documentation (Egerod, 2002; Guttormson et al.). Therefore it is important to understand how sedation, pain, and analgesia in the ICU are related to post-operative delirium in cardiac surgery patients.  The tension among analgesia, sedation and delirium is evident in Burkhart et al.‟s (2010) study where the intra-operative dose of fentanyl and the length of mechanical ventilation were found to be significant risk factors; they could account for 30% of the prevalence of delirium  18   after cardiac surgery. These factors (both considered to be modifiable) become confusing when attempting to see their relationship with delirium because the patients who received higher doses of fentanyl intra-operatively had a longer duration of mechanical ventilation in the post-operative period. When considering the results, it is unclear whether there is a relationship with delirium because, as the authors noted, the patients that were mechanically ventilated required sedation and the effects of the sedative drugs may also have contributed to the development of delirium. Burkhart et al. also mentioned that in previous studies of non-cardiac patients, a significant relationship was found between narcotic exposure and the development of delirium. Highlighting this potential relationship between narcotic analgesia and delirium may lead practitioners to believe that administering analgesia to control pain may be contributing to the development of delirium and may become a hindrance to adequate control of patients‟ pain – a “Catch-22”. Placing opioid analgesia in a risk factor category for delirium conflicts with research evidence that stresses the importance of eliminating pain by administering analgesia as a delirium prevention strategy. Sedation Concerns and Delirium  Effective use of sedative agents in the ICU to enhance patient comfort while limiting exposure to adverse events is a difficult goal to achieve. Sedation practices in ICUs have changed over time, and recommendations and guidelines are available to assist physicians and nurses to make the appropriate decisions. Jacobi et al. (2002) suggested that sedatives should be administered intermittently or “as needed” to achieve individual patient sedation goals and frequent assessment is necessary to determine the degree of sedation or agitation and to facilitate titration. Jacobi et al. also suggested that inappropriate drug regimens for sedation and analgesia may exacerbate delirium symptoms.  19   Individualized titration of sedation in consideration of patient specific factors such as age, concurrent pathology, prior alcohol abuse and concurrent drug therapy is essential for good patient care (Jacobi et al., 2002). These factors affect the intensity and duration of activity of such drugs as benzodiazepines that are commonly used for sedation. Elderly patients have been shown to have a slower clearance of benzodiazepines, longer distribution, and prolonged elimination (Jacobi et al.); it remains debatable whether this drug class contributes to the development of delirium in this patient group. Age and choice of agent are important factors to consider because a majority of cardiac surgery patients are 60 years of age or older and Midazolam is a benzodiazepine frequently used for sedation in the initial post-operative hours following cardiac surgery. It is a drug of choice because of its rapid onset and short duration, but as Jacobi et al. mentioned, paradoxical agitation has been observed with Midazolam administration, which may be a result of drug-induced amnesia or disorientation. Therefore, using Midazolam in cardiac surgery patients may seem to make sense and is preferable for treating acute agitation (Jacobi et al.), but it may also be exposing patients to a higher risk of delirium. The sedation practices of nurses and doctors in an ICU setting and their relationship with the development of delirium are important to study. Pandharipande et al.‟s (2008) study of delirium in surgical and trauma ICU patients examined the rate of delirium in relation to exposure to several drugs. Their study showed a 70% prevalence rate for delirium with Midazolam exposure, which was found to be the strongest independent risk factor for the development of delirium. Interestingly, the median age of the study participants was younger (48 years) than the average cardiac surgery population, a factor that deserves some attention because the generalizability of their results to the cardiac surgery population is questionable. The participants in Pandharipande et al.‟s (2008) study received either morphine or fentanyl for analgesia and Midazolam for sedation. The results showed that patients on morphine  20   experienced delirium but had a lower proportion of time being delirious than those who received fentanyl; however, exposure to Midazolam in both the surgical ICU patients and the trauma ICU patients increased the risk of developing delirium. Pandharipande et al. indicated that it is the benzodiazepine class and not the specific drug, Midazolam that increases the incidence of delirium among this group, and referred to the results of a previous study with medical ICU patients that found lorazepam to be a significant risk factor for delirium (Pandharipande et al., 2006). What is important to consider when reviewing Pandharipande et al.‟s work is that benzodiazepines are a sedative agent commonly used for sedation in ICUs. Their results raise questions about the use of benzodiazepines to prevent delirium among trauma and surgical ICU patients. Although the mean age of the patients in their study was lower than typically found in cardiac surgery patients, the results deserve attention because the effects of age in this sample may not be as profound; younger patients sedated with benzodiazepines experienced a relatively high rate of delirium. Sedation used in the ICU may be contributing to a higher prevalence of delirium experienced by cardiac surgery patients but these drugs are often required for patient comfort and safety. The benefits or risks of sedation need to be considered and the literature appears to recommend that nurses should attempt to limit sedation exposure where possible. Ensuring that patients are comfortable with only the necessary amount of sedation can be accomplished with adequate patient assessments focusing on sedation, delirium and pain. Wier and O‟Neill (2008) studied eight ICU nurses to explore their perceptions and experiences of using a sedation or agitation scoring tool to assess and manage sedation and agitation. All of the nurses in this study believed that sedation practices were improved with a scoring tool and guidelines, and that there was a raised awareness of pain when regularly scoring sedation levels. The nurses also believed that patients who had received guideline-directed sedation required less time being mechanically ventilated and received an improved quality of care. Interestingly, these nurses shared a common  21   goal of minimizing sedation, which is not always a consistent attitude expressed by nurses towards sedation practices, which may stem from knowledge deficits. The study participants reported that a lack of clinical experience often led to over-sedation of patients. The need for additional education and training of both nurses and physicians was identified by both Wier and O‟Neill and their study participants, and that enhanced education about sedative agents and assessing sedation would lead to positive patient outcomes. The results of this study support the importance of incorporating regular assessments of pain and sedation into practice, which could also aid in the early identification of delirium risk factors or symptoms. Pain and Delirium  Pain is a normal physiological response to tissue trauma and damage and is an important and expected stressor in an ICU following cardiac surgery (Puntillo et al., 2009). Post-operative pain is influenced by physical factors as well as psychological, social, cultural, and individual factors (Ferguson et al., 1997). Increased pain perceptions may occur in critically ill patients due to their acute, unstable conditions. The pain experienced by ICU patients has received some attention with studies focusing on patients‟ recall of their pain experiences. The overall pain experience in ICU is influenced by caregivers‟ assessments and interventions, which influences their analgesia administration practices (Ferguson et al., 1997). Puntillo (2009) assessed 152 patients for delirium with the CAM-ICU and equal percentages of CAM-positive and CAM- negative patients were able to report the presence of pain, in spite of experiencing delirium. This shows that even patients with delirium and diminished cognitive capacity are able to report pain (Puntillo, 2009), stressing the importance of assessing and treating pain.  The literature that addresses the risk factors for delirium is contradictory regarding the role of pain. Opioids have been linked to the development of delirium yet uncontrolled pain also appears to place patients at risk, complicating care delivery for nurses. Berger and Waldhorn  22   (1995) mentioned that ICU patients‟ experiences of pain and discomfort affect their sleep-wake cycles, which can further induce anxiety and delirium. Severe pain was associated with a 9-fold risk of developing delirium (Morrison et al., 2003). Drug-related delirium has been linked to the use of opioids to treat pain (Segatore et al., 1998) and many practitioners hold the belief that elderly patients experience confusion after receiving opioids and thus these drugs need to be used sparingly (Robinson et al., 2008). Roa and Cherukuri (2006) presented an alternative view with their study of hip fracture patients. They found that appropriate doses of opioid analgesia did not increase the risk of delirium, and they speculated that uncontrolled post-operative pain may increase the incidence of delirium. Burns (2004) also supported this idea by mentioning that pain is often not recognized as a potential precipitating factor when delirium occurs. According to the literature reviewed, nurses often do not provide adequate pain relief and patients often suffer from high levels of pain (McCaffrey et al., 2000). Cardiac surgery patients pose additional challenges to pain assessment and management for nurses because while they attend to abnormal organ functioning they frequently neglect adequate pain relief (Maxam-Moore et al., 1994). Therefore we know that cardiac surgery patients experience pain, patients who are delirious acknowledge pain, pain is poorly controlled in many acute care settings, barriers to pain assessment are present in the ICU, and uncontrolled pain and opioid administration are both linked to the development of delirium.  Cardiac surgery patients in the ICU are at risk of experiencing pain from surgical sites, invasive monitoring equipment, and procedures related to their care. Gelinas (2007) studied the pain experiences of 93 cardiac surgery patients and found that 77.4% of the patients recalled having pain, and 65.5% recalled being mechanically ventilated. Although the participants all received continuous infusions of fentanyl, many of them described their pain as moderate to severe. These data provide us with evidence that despite attempts made at controlling post- operative pain, the majority of patients recalled experiencing high levels of pain, possibly adding  23   to their risk for delirium. Where this research falls short, however, is that there is no mention of the dosages administered, how the fentanyl was titrated, and to what standards the drug was given. Gelinas‟s study does tell us that although patients experienced high levels of pain, they waited for the nursing staff to ask them about their pain before indicating its presence. These results are worrisome because the nurses caring for these patients may be under the impression that patients who receive a continuous infusion of analgesia will have adequate pain control and thus pain assessment may be viewed as unnecessary. This study clearly demonstrates the importance of assessing pain and tailoring care to address patients‟ individual pain control needs, especially when performing nursing care that may inflict higher levels of discomfort and pain such as turning and repositioning. Pain Assessments in Cardiac Surgery  Thorough pain assessments are crucial to understanding the pain experience of patients and to providing appropriate pain relief. Although the importance of pain assessments are frequently stressed in the literature, there continues to be difficulties incorporating this into nursing practice, as evidenced by the high levels of pain reported by ICU patients (Ferguson et al., 1997). The most reliable and valid indicator of pain is a patient‟s self-report; the pain‟s location, characteristics, aggravating and alleviating factors, and intensity are all important details required for a thorough assessment (Jacobi et al., 2002). Ferguson et al. studied 43 coronary artery bypass graft patients to evaluate their pain intensity and distress and compared their pain scores with the scores assigned by their nurses. They found that the “worst pain” intensity scores of the patients increased over time, and the nurses consistently rated the patients‟ pain lower than did the patients themselves. This research shows that nurses cannot assume that pain after cardiac surgery diminishes and that patients are in less pain as time elapses. Ferguson et al. suggested that patients may experience more pain later in the recovery period compared  24   with immediately after surgery because of an increase in their level of consciousness and mobility. These findings should be considered in conjunction with when delirium occurs because, as Robinson et al. (2008) explained, patients are at an increased risk of developing delirium in the first 3 days of the post-operative period. Therefore, cardiac surgery patients require adequate pain control in all recovery periods to help prevent delirium.  What is of further interest in Ferguson et al.‟s (1997) study is that the nurses rated their patients‟ pain levels lower than did the patients and proceeded to administer smaller doses of analgesia even though the patient reported pain. These actions are puzzling and raise several questions regarding pain assessment, knowledge of adequate dosing of analgesia, the priority and importance given to controlling pain, and nurses‟ decision-making. Ferguson et al. attempted to explain these results by arguing that the initial post-operative period may be viewed as the most painful and nurses may be more concerned with pain control because of the frequent monitoring that occurs in the ICU, allowing them to easily detect the signs and clues that patients are in pain. The researchers also mentioned that nurses are most likely to consider the initial few hours following surgery as the most painful and therefore they are more likely to provide analgesia.  Ferguson et al. (1997) emphasized the gap between patients‟ self-reports of pain and nurses‟ assessments, and the need for patient self-report to be the „gold standard‟ when assessing and treating pain. They recommended that physicians‟ orders for analgesia should be patient specific and should allow for variations in dosage to meet individual needs, and that pre- operative information about patients‟ past experiences of pain, anxiety and coping abilities could increase nurses‟ understanding of patients‟ perceptions of pain. The literature presented about pain and the difficulties that nurses have in adequately assessing and treating pain are of great importance in the context of preventing and minimizing the risk of delirium.   25   Nursing Care:  A Modifiable Risk for Delirium?  The literature presented shows that delirium is a complex disorder with its prevalence linked to a variety of risk factors. The literature reviewed so far establishes the concepts of pain, analgesia, sedation and sedation practices as modifiable risk factors for delirium and attempts to explain some of the barriers related to analgesia and sedation administration in the cardiac surgery ICU. Because analgesia and sedation are given on a PRN basis, it is evident that factors such as knowledge, education, experience, and professional autonomy influence the administration of both analgesia and sedation. Studies that have examined the sedation practices of nurses highlight the concerns, barriers, misconceptions, and attitudes that ICU nurses have that currently guide their practice.  Currey, Browne, and Botti (2006) conducted an exploratory descriptive study of 38 cardiac surgery ICU nurses to describe their perceptions of assuming responsibility for the nursing management of cardiac surgery patients in the initial two-hour post-operative period. They reported that although cardiac surgery is considered to be routine in light of its high frequency, cardiopulmonary bypass and the technology used to monitor patients and their hemodynamic status create a complex and unique circumstance for post-operative patients and the nurses caring for them. Currently, little is known about nurses‟ perceptions of assuming significant decision-making responsibility for critically ill patients, although clinical experience has been found to positively influence their decision making. Understanding Sedation Practices  Nurses are given the responsibility to provide sedation to cardiac surgery ICU patients during the initial post-operative period to provide the requisite care to meet patients‟ physiological needs to recover and return organ functioning to an enhanced or normal state.  26   Nurses are the practitioners who often make the decisions at the bedside about the amount of sedative drug to administer (Weinert et al., 2001), and the expected goal for nurses is to administer an appropriate amount of sedation to allow for nursing and medical care without causing deleterious effects to the patient. The decision to sedate with or without the presence of specific sedation protocols can be linked to common theories of patients‟ needs and safety that are contextually based in the ICU (Egerod, 2002). Unclear sedation goals and a lack of guidelines determined in a collaborative effort by nurses and doctors may result in the mismanagement of sedation and may increase a patient‟s risk for delirium. Safety concerns for patients are evident in the ICU and often surround mechanical ventilation and patient comfort. These are real and relevant concerns of nurses; however, one should question whether these concerns have become exaggerated to the point that they alone guide practice and become the rationale when providing sedation. The danger with this notion of one factor guiding practice is the exclusion of other adverse events or long-term effects as a result of the action taken. One of the potential effects of sedation is the relationship it has to the development of delirium. Walker and Gillen (2006) studied the perceptions of nurses in the role they assume in sedation management in the ICU, and their results help to broaden our understanding of these sedation practices. In their study, 78% of the nurses reported that their input was necessary in the sedation plans for their patients, yet only 51% of these nurses felt confident managing a patient‟s sedation with autonomy. The length of time working in the ICU was a determining factor of the nurses‟ comfort, which may have also explained the results of Guttormson et al.‟s (2010) study wherein only 17.7% of the participating nurses reported that it was easier to care for an awake, alert, mechanically ventilated patient. Both of these studies lead us to the idea that the comfort levels of nurses, in caring for patients in an ICU setting, may be guiding sedation practices. Guttormson et al. reported that patients who were easily rousable and responsive, able to move their trunk or limbs, and had signs of tachypnea or ventilator dysynchrony were believed to be under-sedated  27   by the nurses in their study. Fifteen percent of these nurses reported that a lack of spontaneous movement and no response to noxious stimuli were indicative of appropriate levels of sedation. This may be an indication that comfort as well as knowledge and experience affect nurses‟ sedation practices.  Although assessment, education and knowledge are essential components to providing appropriate sedation to patients, contextual factors seen only in the ICU also seem to influence the sedation administration practices of nurses. Sedation to improve mechanical ventilation and to enhance patients‟ comfort appears to be one of the misconceptions held by nurses and is seen in the results of several studies that have examined the sedation practices of ICU nurses. Patient comfort was a stated concern of the nurses caring for mechanically ventilated patients in Guttormson et al.‟s (2010) study, and 81% of the nurses reported that sedation was necessary for comfort. The studies by Egerod (2000), Walker and Gillen (2006), and Guttormson et al. all described this concept of comfort in relation to ventilator dysynchrony as a concern of nurses, which led to their decision to administer sedation. Egerod found that nurses in her study sedated their patients to improve ventilation and prevent dysynchrony, which was referred to as “fighting the ventilator” (pp. 835). The idea of improving ventilation was also found in Walker and Gillen‟s study - the nurses provided sedation more frequently when they had fears and concerns of self-extubation and ventilator dysynchrony. Egerod explained that the term “ventilator dysynchrony” suggests that a patient is inappropriate or non-compliant with mechanical ventilation. This non-compliance and the fear of adverse respiratory effects is a factor that guides nurses‟ decisions to administer sedation. What is interesting is that the nurses did not consider adjusting the ventilation settings and instead relied on sedation to solve inadequacies in ventilation (Egerod). This is an interesting finding because mechanical ventilation superimposes an artificial ventilation pattern on the natural breathing of a patient, often causing the patient to experience impaired sleep, pain, fear, and an inability to communicate, yet to solve the problem,  28   the nurses were under the impression that sedation was required. These studies tell us that it is these instances when the administration of sedation is thought of as the only possible solution that patients may receive doses that are not necessarily required that exposes them to the risk of delirium.  The results found by Guttormson et al. (2010) leads to the conclusion that nursing judgment and behaviour regarding sedation are often reflected in patients‟ sedative exposure and level of sedation. When considering the importance of nurses and physicians working to establish appropriate patient sedation guidelines, Guttormson et al. showed that 67% of nurses agreed that physicians consider nursing assessment data when they order sedation and that orders are often written with broad parameters allowing for nurses‟ discretion, which increases their autonomy, and in turn greater control regarding sedation practices. Where this becomes problematic is when the knowledge, comfort and assessment skills of nurses are lacking.  The use of objective scoring systems was found to improve clarity for nurses when determining a target sedation level for individual patients (Wier & O‟Neill, 2008). When indications for sedation are unclear, this may result in an inappropriate choice and level of sedation chosen by nurses. Egerod (2002) supported this idea by showing that experienced nurses of the same competency and autonomy levels who worked in areas that did not have formal sedation guidelines or sedation assessment tools used less sedation than did inexperienced nurses. From these findings it can be speculated that more experienced nurses are more comfortable dealing with a conscious intubated patient and are able to provide care that aims at expediting extubation, whereas less experienced nurses keep patients sedated deeper and longer. The studies by Egerod, Guttormson et al.(2010), and Walker and Gillen (2006), are in agreement that undetermined sedation goals and guidelines can possibly lead to nurses administering sedation without regard for its adverse effects such as delirium related to excessive sedation use.  29   Patients’ Pain and Nurses’ Responses  Because nurses‟ perceptions of patients‟ pain appear to directly affect their pain assessment and administration of analgesia it is important to gain an understanding of how these perceptions influence clinical decision-making. McCaffrey, Ferrell, and Pasero‟s (2000) survey of 400 nurses explored the impact of nurses‟ personal opinions about pain intensity and how it influenced their decisions about recorded assessments and titration of opioids to relieve severe pain. The researchers not only found disagreement between the pain levels reported by patients and those reported by nurses, but went further to say that nurses were more likely to believe a patient‟s self-report of pain when the patient was grimacing and not smiling (McCaffrey et al.,). The significant difference found between personal opinion (influenced by patient behaviour) and the pain assessment recorded by the nurses again leads us to question why patients‟ self-reports are not considered to be accurate.  McCaffrey et al. (2000) also examined analgesia titration in response to pain assessments and nurses were found to be more likely to increase the dose of morphine for grimacing patients (62.5%) compared with smiling patients (47.3%), who were more likely to receive either no morphine or half the dose that was previously ineffective. Overall, less than one half (43.8%) of the nurses surveyed by McCaffrey et al. correctly answered the questions regarding pain assessment and opioid dosage for both types of patients. Although most nurses recorded the patients‟ pain ratings accurately, they did not believe the ratings. The study showed that 78.2% of the nurses did not believe a pain rating of 8 on a scale of 10, and 93.8% selected a dosage of analgesia that would cause the patient to continue to suffer pain. The researchers speculated that there are two possible reasons for this behaviour: concern about patient safety and failure to believe the patient‟s report of pain.  30    In McCaffrey et al.‟s (2000) study it appears that nurses‟ personal opinions rather than patients‟ pain ratings have the greatest influence on dosage selection, which results in the under treatment of pain. This highlights the need for further education; nurses need to know when to increase or decrease the doses they administer. Lastly, an important aspect of this research is that the survey was given to nurses who were attending a conference about pain, and who demonstrated an interest in further education about pain and pain control. If nurses who display an interest and concern in pain control have difficulty in believing patients‟ pain ratings and providing adequate analgesia, then one must wonder about the quality of care provided to patients by nurses who do not exhibit special interest in the area.  Another factor that has been shown to influence nurses‟ analgesia administration is patients‟ vital signs. Although Jacobi et al. (2002) stated that vital signs are not specific or sensitive markers of the level of sedation or comfort of critically ill patients, many nurses still use these signs to guide their medication administration practices. Chuk (1999) studied the influence of vital signs on nurses‟ decision making and the titrated dosages of intravenous morphine given for pain relief following cardiac surgery, and found that nurses tended to administer more bolus dosages of analgesia instead of increasing maintenance infusions of analgesia to prevent severe pain. Jacobi et al. also described the problem of patients receiving analgesia on an “as needed basis” in that patients often receive less than the prescribed dose. Analgesia should be administered continuously, or in regular intervals with bolus PRN doses given frequently to control break through pain, to prevent rather than treat pain. It remains unclear why nurses are reluctant to increase continuous analgesia infusions. We can speculate that nurses may believe that they can gain greater control of pain when they deliver PRN boluses, however nurses failing to increase continuous analgesia may be highlighting the larger problem of knowledge deficits regarding pain control and analgesia. Interestingly, Chuk found that nurses feared respiratory depression (reported by 61% of nurses surveyed) and provided this  31   as a rationale for their practice choices, along with factors such as patients‟ body weight (31%) and behaviour such as appearance and mental status, which were used as objective pain indicators by 15% of the nurses.  The results of this study showed the inconsistencies in titrating intravenous morphine for pain relief, and that vital signs were always used as indicators of pain or relief from it. The decision-making process of drug administration used by nurses in this study and the tendency to choose inappropriate boluses as well as maintenance dosages, can lead to patients suffering higher levels of uncontrolled pain and thus requiring more analgesia overall. As well, patients‟ recovery can be impeded with inadequate pain relief and this study demonstrated that an under- rating of pain by nurses and an unreasonable concern about causing opiate-induced respiratory depression impeded care and effective pain control for patients.  The study by Chuk (1999) helps to bring to the foreground the inconsistencies in care when orders for pain relief are written on a PRN basis and are at the discretion of nursing staff. The lack of guidance in clinical decision-making regarding analgesia titration and PRN bolus dosing can generate discrepancies in practice between nurses, leaving patients to possibly suffer unacceptable levels of pain and exposing them to the risk of delirium. It is evident that the concept of pro re nata (PRN) analgesia and sedation practices by nurses needs to be examined because these orders are influenced by the accuracy of assessment, knowledge, experience, and comfort of nurses. Administration of PRN analgesia and sedation often allows nurses to determine dosage and, as Ostermann et al. (2000) mentioned, dosage is often more important than the drug chosen because all drugs present with side effects and risks - it is the manner in which drugs are administered that is of utmost importance. Although there are numerous studies that have examined the effects of nurses providing analgesia and sedation, there is little knowledge regarding the decision-making process necessary to adequately administer PRN medications for analgesia and sedation in the ICU. The critical care literature appears to not  32   address specific circumstances of PRN administration for adult patients. Therefore, to examine the concept of PRN administration, the literature in the field of psychiatry was examined because PRN administration of antipsychotics and sedatives is a common practice in this area that has been studied.  The study by Geffen et al. (2002) regarding the knowledge and beliefs of doctors and nurses in inpatient psychiatry units and the use of PRN medications for psychotic disorders were examined. It is apparent that the decision-making process involved in the use of PRNs is complex. PRN orders often lack key specifications such as indication for administration or maximum dose, and nurses caring for patients are often forced to make these decisions. Geffen et al. found that nurses frequently administered PRN medications and often made decisions regarding the choice of agent, dose, and reason for administration. They reported that when physicians fail to provide indications that specify appropriate times for nurses to initiate PRN antipsychotic medication, disparities occur between the doctors and nurses. These results are comparable to the situations that ICU nurses find themselves in when providing PRN analgesia and sedation without adequate knowledge, support and guidelines. Nurses are capable of providing PRN medications at appropriate times and with appropriate dosing, however due to various levels of nurses‟ comfort, experience and knowledge, these practices may be inconsistent, ineffective, and inappropriate for enhancing the health status of patients, and may expose patients to risks for delirium.  Idemoto and Kresevic (2007) reported that nursing care of patients with delirium requires complex and ongoing assessment of the multiple risk factors and initiation of individual interventions. They also identified the critical need for nurses to incorporate routine assessments of acute confusion into their care to detect and treat delirium. This requires the nurse to be aware of the risk factors of delirium, such as medication use, sleep deprivation, and pain. Idemoto and Kresevic‟s article about nursing sensitive outcomes and evidenced-based practice in post-  33   operative cardiac surgery patients suggests that these critically ill patients with a shortened hospital length of stay require astute assessment skills and early intervention to prevent complications. They also emphasized the need for nursing assessment and reassessment, which are crucial in the immediate post-operative period, to proactively balance activities, pain, and fatigue with appropriate medications and sedation. Finally, the concept of “nurse dose” must be evaluated by considering nurse education, expertise, and experience. Summary  Delirium experienced by cardiac surgery patients in a major setback for recovery that greatly increases health care costs, but more importantly patients suffer potentially long term deficits. It is apparent that although delirium is a complex disorder with inconsistencies in definition, risk factors, assessment tools, and disagreement regarding effective treatment, it is a problem that affects and is affected by nursing care. Nurses have the ability to control some of the risk factors for delirium such as pain, analgesia, and sedation administration by using knowledge and assessment to guide their practice. Although there is no clear picture for delirium or its prevention, issues such as adequate assessment and identification of risk are paramount for nurses to consider when caring for patients. The risk factors of sedation exposure, pain, and proper analgesia are apparent as well as are the barriers that nurses encounter in practice. This literature review has highlighted the prevalence of delirium in cardiac surgery, the risk factors, and the importance of accurate, thorough and continuous assessment of sedation, pain and delirium in light of the array of contextual factors of an intensive care environment. Discrepancies about the relationship among analgesia, sedation and delirium are evident and studies that address the PRN administration practices of ICU nurses when using these drugs are lacking, which indicates that more research in this area is required. Examining the analgesia and sedation practices of nurses and their relationship with the development of post-operative  34   delirium will help to determine where education and changes to practice are necessary. Determining if a relationship exists between the PRN administration of opioid analgesia and benzodiazepines and delirium will help nurses to acknowledge the role that they have to play in preventing this adverse event following cardiac surgery.                  35   Chapter Three: Methods   This chapter provides the research questions that guided this study and influenced the research design chosen to best examine the phenomenon of delirium and the variables of analgesia and sedation administration within the study setting of the cardiac surgery. The inclusion and exclusion criteria are presented to describe how the study participants were selected and which patient population is represented in this research. The DSM-IV definition of delirium is provided for the reader to understand the selection of the delirium measurement tool used (Confusion Assessment Method for the Intensive Care, CAM-ICU) and also to explain the need for pre-operative cognitive assessment with the Mini Mental State Examination. This chapter also explains the recruitment process initiated by the research team and the determination of sample size required for the study. Finally, the methods used to obtain the data, the ethical approval for the research, and the data analysis plan are presented for the reader to gain an understanding of how this research was conducted and how the results were derived. Research Question  This research study aimed to answer two questions related to the development of post- operative delirium following cardiac surgery. The following questions guided the development of the study methods and the data analysis. 1. Does a relationship exist between post-operative delirium experienced by cardiac surgery patients (who have undergone cardio-pulmonary bypass) in the post-operative period and the dose and frequency of opioid analgesia given “as needed” by registered nurses?  36   2.  Is there a relationship between the development of post-operative delirium in the post- operative period, in this cohort, and the dosage of benzodiazepine used for sedation and administered by registered nurses on an “as needed” bases? Research Design  A prospective, observational cohort design was used to determine whether a relationship existed between the development of post-operative delirium and the dosage and frequency of opioid analgesia and benzodiazepine sedation administered. The participants were assessed for delirium on post-operative days 1, 2 and 3 by a visit from one of four research nurses who administered the Confusion Assessment Method for the ICU (Ely et. al, 2001). Additional data were collected from the participants following the third post-operative day, or once the participant had scored a positive result on the delirium assessments, including data related to the surgery (e.g., type, time, outcomes, adverse events) and the participants‟ co-morbidities, risk factors for delirium, previous health history and clinical indication of delirium.  This study took place at Vancouver General Hospital (VGH) between April and January 2009. The cardiac surgery program at VGH is one of four provincial centres that offer cardiac surgery to residents of British Columbia. At VGH, patients are scheduled for cardiac surgery from Monday to Friday, with four to six potential patients on a daily basis. A total of 355 patients underwent scheduled cardiac surgery from April 1st 2009 – October 31st 2009.  Of this number, 196 patients underwent coronary artery bypass grafting, 36 had aortic valve replacement, 17 had a mitral valve repair or replacement, and 65 patients had a combined procedure of valve and CABG surgery. Three inpatient units of the Cardiac Sciences Division of this tertiary care centre were accessed during the study: the Coronary Care Unit, the Cardiac Surgery ICU, and the Cardiac Surgical Ward (CP10). The pre-anaesthetic admission clinic where  37   patients present for pre-operative assessment and teaching before their surgery was also used as a site for participant recruitment. Table 2 Inclusion and Exclusion Criteria for Participant Recruitment  INCLUSION CRITERIA EXCLUSION CRITERIA  Non-emergent coronary artery bypass graft surgery patients  Non-emergent valve repair/replacement surgery patients  Patients requiring cardio-pulmonary bypass  Known pre-operative cognitive impairment, or MMSE score <24/30  Non English speaking  Visual impairment not improved by corrective lenses  Haemodialysis pre-operatively  Substance misuse  Self-report of alcohol intake >7 drinks per week.   Table 2 describes the inclusion and exclusion criteria for study participation. Patients scheduled for non-emergent coronary artery bypass grafting (CABG) or aortic and or mitral valve repair or replacement with cardio-pulmonary bypass, were chosen as the target population. Cardio-pulmonary bypass (CPB) was a requirement for participation because previous research has indicated that this procedure has a significant impact on the development of delirium: numerous theories postulate that micro-embolisms from the CPB machine cause delirium (Clark et al., 1995). The type of cardiac surgery was also important because patients requiring thoracic aortic aneurysm surgery may be at increased risk for developing delirium, especially because this is typically a longer operation, more anaesthesia and sedation are administered and the bypass pump times are generally longer. Patients who required emergency surgery (defined as within 24 hours of admission) were not included because their pre-operative medical conditions may not  38   have allowed for informed consent to be obtained. The research team did not approach these patients. Patients were further excluded from the study recruitment if they had pre-operative cognitive impairment. This was assessed by the research team with the Mini Mental State Examination (Folstein, Folstein, & McHugh, 1975). Potential participants were required to obtain a minimum score of 24 out of a possible score of 30 to be eligible for the study. This was an important factor because previous research has indicated that pre-operative cognitive impairment is a significant risk factor for post-operative delirium (Kazmierski et al., 2008). The presence of pre-operative cognitive impairment may have resulted in difficulty for the study team in determining whether the characteristics of post-operative delirium were present. These patients may have presented with signs and symptoms of confusion related to their pre-operative cognitive condition and not to incident delirium. This study was limited to English-speaking patients because trained research assistants who only spoke English administered the delirium assessments, and translation by untrained individuals such as family members would have possibly influenced the reliability of the results. In addition, patients who suffered from visual impairment that was not improved with corrective lenses were excluded from the study because the tool used to determine whether delirium was present, the Confusion Assessment Method for the Intensive Care Unit (CAM – ICU) relies on participants identifying common objects on picture cards. Other study criteria excluded patients that had renal dysfunction that required haemodialysis pre-operatively. These patients were excluded because of the effects of severe renal impairment and the inability of their kidneys to metabolize and clear medications. This population may experience delirium due to drug accumulation, which is not necessarily related to the dosage or frequency of administration of the drugs under investigation.  39   Additional exclusion criteria for the study were related to substance misuse; patients with self-reported alcohol intake of greater than 7 drinks per week or current substance misuse were excluded. This history was usually identified during the pre-operative consultation with the surgeon or anaesthesiologist in the pre-anaesthetic clinic. These patients were not included because alcohol and substance misuse have been shown to significantly increase a patient‟s risk of developing delirium (Rothenhausler et al., 2005). It was decided that this exclusion criterion was required to avoid relating behaviour to delirium when in fact the patient may have been suffering the effects of substance withdrawal in the post-operative period; these two disorders often have similar clinical presentations. The effects of alcohol and substance misuse posed a problem of potentially confounding the results of this study. Theory, Definitions and Measurement  Delirium is a complex disorder that is often broadly defined in the literature. For consistency, this research study defined delirium according the American Psychiatric Association, as published in the DSM-IV. This definition also was used in the development of the delirium testing tool (i.e., the CAM-ICU) selected for use in this study. The presence of delirium is defined by the following DSM-IV criteria for Delirium due to Medical Condition. A. Disturbance of consciousness (i.e., reduced clarity of awareness of the environment) with reduced ability to focus, sustain, or shift attention. B. A change in cognition (e.g., memory deficit, disorientation, language disturbance) or the development of a perceptual disturbance that is not better accounted for by a preexisting, established, or evolving dementia. C. The disturbance develops over a short period of time (usually hours to days) and tends to fluctuate during the course of the day.  40   There is evidence from history, physical examination, or laboratory findings that the disturbance is caused by the direct physiologic consequences of a general medical condition, the result of medication use or substance intoxication, a consequence of a withdrawal syndrome, or related to more than one these sources.    (American Psychiatric Association, 1997, pp. 129). Measurement Tools Mini mental state examination 1 The Mini-Mental State Examination was developed in 1975 to briefly and systematically assess peoples‟ mental status (Kurlowicz, & Wallace, 1999). It is regarded as the foundation of bedside mental status examinations (Tang-Wai et al., 2003) because it is considered to be a valid tool for the initial testing of cognitive function (Folstein, Folstein, & McHugh, 1975). The tool contains 11 items and concentrates on the cognitive aspects of mental functioning such as orientation, concentration, serial subtraction, immediate and delayed verbal memory, language, 3-step praxis, and the ability to copy geometric designs (Folstein et al., 1975). The tool is scored numerically with a perfect score of 30, and is divided into two sections. The first section requires verbal responses and assesses orientation, memory, and attention with a maximum score of 21. The second section tests the ability to name items, follow verbal and written commands, such as writing a sentence spontaneously and copying a complex polygon, and has a possible maximum score of 9.  1 The MMSE was purchased form PAR®. The scale is protected by copyright and thus not provided here.  41   Confusion assessment method for the intensive care unit2 The Confusion Assessment Method for the Intensive Care (CAM – ICU) tool was designed after the initial Confusion Assessment Method (CAM) (1990) and was created for ICU personnel who did not have formal psychiatric training (Ely et al., 2001) to detect delirium in non-verbal, mechanically ventilated patients. The initial CAM is the most widely used instrument for diagnosing delirium by non-psychiatrists and has a good combination of speed, ease, reliability and validity (Rolfson et al., 1999). The CAM-ICU tool uses non-verbal tasks such as picture recognition, simple yes/no logic questions, and simple commands, which takes approximately five minutes to administer (McNicoll, et al., 2005). To determine the presence of delirium, four criteria must be present: (a) acute onset and fluctuation of course, (b) inability to concentrate or pay attention, (c) disorganized thinking, and (d) an altered level of consciousness (Larsson, Axell, & Ersson, 2007). Patients are considered to be delirious if they have the first and second feature and either the third or fourth feature (Ely, et al. 2001). The assessment begins by assessing consciousness with the use of the Richmond Agitation and Sedation Scale (RASS) (Sessler et al., 2000) which is a scale that ranges from negative 5 (the patient does not have any response to verbal or physical stimulation), to positive four (the patient is overtly combative). The patient is given a sedation score based on the criteria presented on the scale and those with a RASS score of -3 to +4 are then eligible to be assessed according to the four key criteria of the CAM-ICU. The second step of the CAM-ICU measure is the Attention Screening Examination to test verbal and visual recognition. Disorganized thinking is assessed using yes or no questioning and verbal commands from the investigator (Ely, & Truman, 2002). To determine the presence or  2 The CAM-ICU tool is available on line for education and research purposes only and is protected by copyright thus is not provided here.  42   absence of delirium, it is only necessary to perform the number of features that confirm the determination, and it is not necessary to use all areas of the assessment tool. Research Protocol Recruitment and participation. Pre-admission patient recruitment.  Patients were first informed of the study during their pre-operative appointment at their cardiac surgeon‟s office where they received a pamphlet that briefly described the study. Patients then attended the pre-admission clinic at VGH, before their surgery, and were asked by the clinic staff if they would like to speak to one of the research staff. If a patient agreed to lean about the study, a nurse employed for the study would visit the patient to explain the study protocol. Willing participants would then be invited to complete a consent form. Once informed consent was obtained, the participant was tested for pre-operative cognitive impairment with the Mini- Mental State Examination. Participants, who received a minimum score of 24, were deemed eligible to participate in the study. To determine whether the potential participant met all the inclusion criteria, the research nurse screened the patient‟s medical record.  Hospital in-patient recruitment.  In-hospital patients awaiting cardiac surgery were identified by the cardiac surgery triage coordinator, and were given the study pamphlet. These patients had been admitted to the Cardiac Care Unit (CCU), the cardiac surgery unit, or a medical unit at the hospital. These patients were approached by the study nurses and asked whether they were interested in learning about a research project for which they might be eligible. If a patient agreed, the study protocol was explained, and informed consent was obtained, if the patient was interested in participating. Once consent was given, these patients were screened with the Mini-Mental State Examination, and if  43   a successful score was obtained, their medical records were examined to determine whether eligibility was met regarding the remaining inclusion criteria. Methods  The study participants were assessed for post-operative delirium with the Confusion Assessment Method of the ICU. To ensure consistency the CAM-ICU, which also includes the Richmond Agitation and Sedation Scale (RASS), was chosen and used in both the cardiac surgery ICU, as well as on the cardiac surgical nursing unit. Although the original version of the Confusion Assessment Method may have been more appropriate and thorough for delirium screening of post-operative participants once they had been transferred to a cardiac surgery in- patient nursing unit, the research team decided that the use of one screening tool would prevent potential administration errors as well as enhance the comparability of the results. The delirium assessments consisted of members of the research team visiting the participants and conducting the first delirium assessment, with the CAM-ICU tool during the first 12-18 hours of their arrival into the CSICU from the operating room. This time period was chosen because it would allow for the participants to have adequate recovery time from the anaesthetic and would have a higher level of consciousness; the large majority of patients are weaned from mechanical ventilation and extubated in the first 12 hours. In this 12 – 18 hour time period, there was a higher probability that the participants would be sufficiently awake to answer the questions, as well as be able to communicate verbally. This period was also chosen since cardiac surgery patients experience an initial period of labiality in terms of their physical status, from both a cardiac and non-cardiac perspective as the body attempts to return to a period of homeostasis from the surgery. The first 6-12 hours following surgery is often the most crucial period of time when surgical problems such as haemorrhage, cardiac rhythm instability and blood pressure problems  44   are common. As well, in this time period, patients may receive larger amounts of sedative agents while the health care team attempt to control and treat any adverse events.  The patients were also assessed on post-operative days 2 and 3 if they had previously scored a negative result for delirium on their previous CAM-ICU assessments. These assessments were either performed in the CSICU (if the participants had not been transferred to the surgical nursing unit), or they were performed on the surgical nursing unit. If the participants had scored a positive result on any of the CAM-ICU assessments, they were not assessed again by the research team and were considered to be positive for delirium.  During the study, the research team indicated that the CAM-ICU delirium tool might not identify all participants experiencing delirium. The specificity of the tool was in question because it appeared that participants were able to obtain a passing score, yet had displayed signs of post-operative delirium, as indicated by the documentation of the medical and nursing staff. Because delirium is a complex process with a fluctuating course of symptoms consisting of both hyper- and hypo-active forms, it was determined that the brief assessment period with the CAM- ICU lasting 2-3 minutes may not have been able to adequately capture all positive cases of delirium.  The medical records were retrieved to determine if the patients had exhibited symptoms of post-operative delirium or had received a diagnosis or treatment of post-operative delirium by the physicians and nurses caring for them. The diagnosis of post-operative delirium by a physician is considered to be the “gold standard”. Therefore a retrospective chart audit of all study participants was conducted once all the delirium assessments with the CAM-ICU were completed.  Participants‟ charts were examined for psychiatry consultations related to delirium, as well the physicians‟ notes were scrutinized for the words, “confusion”, “agitation”, “ altered sleep”, or “delirium” or for the prescribing of the antipsychotic agents loxapine and quetipine  45   (commonly used in this setting to treat delirium) to determine if the patients had exhibited signs and symptoms of post-operative delirium. Measurement  The study sample was described according to the table of risk factors discussed in Chapter 1, and presented on page 11. Because delirium is a complex syndrome with numerous risk factors, the data obtained from the participants‟ charts were examined for the presence of several of these risk factors. Measuring opioid analgesia.  Patients in the CSICU are provided narcotic analgesia for pain control; hydromorphone is the current standard opioid of choice. In this setting, patients receive intravenous as well as oral forms of hydromorphone. Patients in the CSICU may also be given meperidine, another opioid to control post-operative shivering. In rare instances patients may also receive intravenous fentanyl, or morphine, as well as oral oxycodone. Both short acting and long acting forms of oral hydromorphone and oxycodone may be used. Therefore, in order to express the total amount of opioid analgesia administered, all opioid drugs were converted into morphine equivalent values based on a conversion chart provided by the pharmacy services at Vancouver General Hospital (Appendix A). Morphine equivalent values allowed for comparisons to be made among the participants. Measuring Midazolam.  Midazolam dosages were obtained from a review of the participants‟ medical records and summed for the duration of time spent in the CSICU. Midazolam is only administered in this setting in the initial post-operative period (usually only when the patient is mechanically ventilated).  46   Determining the presence of delirium.  Determining the prevalence rate of delirium among the study participants was based on both the CAM-ICU delirium assessment tool and the clinical assessments of delirium determined by a physician, if any (i.e., psychiatrist, anaesthesiologist, cardiac surgeon or cardiologist). The clinical assessment data were collected from the participants‟ medical records through a chart audit. Because post-operative delirium was determined by two methods (i.e., the CAM-ICU tool, and the clinical information), an algorithm to address the disagreements was developed. Where agreement existed between the two assessment methods, the disposition was straightforward. Patients who had a positive CAM-ICU result and who had a physician‟s diagnosis of delirium in their medical record were considered to have suffered from delirium. The participants who scored a negative result on all the CAM-ICU delirium assessments, and who did not have a diagnosis of delirium in the medical record were considered to not have delirium.   Disagreements between the two assessment methods were handled as follows. The participants who had negative results on the CAM-ICU assessments and had a diagnosis of delirium recorded in their medical records were considered to be positive for delirium (i.e., a physician‟s clinical judgment was considered the “gold standard”). When the participants received a positive result on a CAM-ICU assessment that was not supported by a medical diagnosis for delirium, they were treated in two ways. All analyses were conducted twice. In one instance, these participants were treated as positive for delirium (a liberal definition) and in the second, they were treated as negative cases (a conservative definition). Participants who scored a positive CAM-ICU result yet did not have a physician‟s diagnosis may have been experiencing a hypoactive form of delirium that was not expressed by changes in behaviour and thus not diagnosed by the clinical staff.   47   Data Analysis  Descriptive statistics of the study participants were obtained. The univariate characteristics of the drug administration and prevalence rates of delirium were examined. The morphine equivalent values and the Midazolam dosages were compared between the participants who had been classified with delirium and those who had not using either parametric or non- parametric tests (depending on the distribution of the data). If the data failed to follow the assumptions of normality for parametric analysis, the Mann-Whitney U test was used to explore the relationship between delirium and the amount of opioid analgesia and Midazolam that the study participants received. Determining Sample Size  When considering sample size, in order to achieve 80% statistical power (alpha criterion = 0.05) with an estimated odds ratio of 2.0 (indicating that patients exposed to the above average amounts of these drugs would be twice as likely to experience delirium) 184 participants were required. This estimate is based on the assumption that 25% of all cardiac surgery patients experience post-operative delirium, and that 60% of patients who are exposed to above average doses of Midazolam or opioid analgesia do not experience delirium. Ethics  Ethical approval for this study was granted from the University of British Columbia Clinical Research Ethics Board in February of 2009. Ethical approval was also obtained from Vancouver Coastal Health.    48   Chapter Summary  This study was designed to determine if a relationship exists between the development of post-operative delirium following cardiopulmonary bypass surgery, and the administration (dosage and frequency) of opioid analgesia and sedation with Midazolam. The research questions and design were specified to examine the opioid analgesia and benzodiazepine administration practices of nurses in a cardiac surgical setting. The specified inclusion and exclusion criteria explicitly delineated the characteristics of the study sample as well as the factors that were considered to potentially confound the study results. The DSM-IV definition of delirium was described to provide a profile of the clinical signs of this disorder, to articulate the need for pre-operative cognitive screening with the MMSE, and to describe how the DSM-IV diagnostic criteria are linked to the CAM-ICU assessment tool. The research protocol detailed how patients were recruited from the various hospital units, and the methods that were used to undertake the delirium assessments of the participants. Finally, the measurement and data analysis procedures used to determine the prevalence of delirium and the exposure to the study variables were provided.           49   Chapter Four: Analysis and Results   This chapter describes the methods and results of the research study. The number of participants recruited from the various hospital units is presented with a flow chart to describe the sampling efficiency of the study protocol. The participants‟ demographics and other characteristics are presented to characterize the presence of, or exposure to, relevant pre- operative, intra-operative and post-operative risk factors; a total risk factor score is provided to characterize the overall risk of potentially confounding variables. The prevalence of post- operative delirium is provided with an explanation of how conservative and liberal estimates of the prevalence rate of delirium were derived. The data pertaining to opioid analgesia and Midazolam exposures are examined for both study groups; those with and without delirium. The statistical analysis used non-parametric tests with the Mann-Whitney U test to determine whether a relationship existed between opioid analgesia and Midazolam administration and the incidence of post-operative delirium. Sample Characteristics  The study population included patients undergoing cardiac surgery (coronary artery bypass grafting with or without aortic valve replacement, or mitral valve repair or replacement) with cardiopulmonary bypass between March 31, 2009 and October 31,  2009. The target population included both men and women who understood both spoken and written English. All patients who were considered elective cases, or those not requiring emergency surgery (i.e., did not have a need for surgery within 24 hours) were eligible for the study (see Table 2, Chapter 3, page 37).  50    During the duration of the study, 368 patients underwent cardiac surgery; 348 were considered elective cases and 20 were considered emergencies. The following table (Table 3) indicates how many patients were approached, gave informed consent, met all the inclusion criteria, and were enrolled into the study. Figure 1 depicts the process in which the study patients were recruited.            Table 3 Recruitment of Participants  Cardiac Care Unit Cardiac Surgery Nursing Unit Pre- admission Clinic Other Nursing Unit Total Number of patients approached  16 (8.6%)  81 (43.3%)  89 (47.6%)  1 (0.5%)  187 (100%)  Figure 1 (page 51) shows that 187 patients were identified as eligible for the study and were approached in the pre-admission clinic, the CCU, or the surgical nursing unit. After the study was explained to them, 17.6% (n = 33) of the patients refused to participate, and 82.4% (n = 154) agreed to participate and completed the consent form. Once the patients had completed the consent form the research team reviewed their medical records and 9.7% (n = 15) of the participants who had consented were found to not meet all the inclusion criteria and therefore were excluded. Of those who consented, almost all (90.3%; n = 139) underwent surgery. In the post-operative phase, a further 10% (n = 14) of the participants were excluded when a more extensive chart review was conducted and an exclusion criterion was identified, one participant withdrew from the study. In summary, 124 participants were followed in the post-operative period following cardiac surgery, were assessed for delirium, and had their medical record data collected. Of this total, 118 participants (95.2%) had complete delirium assessments with the CAM-ICU and complete data available for analysis.  51   Figure 1 FLOWCHART of Participant Recruitment   52 Study Sample  A summary of the participants‟ demographic characteristics and health status is presented in Table 4. The study consisted of 124 participants (including those with incomplete data); they had an average age of 66.8 years (SD = 9.3, range 43 - 87). Of the participants, 92 were male (74%) and 32 were female (26%). Within this sample, 69 participants (55.6%) had coronary artery bypass (CABG) surgery, 33 participants (26.6%) had aortic or mitral valve surgery, and 22 participants (17.7%) had combined CABG and valve surgery.             Table 4 Patient Demographics and Health Status CHARACTERISTIC       FREQUENCY (%)           (N = 124)  Age (Years, Mean  SD)       66.8  9.3 Gender  Male                                                        92 (74.2)  Female        32 (25.8) Type of Surgery  CABG         69 (55.6)  Valve         33 (26.6)  CABG and valve       22 (17.7) Pre-operative Serum Creatinine (Mean  SD)    95.9  25.3 _______________________________________________________________  The prevalence rates of the participants‟ pre-operative, intra-operative and post-operative risk factors for delirium are provided in Table 5 (page 53).       53 Pre-operative Risk Factors 3Table 5   Pre-operative Medical Characteristics and Risk Factors  CONDITION           FREQUENCY (%)                  (N = 124) Hypertension (SBP >140)      109 (87.9) Lipid Dysfunction           82 (66.1) Coronary Artery Disease  Angina or Coronary Artery Disease    79 (63.7)  Previous STEMI or NSTEMI                               14 (11.3) Former or Current Smoker      64 (51.6) Renal Failure        34 (27.4)  Pre-operative Serum Creatinine (Mean  SD)   95.9  25.8 Congestive Heart Failure       31 (25.0)  Ejection fraction (%, Mean  SD)     55.2  12.6 Diabetes  NIDDM        31 (25.0)  IDDM           3 (2.4) Gastroesophogeal Reflux       24 (19.4) History of Cancer        20 (16.1) Peripheral Vascular Disease      16 (12.9) Atrial Fibrillation        16 (12.9) Chronic Obstructive Pulmonary Disease  Emphysema        15 (12.1)  Obstructive Sleep Apnea      9 (7.3) Cerebral Vascular Accident or Transient Ischemic Attack      8 (6.5) Previous Cardiac Intervention  Percutaneous Coronary Intervention    6 (4.8)  Previous Cardiac Surgery                                                      4 (3.2)   3 BOLD font represents pre-operative risk factors for delirium as shown in Table 1, Chapter 1, page 11.   54 Table 5 (page 53) provides the medical histories of the study participants collected from their medical records during the post-operative period. Table 1 (Chapter 1, page 11) lists the pre- operative conditions that are established risk factors for delirium. Hypertension and a history of smoking were the most common pre-operative conditions and were present in over 50% of the study sample. Renal failure and diabetes were also prevalent in this sample at a rate of 25% or greater. Peripheral vascular disease, atrial fibrillation, and a history of TIAs were present in less than 20% of the study sample.  The pre-operative health status of the study sample was analyzed to determine the prevalence of eight pre-operative risk factors for delirium, when aggregated. Table 6 shows the number of pre-operative risk factors the study participants had. Six participants (4.8%) did not have any of the pre-operative risk factors listed. Approximately one half (n=64; 51.6%) of the participants had 1 or 2 pre-operative risk factors and over one third (n = 45; 36.3%) had 3 or 4 risk factors. Nine participants (7.2%) had 5 or 6 risk factors for delirium.           Table 6  Pre-operative Risk Factors for Delirium Present in Study Sample  (N = 124) Number of Pre-operative Risk Factors Frequency Percent (%) None 1 2 3 4 5 6 Total (N) 6 17 47 26 19 7 2 124 4.8 13.7 37.9 21.0 15.3 5.6 1.6 100    55 Table 6 (page 54) shows that 119 (95.0%) of the sample had one or more pre-operative risk factors that could have potentially increased their risk of developing delirium following their cardiac surgery. Intra-operative Risk Factors  The frequency of intra-operative risk factors was also obtained. Figure 2 shows the total cardiopulmonary bypass pump (CPB) time, and duration of both the surgery and anaesthesia, which were estimated by subtracting the CSICU admission time from the surgical and anaesthetic start times, respectively.          Figure 2 Intra-operative Times Bypass Pump Time (minutes, mean  SD)   130.7 (  60.3) Anaesthetic Duration (hours, mean  SD)   5:56  (  1:41) Duration of Surgery (hours, mean  SD)   4:48 (  1:35)   Intra-operative adverse events that occurred included one (0.8%) participant experiencing a myocardial infarction during surgery and 24 (19.4%) participants experiencing intra-operative bleeding that required the administration of blood products, indicated on the anaesthetic record.  Table 7 (page 56) shows the frequency of patients‟ intra-operative risk factors. Hypothermia on admission was defined as a body temperature of less than 36.0 C, which was measured orally or internally via a pulmonary artery catheter.         56            Table 7    Intra-operative Risk Factors (N = 124) Intra-operative Risk Factor Frequency Percentage Use of Pulmonary Artery Catheter  Intra-operative Bleeding  Hypothermia 110   24   79 88.7%   19.4%   63.7%            Table 8 Number of Intra-operative Risk Factors (N = 124) Number of Intra-operative Risk Factors Frequency % 0  1  2  3  4  5  6 2  11  38  35  28  8  2  1.6  8.8  30.4  28.0  22.4  6.4  1.6    Over one half (n = 73; 58.4%) of the participants had 2-3 intra-operative risk factors for developing post-operative delirium, and only 2 of the study participants (1.6%) were not exposed to any intra-operative risk factors (Table 8).    57 Post-operative Risk Factors  The data were also examined for post-operative complications and risk factors for delirium. Figure 3 shows the post-operative events that affected the study participants. Two participants (1.6%) experienced seizures in the immediate post-operative period and one participant (0.8%) received hyperbaric treatment for air emboli. One patient was diagnosed with myasthenia gravis and one participant died on the fifth post-operative day (after the study was completed) from an unwitnessed cardiac arrest.             Figure 3 Post-Operative Events Seizures         2 (1.6) Myasthenia Gravis        1 (0.8) Transient Ischaemic Attack      1 (0.8) Death          1 (0.8)  Table 9 (page 58) shows the post-operative risk factors examined. The determination of hypotension was made on the basis of whether the patient received vasoconstricting agents in the CSICU. If norepinephrine or vasopressin were required, then the patient was considered to have experienced post-operative hypotension. The patients‟ ejection fractions were considered in relation to their cardiac index scores from the first cardiac output measurement taken in the CSICU. If the first cardiac index was less than 2.2 litres per minute, the patient was considered to have experienced decreased cardiac function and thus a low ejection fraction. The need for inotropes was evaluated by determining whether the participants required inotropic agents such as milrinone, dopamine, or dobutamine in the post-operative period. Finally, the presence of hypoxemia and hypercarbia was evaluated by examining the first arterial blood gas results for   58 indications of respiratory acidosis or alkalosis. All the study participants required mechanical ventilation and a urinary catheter and thus these risk factors were not included in the analyses.          Table 9 Post-operative Risk Factors Post-operative Risk Factor Frequency Percentage Hypotension Respiratory acidosis or alkalosis Use of Inotropes Atrial Fibrillation Low Ejection Fraction 92 44  34 22 17 74.2 35.5  27.4 17.7 13.7   Table 10 Number of Post-operative Risk Factors (N= 124) Number of Post- operative Risk Factors Frequency % 0 1 2 3 4 12 42 48 17 5 9.7 33.9 38.7 13.7 4.0  Table 10 shows that 12 participants were not exposed to any of the post-operative risk factors included in this analysis. Ninety participants (72.6%) were exposed to 1 or 2 post- operative risk factors, and 5 participants (4.0%) were exposed to all four post-operative risk factors for delirium (in addition to mechanical ventilation and a urinary catheter).     59  The pre-operative, intra-operative, and post-operative risk factors for post-operative delirium were summed to obtain a risk factor score for each participant.          Table 11 Sum of Risk Factors (all periods)  Number of Risk Factors Frequency %  2 3 4 5 6 7 8 9 10 11 12 Total 5 3 12 14 19 18 15 17 9 7 5 124 4.0 2.4 9.7 11.3 15.3 14.5 12.1 13.7 7.3 5.6 4.0 100    Of the 16 risk factors for delirium identified (excluding mechanical ventilation and urinary catheterization), 69 participants (55.6%) had 6 or 9. All study participants were exposed to at least two risk factors, and the maximum number of risk factors was 12.        60  Prevalence of Post-operative Delirium  The prevalence rates of delirium, as determined by the CAM-ICU or the clinical assessment are presented in Tables 12, and 13.           Table 12 Confusion Assessment Method for the ICU Delirium Assessment Assessment Outcome Frequency (N = 122) % Negative CAM-ICU Assessment  Positive CAM-ICU Assessment (Anytime during post-op days 1-3)   Missing data 95    27    2 77.9    22.1    1.6   The CAM-ICU delirium-screening tool detected 27 participants (22.1%) with delirium on one of the three post-operative days in which assessments were carried out, and 95 participants (77.9%) obtained negative screenings.           Table 13 Clinical Indication of Delirium as Determined by a Physician Assessment Outcome Frequency (N = 123) % Negative for Delirium  Positive for Delirium (by psychiatrist)  Positive for Delirium (by other physician)  Missing data 76  35   12   1 61.3  28.2   9.7   0.8  Of the 124 enrolled participants, one case had missing data for the clinical delirium assessments performed by a physician (Table 13). Most of the participants (n = 76; 61.3%) were   61 not diagnosed with delirium by a physician; 35 (28.2%) participants had a positive report for delirium as recorded by a psychiatrist and a further 12 participants (9.7%) were noted to be suffering from delirium by another member of the medical team (cardiologist, anaesthesiologist, or cardiac surgeon). Table 14 provides the agreement between the CAM-ICU delirium assessments and the clinical assessments performed by physicians.           Table 14 Agreement of Delirium Assessments     Clinical Assessment for Delirium CAM-ICU Delirium Assessment Scores  Delirium  No Delirium  Delirium  19 (70.4%)  28 (29.5%)  No Delirium  8 (29.6%)  67 (70.5%) Total (N = 122) 27 95  When comparing the agreement of delirium assessments completed with the CAM-ICU assessment tool and by a physician‟s clinical assessment, 67 participants (54.9%) were found to have no indication of delirium in both assessments. The results show that there was agreement for positive screens in 19 patients (15.6%). The CAM-ICU assessment failed to detect delirium in 28 (23.0%) participants that were diagnosed with delirium by physician. The CAM-ICU assessment detected delirium in 8 (6.6%) participants that did not have any indication in the medical record that a physician had diagnosed delirium. Tables 15 and 16 (page 62) provide the overall liberal and conservative estimates of the prevalence rate of delirium in the study sample.     62 Conservative and liberal estimates of the prevalence rate of delirium.  Table 15 Prevalence Rate of Delirium: Conservative Estimate  Frequency  Percentage No Delirium  Positive for Delirium  76  46 62.3%  37.7%  Note. A conservative estimate of delirium is one in which there was disagreement between the two assessment methods and the patients with positive CAM-ICU assessments and negative clinical assessments were classified as not having delirium.             Table 16 Prevalence Rate of Delirium: Liberal Estimate  Frequency  Percentage No Delirium  Positive for Delirium 68  54 55.7%  44.3%   Note. A liberal estimate of delirium is one in which there was disagreement between the two assessment methods and the patients with positive CAM-ICU assessments and negative clinical assessments were classified as having delirium.       63 4Table 17 Frequency Distribution of Risk Factors Stratified by Delirium  Risk Factor   Delirium  (liberal classification)  (N = 54) Frequency (%) No Delirium  (N = 68) Frequency (%) Pre-operative Risk Factors  Hypertension Smoking History Renal Failure Heart Failure Diabetes Peripheral Vascular Disease Atrial Fibrillation CVA / TIA  Intra-operative Risk Factors  Use of PA catheter Bleeding Hypothermia  Type of Surgery CABG Valve CABG & Valve  Post-operative Risk Factors  Hypotension Respiratory Problems Use of Inotropes Atrial Fibrillation Low Ejection Fraction     51(94%) 29(54%) 21(39%) 18(33%) 19(35%)  12(22%) 8(15%) 5(9%)     49(91%) 10(19%) 40(74%)   30(56%) 12(22%) 12(22%)     40(74%) 23(43%) 15(28%) 12(22%) 8(15%)     57(84%) 34(50%) 13(19%) 14(21%) 15(22%)  4(6%) 8(12%) 3(4%)     60(88%) 13(19%) 39(57%)   37(54%) 21(31%) 10(15%)     51(75%) 20(29%) 19(28%) 9(13%) 9(13%)  4  For the post-operative risk factor of Low Ejection fraction n = 109 since not all study patients had a PA catheter to measure cardiac output.   64  Table 17 (page 63) provides a comparison of the risk factors experienced by the patients with delirium (with the liberal classification applied) and the patients who were not considered to have delirium.  The pre-operative risk factors all had a greater frequency of occurring in the group with delirium. The examination of intra-operative risk factors also showed that patients in the delirium group were more likely to have these risk factors. Intra-operative bleeding was the only risk factor in this category that occurred equally between both groups.  The types of surgery performed were distributed equally across the two study groups for CABG surgery, and fewer patients in the delirium group had valve surgery compared with the no delirium group, although 22% of those with delirium had combined CABG and valve replacement procedures compared with 15% of the group without any indication of delirium. When reviewing the post-operative risk factors, the delirium group had a higher prevalence of respiratory risk factors and incidence of atrial fibrillation post-operatively when compared with the no delirium group.             Table 18 Association between Intra-operative Risk Factors and Delirium Variable Delirium Group (Mean  SD) No Delirium Group (Mean  SD) Age Bypass Pump Time (minutes) Anaesthesia Time (hours) Surgical Time (hours) 69.7 years  8.3 133.6 minutes  12.7 5:51 hours  1:40 4:44 hours  1:34 64.5 years  9.5 128.5 minutes  65.9 5:59 hours  1:43 4:51 hours  1:36    65  Table 18 (page 64) shows the average age and bypass pump, anaesthesia and surgical times of the two groups. The patients with delirium were slightly older, with a mean age of 69.7 years in comparison with the no delirium group at 64.5 years.           Table 19 Participant Ages and Delirium Age NO Delirium Group (N = 68) Delirium Group (N = 54) 40  - 49 years 50 – 59 years 60 – 69 years 70 – 79 years 80+ years 5 (7.4%) 15 (21.1%) 29 (42.6%) 16 (23.5%) 3 (4.4%) 1 (1.9%) 4 (7.4%) 24 (44.4%) 18 (33.3%) 7 (13%)   Table 19, shows the breakdown of ages in both the “no” delirium and the delirium groups, and Table 20 shows the distribution of gender and type of surgery in both groups (liberal estimate with all positive cases; N = 54).          Table 20 Distribution of Gender and Surgery Type in Delirium & No Delirium Groups Variable No Delirium Group (N=68) Delirium Group (N=54) Gender                                  Male                              Female  56 (82.2%) 12 (17.6%)  34(63%) 20(37%) Type of Surgery       CABG Valve CABG + Valve  37(54.4%) 21(30.9%) 10(14.7%)  30(55.6%) 12(22.2%) 12(22.2%)      66 Table 21 Comparison of Total Number of Risk Factors between Delirium Groups  Total Number of Risk Factors (pre-, intra-, post- operative) Delirium Group Frequency (%) (N = 54) No Delirium Group Frequency (N = 68) 2 Risk Factors 3 Risk Factors 4 Risk Factors 5 Risk Factors 6 Risk Factors 7 Risk Factors 8 Risk Factors 9 Risk Factors 10 Risk Factors 11 Risk Factors 12 Risk Factors 0 0 4 (7.4) 6 (11.1) 6 (11.1) 9 (16.7) 7 (13.0) 9 (16.7) 5 (9.3) 5 (9.3) 3 (5.6) 4 (5.9) 3 (4.4) 8 (11.8) 8 (11.8) 12 (17.6) 9 (13.2) 8 (11.8) 8 (11.8) 4 (5.9) 22 (2.9) 2 (2.9)   Overall, Table 21 shows that the group of participants who obtained positive assessments for delirium had a slightly greater number of risk factors for delirium than did the group that was not positive for delirium. Table 22 Risk Factors of the Delirium and No Delirium Groups by the Phase of Surgery Risk Factors By Phase of Surgery Delirium Group (N = 54) No Delirium Group (N = 68) Pre-operative 0 – 3 Risk Factors 4 – 6 Risk Factors  Intra-operative 0 – 3 Risk Factors 4 – 6 Risk Factors  Post-operative 0 – 2 Risk Factors 3 – 4 Risk Factors  35 (64.8%) 19 (35.2%)   33 (61.1%) 21 (38.9%)   43 (79.6%) 11(20.4%)  59 (86.7%) 9 (13.3%)   51 (75%) 17 (25%)   57 (83.8%) 11 (16.2%)    67  Table 22 (page 66) shows that the delirium group had more patients with a greater number of risk factors in all phases of surgery in comparison with the no delirium group. Opioid Analgesia and Post-operative Delirium  The total amount of opioid analgesia administered was calculated as morphine equivalent (ME) dosages in milligrams, and was analyzed by delirium group. Figure 4 shows the amount of opioid analgesia the participants received for the first 72 hours following surgery. The amounts are displayed in consecutive 12-hour intervals. Figure 4   68 Morphine equivalent doses.  Figure 4 (page 67) also shows that the morphine equivalent data do not have normal distributions; extreme outliers resulted in a positive skew at each time interval. This figure, however, shows that during the second post-operative interval of 12-24 hours, the study participant received the most opioid. It is also apparent that some study participants received very little opioid and others received extremely large amounts. Table 23 shows the descriptive statistics of the opioid dosages administered.           Table 23 Total Morphine Equivalents (mgs) Administered: Post-operative Period (0 – 72 hours) N =122 0 – 12 hours 12 – 24 hours 24 – 36 hours 36 – 48 hours 48 – 60 hours 60 – 72 hours  MEAN  MEDIAN  ST. DEV.  MINIMUM  MAXIMUM  18.4  15.3  10.61  2.5  59.83  22.6  21.0  13.61  0.0  63.50  13.9  11.4  11.31  0.0  56.43  12.5  9.4  11.84  0.0  62.14  9.53  5.7  9.70  0.0  45.00  7.8  5.7  11.10  0.0  85.0  During the first 12 hours of admission into the CSICU, the median amount of opioid analgesia the participants received was 15.3 mg in morphine equivalents (approximately 1.3 mg per hour). One participant received only 2.5mg of opioid in the first 12 hours; another participant received approximately 60 mg of opioid analgesia (the maximum amount given; approximately 5 mg of opioid per hour). In the second time period, 12 – 24 hours post-operative, the participants received the greatest amount of opioid analgesia in comparison with the other 12-hour intervals. The median of the dosages received was 21.0 mg (approximately 1.8 mg per hour). The minimum amount administered was 0.0 mg indicating that some participants did not receive any   69 opioid analgesia, and the greatest dose administered was 63.5mg (approximately 5.3 mg per hour). The third time interval, 24 – 36 hours following surgery showed a large reduction in the amount of opioid analgesia administered with a median of 11.4 mg (the minimum amount given was no analgesia and the maximum amount given was 56.4 mg, 4.7 mg per hour).  Table 24 Total Opioid Given in First 72 Hours Following Surgery  Total Opioid (Morphine Equivalent: mg) Total Opioid Administered per Hour (mg/hour) Mean Median Standard Deviation Minimum Maximum 84.8 77.7 48.67 9.50 288.0 1.17 1.08  0.13 4.00  Table 24 shows the total amount of opioid the participants received in the 72 hours following their admission into the CSICU, as well as the total hourly dosage of opioid given. The total 72-hour opioid data were positively skewed because some patients received very large dosages. As evident, the per hour range of opioid analgesia was between 0.1 mg per hour and 4.0 mg per hour showing a large variation in what the study patients received.           70 The relationship between opioid dosages and delirium.            Table 25 Morphine Equivalents (mg) Given in 72 - Hour Study Period: Delirium Group  (N = 54) 0-12 hours 12-24 hours 24-36 hours 36–48 hours 48-60 hours 60-72 hours Mean Median Std. Dev. Minimum Maximum 18.7 14.8 11.0 3.0 50.5 21.4 17.7  14.7 1.5 59.9 14.0 12.7 10.6 0 48.5 12.2 10.7 11.4 0 52.0 9.3 5.7 10.8 0 45.0 8.6 5.0 14.4 0 85.0  Table 26 Morphine Equivalents (mg) Given in 72 - Hour Study Period: No Delirium Group  (N = 68) 0-12 hours 12-24 hours 24-36 hours 36–48 hours 48-60 hours 60-72 hours Mean Median Std. Dev. Minimum Maximum 18.2 16.3 10.34 2.5 59.8 23.7 22.3 12.7 0 63.5 13.8  11.0 12.0 0 56.4 12.8 8.9 12.3 0 62.1 9.7 5.7 8.8 0 45.0 7.1 5.7 7.7 0 33.9  Tables 25 and Table 26, show opioid analgesia given to both the delirium and no delirium groups. During the first 12 hours of the post-operative period, the patients in the no delirium group received slightly more opioid analgesia (Median = 16. 3 mg), and in the second time period of 12-24 hours patients in the no delirium group also received more analgesia (Median = 22.3 mg). Time periods 3 and 4 show a slightly greater amount of opioid analgesia given to the   71 delirium group, and time period 6 shows that a patient in the delirium group received a maximum 12- hour dose of 85 mg of opioid. It is also important to note that of the 6 times periods studied, the no delirium group had the greatest maximum amount of opioid analgesia administered in the first 4 time periods. None of the differences were found to be statistically significant with t-test and Mann – Whitney U analyses. Figure 5              Time of Onset of Delirium and Morphine Equivalents Administered    Figure 5 is a scatter-plot showing the dosages of morphine equivalents the participants were administered and the number of hours that had passed in the post-operative period before the diagnosis of delirium was made. It appears that there is an association between the amount of   72 analgesia given over the post-operative period and when the diagnosis of delirium was made. Participants who received smaller amounts of opioid analgesia experienced delirium within the first 24 hours following surgery.  Because the opioid data distribution did not follow a normal distribution in any time period, the non-parametric Mann-Whitney U test was used to determine whether a relationship existed between the opioid dosage administered and the development of delirium. Because delirium was diagnosed with two methods, statistical testing was conducted with both the conservative and liberal classifications of patients with delirium (Table 27 and Table 28, page 73). Table 27 The Relationship between Opioid Analgesia and Delirium Group  (Conservative Classification)    The participants who were positive for post-operative delirium did not differ in the amount of opioid analgesia (Median = 76.8) they received in the first 72 hours following surgery (Mann-Whitney U = 1654.0, p = .62) compared with the group with no delirium.         Opioid Analgesia Total (72 hours) Mann-Whitney U  Z  Asymp. Significance (2-tailed) 1654.0  -0.479  .619   73          Table 28 The Relationship between Opioid Analgesia and Delirium Group (Liberal Classification)  Opioid Analgesia Total (72 hours) Mann-Whitney U Z Asymp. Significance (2- tailed) 1697.0 -0.716 0.474   The participants who were classified with a liberal definition as having delirium did not differ in the amount of opioid analgesia (Median = 76.8) they received in the first 72 hours following surgery (Mann-Whitney U = 1697.0, p = .47) compared with the group with no delirium (Table 27, page 71).  Tables 27 (page 71) and Table 28 show that there were no statistically significant relationships between the amount of opioid analgesia the participants received and the development of post-operative delirium regardless of the criteria employed to determine the incidence of delirium.          74 Midazolam and Post-operative Delirium  Midazolam dosages (in milligrams) given post-operatively to the study participants in the CSICU were summed. Figure 6 shows the distribution of the Midazolam dosages given to the study sample. Figure 6    Figure 6 shows that the dosages of Midazolam in milligrams given to the sample did not follow a normal distribution. Several participants received large amounts of Midazolam. Table 29 (page 75) shows the descriptive statistics for the total amount of Midazolam administered to the participants.   75           Table 29 Midazolam Dosages Received by Study Participants  No Midazola m 1.0 – 4.0 mg of Midazolam 4.1 – 8.0 mg of Midazolam 8.1 – 20.0 mg of Midazolam > 20.1 mg of Midazolam Number of Study Participants (N = 122)  30 (24.6%)  52 (42.6%)  28 (23.0%)  9 (7.4%)  3 (2.5%)   Table 29 shows that within this study, 24.6% of participants did not receive any Midazolam, 42.6% received 1.0 – 4.0 mg, 23.0% received 4.1 – 8.0 mg, 7.4% received 8.1 – 20.0mg, and 2.5% of participants received dosages of Midazolam that were greater than 20.1 mg.         Table 30 Total Dosage of Midazolam Received in CSICU Mean Median Standard Deviation Skewness Standard Error of Skewness Minimum Maximum 4.8 3.0  9.4 5.8 0.2 0.0 83.0   Table 30 shows that the mean dosage of Midazolam given to the participants was 4.8 mg in a 72-hour period. However, it is apparent that the data had a strong positive skew. Some study participants did not receive any Midazolam, while there were cases that received large amounts as indicated by the maximum dosage of 83 mg. Because these data did not meet the assumptions of a normal distribution, the non-parametric Mann–Whitney U test was used to determine   76 whether a relationship existed between the total dosage of Midazolam received by the participants and the development of post-operative delirium.  Total dosage of Midazolam administered and delirium. Table 31 Total Amount of Midazolam Received by “No” Delirium Participants   0 mg Midazolam 1.0 – 4.0 mg Midazolam 4.1 – 8.0 mg Midazolam > 8.0mg of Midazolam Number of Participants with No Delirium N= 68  19 (27.9%)  33 (48.5%)  13 (19.1%)  3 (4.4%)   Table 32 Total Amount of Midazolam Received by Delirium Positive Participants  χ² = 8.4 (df = 3), p = 0.039* Tables 31 and Table 32 show that 81.2% of the patients who were positive for delirium (n = 44) received Midazolam, and 70.5% (n = 48) that did not have delirium received Midazolam in the CSICU post-operatively. There were more patients in the no delirium group that did not receive any Midazolam (29.4% compared to the group that was positive for delirium 18.5%). As well, more patients in the delirium group received larger doses of Midazolam (43.6%) than patients in the no delirium group (22.0%).  0 mg Midazolam 1.0 – 4.0 mg Midazolam 4.1 – 8.0 mg Midazolam > 8.0mg of Midazolam Number of Participants Positive for Delirium N= 54  10 (18.5%)  19 (35.2%)  16 (29.6%)  9 (16.7%)   77 Figure 7    Figure 7 shows the conservative classification of patients with delirium and the total dosage of Midazolam administered by delirium status. The box plots show that the patients who had a positive result for delirium received larger dosages of Midazolam in the CSICU.     78 Figure 8     Figure 8 shows the liberal classification of patients with delirium and the total dosage of Midazolam given in the CSICU by delirium status. It is apparent from the box plots that the participants who had positive screenings for delirium received larger dosages of Midazolam.     79 Figure 9    Figure 9 shows the total amount of Midazolam received (in milligrams) and the time of the delirium diagnosis. From this figure it is apparent that the participants were diagnosed with delirium at various times throughout the 72-hour study period, and that the larger dosages were not simply the result of an accumulation over time.  The relationship between the development of post-operative delirium and the administered dosage of Midazolam was analyzed with the Mann-Whitney U test because the distribution did not meet the assumption of normality. Tables 33 and 34 (page 80) show the results of this analysis.   80          Table 33 The Relationship between Midazolam and Delirium Group (Liberal Classification)   Midazolam Dose Received in CSICU Mann-Whitney U Z Asymp. Significance (2- tailed) 1393.0 -2.31 .021   The participants who were liberally classified as having post-operative delirium received a statistically significantly greater dosage of Midazolam in the CSICU, post-operatively (Median = 3.0).  Table 34  The Relationship between Midazolam and Delirium Group (Conservative Classification)   Midazolam Dose Received in CSICU Mann-Whitney Z Asymp. Significance (2- tailed) 1251.5 -2.65 .008   The participants who were positive (conservative classification) for post-operative delirium received a statistically significantly larger dosage of Midazolam in the CSICU, post- operatively (Median = 3.0).      81 Summary of Chapter Results  In this study, data were collected from 124 participants, 74% of whom were men. The mean age of the sample was 66 years of age and 55% of the sample underwent coronary artery bypass grafting, 27% underwent valve repair or replacement, and 18% underwent both coronary artery bypass surgery and valve repair or replacement. About 56% of the study participants had 6 - 9 risk factors for post-operative delirium.  The prevalence rate for delirium, based on the Confusion Assessment for the ICU (Wesley et al., 2000) alone, was 22% and a chart audit indicated that a physician diagnosed 38% of the participants with delirium. The methods for determining delirium were combined and resulted in agreement with 54% of the participants having negative screening results and 15% of the sample having both a positive CAM-ICU and clinical assessment for delirium. With the clinical assessment for delirium being considered the “gold standard,” all patients with a positive clinical assessment were classified as having delirium. In the situation where 8 cases that had positive CAM-ICU scores and no clinical notation of delirium in the medical record, two classifications were applied: they were treated conservatively as not having delirium and liberally as having delirium. Therefore the estimated prevalence rates of delirium ranged between 37.0% and 43.5% of the study participants.  The pre-operative and intra-operative risk factors for delirium were more prevalent in the delirium group than in the group without delirium; the participants in the delirium group were more likely to have undergone combined CABG and valve surgeries. As well, the delirium group was found to have more patients with a higher number of risk factors in the pre-, intra-, and post- operative periods. The delirium group also had more post-operative respiratory events, including hypercarbia and hypoxia, compared with the no delirium group, and also had a higher incidence of atrial fibrillation post-operatively compared with the no delirium group. Lastly, the   82 participants in the delirium group were found to be slightly older (69.7 years of age). The average bypass pump, and anaesthesia and surgical times were comparable in the two study groups.  The participants received the greatest amount of opioid analgesia in the first 12-24 hours following surgery. Some participants received very little opioid analgesia and others received very large dosages. Some patients received no opioid analgesia after the first 12 hours of the post-operative period. The median opioid dosage administered over the 72-hour study period was 76.8 mg or 1.2 mg of opioid per hour. These results showed that there was wide variation in the amount of opioid analgesia received by the participants. There was no relationship found between the amount of opioid analgesia administered and delirium with either the conservative or liberal classification approaches applied.  About one half of the participants received 3mg of Midazolam or more in the first 72 hours of their being in the CSICU (median = 3.0 mg). There was a statistically significant difference in the amount of Midazolam given to the study groups (regardless of classification rules for delirium); the participants who were positive for delirium had received larger doses of Midazolam, on average.    83 Chapter Five: Discussion   This final chapter summarizes the key findings of the study. The prevalence of post- operative delirium in the sample is discussed in consideration of the pre-operative, intra- operative and post-operative risk factors the participants were exposed to or experienced. Comparisons are made between the patients who did not suffer from delirium and those who tested positive either by the CAM-ICU tool or through a clinical assessment of delirium to determine the impact of all the risk factors examined in the study and the potential differences between the two groups. The findings from the analysis of opioid analgesia and Midazolam are further examined and explained in the context of the cardiac surgery ICU to gain a better understanding of the administration practices of nurses and their implications for patient care. The prevalence rates of delirium, obtained in the conservative and liberal assessment approaches are explained, as is the role of Midazolam and opioid analgesia administration in the development of post-operative delirium in this sample of cardiac surgery patients. A discussion of the study limitations, the implications for nursing care, and areas where future research can focus concludes the chapter. Delirium Risk Factors  Post-operative delirium in cardiac surgery patients is linked to a variety of factors that occur or are present in the pre-operative, intra-operative or post-operative phases of cardiac surgery. Although there are inconsistencies and disagreement in the literature regarding the influence of these factors in the development of delirium, it is important to discuss the prevalence of several specific factors in relation to the sample population of this study. Age, gender, and type of surgery are discussed, and then the pre-operative, intra-operative, and post- operative periods are examined for the total sample as well as for the delirium and no delirium   84 groups. The final, cumulative risk factor score is examined to shed some light on the confounding factors and variables that may require further consideration when analyzing delirium data with this population. Risk Factors Age has been studied as a risk factor for delirium in numerous patient populations and there is some consensus that older age is linked to a higher rate of delirium. Older age may increase a patient‟s risk for delirium because of reductions in cerebral neuronal density, blood flow, metabolism, and levels of brain neurotransmitters (Lipowski, 1984 as cited in Santos et al., 2004). Aging is often accompanied by changes in cognitive capacity and elderly patients are more likely to have multiple co-morbidities, which, in combination, increase a patient‟s risk of experiencing delirium (Norkeine et al., 2007). Older age has many effects on organ functioning and is especially important when considering drug absorption, distribution and metabolism by the kidneys. Patients requiring cardiac surgery are often in the later stages of life with a majority of patients over the age of 60 years, making this a relevant risk factor to explore. In the current study, the average age of the participants was 66.8 years with patient range of 43 to 87 years of age. The participants in the delirium group had an average age of 69.7 years with a range of 49 to 87 years. In comparing the delirium and no delirium groups it appears that age was equally distributed in both groups. The delirium group had a slightly larger number of patients over the age of 80 years (n = 7; 13%) compared with the no delirium group (n = 3; 4.4%).  In some delirium studies, gender has been hypothesized to be a risk factor for delirium. The delirium group in this study had a larger percentage of female patients (37%); however this finding is difficult to fully explain because of the relatively small number of female patients in the total sample. As a result, we are not able to precisely determine whether gender is a predictor of delirium; this relationship deserves more attention.   85  The type of surgery also has been mentioned as a risk factor for post-operative delirium in cardiac surgery patients, and valve surgery patients have been shown to experience higher rates of delirium (Burkhart et al., 2010; Herrmann et al., 1999). Herrmann et al. have explained this as neuronal damage from the release of isoenzymes, and by Bucerius et al. (2004) as the result of the embolization of trapped air in the cardiac chambers during valve surgery. Type of surgery was analyzed for this study and the number of patients undergoing coronary artery bypass was fairly equal in both groups (no delirium = 54.4% vs. delirium = 55.6%). Valve surgery was more common in the no delirium group (30.9%) than in the delirium group (22.2%), which is inconsistent with the claim in the literature that valve surgery is a risk factor for delirium. A difference found between the delirium and no delirium groups was that there was a higher prevalence of delirium with patients undergoing both coronary artery bypass and valve surgery combined (22.2%), in comparison with those in the no delirium group (14.7%). Pre-operative Risk Factors Various pre-operative risk factors consist of the health challenges patients present with before surgery and which are often considered non-modifiable because they are linked to the disease process that brings patients to require cardiac surgery. In this study, eight pre-operative risk factors were analyzed in relation to delirium: hypertension, smoking status, renal failure, congestive heart failure, diabetes, peripheral vascular disease, atrial fibrillation, and previous cerebral vascular accident or transient ischemic attack.  Hypertension (i.e., a systolic blood pressure exceeding 140 mmHg) was the most prevalent risk factor; 87.9% (n = 109) of the sample had it. Hypertension is recognized as a risk factor for delirium, in the literature. It is believed high systemic pressures that exceed the body‟s ability of autoregulation produce perfusion changes and tissue damage (Santos et al., 2004). Cardiac surgery patients with hypertension have been shown to have higher rates of   86 thromboembolic infarcts, lacunas, and brainstem lesions (Schmidt et al., 1993 as cited in Santos et al. 2004), that affect the brain and can be exacerbated by the effects to perfusion that occur during cardiac surgery. The prevalence of hypertension was only slightly greater in the delirium group (94%) compared with the no delirium group (84%) so we are uncertain of the impact of this difference.  Smoking was identified as a risk factor for delirium in Van Rompaey et al.‟s (2009) study; they speculated that the disruption in the consumption of nicotine may cause a withdrawal delirium because nicotine has been linked to memory and attention tasks (Van Rompaey et al.). Abrupt cessation of nicotine may cause deleterious effects on memory and attention (Rezvani & Levin, 2001). Smoking is also related to atherosclerotic changes that may induce cerebral damage (Bolinder et al., 1997). Patients who smoke are also vulnerable to chronic pulmonary disorders and lower oxygen saturation in the brain, which may influence the onset of delirium (Van Rompaey et al.). Smoking status was the second most prevalent risk factor in this sample, with 64 participants (51.6%) either having a previous history of smoking or being current smokers at the time of surgery. The delirium group had a slightly larger percentage of participants with a smoking history compared with the no delirium group. The patients were also assessed for pre-operative renal impairment based on the creatinine values that were recorded on the pre-anaesthetic record. Patients with a pre-operative serum creatinine level greater than 100 (μmol/litre) were considered to have this risk factor. In this sample, 34 patients (27.4%) were described as having renal failure or impairment prior to surgery. Tan et al. (2008) stated that renal failure often is a reflection of poorly controlled diabetes, hypertension or vascular disease, which are also risk factors for delirium. On average, the study participants had a relatively high mean serum creatinine level of 95.5 ( mol/L), although the distribution was positively skewed and a large variance was seen in the levels. Renal function is an important factor to consider because it has dramatic effects on the ability of   87 the body to secrete metabolites from various medications administered. Opioids and benzodiazepines are both metabolized by the liver, but may cause prolonged effects when renal clearance is impaired since these metabolites are excreted by the kidneys (Berger & Waldhorn, 1995).  The delirium groups had a higher prevalence of impaired renal function (39%) compared with the no delirium group (19%).  Congestive heart failure (CHF) was another pre-operative condition that was present in 25% of the sample. This condition has been found to be a risk factor for delirium because of the potential fluctuations and reduction in perfusion leading to hypoxemia (Giltay et al., 2006) that affects blood flow and oxygenation to the brain. In this study, there were approximately 10%more participants who presented with CHF in the delirium group than in the no delirium group.  Diabetes has been linked to delirium because of its association with atherosclerosis in which narrowing of the arteries and veins occur and impede blood flow and organ perfusion. Patients with diabetes have an increased risk of microemboli and ischemia from atherosclerotic plaques (Chang et al., 2008; Norkeine et al., 2007). In this study, 25% of the participants were classified as having non-insulin-dependent diabetes mellitus (NIDDM) and either required dietary restrictions or oral hypoglycemics to control their blood glucose levels. Insulin-dependent diabetes mellitus (IDDM) was less common in this sample (2.4% were affected). A larger percentage of patients with diabetes were found in the delirium group.  Peripheral vascular disease (PVD) is a risk factor for delirium because of the narrowing of cerebral blood vessels and resultant diminished blood flow to the brain that is further exaggerated with hypoperfusion during surgery (Norkeine et al., 2007). Peri-operative hypotension or hypoxemia, which occurs frequently during and after cardiac surgery, also causes decreased oxygen supply to the brain (Norkeine et al., 2007) posing a risk for delirium. In this   88 study PVD was found in 12.9% of the study participants with a larger percentage of patients with PVD in the delirium group.  Yet another pre-operative health condition that has been found to be a risk factor for post- operative delirium is atrial fibrillation. Patients who suffer from atrial fibrillation are at increased risk of delirium because of impaired cerebral blood flow that results from low cardiac output or emboli (Santos et al., 2004). Atrial fibrillation was found in 12.9% of the study participants and had almost equal prevalence rates in the delirium and no delirium groups.  The last pre-operative risk factor for delirium to consider is the presence of a history of cerebral vascular accident (CVA more commonly known as stroke) or transient ischemic attacks (TIA). These conditions are thought to increase the risk of delirium because of peri-operative stressors, such as hypoperfusion and or atrial fibrillation that reduce the availability of acetylcholine (a cerebral neurotransmitter), which is detrimental to the brains of stroke patients (Tan et al., 2008). The risk factor of CVA and or TIA was present in only 6.5% of the study participants; the prevalence was slightly greater in the delirium group. Intra-operative Risk Factors Duration of surgery, anaesthesia, and cardiopulmonary bypass. When reviewing the intra-operative risk factors for delirium, data pertaining to anaesthetic time, duration of surgery and bypass pump time were obtained from the operating room and perfusion records. The average length of anaesthesia for the sample was 5 hours and 56 minutes (SD 1:41). When comparing the delirium and no delirium groups, the mean times differed by 9 minutes. The duration of the surgery was calculated in relation to the cardiac surgery ICU admission time and was not an accurate representation of the actual surgical time because it was not based on chest closure. The average surgical time for the study population was 4 hours 48 minutes and the difference between the delirium and no delirium group was minimal.   89 The delirium group had a slightly shorter average surgical time at 4 hours 44 minutes compared with 4 hours 51 minutes for the no delirium group. All patients included in this study underwent cardiac surgery with CPB, which is strongly linked to delirium because of its effects on neurotransmitters and catecholamines (Van der Mast et al., 2005), as well as intra-operative cerebral embolization (Sockalingham et al., 2005). The bypass pump time for the sample population was 130.7 minutes and the delirium group had a slightly longer duration of CPB time at 133.6 minutes compared with 128.5 minutes for the no delirium group, which was neither statistically significant nor clinically relevant in this study. Hypothermia, pulmonary artery catheters, and bleeding. Hypothermia, the use of a pulmonary artery catheter (PA catheter) and intra-operative bleeding where patients required the administration of blood products were the three intra- operative risk factors analyzed to determine the frequency in which they occurred within the study population and how these results differed between the delirium and no delirium groups. Hypothermia has been linked to delirium because of lowered cerebral oxygenation due to increased catecholamines, increased shivering and a left shift in the oxyhemoglobin dissociation curve (Grimm et al., 2000). In the study sample, 79 patients (63.7%) fit the criteria for intra- operative hypothermia with 40 (74%) of the delirium group exposed to this risk factor and 39(57%) of the no delirium group exposed. The use of a pulmonary artery (PA) catheter was found to be a risk factor for delirium in Santos et al.‟s (2004) study. This is not a common risk factor identified by delirium researchers however it was added into the risk factor analysis because PA catheters are commonly used in most cardiac surgery patients as a means of evaluating post-operative cardiac output and titrating inotropes to achieve satisfactory cardiac output. Overall, 88.7% (110) of the study sample had a   90 PA catheter. This risk factor was found to be equally prevalent in both the delirium and no delirium groups (91% vs. 88%), and therefore was not likely a risk factor for delirium.  The last intra-operative risk factor evaluated was blood loss during surgery that required blood product administration. Blood loss requiring transfusion is a risk factor for delirium because it is often associated with hypoperfusion and insufficient delivery of oxygen to the brain (Norkeine et al., 2007). In this study, 19.4% (24) of the sample experienced intra-operative bleeding and required the administration of blood products. The delirium group and the no delirium group had an equal percentage of patients who experienced  intra-operative bleeding requiring administration of blood products .  In evaluating the intra-operative risk factors, two of the study participants were found to have had none of these risk factors, 59% (n = 73) of the sample were exposed to 2-3 risk factors, 29% were exposed to 4-5 risk factors, and 1.6% (n = 2) participants were exposed to all six risk factors. Overall, the delirium group had a greater exposure to intra-operative hypothermia compared with the no delirium group. Post-operative Risk Factors and Adverse Events The following section discusses the post-operative risk factors of hypotension, respiratory acidosis or alkalosis, need for inotropic medications, low ejection fraction, and atrial fibrillation. It is important to mention that several adverse events occurred to some of the study participants post-operatively: seizures (2), a new diagnosis of myasthenia gravis (1), transient ischemic attack (1), and death (1). These are not common post-operative events but it was felt that they should be mentioned.   91 Hypotension, inotropes, and low ejection fractions as risks. Hypotension in the post-operative period has been reported as a risk factor for delirium because low blood pressure often results in inadequate organ and tissue perfusion, which jeopardizes cerebral oxygenation (Giltay et al., 2006). Hypotension during the post-operative period was determined on the basis of whether a patient received the vasoconstricting agents of norephinephrine or vasopressin. In the study, 92 participants (74.2%) required these agents to treat hypotension. The delirium group and no delirium group had the same frequency of hypotension at approximately 75%%.  Both the need for inotropes and a low ejection fraction are two related risk factors that have been shown to be a post-operative risk factor delirium. Low ejection fractions result in reduced cardiac output and a reduction in organ perfusion (Norkeine, et al., 2007; Bucerius et al., 2004). The participants‟ ejection fractions were evaluated from the first cardiac output measurement from the pulmonary artery catheter, and were considered problematic if the cardiac index was less than 2.2 (L/min/square metre). In this study, 13.7% (n = 17) of the participants had a recorded cardiac index of less than 2.2. When comparing the delirium and no delirium groups,  both groups had equal amounts of patients (28%) with a cardiac index measurement of less than 2.2.  The use of inotropes in the post-operative period is also a potential risk factor and participants were considered to have been exposed if they required any of the three inotropic agents: milrinone, dopamine, or dobutamine. Overall, 27.4% (n= 34) of the participants required intravenous infusions of these inotropic agents alone or in combination; 28% (n = 15) of the delirium group and 28% (n = 19) of participants in the no delirium group required inotropes in response to low cardiac output readings showing that there was no difference in this risk factor between groups.   92 Risks of atrial fibrillation and respiratory acidosis / alkalosis. Atrial fibrillation is a risk factor for delirium that may be present in both the pre-operative and post-operative phases. Patients who undergo cardiac surgery are at risk for atrial fibrillation following surgery because of the irritability to the atria caused by factors such as altered electrolytes, and encephalopathy (Chang et al., 2008). Atrial fibrillation during the post-operative period presents as a risk for delirium because of its association with thromboembolization (Bucerius et al., 2004) and because it potentially alters cardiac output and therefore organ perfusion (Chang et al., 2008; Santos et al., 2004). In this study, 17.7% (n = 22) of the study participants experienced atrial fibrillation (after surgery) with 22% (n = 12) of the delirium group and 13% (n = 14) of the no delirium group experiencing this cardiac arrhythmia.  Respiratory acidosis or alkalosis was also a risk factor for delirium that was examined because respiratory alterations such as hypercarbia and hypoxia cause a decrease in oxygen to major organs such as the brain (Norkeine et al., 2007). Respiratory acidosis or alkalosis was determined by evaluating the first arterial blood gas drawn within 20 minutes of the patient arriving in the CSICU. In this study, 35.5% (n = 44) of the participants experienced respiratory acidosis or alkalosis (43% (n = 23) in the delirium group vs. 29% (n = 20) in the no delirium group.  The overall post-operative risk factor scores showed that 9.7% (n = 12) of the sample did not have any of the post-operative risk factors, 72.6% (n = 90) had 1-2 post-operative risk factors, and 17.7% (n = 21) had 3-4 risk factors. Hypotension was the most frequently occurring risk factor, which occurred equally in the two study groups. The use of inotropes was also found in approximately one third of the sample, and was equal between groups. The major difference between the two delirium groups was that atrial fibrillation was more prevalent in the delirium group, as was respiratory acidosis and alkalosis which was found in approximately one third of the study sample and was also more prevalent in the delirium group.   93  Overall risk factor profile. All of the study participants had a minimum risk factor score of 2 (the maximum risk factor score was 12, and 7 was the modal number). When comparing the delirium and no delirium groups, there was a significant difference in their risk factor scores; the delirium group had more patients exposed to a higher number of risk factors in all phases of surgery. Overall, the delirium group had a higher frequency of 12 of the 19 risk factors listed in Table 17 (Chapter 4, page 65). This is an important factor to consider since there are numerous various that could potentially confound the analysis and results. Prevalence of Post-operative Delirium  Assessments with the CAM-ICU delirium tool detected delirium in 27 participants (n = 122), a prevalence rate of 22.1%. These findings are consistent with other studies that have reported a wide range of prevalence rates of post-operative delirium in cardiac surgery patients (i.e., 28% – 83%) (Idemoto & Kresevic , 2007). Of the 122 patients studied, 77.9% did not display features that would classify them as suffering from delirium according to the CAM-ICU delirium assessment at any of the three assessment intervals.  Despite having negative CAM-ICU delirium assessments, many of the study participants exhibited signs or behaviours that were consistent with delirium. For example, some had been given loxapine or quetipine (two antipsychotic medications used to treat delirium) or the physicians‟ and nurses‟ progress notes described agitation, confusion, altered thinking, hallucinations or paranoia. These observations raised questions about the CAM-ICU and whether a five-minute visit by a research study nurse was able to detect delirium in some participants. This inconsistency jeopardizes the validity of the study findings, although there may be several explanations. First, the participant was assessed for delirium once a day in a five-minute time span, and thus the fluctuating nature of delirium was probably not adequately captured. Further,   94 there appears to be a „sun-downing‟ effect of delirium in which, as the day progresses into the later afternoon, patients begin to exhibit delirium symptoms, even though they may have been fully oriented and symptom free earlier in the day. The CAM-ICU measuring tool was chosen to ensure that the results could be comparable across the sample, because it was developed for use in ICU patients who are non-verbal due to tracheal intubation. Because the first delirium assessment occurred within 12-18 hours of surgery, we believed that this tool would be appropriate for participants who would be weaning from mechanical ventilation.   The clinical assessments recorded in the participants‟ records revealed a greater prevalence rate of delirium 37.9% (n = 47). This clearly demonstrates that the CAM-ICU tool did not effectively identify all of the participants with delirium in the first 72 hours of surgery. In reviewing the agreement between the CAM-ICU and clinical assessments for delirium, 68 cases did not have any indication of delirium. Agreement was also present for 19 cases; both the CAM- ICU and clinical assessments indicated positive results for delirium. The CAM-ICU and clinical assessments differed in 28 cases where a clinical assessment was positive for delirium that the CAM-ICU was negative. In a further 8 cases, the CAM-ICU results were positive and there was no clinical assessment suggestive of delirium in the record.  Positive CAM-ICU scores with no indication of positive clinical assessments may occur when patients suffer from a hypoactive form of delirium. Idemoto and Kresevic (2007) suggested that hypoactive delirium differs in its presentation; patients with it present with decreased activity, alertness and depression like symptoms. The hyperactive form is more easily distinguishable with features such as agitation, restlessness, hypervigilance and psychomotor reactivity. It is possible that the CAM-ICU tool sufficiently detected the subtle symptoms of hypoactive delirium, which were not identified by the nursing staff or physicians.  The conservative prevalence estimate (treating the patients with positive CAM-ICU scores and negative clinical assessments as not having delirium) of delirium was 37.7% (46   95 cases). A more liberal estimate that treated the eight participants with positive CAM-ICU scores and negative clinical assessments as delirious produced a prevalence rate of 44.3% (54 positive cases). Therefore the prevalence rate of post-operative delirium found in this study was 37.7% - 44.3%. This range is similar to what is found in the literature, and makes evident that delirium is a significant problem. The prevalence of delirium and the presence of risk factors reinforce the importance of creating standards and protocols that incorporate delirium screening to ensure that patients are identified and treated in a timely manner. It is also imperative that nurses and physicians are made aware of the different presentations of hypo- and hyper-active delirium and that thorough assessments need to be ongoing. Administration of Opioid Analgesia to Cardiac Surgery Patients  A principal research question that guided this study was whether a relationship exists between the development of post-operative delirium and the dosage and frequency of opioid analgesia administered PRN to cardiac surgery patients. The administration of opioid analgesia in the cardiac surgery intensive care, which served as the setting for this study, is based on standard PRN orders for hydromorphone. These administration practices are influenced by various factors including the knowledge and understanding of pain, and the assessment skills of the nurses. Through pre-printed orders, the cardiac surgery patients included in this study were administered hydromorphone intravenously to control their post-operative pain. The dosages ranged from 0.1 – 0.6 mg of hydromorphone. Occasionally, the patients may have also received meperidine for post-operative shivering (an effect of anaesthetic), which is treated to reduce myocardial demand. That is, although meperidine is in the opioid family, it is not given for its analgesic properties. Nonetheless, its use was included in the opioid calculation. During the study there were also instances wherein the patients received bolus doses of fentanyl and   96 morphine; they were also included in calculation of opioid exposure. The patients also receive oral forms of hydromorphone and oxycodone once extubation had occurred.   When comparing the morphine equivalent dosages of the patients in the six time periods of the study, the research team was able to describe the PRN opioid administration practices of the patients‟ nurses. From these data it is apparent that there was a wide variance of opioid administration in all 12 post-operative time periods, which, to some extent, would be expected because opioid dosages are based on patients‟ responses, the time of the last dose, the ability of patients to display signs of pain, and the knowledge and comfort levels of the nurses who administered these agents (Ndosi & Newell, 2008). Some patients received large amounts of opioids, whereas others received very little (the median amount of morphine equivalent analgesia that the patients received were 15.3 mg in the first 12 hours following surgery; the lowest dosage was 2.5 mg and the highest was 59.8 mg). It is concerning that in the first 12 hours following surgery these patients appeared to receive approximately 1.3 mg of opioid per hour, which is a rather small amount and which may not have met their pain needs. This occurred at a time when the patients were in the highest level of sedation and were not able to verbally communicate their pain needs - they may have suffered from uncontrolled post-operative pain.  The patients may have received large dosages of analgesia from the anaesthesiology staff before leaving the operating room, which may have sustained their pain needs for the first few hours. However, intravenous forms of opioids have a significantly shorter half-life than other routes. Although it is impossible to determine the effectiveness of opioids without patients‟ reports of pain, the dosages of opioid given, on average, was fairly small in comparison with the amount of intramuscular morphine once given to general surgery patients (usually given in dosages of 10 mg every 4 hours). It was in the second 12-hour interval after surgery that the patients were given the largest dosages of opioid, on average (22.6 mg or approximately 1.8mg per hour). The significance of this time period related to analgesia administration practices may   97 be that: (a) patients are usually sufficiently awake to verbalize the presence of pain either independently or when assessed and the nursing staff are then more likely to increase their opioid administration and (b) patients who are awake, extubated and in pain are likely to receive oral forms of hydromorphone, which are provided in dosages of 2, 3, or 4 mg. These standard tablet doses leave less flexibility in a nurses‟ determination of an appropriate dose, these standard dosages may lead to larger amounts of total opioid administered.  In the second time interval, 12-24 hours following surgery, although the patients received a larger dosage, on average, some patients received no opioids (i.e., morphine equivalents of 0 mg). Although patients recovering from cardiac surgery are given regular doses of acetaminophen and some also receive regular non-steroidal anti-inflammatories (depending on whether there is any bleeding and patients‟ kidney function measured by serum creatinine), which will provide some pain control, it is still surprising and concerning that some patients did not receive any opioid analgesia. We are left to wonder whether these patients had been adequately assessed for the presence of pain because we know that many patients do not report pain until they are asked (McCaffrey et al., 2000). The amount of analgesia given in the third time interval (i.e., 24-36 hours after surgery) fell to a median dosage of 13.9 mg of morphine equivalent, which may be explained by the use of oxycodone. Although hydromorphone (IV and oral forms) is used in the CSICU, oral oxycodone (both instant and sustained release) is administered to patients on the cardiac surgery ward. Again this time period is of interest, however, because some patients did not receive any opioid analgesia although they very likely had pain so soon after heart surgery.  The final three time periods of study (36-48 hours, 48-60 hours, and 60 -72 hours following surgery) presented further declines in the amount of analgesia administered. It is expected that patients‟ pain will decrease in the recovery period. As in the previous two time periods, the minimum amount of analgesia administered was 0 mg, which may indicate that oral   98 acetaminophen and non-steroidal anti-inflammatory agents adequately controlled the patients‟ pain. The important factor to consider is that the patients may have actually experienced more pain during this time, however, because of their increased mobility and increased use of limbs with incisions from radial artery or saphenous vein harvesting. We are left to question whether the nurses believed that their patients should have had less pain in these latter time periods and whether holding this belief affected their pain assessments and analgesia administration practices. The variance in the amount of opioids administered may be explained, in part, by the different pain control needs of the patients. They require further attention, however, because it may also be indicative of deficits in the nurses‟ knowledge regarding decision making related to the appropriate frequency and dosage of drug administration, and of the need for standard approaches to pain assessment after cardiac surgery. Delirium and Opioid Exposure  The question for this study was whether a relationship exists between opioid dosing and the development of post-operative delirium. The amount of analgesia given before the time of delirium diagnosis suggests that patients who received relatively small amounts of opioid analgesia had higher rates of delirium within the first 24 hours following surgery. Patients who received larger amounts of opioid analgesia experienced delirium in the latter part of the 72 hours of the study. This information suggests that small amounts of opioid analgesia, and thus uncontrolled post-operative pain, may result in early onset delirium and patients that receive larger amounts of opioid analgesia, and who possibly have their pain controlled, experience later onset delirium, which may arise because of other factors.  Although we identified potential problems in the current practices of PRN administration of opioid analgesia in these cardiac surgery patients, we were not able to detect a statistically significant relationship between the amount of opioid analgesia administered to the patients and   99 the development of post-operative delirium. This may be the result of an inadequate sample size or that opioid dosage does not significantly influence the development of post-operative delirium in cardiac surgery patients. Midazolam Administration The patients in this study received an average of 4.8 mg of Midazolam during the observation period (median dose of 3.0 mg). There was wide variation in the amount given, with a positive skew in the distribution of dosages; 24.6% of the study participants did not receive Midazolam (i.e., 0 mg). Some participants received relatively large doses of Midazolam (the maximum dose was 83 mg). The variation in PRN Midazolam administered, can only be explained by different nurses‟ practices related to the drug. As mentioned above, cardiac surgery patients, in the initial post-operative period, experience fluctuations in organ functioning which affects their ventilation, haemodynamics, organ perfusion, electrolyte balance, hemoglobin levels, and coagulation. The CSICU has acceptable parameters for these values that determine the initiation of weaning from mechanical ventilation. Patients awaken from anaesthesia at various times, which may not be conducive to their recovery if they are experiencing problems in organ functioning, and a higher level of consciousness may further complicate their recovery status. Therefore, the intent to sedate patients with Midazolam is based on prioritizing their physiological needs and reaching the goal of stability before allowing them to awaken and be extubated from mechanical ventilation. In a sense, Midazolam allows the nursing staff to control the level of consciousness of their patients so as not to jeopardize the recovery of the heart and other organ systems after surgery. This may explain why there is wide variation in the amounts of Midazolam administered to these patients.  Another reason for the variation in the amount of Midazolam administered may not be based on the physiological status of the patients, but rather on their response to awakening from   100 surgery. Some patients cross over from the anaesthetic state to a higher level of consciousness rapidly and awaken suddenly; they find themselves in a strange environment with an endotracheal tube and other invasive equipment in place. Therefore, it is not uncommon for many patients to experience fear upon awakening, which evokes a stress response that is often accompanied by agitation. It is in these instances that sedation, along with verbal reassurances from nurses are used to calm patients and aid their adjustment to the recovery situation. In this instance, it can be argued that sedation is used to control patients‟ behaviour if it is not conducive to recovery. These factors may also explain the wide variation in Midazolam administration observed here.  Midazolam administration only occurs in the CSICU and is not used on the cardiac surgery ward because of the risk of inducing sedation that would affect patients‟ respiratory drive and compromise their airways. Therefore, the variation in Midazolam administration is also affected by the time spent mechanically ventilated and in the CSICU. Patients who recover rapidly from cardiac surgery and anaesthesia and who are transferred to the cardiac surgery ward promptly on the first post-operative day are less likely to receive large dosages of Midazolam. Delirium and Midazolam exposure.  A larger percentage of patients in the no delirium group received no Midazolam in the study period compared with the group that had delirium. Although the no delirium group had a larger percentage of patients who received between 1.0 mg and 4.0 mg of Midazolam then the delirium group, there was a larger percentage of patients in the delirium group who received 4.1 mg of Midazolam or more. There are at least two interpretations for these findings. Midazolam may have been used to control behaviour that presented as restlessness and agitation as would have been seen in the delirium group. Alternatively, it may be the case that there is a relationship   101 between the amount of Midazolam administered to post-operative cardiac surgery patients and the risk of developing post-operative delirium.  The total amount of Midazolam administered and the time at which delirium was first detected or noted in the patients revealed that delirium was diagnosed throughout the 72-hour time period of the study, and that the total amount of Midazolam received by patients was not a result of accumulation over time (that is, large amounts were given in relatively short periods of time). It was evident that delirium in these patients occurred in conjunction with exposure to Midazolam. This may have resulted because these patients were being sedated with Midazolam, were not being given analgesia, and were not achieving adequate pain control. The sedation would not allow the patients to express their pain control needs to their nurses.  We found a statistically significant relationship between the amount of Midazolam received and the presence of delirium. This leads us to conclude that the PRN administration of Midazolam is a significant post-operative and modifiable risk factor for delirium. With this in mind, nursing practice related to the PRN administration of Midazolam in the CSICU may need to be altered and nurses may need to modify their sedation practices to reduce the prevalence of, if not prevent, delirium after cardiac surgery. Implications for Nursing Care The findings of this study reveal that nurses providing care in the cardiac surgery intensive care unit play a major role in the management of sedation and pain control. The PRN analgesia and sedation orders they follow allow them to determine the dosage, frequency, and reason for administration. With this professional autonomy there is also a greater responsibility to use critical thinking and decision-making skills to meet the pain control and sedation needs of cardiac surgery patients and to identify a patient‟s risk for delirium. The results of this study lead us to emphasize the importance of nurses recognizing the multi-factorial nature of delirium and   102 decision-making; analgesia administration practices should not be based solely on age or body size, since this potentiates the risk of patients suffering unnecessarily with uncontrolled pain. An emerging theory that has received limited attention, to date, identifies a potential link between uncontrolled pain and the development of post-operative delirium (Robinson et al., 2008) and requires further examination beyond this study.  Regardless of the presence of delirium it is evident that uncontrolled post-operative pain continues to be a problem experienced by patients. Research has shown that the factors of education, adequate training, and personal opinion are those most likely to influence opioid dosages (Idemoto et al., 2007; McCaffrey et al., 2000). An underestimation of pain by nurses in the CSICU may lead to the administration of ineffective dosages of analgesia, and may contribute to higher levels of pain experienced by patients, predisposing them to delirium. It is apparent that the critical nature of the first several hours following cardiac surgery may easily overshadow the concept of pain control, as seen by the relatively low amounts of opioid analgesia administered to patients in this study. This is where knowledge and experience are crucial to the decision-making process involved in treating patients‟ pain with analgesia ordered as PRN.  The administration of PRN Midazolam in the CSICU should involve accurate assessments and goals for sedation. When nurses care for agitated, restless, and uncomfortable patients, Walker and Gillen (2006) stated, PRN sedation may be used extensively to deal with the behaviour. This may be a factor that determines Midazolam administration in the CSICU and can become problematic if the agents are given simply to control undesired patient behaviour and thorough assessments and treatment of the underlying causes are ignored. Administering sedation ad lib to cardiac surgery patients in the ICU without consideration of the risk factors and potential causes of the presenting problem may result in immediate but short-lived cessation of   103 symptoms and behaviour, however adverse reactions and longer duration or worsening of symptoms may result.   PRN analgesia and sedatives are common means of treating pain and providing sedation in the cardiac surgery intensive care unit. Although they assist nurses in providing immediate care to their patients, pain and sedation assessment is necessary, as well as knowledge of the pharmacokinetics and pharmacodynamics of the drugs being administered to provide safe care. Adequate knowledge of these agents, as well as having patient-specific goals with an individualized approach that considers patients‟ risk factors for delirium, and knowing how pain, sedation and analgesia are linked to the development of delirium can serve to minimize adverse events.  Clinical decision-making skills are used by cardiac surgery nurses to plan, prioritize and determine a course of action that depends on the clinical presentation and assessment of their patients. It is imperative that nurses recognize that clinical decision making is an aspect of nursing care that is included in every interaction between nurses and patients. When examining the cardiac surgery intensive care unit, empirical knowledge, experience and expertise are all factors that contribute to this decision-making process (Guttormson et al., 2010; Wier & O‟Neill, 2008). The decisions that nurses make related to patient care in the cardiac surgery ICU consist of determining when and how much sedation and opioid analgesia to provide following cardiac surgery. In this context, we now have a better understanding of how these decisions made by nurses when administering opioids and benzodiazepines potentially contribute to the development of post-operative delirium. Bourne (2008) suggested that improvements in the use of both sedative and analgesic agents have a role to play in the prevention of delirium in an intensive care setting and evidenced-based practice must be incorporated in the decision-making process.   104  Decisions in the cardiac surgery ICU related to pain control and sedation are important to examine because nurses must make informed decisions about their patients‟ needs, PRN analgesia requirements, and potential risks for the development of delirium. Literature about patients‟ experiences with pain continues to report that it is not well controlled in surgical settings in spite of the use of adjunctive therapies, a variety of opioids available, and advances in less invasive surgical techniques (Ferguson et al., 1997). As well, it is unknown whether the risk factors for post-operative delirium in cardiac surgery patients are considered when nurses make their decisions of how much and how often they will administer analgesia and sedation.  In an ICU environment, factors that affect decisions about sedation and opioid analgesia administration consist of personal and professional judgment. If nurses are unsure of specific and individualized patient goals and their decision making is not guided by a clear end-point, inconsistencies in practice may occur. Awake intubated patients may experience fear and anxiety related to the endotracheal tube or other invasive equipment present, and without support, reassurance and communication from the bedside nurse, patients may become increasingly agitated and restless, which in turn can affect their haemodynamics and oxygenation (Siegel, 2003) and can lead to decisions to administer larger than necessary dosages of sedation. Walker et al. (2006) stated that the ability of bedside nurses to balance the risks and benefits of sedation influences their use of alternative methods to address sedation-related issues, and is an important aspect of clinical decision making.  With the many different approaches to PRN administration and the multitude of factors to consider regarding the use of sedation and analgesia, one can clearly see how inconsistencies develop in sedation practices and how decision-making skills differ according to practitioner skill and experience levels. Benzodiazepines and uncontrolled pain may increase a patient‟s risk for delirium, and consideration of all risk factors must be a part of the decision-making process concerning PRN administration practices. Although both patient comfort and time to extubation   105 are important factors to consider, the decisions made by nurses regarding the frequency and dosage of analgesia and sedation in the cardiac surgery ICU must consist of empirical knowledge, accurate assessment data and clear rationale with consideration of how these actions may be risk factors for the development of post-operative delirium. Study Limitations A major limitation of this study relates to the sample size. Although 138 patients were recruited and initially met the study criteria, only 122 participants had delirium assessments and data collected. The statistical power of this study was limited because of the small sample size which may have led to important clinical issues not being detected.  Another limitation of the study was that the inter-rater reliability of the four researchers that conducted the delirium assessments was not determined. Without an analysis of inter-rater reliability it is uncertain whether the delirium assessments were conducted in a consistent and uniform manner. The sedation rating on the RASS is fairly straightforward; however, it is open to some interpretation by the person conducting the assessment. An analysis of inter-rater reliability would have strengthened the credibility of the results of this study.  Another limitation of the study was the use of the CAM-ICU tool in both the intensive care unit and the surgical ward. The CAM-ICU is an adapted tool from the initial Confusion Assessment Method and is meant for use in environments where patients are non-verbal due to mechanical ventilation. Although the tool was necessary for use when the participants were in the CSICU, the original CAM may have been more appropriate once the patients were transferred to the surgical floor. The original CAM has more questions and possibly allows for a more thorough assessment of delirium. The CAM-ICU tool was chosen for assessment in both areas to ensure ease and accuracy of tool administration.   106  Finally, a further limitation of the study refers to the calculation of the opioid analgesia totals. Although this study was designed to reflect nursing care related to analgesia administration, the amount of opioid analgesia the participants received in the operating room, administered by the anaesthetic team, was not taken into account and is a factor when considering the sedation levels and analgesia needs within the patients‟ first 12-hours following surgery. Opioid dosages received in the operating room were likely to have affected the amount of opioid received in the CSICU and may have explained the relatively small amounts of drug some participants received in the first 12 hours of the post-operative period.                     107 Chapter Six: Conclusion   Patients undergoing cardiac surgery rely on the knowledge, experience, and decision- making skills of their nurses and physicians to ensure that safe, competent and adequate care addresses all of their physiological needs in the hours following surgery. This responsibility involves ensuring not only that issues pertaining to post-operative bleeding, heart rate and rhythm, and hemodynamic disturbances are effectively managed and treated, but also that the nursing care provided to patients does not fail to neglect the consideration of comfort and prevention of adverse events such as post-operative delirium. Pain and prolonged ventilation can lead to the development of pneumonia, atrial fibrillation, agitation and delirium and can have significant prolonged effects on patients‟ health. Delirium in the context of cardiac surgery is a complex phenomenon, but nurses and physicians need to examine current practices related to sedation and analgesia administration to determine whether patients are being unnecessarily exposed to risk factors for delirium. With what is known about the potential relationship between post-operative pain, sedation and delirium, it is also imperative that pre-operative, intra-operative and post-operative risk factors are considered when nurses are given the autonomy to use their decision-making skills in the PRN administration of both analgesia and sedation.  The research questions regarding whether a relationship exists between the PRN administration of Midazolam and opioids and post-operative delirium in the cardiac surgery setting were posed. What we can take from this study is that awareness and education of risk factors for delirium as well as the role that nursing care and practice have in the prevention and treatment of post-operative delirium. This study brings forth several key concepts related to pain control, sedation practices, the need for assessment protocols for sedation as well as pain, and delirium that nurses, physicians and patients can benefit from. Although the CSICU was the backdrop for this particular study, the problems and deficiencies in care that were seen here are   108 not unique to this setting or unit, but appear in many different areas of acute care. This research highlights the need for further investigations regarding how the pain of patients is perceived by nurses and physicians and how this translates into practice decisions, as well as the effectiveness of initiating standardized assessments for pain, sedation and delirium and how this may affect patient outcome. More work is needed to enhance our understanding of delirium with the development of strategies to address risk factors that are modifiable such that prevention can be achieved.          109 References  American Psychiatric Association (1997). Diagnostic and statistical manual of mental  disorders. 4 th  ed. Washington (DC): American Psychiatric Association APA  Press.  Berger, I., & Waldhorn, R.E. (1995). Analgesia, sedation and paralysis in the intensive care unit. American Family Physician, 51(1), 166-172.  Bolinder, G., Noren, A., de Faire, U., & Wahren, J. (1997). Smokeless tobacco use and  atherosclerosis and ultrasonographic investigation of carotid intima media  thickness in health middle-aged men.  Atherosclerosis, 132, 95-103.  Bourne, R.S. (2008). Delirium and use of sedation agents in intensive care.  Nursing in Critical  Care, 13(4), 195-202.  Burkhart, C.S., Dell-Kuster, S., Gamberini, M., Moeckli, A., Grapow, M., Filipovic, M.,  Seeberger, M.D., Monsch, A.U., Strebel, S.P., & Steiner, L.A.(2010). Modifiable  and nonmodifiable risk factors for post-operative delirium after cardiac surgery  with cardiopulmonary bypass. Journal of Cardiothoracic and Vascular  Anesthesia,24(4), 555-559.  doi:10.1053/j.jvca.2010.01.003  Bucerius, J., Gummert, J.F., Borger, M.A., Walther, T., Doll, N., Falk, V., Schmitt, D.V.,  & Mohr, F.W. (2004). Predictors of delirium after cardiac surgery delirium:  Effect of beating-heart (off-pump) surgery. The Journal of Thoracic and  Cardiovascular Surgery,127(1), 57-64.  doi:10.1016/S0022-5223(03)01281-9  Burns, A., Gallagley, A., & Byrne, J. (2004). Delirium. Journal of Neurology, Neurosurgery & Psychiatry, 75, 362-367. doi:10.1136/jnnp.2003.023366  Chang, Y, Tsai, Y., Lin, P., Chen, M., & Liu, C. (2008). Prevalence and risk factors for post-  operative delirium in a cardiovascular intensive care unit. American Journal of Critical  Care, 17(6), 567-575.  Chuk, P.K-C. (1999). Vital signs and nurses‟ choice of titrated dosages of intravenous  morphine for relieving pain following cardiac surgery. Journal of Advanced  Nursing, 30(4), 858-865.  Clark, R.E., Brillman, J., Davis, D.A., Lovell, M.R., Price, T.R., & Magovern, G.J.  (1995). Microemboli during coronary artery bypass grafting: genesis and effect on  outcome. Journal of Thoracic and Cardiovascular Surgery, 109, 249-257.      110 Currey, J., Browne, J., & Botti, M. (2006). Haemodynamic instability after cardiac  surgery: nurses‟ perceptions of clinical decision-making. Journal of Clinical  Nursing, 15, 1081-1089.  doi:10.111/j.1365-2702.2006.01392.x  Devlin, J.W., Fong, J.J., Howard, E.P., Skrobik, Y., McCoy, N., Yasuda, C., & Marshall,  J. (2008). Assessment of delirium in the intensive care unit: nursing practices and perceptions. American Journal of Critical Care, 17(6), 555-566.  Egerod, I. (2002). Uncertain terms of sedation in ICU. How nurses and physicians  manage and describe sedation for mechanically ventilated patients.  Journal of Clinical Nursing, 11, 831-840.  Ely, E.W., Inouye, S.K., Bernard, G.R., Gordon, S., Francis, J., May, L., Truman,  S., Speroff, T., Gautam, S., Margolin, R., Hart, R.P., & Dittus, R. (2001). Delirium in mechanically ventilated patients: Validity and reliability of the confusion assessment method for the intensive care unit (CAM-ICU). JAMA, 286(21), 2703-2710. doi:10.1001/jama.286.21.2703  Ely, E.W., & Truman Pun, B. (2002). The Confusion Assessment Method for the  ICU (CAM-ICU): Training Manual. Vanderbilt University.  Ely, W.E., Truman, B., Shintani, A., Thomason, J.W.W., Wheeler, A.P., Gordon, S.,Francis, J., Speroff, T., Gautam, S., Margolin, R., Sessler, C.N., Dittus, R.S., & Bernard, G.R. (2003). Monitoring sedation status over time in ICU patients:  Reliability and validity of the Richmond Agitation – Sedation Scale (RASS).  JAMA, 289,(22), 2983-2991.  Ferguson, J., Gilroy, D., & Puntillo, K. (1997). Dimensions of pain and analgesia  administration associated with coronary artery bypass grafting in an Austrailian  intensive care unit. Journal of Advanced Nursing, 26, 1065-1072.  Folstein, M.F., Folstein, S.E. & McHugh, P.R. (1975). “Mini-Mental State” A practical  method for grading the cognitive state of patients for the clinician. Journal of  Psychiatric Research, 12, 189-198.  Fong, H.K., Sands, L.P., & Leung, J.M. (2006). The role of post-operative analgesia in  Delirium and cognitive decline in elderly patients: a systematic review.  Anaesthesia Analgesia, 102, 1255-1266.  doi:10.1213/01.ane.0000198602.29716.53  Geffen, J., Cameron, A., Sorensen, L., Stokes, J., Roberts, M.S., & Geffen, L. (2002). Pro  re nata medication for psychoses: the knowledge and beliefs of doctors and nurses. Australian and New Zealand Journal of Psychiatry, 36, 642-648.  Gelinas, C. (2007). Management of pain in cardiac surgery ICU patients: Have we  improved over time?  Intensive and Critical Care Nursing, 23, 298-303.  doi:10.1016/j.iccn.2007.03.002    111 Giltay, E.J., Huijskes, V.H.P., Kho, K.H., Blansjaar, B.A., & Rosseel, P.M.J.(2006).  Psychotic symptoms in patients undergoing coronary artery bypass grafting and  heart valve operation. European Journal of Cardio-Thoracic Surgery, 30, 140-  147.  doi:10.1016/j.ejcts.2006.03.056  Gao, L., Taha, R., Gauvin, D., Othmen, L.B., Wang, Y., & Blaise, G. (2005). Post-  operative cognitive dysfunction after cardiac surgery. Chest, 128(5), 3664 – 3670.  Grimm, M., Czerny, M., Baumer, H., Kilo, J., Mald, C., Kramer, L., Rajek, A., &  Wolner, E. (2000). Noromthermic cardiopulmonary bypass is beneficial for  cognitive brain function after coronary artery bypass grafting – a prospective  randomized trial. European Journal of Cardio-thoracic Surgery, 18, 270-275.  Guttormson, J.L., Chlan, L., Weinert, C., & Savik, K. (2010). Factors influencing nurse  sedation practices with mechanically ventilated patients: A US national survey.  Intensive and Critical Care Nursing, 26, 44-50. doi:10.1016/j.iccn.2009.10.004  Herrmann, M., Ebert, A.D., Tober, D., Hann, J., & Huth, C. (1999). A contrast analysis  of release patterns of biochemical markers of brain damage after coronary artery  bypass grafting and valve replacement and their association with the  neurobehavioral outcome after cardiac surgery. European Journal of Cardio-  Thoracic Surgery, 16, 513-518.  Idemoto, B.K., & Kresevic, D.M. (2007). Emerging nurse-sensitive outcomes and  evidence-based practice in post-operative cardiac patients. Critical Care Nursing Clinics of North America, 19, 371-384. doi:10.1016/j.ccell.2007.07.005  Inouye, S.K., Foreman, M.D., Mion, L.C., Katz, K.H., & Cooney, L.M.(2001). Nurses‟  recognition of delirium and its symptoms. Archives of Internal Medicine, 161,  2467-2473.  Jacobi, J., Fraser, G.L., Coursin, D.B., Riker, R.R., Fontaine, D., Wittbrodt, E.T., Chalfin,  D.B., Masica, M.F., Bjerke, S., Coplin, W.M., Crippen, D.W., Fuchs, B.D., Kelleher, R.M., Marik, P.E., Nasraway, S.A., Murray, M.J., Peruzzi, W.T., & Lumb, P.D. (2002). Clinical practice guidelines for the sustained use of sedatives and analgesics in the critically ill adult. Critical Care Medicine, 30(1), 119-141.  Kazmierski, J., Kowman, M., Banach, M., Fendler, W., Okonski, P., Banys, A.,  Jaszewski, R., Sobow, T., & Kloszewska, I. (2008). Clinical utility and use of  DSM-IV and ICD-10 criteria and the memorial delirium assessment scale in  establishing a diagnosis of delirium after cardiac surgery. Psychosomatics, 49(1),  73-76.  Kurlowicz, L., & Wallace, M.(1999). The mini-mental state examination (MMSE).  Journal of Gerontological Nursing, 25(5), 8-9.    112 Laferlita, BW, Klein, RR, Choi, PT, Serfontein, L. Delirium after coronary artery bypass  graft surgery: Retrospective analysis (unpublished data) 2005.  Larsson, C., Axell, A.G., & Ersson, A. (2007). Confusion assessment method for the  intensive care unit (CAM-ICU): translation, retranslation and validation into  Swedish intensive care settings. Acta Anaesthesiology Scandanavia, 51, 888-892.  Marcantonio, E.R., Goldman, L., Mangione, C.M., Ludwig, L.E., Muraca, B., Haslauer,  C.M., Donaldson, M.C., Whittermore, A.D., Sugarbaker, D.J., Poss, R., Haas, S.,  Cook, E.F., Oraz, J., & Lee, T.H. (1994). A clinical prediction rule for delirium  after elective noncardiac surgery. JAMA, 217(2), 134-139.  Maxam-Moore, V.A., Wilkie, D.J., & Woods, S.L. (1994). Analgesics for cardiac surgery  patients in critical care: Describing current practice. American Journal of Critical  Care,3(1), 31-39.  McCaffrey, M., Ferrell, B.R., & Pasero, C. (2000). Nurses‟ personal opinions about  patients‟ pain and their effect on recorded assessments and titration of opioid doses. Pain Management Nursing, 1(3), 79-87. doi:10.1053/jpmn.2000.9295  McNicoll, L., Pisani, M.A., Wesley, E., Gifford, D., & Inouye, S.K. (2005). Detection of  delirium in the intensive care unit: Comparison of Confusion Assessment Method  for the Intensive Care Unit with Confusion Assessment Method ratings. Journal  of the American Geriatrics Society, 53,  495-500.  Meagher, D.J.(2001). Regular review: Delirium: Optimising management. British  Medical Journal, 322, 144-149.  doi:10.36/bmj.322.7279.144  Mehta, S., McCullagh, I.M., & Burry, L. (2009). Current sedation practices: Lessons learned  from international surveys. Critical Care Clinics, 25, 471-488.  doi:10.1016/j.ccc.2009.04.001  Miller, R.R., & Ely, E. (2006). Delirium and cognitive dysfunction in the  intensive care unit. Seminars in Respirology and Crtical Care Medicine,27, 210-  220.  Morrison, R.S., Magaziner, J., Gilbert, M., Koval, K.J., McLaughlin, M.A., Orosz, G.,  Strauss, E., & Siu, A.L. (2003). Relationship between pain and opioid analgesics on the development of delirium following hipfracture.  Journal of Gerontological Medical Sciences, 58A, 76-81.  Ndosi, M.E., & Newell, R. (2008). Nurses‟ knowledge of pharmacology behind drugs  they commonly administer. Journal of Clinical Nursing, 18, 570-580.  doi:10.1111/j.1365-2702.2008.02290.x     113 Norkeine, I., Ringaitiene, D., Misiuriene, I., Samalavicius, R., Bubulis, R., Baublys, A.,  & Uzdavinys, G. (2007). Incidence and precipitating factors of delirium after  coronary artery bypass grafting. Scandinavian Cardiovascular Journal, 41,  180185.  doi:10.1080/14017430701302490  Ostermann, M.E., Keenan, S.P., Seiferling, R.A., & Sibbald, W.J. (2000). Sedation in the  intensive care unit. JAMA, 283, 1451-  1459.  Pandharipande, P., Shintani, A., Peterson, J., Truman Pun, B., Wilkinson, G.R., Dittus,  R.S., Bernard, G.R., Ely, E. (2006). Lorazepam is an independent risk  factor for transitioning to delirium in intensive care unit patients. Anesthesiology,  104, 21-26.  Pandharipande, P., Cotton, B.A., Shintani, A., Thompson, J., Truman Pun, B., Morris, J.  A., Diitus, R., & Ely, E. (2008). Prevalence and risk factors for  development of delirium in surgical and trauma intensive care patients. The  Journal of Trauma, 65(1), 34-41.  doi:10.1097/TA.0b013e31814bc2c4d  Puntillo, K., Pasero, C., Li, D., Mularski, R.A., Grap, M.J., Erstad, B.L., Varkey, B.,  Gilbert, H.C., Medina, J., & Sessler, C.N. (2009). Evaluation of pain in ICU.  patients. Chest, 135(4), 1069-1047.  Ramsay, M.A. (2000). Measuring level of sedation in the intensive care unit. JAMA, 284,  441- 442.  Rasmussen, L.S., Steentoft, A., Rasmussen, H., Kristensen, P.A., Moller, J.T., & ISPOCD group. (1999). Benzodiazepines and post-operative cognitive dysfunction in the elderly. British Journal of Anaesthesia, 83(4), 585-589.  Rezvani, A.H., & Levin, E.D. (2001). Cognitive effects of nicotine. Biological  Psychiatry, 49, 258-267.  Roa, S.S., & Cherukuri, M. (2006). Management of hip fracture: The family physician‟s  role. American Academy of Family Physicians, 73(12), 2195-2200.  Robinson, S., Vollmer, C., Jirka, H., Rich, C., Midiri, C., & Bisby, D. (2008). Aging and  delirium: Too much or too little pain medication. Pain Management Nursing, 9(2), 66-72.  doi:10.1016/j.pmn.2007.12.002  Rolfson, D.B., McElhaney, J.E., Jhangri, G.S., & Rockwood, K. (1999). Validity of the  confusion assessment method in detecting postoperative delirium in the elderly.  International Psychogeriatrics, 11(4), 431-438.      114 Rothenhausler, H.B., Grieser, B., Nollert, G., Reichart, B., Schelling, G., & Kapfhammer,  H. (2005). Psychiatric and psychosocial outcome of cardiac surgery with  cardiopulmonary bypass: a prospective 12-month follow-up study. General  Hospital Psychiatry, 27, 18-28.  doi:10.1016/j.genhosppsych.2004.09.001  Rudolph, J.L., Jones, R.N., Levkoff, S.E., Rockett, C., Inouye, S.K., Selke, F.W., Khuri, S.F., Lipsitz, L.A., Ramlawi, B., Levitsky, S., Marcantonio, E.R. (2009). Derivationand validation of a pre-operative prediction rule for delirium after cardiac surgery. Circulation, 119, 229-236. doi:10.1161/CIRCULATIONAHA.108.795260  Rudolph, J.L., Marcantonio, E.R., Culley, D.J., Silverstein, J.H., Rasmussen, L.S.,  Crosby, G.J., & Inouye, S.K. (2008). Delirium is associated with early  post-operative cognitive dysfunction. Anaesthesia, 63, 941-947.  doi:10.1111/j.1365-2044.2008.05523.x  Santos, F.S., Velasco, I.T., & Fraguas, R. (2004). Risk factors for delirium in the elderly after coronary artery bypass graft surgery. International Psychogeriatrics, 16(2), 175-193.  doi:10.1017/S1041610204000365  Segatore, M., Dutkiewicz, M., & Adams, D. (1998). The delirious cardiac surgical patient:  Theoretical aspects and principles of management. Journal of Cardiovascular Nursing,  12(4), 32-48.  Sessler, C.N., Gosnell, M.S., Grap, M.J., Brophy, G.M., O‟Neal, P.V., Keane, K.A., Tesoro, E.P., & Elswick, R.K. (2002). The Richmond agitation-sedation scale: validity and reliability in adult intensive care unit patients. American Journal of Respiratory Critical             Care Medicine, 166, 1338-1344.  Siegel, M.D. (2003). Management of agitation in the intensive care unit. Clinics in Chest  Medicine, 24, 713-725. doi:10.1016/S0272-5231(03)00104-7  Socklingham, S., Parekh, N., Bogoch, I.I., Sun, J., Mahtani, R., Beach, C., Bollegalla, N.,  Turzanski, S., Seto, E., Kim, J., Dulay, P., Scarrow, S., & Bhalerao, S. (2005).  Delirium in  the post-operative cardiac patient: A review. Journal of Cardiac  Surgery, 20, 560-567.  doi:10.1111/j.1540-8191.2005.00134.x  Tan, M.C., Felde, A., Kuskowski, M., Ward, H., Kelly, R.F., Adabag, A.S., & Dysken,  M. (2008). Incidence and predictors of post-cardiotomy delirium. American  Journal of Geriatric Psychiatry, 16(7), 575-583.  Tang-Wai, D.F., Knopman, D.S., Geda, Y.E., Edland, S.E., Smith, G.E., Ivnik, R.J.,  Tangalos, E.G., Boeve, B.F., Petersen, R.C. (2003).  Comparison of the short test  of mental status and the Mini-Mental State Examination in mild cognitive  impairment. Archives in Neurology, 60, 1777-1781. Retrieved December 2, 2008  from http://www.archneurol.com.    115 Van Der Mast, R.C., & Friths, H.J.R. (1996). Delirium after cardiac surgery: A critical review.  Journal of Psychosomatic Research, 41(1), 13-30.  Van Rompaey, B., Shuurmans, M.J., Shortridge-Baggett, L.M., Truijen, S., & Bossaert, L. (2008). Risk factors for intensive care delirium: A systematic review. Intensive and Critical Care Nursing, 24, 98-107. doi:10.1016/j.iccn.2007.08.005  Van Rompaey, B., Elseviers, M.M., Schururmans, M.J., Shortridge-Baggett, L.M.,  Truijen, S., & Bossaert, L. (2009).  Risk factors for delirium in intensive care  patients: a prospective cohort study. Critical Care, 13(3), 1-12.  doi:  10.1186/cc7892  Walker, N., & Gillen, P. (2006). Investigating nurses‟ perceptions of their role in managing  sedation in intensive care: An exploratory study. Intensive and Critical Care Nursing, 22,  338-345. doi:10.1016/j.iccn.2006.03.008  Weinert, C.R., Chlan, L., & Gross, C. (2001). Sedating critically ill patients: factors  affecting nurses‟ delivery of sedative therapy. American Journal of Critical Care,  10(3), 156-165.  Weir, S., & O‟Neill, A. (2008). Experiences of intensive care nurses assessing sedation/agitation in critically ill patients. Nursing in Critical Care, 13(4), 185-194.                  116 Appendices  APPENDIX A: Equianalgesic Dosing (Morphine Equivalents)  EQUIANALGESIC DOSING FOR MANAGEMENT OF ACUTE OR CHRONIC PAIN (Equivalent to morphine 10mg IV/IM/subcutaneous) Drug IV/IM/ subcutaneous (mg) PO/rectal/sublingual (mg) Duration of Action (hours) morphine 10 20 to 30 3 to 4 Codeine 120 200 3 to 4 HYDROmorphone 2 4 3 to 4 meperidine 75 300 3 fentanyl 100 mcg (0.1) - 2 to 3 oxyCOdone - 15 to 20 3 to 4   In round the clock administration of opioids, the IV/IM/subcutaneous routes are essentially equianalgesic   Dosing Frequency is usually based on the duration of action of the analgesic used; certain individuals may obtain control with longer dosing intervals, and some patients need more frequent analgesic administration   Intravenous opioids may have shorter duration of action compared to other routes of administration   Equianalgesic doses and intervals are based on available published data and clinical experience    (Vancouver Coastal Health (VGH) Pharmacy, May 2010)      117 APPENDIX B: VCHRI Ethics Approval Certificate      118 Appendix C: UBC CREB Ethics Approval Certificate   The University of British Columbia Office of Research Services Clinical Research Ethics Board – Room 210, 828 West 10th Avenue, Vancouver, BC V5Z 1L8   ETHICS CERTIFICATE OF EXPEDITED APPROVAL  PRINCIPAL INVESTIGATOR: INSTITUTION / DEPARTMENT: UBC CREB NUMBER: Pamela A. Ratner  UBC/Applied Science/Nursing  H09-00241 INSTITUTION(S) WHERE RESEARCH WILL BE CARRIED OUT: Institution Site Vancouver Coastal Health (VCHRI/VCHA) Vancouver General Hospital Other locations where the research will be conducted: N/A  CO-INVESTIGATOR(S): Paul Galdas Priscilla G. Taipale Carol Jillings SPONSORING AGENCIES: - Vancouver Hospital and Health Sciences Center PROJECT TITLE: The Relationship Between Postoperative Risk Factors for Delirium After Cardiac Surgery and Nursing Care. THE CURRENT UBC CREB APPROVAL FOR THIS STUDY EXPIRES:  February 24, 2010 The UBC Clinical Research Ethics Board Chair or Associate Chair, has reviewed the above described research project, including associated documentation noted below, and finds the research project acceptable on ethical grounds for research involving human subjects and hereby grants approval.  DOCUMENTS INCLUDED IN THIS APPROVAL: APPROVAL DATE: Document Name Version Date Protocol: Research Proposal 3 December 11, 2008 Consent Forms: Consent Form 3 February 9, 2009 Advertisements: Brochure 1 January 1, 2009 Questionnaire, Questionnaire Cover Letter, Tests: Mini Mental State Exam tool 2 February 9, 2009 CAM - ICU tool 2 May 23, 2005 Data Collection tool 2 February 9, 2009  February 24, 2009 CERTIFICATION: In respect of clinical trials: 1. The membership of this Research Ethics Board complies with the membership requirements for Research Ethics Boards defined in Division 5 of the Food and Drug Regulations. 2. The Research Ethics Board carries out its functions in a manner consistent with Good Clinical Practices. 3. This Research Ethics Board has reviewed and approved the clinical trial protocol and informed consent form for the trial which is to be conducted by the qualified investigator named above at the specified clinical trial site. This approval and the views of this Research Ethics Board have been documented in writing.  The documentation included for the above-named project has been reviewed by the UBC CREB, and the research study, as presented in the documentation, was found to be acceptable on ethical grounds for research involving human subjects and was approved by the UBC CREB.  Approval of the Clinical Research Ethics Board by:            Dr. Caron Strahlendorf, Associate Chair     119 Appendix D: Participant Consent Form                                                          UBC CREB Number: H09-00241                 VCHA Reserch Institute #V09-0067          April 1 – October 31, 2009 CONSENT FORM Title of Study: The Relationship Between Postoperative Risk Factors for Delirium after Cardiac Surgery and Nursing Care.  Principal Investigators: Pamela A. Ratner, PhD, RN Professor, School of Nursing, UBC. Priscilla G. Taipale BScN, RN Master of Science in Nursing Student, UBC Co-Investigators: Connie Fernandes BScN, RN, Vancouver General Hospital Cardiac Sciences Deborah Manning, RN, Vancouver General Hospital Cardiac Sciences Jamie Gallaher, RN, BSN, Vancouver General Hospital Emergency Department Carol Jillings PhD, RN, Assistant Professor, School of Nursing, UBC Paul Galdas PhD,  Associate Professor, School of Nursing, UBC Nursing researchers from the School of Nursing at UBC and Vancouver General Hospital Cardiac Sciences are conducting research to examine postoperative risk factors and nursing care that may be related to patients experiencing delirium after heart surgery.  The ultimate aim of this study is to identify if specific medications for pain control and sedation are increasing the incidence of postoperative delirium in heart surgery patients and to examine current nursing care regarding the administration of these medications.  You are being invited to take part in this research study because you have agreed, as a part of standard of care, to be assessed for delirium after your heart surgery and for your personal health information to be collected after your surgery in order to determine if a link exists between these factors. Your participation is entirely voluntary.  You have the right to refuse to participate in this study.  If you decide to participate, you will be asked to sign this form.  The information obtained by the researchers will be used for study purposes only and will not in any way be identified with you or members of your family.  Please take time to read the following information carefully.    120 Appendix D STUDY PROCEDURES Participation in this study will involve a brief (5 minute) face to face interview with the researcher where you will be asked to complete several tasks related to memory, orientation, concentration and writing.  Once this is complete you will have your surgery as it is scheduled. When the surgery is finished you will be cared for in the Cardiac Surgery Intensive Care Unit by nurses who are not involved in the study.  At 12-18 hours after you enter the CSICU, one of the study researchers will perform an assessment to determine if you have any of the signs of delirium.  This assessment will take 5-10 minutes and will involve a series of questions related to attention, memory, orientation, and thinking.   This assessment will also occur on your postoperative day 2 and 3 and will involve the same questions.  These assessments will be performed while you are in the Cardiac Surgery ICU, or when you are on the Cardiac Surgery ward (Centennial Pavilion, 10th floor).  After the third assessment, data related to your health history, surgery and current health status will be obtained by the researchers from your hospital chart. RISKS & BENEFITS There are no known risks to you from participation in this study. During the assessments for delirium you may find some of the test questions to be upsetting if you are unable to answer them.   You will receive the standard postoperative care from all members of the health care team regardless of your participation in this study. Although you may not receive any direct benefits from participating in this study however; your participation will help to increase our understanding of possible modifiable risk factors for delirium following heart surgery and may influence changes to care for future patients. CONFIDENTIALITY Your confidentiality will be respected throughout the study. 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 Health Canada, and the UBC Research Ethics Board for the purpose of monitoring the research. However, no records that identify you by name or initials will be allowed to leave the Investigators’ offices. Research records will be coded numerically and only your consent form will have your name on it.  The research data collected from your hospital chart will be inspected by the investigators, data entry designates, and possibly by representatives of the UBC and / or Vancouver Coastal Health Ethics Board for monitoring purposes.  Hard copy or electronic study data will be stored securely at the VGH site, principal investigator’s residence, or the UBC School of Nursing. 1. By signing this consent form, you are authorizing this access. 2.   If results of this study are published, you will not be identified in any way.    121 Appendix D FOR MORE INFORMATION If you have any questions or desire further information, please contact the Principal Investigators Dr. Pamela Ratner at 604-822-7427 if you have any concerns about your rights as a research subject and / or your experience 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 or the Chair of the UBC / PHC Research Ethics Board at 604-682-2344 ext 62325. By signing this form, you do not give up any of your legal rights.  There will be no costs to you for your participation in this study. CONSENT I have read and understood the consent form. I have been given a copy of this consent form. I have had sufficient time to consider the information provided and to ask questions, and have received satisfactory responses to my questions. I understand that all of the information will be kept confidential and will only be used for scientific purposes. I understand that my participation is voluntary and that I have the right to refuse to participate or withdraw from this study at any time without any negative consequences to the medical care, or other services that I receive. I understand that this study will not provide any direct benefits to me. I understand that by signing this form I do not give up any of my legal rights. I have read this form and freely consent to participate in this study.  Printed Name of Participant   Signature    Date _________________________________  ______________________________ ______ Printed Name of Witness   Signature    Date __________________________________  _______________________________ ______   122 Appendix E: Data Collection Tool              Participant # ___ Title of Study: The Relationship between Postoperative Risk Factors for Delirium after Cardiac Surgery and Nursing Care.   DATA COLLECTION TOOL PREOPERATIVE DATA:  AGE: __________  SEX: _____________ SURGERY DATE: ____________________________   SURGEON:________________________________ PREOPERATIVE DIAGNOSIS: _________________________________________________________________________________________________ Co-morbidities: _________________________________________________________________________________________________ _________________________________________________________________________________________________ PREOP CREATININE :______________________      ESTIMATED EF: _________________________ PREOPERATIVE MEDICATIONS: Drug Dose Frequency         SMOKING HISTORY: _______________________________________________________ MINI MENTAL STATE EXAM SCORE: ___________________________ OPERATIVE DATA SURGICAL PROCEDURE: _____________________________________________________________________________ ANAESTHETIC START TIME: _________________  SURGICAL START TIME:_________________ CROSS CLAMP TIME :_______________________ BYPASS PUMP TIME: _______________________ BLOOD TRANSFUSIONS INTRAOP :________________________________________________________   123  Appendix E Complications: _______________________________________________________________________________________                               _______________________________________________________________________________________                               _______________________________________________________________________________________  POST OPERATIVE DATA TIME ADMITTED TO CSICU: ________________________________ GCS: _____________           CORE TEMP: _______________________________ PROPOFOL: _____________________________________________________________________ CARDIAC RHYTHM :_____________________________________________________________ BP: _______________   PAP: _______________  CVP: _______________   SPO2:______________ MEDICATIONS ON ARRIVAL TO CSICU: Drug / Classification Dose Frequency Time Started Time stopped          COMMENTS: ________________________________________________________________________________________________ _________________________________________________________________________________________________ _________________________________________________________________________________________________ FIRST ABG RESULTS:      PH: ____________ (02 VIA VENT): __________ TIME: _________________ P02: ___________  PC02: __________  BICARB: _______                                               HGB: _______________   BLOOD SUGAR: _________________ CHEMISTRY RESULTS: NA: _________ K+:________  CL: ______________     CR: _________ BUN: ________  MG: ______________   124 Appendix E COAGULATION:  INR: ________ PTT: _________ CBC:     WBC: _______ HGB:_________   PLATELETS: __________  TIME OF FIRST RESPONSE (EYES OPENING): ___________  = __________ HRS POST OP TIME WHEN OBEYING COMMANDS: ______________  = ____________ HRS POST OP Glascow Coma Scale: _________________  CARDIAC OUTPUT RESULTS DATE / TIME: _______  DATE/TIME: _________ DATE/TIME__________ CO ________   CO____________   CO__________ CI _________   CI_____________   CI___________ SVR________   SVR__________   SVR_________ PVR _______   PVR _________   PVR_________ FIRST DOSE OF HYDROMORPHONE:  AMOUNT _____________   TIME ___________ FIRST DOSE OF MIDAZOLAM:               AMOUNT _____________   TIME ___________ HYDROMORPHONE TOTAL FOR FIRST 18 HOURS POST OP: Time Dose Route      Total Hydromorphone (mg) FROM ADMISSION TO CSICU TO 2400: ____________ 2400 – 0730: ___________________ OTHER ANALGESIA: _________________________________________________________________ MIDAZOLAM ADMINISTERED IN CSICU: Time Dose Route       125 Appendix E Total Midazolam (mg): ____________ EXTUBATION DATE / TIME: ___________________________________________ TOTAL VOLUME REPLACEMENT 0-12 HOURS POST OP: _______________________________ POSTOPERATIVE EVENTS (FIRST 12 HOURS):  _________________________________________________________________________________________________________ POSTOPERATIVE DAY 1 (12-18 HOURS)       DATE: _____________ RASS SCORE: ______________________________ CAM-ICU SCORE: _______________________________ PAIN SCORED? (0-10 SCALE): _______________ REGULAR & PRN MEDICATIONS: Drug Dose Frequency        ANALGESIA  (DRUG) DOSE TIME        02 REQUIREMENTS: ________________________________________ AM ABG TIME _____________ PH___________ P02__________ PC02__________ HCO3___________ HGB__________ K+ _________ CR _______ CARDIAC RHYTHM: ____________________________________________   BP ______  SP02 ______ URINE OUTPUT >0.5mls /kg/hr:        YES   NO   126 Appendix E TRANSFER TO WARD:          YES  NO DATE/TIME: _________________________  POSTOPERATIVE DAY 2       DATE: ____________ RASS SCORE: __________________________________ CAM-ICU SCORE: ___________________________________ PAIN SCORED? (0-10 SCALE): __________________  REGULAR & PRN MEDICATIONS POD 2: ANALGESIA GIVEN DAY 2: DRUG DOSE FREQUENCY      02 REQUIREMENTS: ________________________________________________ CARDIAC RHYTHM: _________________________________________________  POSTOPERATIVE DAY 3        DATE: _________________ RASS SCORE: _________________________________________ CAM-ICU SCORE: _____________________________________ PAIN SCORE (0-10 SCALE): __________________________   Drug Dose Frequency          127 Appendix E REGULAR & PRN MEDICATIONS POD 3: Drug Dose Frequency        ANALGESIA POST OP DAY 3: ANALGESIA  (DRUG) DOSE TIME        02 REQUIREMENTS: ________________________________________________________________________ CARDIAC RHYTHM: _________________________________________________________________________ OTHER COMMENTS / INFORMATION:     DATE OF DISCHARGE: _________________________       128 Appendix F: Chart Review Form CHART REVIEW The Relationship Between Postoperative Risk Factors for Delirium after Cardiac Surgery and Nursing Care.   PARTICIPANT # __________________ Chart reviewed date: _____________________________ Chart reviewer: ______________________________________________________________ Psychiatry Consult found in chart:                YES   NO Date of Psych consult: ______________________ Diagnosis of Delirium indicated by Psychiatrist:    YES   NO Patient prescribed antipsychotics, benzodiazepines and hs sedation by   YES   NO Psychiatrist Diagnosis of Delirium indicated by other MD:   YES   NO Date of diagnosis: ___________________________ Doctor diagnosing delirium (Name & Specialization): __________________________________________________ _________________________________________________ Nurses Notes indicate Signs of Delirium (as per CAM)  YES  NO fluctuation in LOC,          agitation,         hallucinations, inability to follow directions,           altered sleep         use of restraints prn antipsychotics DATE & TIME of first indication of S&S of Delirium in chart: __________________________ Other: ________________________________________________________________________________________________ _________________________________________________________________________________________________

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}]}"
                            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:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0071390/manifest

Comment

Related Items