Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Total joint arthroplasty patients' adherence to a pre-operative staphylococcus aureus decolonization… Martino, Daniela 2013

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

Item Metadata

Download

Media
24-ubc_2014_spring_martino_daniela.pdf [ 303.01kB ]
Metadata
JSON: 24-1.0103388.json
JSON-LD: 24-1.0103388-ld.json
RDF/XML (Pretty): 24-1.0103388-rdf.xml
RDF/JSON: 24-1.0103388-rdf.json
Turtle: 24-1.0103388-turtle.txt
N-Triples: 24-1.0103388-rdf-ntriples.txt
Original Record: 24-1.0103388-source.json
Full Text
24-1.0103388-fulltext.txt
Citation
24-1.0103388.ris

Full Text

TOTAL JOINT ARTHROPLASTY PATIENTS? ADHERENCE TO A PRE-OPERATIVESTAPHYLOCOCCUS AUREUS DECOLONIZATION PROTOCOLbyDaniela MartinoB.A., The University of British Columbia, 1999B.S.N., The University of British Columbia, 2003A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFMASTER OF SCIENCE IN NURSINGinTHE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES(Nursing)THE UNIVERSITY OF BRITISH COLUMBIA(Vancouver)December 2013? Daniela Martino, 2013iiAbstractBackground: Staphylococcus aureus surgical site infections and their treatment, in the total jointarthroplasty population, can significantly affect patients? recovery and their quality of life, andcan generate considerable economic cost for the health care system. The use of a pre-operativescreening and decolonization protocol has shown promising eradication rates of Staphylococcusaureus and a decreased incidence of surgical site infections, however, the results have lackedstatistical significance. Adherence to the decolonization protocol has been identified as apossible missing link or explanation for these equivocal findings.Objective: The purpose of this study was to examine the relationships between age, self-efficacy, and adherence to a Staphylococcus aureus screening and decolonization protocol in thetotal joint arthroplasty population.Methods: A descriptive correlational study design was conducted. The study sample included 40participants who underwent primary total joint arthroplasty surgery between May 1, 2013 andOctober 1, 2013 at a hospital in Vancouver, British Columbia. Self-efficacy and adherence wereassessed using two self-report measures: the Self-efficacy Survey and the AdherenceQuestionnaire. Data were analyzed using correlational and multiple linear regression analyses.Results: The findings suggest that there was a positive relationship between age and adherenceto the use of chlorhexidine gluconate cloths, and a negative relationship between age andadherence to the use of nasal Mupirocin. These results were not statistically significant. Therewas a statistically significant and strong positive relationship between the patients? level of self-efficacy in applying Mupirocin and their adherence to its use, timing, and application. Little if norelationship was found between the patients? level of self-efficacy to chlorhexidine gluconatecloths and their adherence to its use, timing, and application. Age, and not self-efficacy,iiicontributed significantly to the outcome, adherence.Conclusions: The study found inconclusive results with respect to the relationships between age,self-efficacy, and adherence. In light of these results, this study highlights the many ways inwhich age and self-efficacy can influence adherence in adults. This information can be usefulwhen evaluating the effectiveness of a decolonization protocol and for nurses in their attempts todesign, implement, and evaluate patient education materials relevant to the protocol.ivPrefaceThis dissertation is original unpublished work by the author, Daniela Martino. All aspectsfrom conceptual formulation, to study design, data collection, and analysis, as well ascomposition are original work. Professor Pamela Ratner was the supervisor of this researchproject and was involved throughout the project in planning the study, supervising the analysis,and guiding the development of the thesis. Ms. Laurel Archer, a thesis committee member,assisted with the data collection at the hospital. All thesis supervisory committee members readdrafts of the thesis and provided suggestions for content, organization, and style.Approval for this research project, ?Total Joint Arthroplasty Patients? Adherence to aPre-operative Staphylococcus Aureus Decolonization Protocol,? was granted by the Universityof British Columbia, Behavioural Research Ethics Board, under certificate number H13-01011.vTable of ContentsAbstract ........................................................................................................................................... iiPreface............................................................................................................................................ ivList of Tables ................................................................................................................................ viiList of Figures .............................................................................................................................. viiiAcknowledgements........................................................................................................................ ixChapter 1: Introduction ................................................................................................................... 1Background ................................................................................................................................. 1Significance of the problem.................................................................................................... 1Problem statement and purpose .............................................................................................. 4Research Questions..................................................................................................................... 5Theoretical Framework............................................................................................................... 5Chapter 2: Literature Review.......................................................................................................... 8Search Strategy ......................................................................................................................... 10Search Results........................................................................................................................... 14Methods of the Selected Studies ............................................................................................... 14Study design.......................................................................................................................... 14Study target populations and settings ................................................................................... 15Variables examined............................................................................................................... 17Statistical analyses ................................................................................................................ 20Instruments Used for the Measurement of Relevant Concepts................................................. 21Self-efficacy.......................................................................................................................... 21Measuring adherence ............................................................................................................ 25Research Findings..................................................................................................................... 28Age and adherence................................................................................................................ 28Self-efficacy and adherence.................................................................................................. 34Limitations of the Research ...................................................................................................... 39Chapter 3: Methods....................................................................................................................... 42Study Design............................................................................................................................. 42Sampling Plan ........................................................................................................................... 42viSample population ................................................................................................................ 42Inclusion and exclusion criteria ............................................................................................ 43Recruitment of participants................................................................................................... 44Power analysis ...................................................................................................................... 45Procedures and Data Collection................................................................................................ 45The established decolonization protocol............................................................................... 45The study protocol ................................................................................................................ 48Data collection ...................................................................................................................... 48Data Analysis ............................................................................................................................ 50Ethical Considerations .............................................................................................................. 54Overcoming Potential Challenges............................................................................................. 54Chapter 4: Results ......................................................................................................................... 56Eligibility and Participants? Response Rates ............................................................................ 56Sample Description................................................................................................................... 57Self-Efficacy and Perceived Infection Risk.............................................................................. 57Adherence ................................................................................................................................. 59Correlations: Age, Self-Efficacy, and Adherence..................................................................... 60Chapter 5: Discussion, Implications, and Conclusions................................................................ 62Discussion ................................................................................................................................. 62Age and adherence................................................................................................................ 65Self-efficacy and adherence.................................................................................................. 68Clinical Implications................................................................................................................. 74Methodological Limitations...................................................................................................... 75Recommendations for Future Research .................................................................................... 77Conclusions............................................................................................................................... 79Bibliography ................................................................................................................................. 81Appendix A................................................................................................................................... 90viiList of TablesTable 1. Search Strategy for CINAHL via EBSCOHost ............................................................. 10Table 2. Search Strategy for CINAHL via EBSCOHost (expanded) .......................................... 10Table 3. Search Strategy for CINAHL via EBSCOHost (expanded) .......................................... 11Table 4. Search Strategy for CINAHL via EBSCOHost (expanded) .......................................... 11Table 5. Search Strategy for PubMed .......................................................................................... 12Table 6. Search Strategy for PubMed (expanded) ....................................................................... 12Table 7. Search Strategy for PubMed (expanded) ....................................................................... 13Table 8. Search Strategy for PubMed (expanded) ....................................................................... 13Table 9. Self-Efficacy and Perceived Infection Risk Scores ....................................................... 58Table 10. Self-Efficacy Survey Response Rates.......................................................................... 58Table 11. Adherence Scores ........................................................................................................ 59Table 12. Pearson?s Correlations: Age, Self-Efficacy, and Adherence....................................... 60Table 13. Regression Analysis Predicting Adherence to the Application of ChlorhexidineGluconate ...................................................................................................................................... 61viiiList of FiguresFigure 1. Assigned Body Areas for the Six Chlorhexidine Gluconate Cloths............................. 47Figure 2. Score Allocation for Adherence Questionnaire............................................................ 53Figure 3. Eligibility and Response Rates..................................................................................... 56ixAcknowledgementsI would like to express my gratitude to the faculty, staff, and colleagues of the UBCSchool of Nursing who provided endless support and who inspired me to continue striving forexcellence in the field of nursing and health care. I will forever be grateful to my thesissupervisor, Dr. Pamela Ratner, for all her support, guidance, expertise, and encouragementthroughout this process. Her attentiveness and kindness will always be remembered andappreciated. I owe particular thanks to my thesis committee members, Laurel Archer and TarniaTaverner. Without Laurel Archer?s willingness to assist with this research project, none of thiswould be possible. I will forever be thankful for her assistance, dedication, and commitment tothis research. I thank Tarnia Taverner for all her support, insight, and feedback throughout thisprocess.I owe a special thanks to the manager of the pre-surgical screening and surgical day caredepartments for supporting this research project. I thank all the staff in the pre-surgicalscreening and surgical day care units for their assistance with the data collection. I also owe aspecial thanks to Felicia Laing, Quality and Patient Safety Project Manager, for introducing meto the decolonization protocol and for providing me with the necessary resources to begin thisresearch project.Special thanks to Leanne Appleton, and managers, Michelle Stanton, Barb Ferreira, andHeather Speirs for supporting my education and for being flexible with my work hours over thelast few months.I offer my enduring gratitude to my husband, Brad, parents, Anna and Carmen, andbrothers, Nick and Giancarlo, for all their support and encouragement throughout this process.Without them, none of this would have been possible. I truly thank them for their patience, kindxwords, and for keeping me focussed. I thank, Brad, for his attentive ears, for all his hugs andkisses, but mostly for just being him. I share this success with each and every one of them.1Chapter 1: IntroductionStaphylococcus aureus has been identified as the main perpetrator of surgical siteinfections in many tertiary health care centres (Anderson, Sexton, Kanafani, Auten, & Kaye,2007). Surgical site infections and their treatment, especially in the total joint arthroplastypopulation, have been shown to significantly affect patients? recovery from surgery and theirquality of life, and generate considerable economic cost for the health care system (Rao et al.,2011). Consequently, global efforts have been made to find ways to eradicate or minimize thedetrimental effects of this pathogen. Van Rijen, Bonten, Wenzel, and Kluytmans (2008a)conducted a systematic review of randomized control studies that examined the effects of a pre-operative screening and decolonization protocol on patients undergoing total joint arthroplasty.They found promising eradication rates of Staphylococcus aureus and a decreased incidence ofpost-operative surgical site infections, although the results were not statistically significant.Some methodological limitations of the studies included in the systematic review likelycontributed to these findings; however, a recognizable gap in the research literature related topatients? adherence to the decolonization protocol may also have been a contributing factor(Hansen et al., 2007).BackgroundSignificance of the problemAccording to the Antibiotic Resistant Organisms Prevention and Control Guidelines,recommended by the Provincial Infection Control Network of British Columbia (2008), a pre-operative screening and Staphylococcus aureus decolonization protocol is warranted in acutecare facilities and should be implemented under the guidance of an infection control practitioner.A particular screening and decolonization protocol, introduced in March 2012 at a community2hospital, in British Columbia, requires that all patients undergoing total joint arthroplasty bescreened for nasal carriage of Staphylococcus aureus before surgery. All patients are advised towash with six Sage 2% chlorhexidine gluconate cloths (free of latex, alcohol, and rinse), theevening before surgery and repeated on the day of admission for surgery. Chlorhexidinegluconate is a chemical antiseptic that is effective against gram-positive and gram negativebacteria. It is both bactericidal and bacteriostatic in action. In addition to the use of chlorhexidinegluconate cloths, those who test positive for nasal carriage of Staphylococcus aureus areprescribed nasal Mupirocin twice daily for five days before surgery to eradicate the nasal area ofthis pathogen. Mupirocin calcium ointment, 2% (Bactroban? Nasal), is an antibiotic of themonoxycarbolic acid class. It is bacteriostatic in low concentrations and bactericidal at highconcentrations.Positive nasal carriage of Staphylococcus aureus, more so than carriage in any other bodypart, is a significant risk factor for the development of surgical site infections (Kim et al., 2010).Wenzel and Perl (1995) estimated the risk to be two to nine times greater in positiveStaphylococcus aureus nasal carriers than in non-carriers. Similarly, Perl et al. (2002) found thatin surgical site infections where Staphylococcus aureus was identified as the isolate, 85% of theinfections were associated with nasal carriage of the same isolate. With the use of nasalMupirocin as a preventive measure, a risk reduction of 1% has been shown to translate into costsavings in the millions of dollars for every ten thousand patients screened (Van Rijen et al.,2008a). Based on these findings it is easy to appreciate the importance of patient adherence to adecolonization protocol to minimize the adverse outcomes and health care costs arising fromStaphylococcus aureus infections.Patient adherence is a complex and dynamic concept rooted in multiple disciplines;3consequently, many variations of a definition exist. For the purpose of this study, patientadherence was defined as the extent to which patients? actions coincide and follow theinstructions they are given for a prescribed treatment or regimen (Christensen, 2004; Haynes etal., 2005). It is worth noting that, in the literature, the word ?adherence? is often usedinterchangeably with the word ?compliance? (Bissonnette, 2008). Because the term complianceis often strongly associated with an underlying notion of blame or an unwillingness to follow orconsent to orders, the word adherence rather than compliance is used here (Bissonnette, 2008).Patient adherence in the health care literature has predominantly been examined in areasof chronic and cardiovascular disease, HIV/AIDS, and cancer treatment and has focussedprimarily on medication compliance and compliance with recommended exercise and lifestylechanges (Bissonnette, 2008). An examination of patients? adherence to a pre-operative screeningand decolonization protocol, specifically, is lacking. The orthopaedic specialty, more so than anyother specialty, has studied the effects of such protocols on the rates of surgical site infectionsand has found some positive results (Van Rijen et al., 2008a). However, despite these promisingresults, the overall findings have not been statistically significant, which has left room for muchspeculation as to the reasons why (Van Rijen et al., 2008a). Could a lack of adherence to thedecolonization protocol be an important missing link?Just as there is a lack of research about patients? adherence to a pre-operativedecolonization protocol, there has been an equivalent lack of objective measurement oroperationalization of adherence in health care research (Bissonnette, 2008). These gaps make theevaluation of treatment protocols difficult and they prevent the implementation of necessaryquality control measures in hospitals (Hansen et al., 2007). Although adherence in the total jointarthroplasty population has not been examined, it has been examined in other patient4populations. Research findings from these other areas lend themselves to enhancing ourknowledge and understanding of patients? adherence in the orthopaedic population. Researchconcerning patients? adherence, specifically in the areas of chronic and cardiovascular disease,and HIV/AIDS, has examined the influence of socio-demographic variables (e.g., income,education, and age), health literacy, and social support (Vaughn-Cooke, 2009). This research hasalso included the examination of self-efficacy as a predictor of adherence (Vaughn-Cooke,2009). One would expect the factors that influence patients? adherence in the total jointarthroplasty population to be similar to those of other patient populations, but without adequateresearch in the orthopaedic population this assumption cannot be verified. The examination ofpatients? adherence to an established decolonization protocol, in the joint arthroplastypopulation, could have significant implications for the design, implementation, and evaluation ofrelevant patient education materials.Problem statement and purposeStaphylococcus aureus has been identified as the main perpetrator of surgical siteinfections in the total joint arthroplasty population, which has motivated researchers to designpre-operative screening and decolonization protocols aimed at eradication (Anderson et al.,2007). Non-adherence to these protocols has significant implications for patients? quality of lifeand for health care costs (Rao et al., 2011). There also is an important need to understand thefactors that influence patients? adherence, particularly for this population, which could enhancethe design and implementation of patient education and the evaluation of a protocol?seffectiveness. The purpose of this study was to examine the relationship between socio-demographic variables and levels of self-efficacy on patients? adherence to a Staphylococcusaureus decolonization protocol in the total joint arthroplasty population.5Research QuestionsThe following research questions guided this study: What is the rate of adherence to apre-operative screening and decolonization protocol, in the total joint arthroplasty population? Ispatients? adherence influenced by their age? Specifically, is there a relationship between age andadherence? Is there a relationship between patients? level of self-efficacy and adherence? Thefollowing hypotheses were tested:Hypothesis 1: There is a relationship between patients? age and adherence.Hypothesis 2: There is a positive relationship between patients? level of self-efficacy andadherence.Theoretical FrameworkThe Health Belief Model is one of the most commonly used theoretical frameworks byresearchers that have examined adherence (Rosenstock, Strecher, & Becker, 1988). This modelproposes that personal health beliefs or perceptions of an illness or disease can predict healthbehaviour, which is multifaceted and can be conceptualized in many different ways. For thepurpose of this study, health behaviour was conceptually defined as adherence to a prescribedregimen. Similarly, rather than examining health beliefs or perceptions of an illness or disease,this study was based on a conceptual definition of illness or disease as meaning the risk ofStaphylococcus aureus infection. The Health Belief Model predicts that if people perceive theirvulnerability to develop an illness or disease as being high, they perceive the illness or disease tobe serious, and they perceive the benefits of engaging in particular health behaviour asoutweighing the risks, they are more likely to adhere to that health behaviour (Sirur, Richardson,Wishart, & Hanna, 2009). There is a large body of literature that supports the theory thatperceptions and personal beliefs about illnesses or diseases stem from prior experiences and6personal interactions with the environment (Bandura, 1977; Vaughn-Cooke, 2009). Hence,embedded in the Health Belief Model are components of Bandura?s (1977) self-efficacy theory,specifically, the influence of vicarious and mastery experiences and verbal persuasion onadherence.Self-efficacy, the belief in one?s ability to perform a specific task, is determined bypersonal experiences with adherence (mastery experiences), the observed successes and failuresthat others have had with performing the same task (vicarious experiences), the ability to bepersuaded by others to adhere (verbal persuasion), and the perception of one?s capability toadhere (one?s physiological state, such as anxiety) (Sirur et al., 2009). There is strong support forthe use of both the Health Belief Model and self-efficacy theory, as suitable theoreticalframeworks, for the examination of patients? adherence in relation to socio-demographicvariables and levels of personal self-efficacy (Granger, Moser, Harrell, Sandelowski, & Ekman,2007). The high degree of interconnectivity between these two theoretical frameworks is furthersupported by the Social Personal Organizational Technological (SPOT) patient compliancemodel (Vaughn-Cooke, 2009), which identifies socio-demographic variables and perceivedtreatment complexity and self-care capability, two components of self-efficacy theory, as three offourteen performance shaping factors that affect compliance. The SPOT model proposes thatperformance shaping factors play a large role in the development of patients? attitudes andknowledge acquisition, which in turn affect their compliance (Vaughn-Cooke, 2009). Patients?attitudes and knowledge shape their personal health beliefs and perceptions about a particularhealth phenomenon. They often guide the type of personal interactions they will have with theirenvironment and they can often predict the types of experiences that will stem from thoseinteractions (Vaughn-Cooke, 2009). It is clear that the principles and components of both the7Health Belief Model and self-efficacy theory can assist the quest for a clearer understanding ofthe factors that influence and predict patients? adherence.8Chapter 2: Literature ReviewVan Rijen, Bonten, Wenzel, and Kluytmans (2008b) conducted a systematic review ofrandomised controlled trials and found a statistically significant reduction in the rate ofStaphylococcus aureus infections with the use of intranasal Mupirocin in diverse surgicalpopulations. Similar results were found with the use of chlorhexidine gluconate cloths appliedthe evening before surgery and the day of surgery in hip and knee arthroplasty patients (Johnson,Daley, Zywiel, Delanois & Mont, 2010; Zywiel et al., 2011). These findings provide strongsupport for the use of a decolonization protocol to prevent the development of post-operativeStaphylococcus aureus surgical site infections. The efficacy and success of a decolonizationprotocol, however, like any treatment, is heavily reliant on adherence. Bearing this latter point inmind, a closer examination of adherence and the factors that influence adherence was warrantedto ensure optimal quality control in clinical practice.While the success and effectiveness of a decolonization protocol, like any treatment, isheavily influenced by adherence, there is a lack of research that has examined patients?adherence to this protocol in the total joint arthroplasty population specifically. In a study thatspecifically examined adherence to outpatient preoperative methicillin-resistant Staphylococcusaureus therapy across diverse surgical populations, a complete adherence rate to both nasalMupirocin and the use of chlorhexidine gluconate cloths was determined to be only 31.1%(Caffrey et al., 2011). Bailey et al. (2011) explained that even with low patient compliance ratesthere is still great economic value in the decolonization protocol when one considers that a smallreduction in the incidence of surgical site infections can translate into cost savings in themillions. In addition to unrealized cost savings, a low adherence rate is concerning and has manyimplications for the health status of patients and for clinical practice, in general. Non-adherence9is a poorly understood phenomenon. A better understanding of the factors that influenceadherence is necessary, as is further elaboration of the theories that explain it.Although adherence in the total joint arthroplasty population has not been thoroughlyexamined, it has been examined in other patient populations, particularly in relation tomedication compliance and compliance with recommended exercise and lifestyle changes(Bissonnette, 2008). Research findings from these other areas lend themselves to enhancing ourknowledge and understanding of patient adherence in the orthopaedic population. Research inthese areas has examined the influence of socio-demographic variables (e.g., income, education,and age), health literacy, and social support (Vaughn-Cooke, 2009). This research also hasincluded the examination of self-efficacy as a predictor of patient adherence (Vaughn-Cooke,2009). One would expect the factors that influence patient adherence in the total jointarthroplasty population to be similar to those of other patient populations, but without adequateresearch in the orthopaedic population this assumption cannot be verified.This review of the literature summarizes the findings of research that has examined therelationships between socio-demographic variables, namely age, and self-efficacy andmedication compliance, adherence to pre-operative guidelines, adherence to recommendedexercise and lifestyle regimens, and adherence to prophylactic treatments in diverse patientpopulations. Prior to discussing the overall research findings, this review first summarizes themethods applied in the studies, including the study designs, methods of subject recruitment,sample populations, variables examined, the instrumentation or measures used within the studies,and the statistical analyses used to examine the relationships between the variables. The reviewconcludes with a discussion of the limitations and gaps in the current literature and somerecommendations for future research.10Search StrategyA systematic literature search was conducted using the CINAHL and PubMed databasesto find research articles published between 2003 and September 2013 in the English language.To conduct the search, the following search terms were used separately and in combination:?adherence,? ?compliance,? ?Staphylococcus aureus,? ?decolonization,? ?exercise programs,??rehabilitation programs,? ?patient guidelines,? ?surgical guidelines,? ?pre-operative guidelines,??medication adherence,? ?age,? ?age associated predictors,? ?age factors,? ?sociodemographicfactors,? ?self-efficacy,? ?health beliefs,? ?health perceptions,? and ?patient attitudes.? The searchwas restricted to adults and human studies. The following tables summarize the search queriesand citation history for each specific database searched.Table 1. Search Strategy for CINAHL via EBSCOHostStep Search Query Citation Count1 TI adherence OR TI compliance 9,9452 Staphylococcus aureus 6,7793 Decolonization 1294 2 AND 3 1065 1 AND 4 1Given that the above search returned only one citation with respect to adherence to aStaphylococcus aureus decolonization protocol, an expanded search of adherence in the contextof recommended rehabilitation or exercise programs, medications, and patient, surgical andpreoperative guidelines, was undertaken.Table 2. Search Strategy for CINAHL via EBSCOHost (expanded)Step Search Query Citation Count6 TI exercise programs OR TI rehabilitation programs 4057 patient guidelines OR TI surgical guidelines OR TI pre-operative guidelines1228 medication adherence 79911Step Search Query Citation Count9 6 OR 7 OR 8 132610 9 AND 1 83111 Published Date: 20030101-20130931; English Language;Research Article; Human; Age Groups: All Adult373The specific question of the relationship between age and adherence in the context of managingStaphylococcus aureus was not addressed in the set of papers identified in the previous searches.A search with respect to the role of age in the context of adherence or compliance, in general,was undertaken.Table 3. Search Strategy for CINAHL via EBSCOHost (expanded)Step Search Query Citation Count12 TI age OR TI age associated predictors OR TI age factorsPublished Date: 20030101-20130931; English Language;Research Article; Human; Age Groups: All Adult4,97513 TI sociodemographic factors Published Date: 20030101-20130931; English Language; Research Article; Human; AgeGroups: All Adult5414 12 and 1 1815 13 and 1 1To address the specific question of the relationship between self-efficacy and adherence, a morespecific search using the terms ?self-efficacy? and constructs of self-efficacy, including ?healthbeliefs,? ?patient attitudes,? and ?health perceptions? was undertaken. This search was combinedwith the previous search results that explored adherence in relation to recommended exerciseprograms, medications, and patient, surgical and preoperative guidelines.Table 4. Search Strategy for CINAHL via EBSCOHost (expanded)Step Search Query Citation Count16 TI self-efficacy OR TI health beliefs OR TI health perceptionsOR TI patient attitudes3,33217 16 AND 11 1912Once the search was concluded, reference lists of relevant papers were scanned for relevantmanuscripts not identified through the search.Table 5. Search Strategy for PubMedStep Search Query Citation Count1 adherence OR compliance 179,8732 Staphylococcus aureus 81,1253 decolonization 3764 2 AND 3 3005 1 AND 4 21Of the 21 citations obtained from the above search, only one examined adherence to aStaphylococcus aureus decolonization protocol. The reference list for this article was scanned forrelevant manuscripts not identified in the search, but no additional articles of relevance wereidentified. An expanded search of adherence in the context of recommended exercise orrehabilitation programs, medications, and patient, surgical and preoperative guidelines, was thenundertaken.Table 6. Search Strategy for PubMed (expanded)Step Search Query Citation Count6 adherence[Title]) OR compliance[Title] Filters: JournalArticle; published in the last 10 years; Humans; English;Adult: 19+ years62677 patient guidelines[Title]) OR surgical guidelines[Title]) ORpreoperative guidelines[Title]) OR rehabilitationprograms[Title] Filters: Journal Article; published in the last10 years; Humans; English; Adult: 19+ years628 exercise programs[Title] Filters: Journal Article; published inthe last 10 years; Humans; English; Adult: 19+ years409 7 OR 8 10210 medication adherence[Title] Filters: Journal Article; publishedin the last 10 years; Humans; English; Adult: 19+ years57611 9 OR 10 67812 11 AND 6 58213The specific question of the relationship between age and adherence in the context of aStaphylococcus aureus decolonization protocol was not addressed in the set of papers identifiedin the previous searches. A specific search with respect to the role of age in the context ofadherence or compliance to recommended exercise or rehabilitation programs, medications, andpatient, surgical and preoperative guidelines, was then undertaken.Table 7. Search Strategy for PubMed (expanded)Step Search Query Citation Count13 age[Title]) OR age associated predictors[Title]) OR agefactors[Title] Filters: Journal Article; published in the last 10years; Humans; English; Adult: 19+ years21,55314 13 AND 12 5The specific question of the relationship between self-efficacy and adherence in the context of aStaphylococcus aureus decolonization protocol was not addressed in the citations identified inthe previous searches. A more specific search with respect to the role of self-efficacy and otherconstructs related to self-efficacy, including health beliefs, health perceptions and patientattitudes, was undertaken. This search was combined with the search results of adherence in thecontext of recommended exercise and rehabilitation programs, medications, and patient, surgicaland preoperative guidelines.Table 8. Search Strategy for PubMed (expanded)Step Search Query Citation Count15 self-efficacy[Title]) OR health perceptions[Title]) OR healthbeliefs[Title]) OR patient attitudes[Title] Filters: JournalArticle; published in the last 10 years; Humans; English;Adult: 19+ years145316 15 AND 12 18Once the search was concluded, reference lists of relevant papers were scanned for relevant14manuscripts not identified through the search.Search ResultsA total of 39 published research studies were returned from the CINAHL database searchand 44 published research articles were returned from the PubMed database search. Thereference lists of relevant research articles were scanned for relevant manuscripts not identifiedthrough the search. Fifteen studies were selected for review based on the following criteria: (i)the study sample included adults ?19 years of age who were outpatients; (ii) the study focused on age and adherence or compliance or on self-efficacy and adherence or compliance; and (iii) thestudy measured adherence or compliance to medication prescriptions, patient, surgical or pre-operative guidelines, or exercise or rehabilitation programs.Methods of the Selected StudiesStudy designWhether examining the relationship between age and adherence or self-efficacy andadherence, there appeared to be consistency in the studies with respect to their study designs.Prospective, cross-sectional observational study designs were the most common designs used toexamine these variables across diverse patient populations whether in relation to medicationadherence, adherence to pre-operative or surgical guidelines, or adherence to patient instructions(Barclay et al., 2007; Birkenfeld et al., 2011; Boeka, Prentice-Dunn, & Lokken, 2010; Curtin etal., 2008; Fawzi et al., 2012; Liang, Yates, Edwards, & Tsay, 2008; Tarantino et al., 2010; Yuhaset al., 2012). Three published systematic reviews also found that cross-sectional, observationaldesigns were the most dominant methodological design used in the examination of medicationadherence in the elderly, patient adherence to cardiac rehabilitation programs, and adherence tochronic treatment (Chia, Schlenk, & Dunbar-Jacob, 2006; Ingersoll & Cohen, 2008; Taylor,15Wilson, & Sharp, 2011). Although not as prevalent a design in the literature examiningadherence, a prospective repeated measures design was used to examine the effects of self-efficacy on adherence outcomes, a relatively strong design in the evidence hierarchy (Ngamvitroj& Kang, 2007). Because cross-sectional, descriptive correlational study designs are positioned atthe lower tier of the evidence hierarchy, they are considered relatively weak in their ability tosupport causal inferences. However, given the nature of the research questions for this study,randomised controlled trials may not be suitable because the variables under examination cannotbe manipulated or controlled (i.e., self-efficacy). Despite the weakness of the study designs, thefindings from descriptive correlational observational studies are informative and can offer awealth of evidence about phenomena for which little is known or understood. They also build astrong foundation from which future research can be conducted and research questionsgenerated.Study target populations and settingsAdherence to a methicillin-resistant Staphylococcus aureus decolonization protocol hasnot been thoroughly examined in the total joint arthroplasty population. To assist in the quest fora clearer understanding of the factors that influence adherence to such protocols, it was necessaryto examine other patient populations where adherence has been more closely examined: the areasof chronic and cardiovascular disease, HIV/AIDS, cancer, and mental health. Across the 15studies selected for review, samples were drawn from these specific patient populations.All 15 studies used non-probability sampling, specifically convenience sampling, torecruit study participants (Barclay et al., 2007; Birkenfeld et al., 2011; Boeka et al., 2010; Chia etal., 2006; Curtin et al., 2008; Fawzi et al., 2012; Ingersoll & Cohen, 2008; Liang, et al., 2008;Ngamvitroj & Kang, 2007; Osterberg & Blaschke, 2005; Peddle et al., 2009; Shi et al., 2010;16Tarantino et al., 2010; Taylor et al., 2011; Yuhas et al., 2012). The use of convenience samplinghas advantages; it is a very rapid, feasible, and inexpensive method of recruitment to ensure thata sufficient number of study participants are enrolled. It lends itself well to research studies thataim to collect data regarding trends and relationships between variables that are poorlyunderstood. However, despite these advantages, the use of convenience sampling likelyintroduces bias. With this method of sampling, subjects are not randomly selected, and thus thesample may not be representative of the population to which the study is intended to begeneralized. Furthermore, similar to other methods of sampling, many factors, which cannot becontrolled, can influence participants? decisions about whether to participate in research. Thesefactors then act as confounding variables when interpreting the study results. Because adherenceis still a poorly understood phenomenon, it is easy to appreciate the wide use of conveniencesampling in studies that have examined it in great depth.A wide range of settings have been used to examine the relationships between age,self-efficacy, and adherence. Outpatient clinical settings specific to a wide range of patientpopulations were the most prevalent locations in which the research has been carried out(Barclay et al., 2007; Birkenfeld et al., 2011; Chia et al., 2006; Schlenk, & Dunbar-Jacob, 2006;Fawzi et al., 2012; Ingersoll & Cohen, 2008; Liang et al., 2008; Ngamvitroj & Kang, 2007; Shiet al., 2010; Taylor et al., 2011; Yuhas et al., 2012). Another common setting was inpatienthospital units. Curtin et al. (2008) examined the relationship between self-efficacy and the self-management behaviour of patients with chronic kidney disease in two inpatient renal units.Similarly, Boeka et al. (2010) based their study in a surgical unit of a major medical centre in thesoutheastern United States to examine the psychosocial predictors of intentions to comply withbariatric surgery guidelines for patients scheduled to undergo gastric bypass surgery. Tarantino17et al. (2010) set their study in a general medicine unit to examine socio-cognitive factorsassociated with non-adherence to medication prescriptions. A unique setting, outside of thehospital environment, was a university research fitness centre in Western Canada in whichadherence to a pre-operative, supervised exercise regimen was evaluated with patients awaitingsurgical removal of malignant lung lesions. It is apparent from this wide array of study settingsthat the relationship between adherence and socio-demographic variables and self-efficacy ismulti-faceted and can be examined in diverse study populations and treatment regimens.Adherence and self-efficacy are such broadly defined concepts in the literature and, with diversepatient populations and settings examined, there is consequently much ambiguity in theinterpretation and meaning assigned to these concepts. Given this state, caution must be used ingeneralizing the results from other patient populations to the total joint arthroplasty population.Despite this limitation, however, there is a wealth of evidence that can be considered whenexamining the documented relationships among these variables in other patient populations; thisevidence can guide and enhance research in the area of orthopaedics.Variables examinedBefore discussing the findings of research that has examined the relationships betweensocio-demographic variables, namely age, self-efficacy, and adherence, it is important to firstreview the types of variables examined in the selected studies. To examine what is known aboutthe relationship between age and adherence it was necessary to review studies that examined therelationship between various socio-demographic variables and adherence or compliance andwherein age was specifically noted and examined. Birkenfeld et al. (2011) conducted a study thatexamined whether socio-demographic factors (i.e., age, gender, socioeconomic status, andimmigrant status) were associated with compliance in faecal occult blood testing in a screening18program for colorectal cancer. Although this study met the inclusion criteria for this literaturereview, it was problematic in that the sample was limited to adults aged 50-75 years of age. It isnot known whether the relationship between age and adherence in those 50-75 years of age is thesame as in those older than 75 or younger than 50 years of age. Barclay et al. (2007) examinedwhether the four dimensions of the Health Belief Model and closely related constructs of themodel, self-efficacy and locus of control, were predictive of adherence to antiretroviral therapyin HIV-positive adults. They hypothesized that these dimensions and constructs predictedadherence in both younger and older adult cohorts. They predicted higher rates of medicationadherence in the older cohort. Although their study included adults of all ages, there was adisproportionate number of men compared with women in their study. Again, it is not knownwhether age-associated predictors of medication adherence vary by gender. Tarantino et al.(2010) examined socio-cognitive factors associated with non-adherence to medication across awide range of ages, 35 to 80 years of age. Using elements of the Health Belief Model, theProtection Motivation Theory, and the Theory of Planned Behavior, they examined whether ageand cognitive factors, including perceived risks and benefits of non-adherence, perceptions ofsusceptibility to disease, and perceived health values were associated with non-adherence tomedication prescriptions. This study drew attention to an important gap in the literature. It isuncertain whether perceived risks and benefits of non-adherence, perceptions of diseasesusceptibility, and the value we assign to health change with age. To close this literature gap,further research is warranted.To examine the relationship between self-efficacy and adherence, it was necessary tobroaden the keyword search to include words or word combinations that captured some of thekey constructs of the Health Belief Model and self-efficacy theory (i.e., health beliefs, health19perceptions, patient attitudes, and health behaviours), two theoretical models with a high degreeof interconnectivity. Fawzi et al. (2012) examined whether health beliefs about medicationspredicted older adults? adherence to prescribed antidepressants. This study, although informative,was limited in the generalizability of its results; it likely is not generalizable to younger agegroups and to the total joint arthroplasty population because the study included only patients 55years of age and older. Furthermore, there may be other confounding factors specific to themental health population that may influence adherence. Chia et al. (2006) examined therelationship between medication adherence and factors such as self-efficacy, medicationefficacy, beliefs in one?s control over one?s health, and illness perceptions. Similar to the work ofChia et al. (2006), Liang et al. (2008) explored the relationships between adherence to prescribedopioid administration and opioid-taking self-efficacy and beliefs. These latter two studies pointto another gap in the literature: the lack of differentiation between personal self-efficacy andmedication administration efficacy. The two concepts are conceptually different and likelyinfluence adherence in different ways. Study outcomes can vary depending on which construct(s)of self-efficacy are examined. For example, Boeka et al. (2010) explored the association betweenadherence and the perceived threat of not following guidelines, as well as the associationbetween perceived self-efficacy and adherence to bariatric surgery guidelines. The results of thisstudy were of particular interest because the context in which adherence was examined (i.e., inrelation to patient guidelines) is similar to the way adherence was examined in the current study,in relation to patient guidelines to a decolonization protocol. Other variables explored in thepublished studies included patients? attitudes and perceptions about a treatment or a disease, andother elements of self-efficacy (e.g., Yuhas et al., 2012).In the 15 studies included in this review, the outcome variable, adherence, was considered20with respect to several therapies: (a) medication use, (b) pre-operative exercise regimens andcardiac rehabilitation programs, (c) prophylactic treatment, and (d) patient and surgicalguidelines (Barclay et al., 2007; Birkenfeld et al., 2011; Boeka et al., 2010; Chia et al., 2006;Curtin et al., 2008; Fawzi et al., 2012; Ingersoll & Cohen, 2008; Liang et al., 2008; Ngamvitroj& Kang, 2007; Osterberg & Blaschke, 2005; Peddle et al., 2009; Shi et al., 2010; Tarantino et al.,2010; Taylor et al., 2011; Yuhas et al., 2012). It is clear from this discussion that there are anumber of approaches that have been taken and numerous outcomes that have been explored inthe selected literature to examine the influence of age and self-efficacy. This reflects thecomplexity of these variables and the dynamic relationships among them.Statistical analysesThe studies selected for this literature review varied in their approaches to examining therelationships between age, self-efficacy, and adherence. Analytical statistics, specificallycorrelational techniques (i.e., Pearson correlations), were the most prevalent statistical tests usedto describe the relationships between self-efficacy and adherence (Chia et al., 2006; Curtin et al.,2008; Liang et al., 2008; Ngamvitroj & Kang, 2006; Peddle et al., 2009; Tarantino et al., 2010;Taylor et al., 2011). In some instances, the researchers used more than one statistical test or acombination of statistical tests. Regression analyses were often conducted after a relationshipwas established to determine the proportion of variance explained or the contribution ofparticular variables to the variability in adherence (Barclay et al., 2007; Boeka et al., 2010; Lianget al., 2008). In studies that explored socio-demographic variables in relation to adherence,descriptive statistics were often used to summarize the data and chi-square statistics were used todetermine the relationships among the demographic variables (Barclay et al., 2007; Curtin et al.,2008; Ngamvitroj & Kang, 2007; Tarantino et al., 2010). In studies that explored differences in21adherence across various age groups and genders or examined differences in demographicvariables between groups categorized as adherent versus non-adherent, statistical tests capable ofmeasuring within and between group differences, such as independent t tests or ANOVA, wereused (Birkenfeld et al., 2011; Fawzi et al., 2012; Peddle et al., 2009; Tarantino et al., 2010).Evident from the discussion thus far is the prevalence of quantitative approaches to examiningthe role of age and self-efficacy in relation to adherence. In one instance, a qualitative approachwas used. Yuhas et al. (2012) examined patients? attitudes and perceptions of implantablecardioverter-defibrillators (ICDs) as primary prophylaxis for those at risk for sudden cardiacdeath. Given that ICDs are underutilized, Yuhas et al. (2012) sought to obtain a betterunderstanding of the potential barriers to consenting to have an ICD implanted. Using groundedtheory to extrapolate some common themes that surfaced from the perspectives of affectedpatients, they strived to gain an understanding of the factors that influence patients? decisions toconsent to the implantation of prophylactic cardioverter-defibrillators. The use of bothquantitative and qualitative research designs to address these research questions has deepenedour understanding of the relationships between age and self-efficacy, and has generated morequestions to drive future research.Instruments Used for the Measurement of Relevant ConceptsSelf-efficacyThe measurement of self-efficacy is plagued with shortcomings and difficulties.Numerous conceptual definitions of self-efficacy exist. Elements and constructs from theoreticalmodels such as the Health Belief Model, Bandura?s (1977) self-efficacy theory, Theory ofPlanned Behavior, and the Protection Motivation Theory have guided the design of self-efficacymeasurement instruments (Ajzen, 1991; Barclay et al., 2007; Boeka et al., 2010; Peddle et al.,222009; Tarantino et al., 2010). Health beliefs, health perceptions, perceived susceptibility or riskof developing an illness or disease, and the belief in one?s capability to perform a particularhealth behaviour are some of the constructs of self-efficacy that stem from the Health BeliefModel and Bandura?s (1977) self-efficacy theory (Sirur, Richardson, Wishart, & Hanna, 2009).Perceived behavioural control, attitude towards performing a behaviour, and the degree ofcontrol patients believe they have over their health are other constructs related to self-efficacythat have been derived from Bandura?s (1977) self-efficacy theory and the Protection MotivationTheory (Rogers & Prentice-Dunn, 1997). Irrespective of patient population or health domain,researchers that have examined the relationship between self-efficacy and adherence haveconceptually defined self-efficacy based on constructs and elements from these theoreticalmodels. Consequently, these constructs and elements have been used as a framework indesigning instruments to measure self-efficacy.Taylor et al. (2011) conducted a systematic review to explore the medical, psychological,and socio-demographic factors associated with adherence to cardiac rehabilitation programs. TheIllness Perceptions Questionnaire-Revised (IPQ-R) (Moss-Morris et al., 2002) was used toexamine illness perceptions, with eight different subscales, in 2 of the 18 studies included in theirsystematic review. Even though a medium to large positive association was found between theperceived consequences of developing an illness and adherence, in their meta-analysis,assessments of the instrument?s reliability and validity were not reported (Taylor et al., 2011).With a lack of evidence of internal consistency, test-retest reliability, and validity (i.e., content orconstruct validity), the validity and robustness of a tool are questionable and caution must beused in the interpretation of the overall study results. Chia et al. (2006) conducted a systematicreview of published research studies to examine the relationships between self-efficacy, and23other belief-laden variables, and medication adherence in the elderly. Four studies in this review,all using different measuring instruments, examined self-efficacy. Two of the studies used a self-efficacy and outcome expectation scale to measure self-efficacy (Brus et al., 1999; Clark &Dodge, 1999), another study used a 12-item questionnaire (Siegel, Karus, & Schrimshaw, 2000),and the fourth study used the Beliefs Related to Medication Adherence Survey (McDonald-Miszczak et al., 2004). No measures of instrument reliability and validity were reported in thisliterature review. Fawzi et al. (2012) also used the Beliefs Related to Medication AdherenceSurvey to examine whether beliefs about medication could predict adherence to antidepressantsin older adults and again no measures of instrument reliability and validity were reported.Barclay et al. (2007) examined health beliefs and attitudes using a questionnaire based on theHealth Belief Model known as the Adherence Determination Questionnaire (DiMatteo et al.,1993). They also examined self-efficacy by exploring patients? locus of control using theMultidimensional Health Locus of Control Scale (Wallston, Wallston, & DeVellis, 1978). Inboth instances, no documentation of the reliability and validity of the instruments was provided.The operationalization of self-efficacy is problematic because (a) multiple definitionsexist and (b) depending on how self-efficacy is conceptually defined, a reliable and validinstrument may not exist to operationalize the construct. Inadequate reporting of instruments?reliability and validity is a widespread issue in the literature exploring self-efficacy andadherence, which is likely an issue in other areas of research, as well. Caution must be used ininterpreting the findings from these studies because, without confirming that an instrument isreliable and valid, there is greater risk for the introduction of bias and error. Depending on thedegree of bias and magnitude of error introduced, the quality of a study can be compromised.Despite these limitations, however, these studies have contributed some knowledge to our24understanding of self-efficacy. Furthermore, in instances where a study has not determined orreported instrument reliability and validity, our focus should instead be directed to studies wherethis information or evidence has been provided. This will allow for a more accurate synthesis ofthe overall findings.With the exception of two studies included in this review, all studies relied on self-reportmethods, either in the form questionnaires or surveys, to measure self-efficacy. Five studiesreported a Cronbach alpha (Boeka et al., 2010; Curtin et al., 2008; Liang et al., 2008; Ngamvitroj& Kang, 2007; Peddle et al., 2009). Using constructs derived from Ajzen?s (1991) Theory ofPlanned Behavior, Peddle et al. (2009) examined the relationships between perceivedbehavioural control, attitudes towards a pre-operative exercise program, intention to comply withthe exercise program, and actual adherence to the program. Subjects completed a questionnairethat generated a personal score for each of these constructs. Cronbach alpha for each constructwas reported in the range of .72 and .97, depending on the construct, an acceptable indicator ofinternal consistency. Similarly, Curtin et al. (2008) measured the relationship between self-efficacy and medication adherence in patients with kidney disease. They administered avalidated questionnaire, known as the Perceived Efficacy in Patient Physician InteractionQuestionnaire, and reported Cronbach alphas of .70 to .92. Comparable values were found in astudy conducted by Ngamvitroj and Kang (2006), who examined the effects of self-efficacy onadherence to peak expiratory flow rate monitoring in asthma patients. They administered avalidated questionnaire, the Knowledge, Attitude, and Self-efficacy Asthma Questionnaire, tomeasure self-efficacy and reported Cronbach alphas of .86 to .92. Liang et al. (2008) reported aCronbach alpha, a test-retest reliability measure, and construct validity evidence for theirmeasurement of opioid taking self-efficacy. The Pain-Opioid Analgesic Beliefs Scale and the25Opioid-Taking Self-efficacy Scale-Cancer had Cronbach alphas of .74 to .84 and .93 to .95,respectively. Test-retest reliability coefficients of .94 and .68 to .82, respectively, were reportedand the construct validity coefficients were .84 and .95, respectively (Liang et al., 2008).In summary, in the five studies that reported reliability assessments, the Cronbach alphavalues ranged from .65 to .92, which are considered acceptable according to establishedmeasurement standards (Boeka et al., 2010; Curtin et al., 2008; Liang et al., 2008; Ngamvitroj &Kang, 2007; Nunnally & Bernstein, 1994; Peddle et al., 2009). Researchers, however, should bemaking greater effort to include evidence of their instruments? reliability and validity, whenappropriate, to establish greater rigor in their studies.Measuring adherenceSimilar to the measurement of self-efficacy, the operationalization of adherence presentsits own set of challenges. Defining and measuring adherence is difficult primarily because of thelack of a consensual definition of adherence (Fawzi et al., 2012). It is important to ascertainwhether adherence is defined in terms of timing, dosing, attendance, or other considerationswhen evaluating patients? responses to treatment recommendations. It is also important toestablish what is meant by adequate adherence. Is a rate of 80% considered adequate? Is a rate of95% a more appropriate indicator of good adherence? (Osterberg & Blaschke, 2005) The answerto these questions may be dependent on the type of patient population being studied and thecontext in which adherence is examined. Measuring adherence to retroviral therapy in HIV-positive adults may warrant an adherence rate closer 95% to be considered adequate forachieving optimal patient outcomes, whereas an adherence rate of 60% or 70% may beconsidered adequate when measuring adherence to baby aspirin in healthy older adults taking itfor prophylaxis. Issues such as these have left room for much ambiguity when it comes to26defining adherence and deciding on the most appropriate method to measure it.In a literature review of 41 articles, Osterberg and Blaschke (2005) identified both directand indirect methods of measuring adherence. Direct methods include measures of drugconcentrations or metabolites in the blood or urine and measurement of biological markers in theblood (Osterberg & Blaschke, 2005; Shi et al., 2010). Direct methods can be very costly, theycan be labour intensive for health providers, and they may be susceptible to distortion by thepatient or other physiological factors (Osterberg & Blaschke, 2005; Shi et al., 2010). They can,however, provide accurate results in studies that examine medication adherence in high riskpopulations. Depending on the research question or the variables being examined, direct methodsare not always suitable. In some instances it may be necessary to consider the use of an indirectmethod of measurement, the most popular method used in the studies selected for this literaturereview.Indirect methods of measurement include the use of electronic monitoring devices, pillcounts, the tracking of prescription refills, patients? medication diaries, self-reported behaviour,and so forth (Osterberg & Blaschke, 2005; Shi et al., 2010). Self-reported questionnaires orsurveys have been the most widely used methods of measurement in the adherence literature(Barclay et al., 2007; Birkenfeld et al., 2011; Boeka et al., 2010; Chia et al., 2006; Curtin et al.,2008; Fawzi et al., 2012; Ingersoll & Cohen, 2008; Liang et al., 2008; Ngamvitroj & Kang,2007; Osterberg & Blaschke, 2005; Peddle et al., 2009; Shi et al., 2010; Tarantino et al., 2010;Taylor et al., 2011). Although self-report methods are convenient, easy to administer, oftengenerate data quickly, and are relatively inexpensive, their accuracy is uncertain (Shi et al.,2010). Self-report methods are subject to many forms of bias including recall and response biasas well as social desirability bias and the Hawthorne effect (Shi et al., 2010). There also are27many confounding factors that influence responses to questionnaires or surveys and thissignificantly limits the precision, credibility, and validity of the data collected (Shi et al., 2010).Another limitation of self-report methods was identified in a literature review that examined therelationship between self-reported adherence questionnaires and electronic medicationmonitoring devices (Shi et al., 2010). Given the moderate to strong correlations found betweenself-administered questionnaires and electronic monitoring devices in terms of measuringadherence, Shi et al. (2010) still found a higher reported rate of adherence through self-reportmethods, approximately 14.9% higher, when compared with the rates derived from electronicmonitoring devices (Shi et al., 2010). These results point to an important limitation of self-reportmethods, the tendency for patients to overestimate their level of adherence. Lastly, similar to themeasurement of self-efficacy, there is incomplete reporting of the reliability and validity of theinstruments used to measure adherence (Fawzi et al., 2012). Only one study, conducted by Boekaet al. (2010), reported evidence related to the reliability of their instrument. They administered anIntention to Comply Questionnaire to assess compliance with bariatric surgery guidelines andreported a Cronbach alpha of .66 with respect to that questionnaire (Boeka et al., 2010).Electronic medication monitoring devices, another indirect method to measure adherence,can be used in studies that examine medication adherence. Barclay et al. (2007) used amedication event monitoring system (MEMS) to measure adherence in HIV-positive adults. Thisdevice attaches to a medication vial and records the date and time of each bottle opening (Shi etal., 2010). Although a good method for providing precise information with regard to patients?medication-taking behaviour, it is limited in its ability to monitor whether patients actually ingestthe medication and, if so, whether they ingest the correct dose (Osterberg & Blaschke, 2005).In five of the studies included in this review, adherence was measured simply by28documenting whether a patient adhered to a regimen (Birkenfeld et al., 2011; Curtin et al., 2008;Ngamvitroj & Kang, 2007; Peddle et al., 2009; Taylor et al., 2011). Adherence rates were thencalculated by dividing the observed level of adherence by the expected level of adherence.Although this method of measurement lacks objectivity, and caution must be used in interpretingthe overall study results, this information can enhance our understanding of adherence.Research FindingsAge and adherenceThe relationship between age and adherence is one of complexity. Contributing to thiscomplexity are: (a) the existence of co-variables that can influence the magnitude and directionof the relationship, (b) the overall inconclusive results with respect to the magnitude of therelationship, and (c) the overall inconsistency in the direction of the relationship when arelationship between age and adherence has been detected (Birkenfeld et al., 2011; Chia et al.,2006; Tarantino et al., 2010; Taylor et al., 2011). As mentioned previously, adherence to adecolonization protocol has been largely understudied in the total joint arthroplasty populationmaking it necessary to consider other health domains and patient populations where adherencehas been examined. Adherence to medication, patients? instructions, surgery guidelines, andphysician-recommended exercise programs are just a few examples of the types of healthdomains where the relationship between age and adherence has been examined. Although thegeneralizability of the results to the orthopaedic population and the decolonization protocol,specifically, is limited, these areas have contributed to our understanding of the relationshipbetween adherence and age.Taylor et al. (2011) conducted a systematic review of the medical, psychological,and socio-demographic factors associated with adherence to cardiac rehabilitation programs.29They found a statistically significant relationship between age and adherence in 6 of the 18studies examined, but the direction of the relationship was inconsistent across those studies(Taylor et al., 2011). Why such variation in the results? Inconsistencies in how adherence wasdefined and measured within those studies, the research design that was used, the characteristicsand size of the samples, and the presence of co-variables could explain the equivocal results.Although this review did not shed light on the adherence behaviour of older and younger adults,the authors discussed some of the factors that might influence adherence across various agegroups. Co-variables, such as the perception of personal control over an illness or disease or thepresence of life commitments to work or family, could be mediating factors in the relationshipbetween age and adherence (Taylor et al., 2011). Further research is needed with morerepresentative samples, an explicit definition and reliable and valid measurement of adherence,and more robust and rigorous methods to further examine the contribution of these co-variablesto the overall interpretation of study outcomes (Taylor et al., 2011). Similarly, more researchabout age-related differences in adherence behaviour across various health fields is needed tobetter understand the relationship between age and adherence and to improve the generalizabilityof the study outcomes. Although the findings from this study did not generate conclusive resultswith respect to the magnitude and direction of the relationship between age and adherence, thefindings do reflect the state of the literature. In other words, there is considerable variance amongthe outcomes of studies that have examined age and adherence.Barclay et al. (2007) examined age-associated predictors of medication adherence inHIV-positive adults and found that adults older than 50 years of age had significantly greaterrates of adherence to antiretroviral therapy than did those under 50 years of age. Theparticipants? ages ranged from 25 to 69 years. When the researchers examined the poor adherers30in each age group, they found a greater percentage of non-adherers in the under-50-year-agegroup compared with the over-50-year-age group (Barclay et al., 2007). They found positiverelationships between non-adherers in the under-50-age group and drug dependency and a lackof independent financial resources. In other words, younger HIV-seropositive adults who werereliant on family, friends, and government programs for income, were less likely to adhere. Theyalso found lower levels of social support, weaker internal locus of control, and more perceivedbarriers to treatment in this younger age group (Barclay et al., 2007). Poor adherence in the oldergroup was significantly and solely related to cognitive changes (Barclay et al., 2007). Theyspeculated, like Taylor et al. (2011), that adherence rates may be greater in older adults becausemedication adherence in older individuals does not require as many lifestyle changes and, ifalterations are necessary, they find incorporating these changes to be less of an inconvenience(Barclay et al., 2007). Furthermore, older individuals may perceive the risk of mortality withnon-adherence to be greater than those of a younger age (Barclay et al., 2007). Thegeneralizability of these findings to the elective orthopaedic population and to the decolonizationprotocol, specifically, is questionable. The HIV population is unique in its attributes. The socialstigma and barriers encountered by this population in terms of accessing health resources andmedication are not attributes necessarily shared by other patient populations. Hence, the factorsthat influence medication adherence in this population are not easily generalized to other patientpopulations. Nonetheless, this study identifies some very important psychological determinantsof adherence that will be discussed later in this review, including the influence of health beliefs,self-efficacy, and perceived risk on adherence (Barclay et al., 2007).Birkenfeld et al. (2011) examined factors affecting compliance in faecal occult bloodtesting for colorectal cancer screening in Israeli adults aged 50 to 75 years of age. They found31that those aged 60 years of age and older had significantly greater adherence rates and morewomen than men adhered (Birkenfeld et al., 2011). The generalizability of these results,however, is questionable because of the potential influence of cultural factors. Replicating thisstudy in Canada could generate very different results to the ones obtained in Israel due, in part, tocultural differences between the two countries. The influence of cultural differences wasapparent in the Netherlands when a replication of this study did not generate similar results(Birkenfeld et al., 2011). Ethnicity and cultural differences should be examined to a greaterextent in future studies to ascertain the generalizability of study results across ethnically diversepatient populations globally (Chia et al., 2006).Tarantino et al. (2010) examined socio-cognitive factors in relation to non-adherence tomedication prescriptions in adults aged 35 to 80 years of age, who were recently discharged fromhospital. Similar to Birkenfeld et al. (2011), they found greater adherence rates in the over-65age group when compared with the under-65 age group, but unlike Birkenfeld et al. (2011), theseresults were not modified by gender. Similar to Taylor et al. (2011) and Barclay et al. (2007),Tarantino et al. (2010) studied lifestyle factors as possible explanations for these findings.Younger patients who are employed and raising a family often have very busy lifestyles. Workor family commitments place relatively younger patients, aged 35 to 65 years, at greater risk fornon-adherence or ineffective and inaccurate management of medication (Tarantino et al., 2010).To speculate that lifestyle alone drives non-adherence may be short sighted, but it is nonethelessan important factor to consider when interpreting the findings of research that has explored therelationship between age and adherence.The discussion thus far has focussed on research domains that have shown greateradherence rates in older adults. There is however a body of literature that reveals the converse,32the vulnerability of the older population and the multitude of factors that often put thispopulation at risk for non-adherence. The manner in which psychosocial, cognitive, andcontextual factors play out in relation to adherence is strongly linked to age (Park & Meade,2007). Psychosocial factors, such as personal beliefs and motivational states, and cognitivefactors, such as memory and comprehension, often change with age (Park & Meade, 2007). Thebeliefs and perceptions that people have of their health and the factors that motivate and drivepersonal behaviour look very different at the age of 40 than they do at the age of 80. Similarly,contextual factors, such as daily commitments and routines or level of busyness, also changewith age (Park & Meade, 2007). As one ages, these changes may not favour adherence,especially when the regimen to which they are adhering is complex (Park & Meade, 2007).?Effortful cognitive processes,? Park and Meade (2007) explained, become more difficult as oneages and becomes more cognitively frail. The ability to consciously interpret, comprehend, andtranslate instructions into action diminishes with age, and this, more so than memory, is thegreatest barrier to adherence in older adults (Park & Meade, 2007). These factors were found tobe most significant in those aged 71 years of age and older (Park & Meade, 2007).Although Park and Meade (2007) did not identify memory as the greatest barrier toadherence in the older population, many researchers that have examined the relationshipbetween age and adherence have identified memory, particularly prospective memory decline, asa strong determinant of adherence in older adults (McDaniel & Einstein, 2007; Tarantino et al.,2010). Prospective memory is what allows one to remember to perform a particular action at atime when it is intended to be done (McDaniel & Einstein, 2007). Time-based prospectivememory relates to the memory of performing everyday tasks, such as remembering to take amedication, which tends to diminish with age-related cognitive decline in adults aged 61 to 7633years of age, but not in those under the age of 61 years (McDaniel & Einstein, 2007). Given thatfollowing physicians? instructions is a prospective memory task, specifically a time-based task, itis easy to appreciate the relevance of these findings in the examination of age and adherence topatient instructions or guidelines for a decolonization protocol.When presented with a time-based task, such as following patient instructions, Liu andGonzalez (2007) explained that one of the ways that older adults often compensate for age-related cognitive decline is by placing greater reliance on ?cognitive shortcuts.? An example of a?cognitive shortcut? might be the memory of prior successes and failures with performing thesame task in the past, a construct of self-efficacy referred to as ?vicarious experience? inBandura?s (1977) self-efficacy theory. Understanding how older adults compensate for age-related decline and use their past experiences to make decisions and judgments has significantclinical implications when designing instructions for older adults and expecting that they willadhere to them (Liu & Gonzalez, 2007).In summary, there is considerable variance among the outcomes of studies that haveexamined the relationship between age and adherence. There is a large body of literature thatshows greater adherence rates, to medication and physician-recommended exercise programs andscreening guidelines, in adults over 50 years of age compared with those under 50 years of age(Barclay et al., 2007; Birkenfeld et al., 2011; Tarantino et al., 2010; Taylor et al., 2011). There ishowever a body of literature that reveals the converse, the vulnerability of the older adultpopulation and the multitude of factors that often put this population at risk for non-adherence(Liu & Gonzalez, 2007; McDaniel & Einstein, 2007; Park & Meade, 2007). The potentialinfluence of co-variables, such as contextual or lifestyle factors and cognitive, psychological, andsocial factors, on the relationship between age and adherence remains unclear. Similarly, the34social stigma and barriers encountered by certain patient populations, (i.e., the HIV population),in terms of accessing health resources and medication are not attributes necessarily shared byother patient populations. These factors can limit the generalizability of study outcomes fromthese patient populations to other patient populations, making the interpretation of researchfindings from studies that have examined the relationship between age and adherence, morecomplex.Self-efficacy and adherenceUnlike the relationship between age and adherence, where the overall findings from theliterature yield inconclusive results with respect to the magnitude and direction of therelationship, the literature about the relationship between self-efficacy and adherencepaints a very different picture. As previously mentioned, elements and constructs fromtheoretical models such as the Health Belief Model, Bandura?s (1977) self-efficacy theory,the Theory of Planned Behavior, and the Protection Motivation Theory have been the basis ofmost research studies that have explored the relationship between self-efficacy and adherence.They also have guided the design and selection of instruments used to measure self-efficacy(Ajzen, 1991; Barclay et al., 2007; Boeka et al., 2010; Peddle et al., 2009; Tarantino et al., 2010).Chia et al. (2006) conducted a systematic review of the influence of self-efficacy, thebelief in one?s ability to perform a particular behaviour under diverse conditions, and five otherbelief-laden variables on medication adherence (Chia et al., 2006). They found a statisticallysignificant positive relationship between adults? beliefs in their ability to take medication (i.e.,medication-taking self-efficacy) and medication adherence in four of the fourteen studiesincluded in the review. In other words, adults who were confident in their ability to take amedication were more likely to adhere to the prescribed medication (Chia et al., 2006). Of the six35variables examined, self-efficacy was the only variable that significantly differed betweenadherent and non-adherent subjects (Chia et al., 2006). Similar findings were found by Curtin etal. (2008), who conducted a study of the association between patients? perceived self-efficacyand five self-management behaviours, one of which was medication adherence in patients withchronic kidney disease. After controlling for socio-demographic and health-related variables,they concluded that self-efficacy was positively and significantly associated with four of the fiveself-management behaviours, including medication adherence (r = .37) (Curtin et al., 2008). Tofurther support there being a positive relationship between self-efficacy and adherence,Ngamvitoj and Kang (2007) found similar results when they examined the effects of asthma self-efficacy, social support, and knowledge on adherence to peak expiratory flow rate (PEFR)monitoring in asthma patients. They found predictive effects between higher scores of asthmaself-efficacy and adherence to PEFR monitoring when univariate analyses were performed,however, the three variables did not contribute significantly to the total variance on adherence toPEFR self-monitoring (Ngamvitoj & Kang, 2007). Perhaps with a larger sample size and greatervariability between the variables, these results may have been statistically significant (Ngamvitoj& Kang, 2007).Another construct of self-efficacy that has been widely examined in relation to adherenceis the level of perceived risk, commonly referred to as perceived threat. Levels of perceived riskor threat associated with not following physicians? instructions or guidelines or with developingillness or disease are a few examples of measures used to predict personal adherence behaviour.In a systematic review that measured medical, psychological, and socio-demographic factorsassociated with adherence to a cardiac rehabilitation program, a large effect size of 0.77 to 0.82was found between more serious perceived consequences of developing cardiac complications36and greater adherence to cardiac rehabilitation programs (Taylor et al., 2011). In other words,patients who perceived greater health risks associated with not attending the program were morelikely to attend (Taylor et al., 2011). In one particular study included in this review, a positivecorrelation (r = .41) was also found between the level of self-efficacy and adherence to a cardiacrehabilitation program. Similar results were found by Tarantino et al. (2010) who examined theassociation between socio-cognitive factors and non-adherence to medication instructions inpatients discharged from a general medicine unit. They found statistically significant positiverelationships between medication adherence and perceived personal susceptibility to disease,perceived risks of non-adherence, and perceived absence of benefits of non-adherence (Taylor etal., 2011). In other words, patients who believed that they were putting themselves at risk by notfollowing medication instructions and who believed there were greater overall health risks withnot taking a prescribed medication as opposed to taking it, were more likely to take themedication and follow their physicians? instructions. These findings were further supported byBoeka et al. (2010) who conducted a study to examine the psychosocial predictors of intentionsto comply with bariatric post-operative surgical guidelines in adults 21 to 56 years of age. Usingconstructs derived from the Protection Motivation Theory (i.e., perceived threat and self-efficacy), Boeka et al. (2010) designed two questionnaires, one to measure self-efficacy andperceived threat and the other to measure behavioural intentions. The questionnaires werevalidated according to their reported assessments of reliability and validity. They found thathigher levels of personal self-efficacy and a greater degree of perceived threat correlated stronglyand positively with intentions to comply with the post-operative guidelines (r = .37). The extentto which behavioural intentions to adhere translate into actual adherence remains unknown.Further research would be needed before concluding that behavioural intentions could act as37reliable and valid predictors of actual adherence. Barclay et al. (2007) found similar results whenthey examined medication adherence in HIV-positive adults. Greater perceived treatment utility(i.e., belief in the efficacy of a medication or treatment) was predictive of medication adherencein adults of all ages, but heightened feelings of self-efficacy were predictive of medicationadherence only in adults younger than 50 years of age (Barclay et al., 2007). The extent to whichage influences predictors of adherence remains unclear.Health beliefs and illness perceptions, constructs derived from the Health Belief Model(Rosenstock et al., 1988), are other constructs of self-efficacy that have been used by researchersto examine the relationship between self-efficacy and adherence. Similar to the findings ofstudies that have examined self-efficacy and adherence, researchers that have examined healthbeliefs about medication use and health prevention or illness perceptions with respect to diseaseseverity or susceptibility have found strong associations between these variables and adherence(Chia et al., 2006; Fawzi et al,, 2012; Liang et al., 2008). Fawzi et al. (2012) examined whetherhealth beliefs about medications could be used to predict adherence to anti-depressantprescriptions in older adults diagnosed with depression. They found a moderate, but significantlypositive, correlation between beliefs regarding the necessity of such medications for improvingoverall health and adherence to antidepressant prescriptions (r = .34) (Fawzi et al., 2012). Inother words, those who perceived their depression to be a serious health issue and who believedthat taking antidepressant medication could improve their overall health were more likely toadhere to them. They similarly found negative correlations between adherence to anti-depressantmedication and (a) beliefs that the medication could cause harmful side effects (r = -.34), (b)perceptions that anti-depressants are over-prescribed by physicians (r = -.23), and (c) perceptionsthat medications are fundamentally harmful (r = -.21) (Fawzi et al., 2012). In a study that38examined factors influencing opioid-taking self-efficacy and analgesic adherence in Taiwanesecancer outpatients, similar results were found (Liang et al., 2008). Cancer patients with negativebeliefs about opioids demonstrated poor adherence (r = -.30), whereas patients with higherperceived opioid-taking self-efficacy showed greater adherence to the opioid regimen (r = .22)and reported greater pain relief (r = .35) (Liang et al., 2008). The degree to which the findingsfrom these two studies can help predict and explain adherence to a decolonization protocol in anelective orthopaedic population is debatable. Attributes of mental health and cancer patientpopulations, similar to the HIV population, are unique in their attributes. There are likely manyconfounding factors to consider when generalizing results from studies that have examined thesepopulations to other patient populations such as an elective orthopaedic surgery population.Despite this limitation, these studies provide a solid foundation from which researchers can adoptideas, concepts, and themes to conduct further research or replicate studies in other patientpopulations.Lastly, Peddle et al. (2009) applied constructs related to self-efficacy from the Theory ofPlanned Behavior (Ajzen, 1991) to examine adherence to a supervised exercise program inpatients awaiting surgical removal of malignant lung lesions. Constructs such as one?s perceivedconfidence and control in performing a particular action (i.e., perceived behavioural control),one?s attitude towards performing that action, and perceived social pressure to perform an action(i.e., the subjective norm) were examined in relation to adherence to the exercise program. Theyfound statistically significant correlations between adherence to the exercise program andstronger behavioural control (r = .63) and greater subjective norm awareness (r = .51) (Peddle etal., 2009). In other words, patients who were confident in their ability to do the exercise programand who felt more social pressure from their family and friends to do so were more likely to39complete the program. The influence of social pressure on personal levels of self-efficacy (i.e.,verbal persuasion) is one of the key elements of Bandura s (1977) self-efficacy theory.Based on this discussion, it is clear that the body of relevant literature showsoverwhelming support for their being a relationship between self-efficacy and adherence. Higherlevels of personal self-efficacy appear to be significantly and positively correlated withadherence. The relationship however is moderate in strength with Pearson?s product momentcorrelation coefficients (r) reported in the range of .3 to .6 (Boeka et al., 2010; Curtin et al.,2008; Fawzi et al., 2012; Liang et al., 2008; Peddle et al., 2009; Taylor et al., 2011). Theliterature findings related to the relationship between age and adherence have yielded moreinconclusive results with respect to the magnitude and direction of that relationship. Caution,however, must be used in interpreting and generalizing these overall findings because there areseveral methodological limitations that need to be acknowledged, and confounding factors thatneed to be considered, before predictions and conclusions can be made with any degree ofcertainty.Limitations of the ResearchIn studies that have examined the relationships between age, self-efficacy, and adherence,a number of methodological limitations and weaknesses have been noted. First, the field lacks aconsensual or standard definition of adherence (Fawzi et al., 2010; Ingersoll & Cohen, 2008). Itis important to ascertain whether adherence is defined in terms of timing, dosing, attendance, orother considerations when evaluating patients? responses to treatment recommendations. It isalso important to establish what is meant by adequate adherence or what an appropriate indicatorof good adherence is. In certain circumstances, 80% adherence may be considered adequateadherence, whereas in other circumstances, 90% may be a better indicator of good adherence.40These parameters are often stipulated by the type of population being examined and the type oftreatment being evaluated. Ninety-five percent may be considered a better standard or measure ofadherence in studies where high-risk patient populations are being examined for adherence withlife-altering medication prescriptions. Despite there being justification for using differentstandards of measurement of adherence, in certain circumstances, a lack of consensual definitionof adherence makes the operationalization of adherence, and comparisons of studies, difficult.Similarly, as evident in the literature and throughout this discussion, there are numerousconstructs of self-efficacy that exist. The use of these various constructs to examine a particularoutcome has undoubtedly added depth and richness to the literature and enhanced ourunderstanding and knowledge of self-efficacy. On the other hand, they have preventedresearchers from speaking a uniform language when it comes to defining self-efficacy. Second,studies that have examined the relationships between age, self-efficacy, and adherence haverelied heavily on self-reported instruments (i.e., patient questionnaires and surveys) formeasurement. Although there are many advantages to using this method of measurement, thereare many disadvantages as well. Many researchers have reported that there is a tendency forsubjects to overestimate their adherence with this method of assessment in comparison withother more objective methods (Barclay et al., 2007; Curtin et al., 2008; Fawzi et al., 2012;Ngamvitroj & Kang, 2007). There are also many forms of bias introduced in studies that haveused self-reported methods of measurement and caution must be applied in the interpretation ofthose study results (Shi et al., 2010). There is no perfect fit or gold standard of measurement forself-efficacy or adherence. When selecting an appropriate instrument for measurement,researchers must consider several factors: the study design, patient population, sample size,method of sampling, nature of the research question, and study outcomes (Shi et al., 2010).41Third, there is an overall lack of evidence of the reliability and validity of the instruments thathave been used (Ingersoll & Cohen, 2008; Shi et al., 2010; Taylor et al., 2011). Failure toconduct studies that are designed to establish the reliability and validity of instruments, or failureto report the statistical evidence, are major limitations in the literature that has examined age andits relationship with self-efficacy and adherence (Ingersoll & Cohen, 2008; Shi et al., 2010;Taylor et al., 2011). Without evidence of reliable and valid instrumentations, it is difficult toensure that accurate and precise data have been collected, and without accurate and precise data,it is difficult to draw valid conclusions or make valid predictions. Fourth, the use of conveniencesampling to recruit subjects and the use of cross-sectional observational designs to examine therelationships between age, self-efficacy, and adherence, although appropriate for the researchobjectives, are problematic. In their own way, both factors make it difficult to control forconfounding variables that may be present during the particular study period. Without the abilityto control for these variables, it is difficult to draw conclusions with any degree of certainty.Furthermore, the studies chosen for this review examined age, self-efficacy, and adherence invarious contexts: different age groups, diverse patient populations, and a range of treatments orregimens. Consequently, the generalizability of these results to an elective orthopaedic surgerypopulation, and to a decolonization protocol specifically, is uncertain. Nonetheless, these studieshave provided some evidence of the relationships between these variables and they have formeda solid foundation from which further research can be generated.42Chapter 3: MethodsThis study examined whether there were relationships between age and adherence andself-efficacy and adherence to an established pre-operative Staphylococcus aureusdecolonization protocol, in the total joint arthroplasty population, at a single health institution.The study hospital performed approximately 600 elective total joint replacement surgeriesannually with primary hip and knee replacements comprising approximately 90% of that total (L.Archer, personal communication, July 10, 2012). On March 5, 2012, the hospital implemented apre-operative screening and decolonization protocol for their primary hip and knee replacementpatients (L. Archer, personal communication, July 10, 2012). This protocol mandated that allprimary hip and knee replacement patients undergoing surgery be screened for nasal carriage ofStaphylococcus aureus approximately two weeks before surgery and wash with chlorhexidinegluconate cloths pre-operatively. Key aspects of the decolonization protocol included theapplication of chlorhexidine gluconate cloths the evening before surgery and on the day ofadmission for surgery, and the administration of nasal Mupirocin, twice daily for five daysbefore surgery, for those who tested positive for nasal carriage of Staphylococcus aureus. Thisstudy specifically examined whether age and level of self-efficacy influenced the patients?adherence to the established decolonization protocol.Study DesignA descriptive, correlational study design was used to examine the relationships betweenage and adherence and level of self-efficacy and adherence.Sampling PlanSample populationAll patients undergoing a primary total hip or knee replacement surgery between May 1,432013 and October 1, 2013 were eligible to participate in the study. Patients had the freedom andright to accept or refuse the invitation to participate.Inclusion and exclusion criteriaAll patients, men and women, 19 years of age and older, scheduled for a primary total hipor knee replacement surgery between May 1, 2013 and October 1, 2013, at the hospital, wereeligible to participate. Eligible participants had to be competent to make an informed decisionand they had to be proficient in the English language; that is, they were required to read andwrite English proficiently.Patients undergoing other orthopaedic procedures including hip and knee revisions andankle and shoulder replacement surgeries were excluded from this study. Patients who had ahistory of an allergic reaction to either chlorhexidine gluconate or Mupirocin or for whateverreason were unable to administer either of these treatments, perhaps due to a physical or mentaldisability, were excluded. There have been rare reports of patients reacting to topicalchlorhexidine gluconate. Two studies reported that 15% to 27% of sampled patients experiencedskin inflammation and none had a hypersensitivity reaction (Mimoz et al., 2007; Valles et al.,2008). In a study of 586 patients, from 18 facilities, Kallen, Patel, and Hess (2011) reported that17% of hemodialysis patients, who had chlorhexidine gluconate applied for the care of centralline insertion sites, were perceived to be chlorhexidine gluconate intolerant. In clinical trials ofBactroban nasal ointment (Mupirocin) reported by the manufacturer, less than 1% of 2,340participants were withdrawn from the research because of adverse events. The most frequentadverse events reported were headache, rhinitis, respiratory disorder, including upper respiratorytract congestion, pharyngitis, and taste perversion (GlaxoSmithKline, 2009).Eligible participants could not be current patients of another health facility or residents of44an extended care facility. All participants were required to be outpatients. This exclusioncriterion was necessary because hospitals and extended care facilities often prohibit patients fromself-administering medications. Because adherence, in this study, was operationalized in terms ofwhether a participant washed with the six chlorhexidine gluconate cloths the evening beforesurgery, whether they applied those cloths to the correct body areas and, if applicable, whether aparticipant filled the required prescription for Mupirocin, began its application as advised, andself-administered the recommended number of applications, eligible participants were requiredto be autonomous in following the decolonization protocol. For this reason, only outpatientswere considered eligible.Lastly, eligible participants had to be residents of Metro Vancouver or the LowerMainland of British Columbia. Out-of-town patients were eligible to participate in the study aslong as they attended their pre-operative appointments two weeks before surgery. Out-of-townpatients were usually not seen in the hospital?s pre-surgical screening clinic until two or threedays before surgery, whereas in-town patients were seen two weeks before surgery. If out-of-town patients were screened for nasal carriage two to three days before surgery and their cultureswere found to be positive, they would not have had the opportunity to begin the administration ofnasal Mupirocin for the recommended five days before surgery. Therefore, unless out-of-townpatients were willing to attend their pre-operative appointments two weeks before surgery, theywere not eligible to participate in the study.Recruitment of participantsConsecutive sampling was used for the study recruitment. Two weeks before surgery, atthe time of the patients? scheduled pre-operative appointments, nurses in the pre-surgicalscreening clinic gave eligible patients an information package that included an invitation and45consent form. The nurses in the pre-surgical screening clinic advised prospective participants toreview the invitation and, if willing to participate, sign and return the consent form to thesurgical day care unit on the day of surgery. If the prospective participants had any questionsregarding the study, they had the opportunity to have their questions answered by contacting theresearch team via the contact information indicated on the study invitation.Power analysisDiMatteo, Haskard, and Williams (2007) conducted a meta-analysis of the relationshipsbetween patients? adherence and patients? ratings of their health status, perceived diseaseseverity, and beliefs about the threat of their disease, factors reflective of the principles andcomponents of both self-efficacy theory and the Health Belief Model. They calculated the effectsizes to be .25, .22, and .32, respectively. Using the findings from this meta-analysis, a mediumor moderate effect size of .30 was specified for an a priori power analysis (Cohen, 1992). Thepower analysis was conducted to determine the sample size needed to detect a relationshipbetween patients? age and adherence and self-efficacy and adherence using correlationestimation for the statistical analysis. Specifying power of .80, a two-tailed alpha probabilitylevel of .05, and a medium effect size of .30, a total sample size of 84 participants was needed.To account for an estimated attrition rate of 10% because of surgeries being cancelled orpostponed due to medical reasons or other unforeseen circumstances, a sample size of 100participants was sought.Procedures and Data CollectionThe established decolonization protocolThe established decolonization protocol, as described here, was followed; nomodifications were made for the purpose of the study. In the pre-surgical screening clinic, the46patients were screened for nasal carriage of Staphylococcus aureus and were given a prescriptionfor nasal Mupirocin ointment along with a patient instruction pamphlet. The clinic?s nursesreviewed the instruction pamphlet with the patients, who were advised not to fill the prescriptionunless they received a call from a pre-surgical screening nurse advising them to do so. Thepatients had each nostril swabbed by a registered nurse and the specimen was sent to theoutpatient laboratory for testing. The results of the testing were returned to the pre-surgicalscreening clinic within 48 to72 hours. If a patient?s test was positive, a pre-surgical screeningnurse contacted the patient and informed her or him of the results. The patient was then advisedto fill the prescription for nasal Mupirocin ointment and to begin applying the ointment twicedaily into each nostril for five days before surgery, as described in the patient instructionpamphlet. Patients whose test results were negative were not contacted and hence did not fill theMupirocin prescription.In the pre-surgical screening clinic, the patients were given a box of six Sage? 2%chlorhexidine gluconate cloths for pre-operative skin preparation with written instructions fortheir use. The manufacturer supplies each box with three individually wrapped packages withtwo 7.5? x 7.5? applicator cloths in each package. The cloths are alcohol-, rinse-, and latex-free.The nurses in the pre-surgical screening clinic advised the patients to apply the cloths theevening before surgery and reviewed the six body locations to which the cloths must be applied(see Figure 1). The patients received a second box of Sage? 2% chlorhexidine gluconate clothsthe morning of surgery, once they were admitted. Adherence to this last phase of thedecolonization process was not included in the study evaluation because the distribution andapplication of the cloths were supervised by a nurse (i.e., the participant was not fullyautonomous).47As per the established protocol, registered nurses in the pre-surgical screening clinicadvised patients of the findings in the literature that showed that side effects to these twoproducts were for the most part restricted to local irritation of the nose or skin, both of whichhave been shown to resolve after discontinued use of the products (Bode et al., 2010).Figure 1. Assigned Body Areas for the Six Chlorhexidine Gluconate Cloths12 253 34648The study protocolThe principal investigator conducted in-service education with all nursing staff in the pre-surgical screening clinic before the commencement of the study to ensure that the decolonizationprocess was administered by all the nurses in the established fashion and that they understood thepurpose and protocol of the study.Data collectionThe primary outcome for this study was participants? adherence to the decolonizationprotocol; the two explanatory variables of interest were age and level of self-efficacy. Level ofself-efficacy was operationalized in the form of a self-efficacy score that was generated using arating scale. Because a validated instrument to measure self-efficacy in relation to adherence tothe decolonization protocol did not exist, it was necessary to examine other health domainswhere self-efficacy and adherence had been thoroughly examined, the area of medicationcompliance. Sleath et al. (2010) developed a validated instrument to measure glaucomamedication self-efficacy. Specifically, they developed an instrument that measured threedifferent elements of self-efficacy: (a) self-efficacy in overcoming the barriers that interferedwith glaucoma medication compliance, (b) self-efficacy in carrying out specific tasks required touse the medication, and (c) self-efficacy in relation to perceived expectations and outcomes withthe use of the medication. These three elements were used to develop an instrument thatmeasured self-efficacy in relation to the use of chlorhexidine gluconate cloths and Mupirocin(see Appendix A).At the end of the patients? pre-operative appointments in the pre-surgical screening clinic,two weeks before surgery, eligible patients were given an information package. Included in thispackage were a study invitation, a consent form, and a self-efficacy survey. If willing to49participate, the participant was asked to sign the consent form and complete the 15-minute self-efficacy survey at home, before surgery. The participants were asked to document their age,gender, and date of surgery on the questionnaire. They were asked to place the completedquestionnaire and signed consent form in an envelope labelled ?consent form and survey?provided in the information package. They were then instructed to seal the envelope and give itto the admitting nurse in the surgical day care unit on the day of surgery. To ensure thatparticipants remained anonymous to the research team, these envelopes had an outside labelcontaining a unique code number.Adherence to the decolonization protocol was operationalized in the form of anadherence score that was generated from the results of an adherence questionnaire. The purposeof the questionnaire was to collect information about whether the participants applied thechlorhexidine gluconate cloths, the number of cloths applied, whether the cloths were applied tothe appropriate areas, and the time of day the cloths were applied. In addition, for participantswho tested positive for nasal carriage of Staphylococcus aureus, items related to whether theyfilled the required prescription for Mupirocin, when they started applying the Mupirocin, thenumber of applications per day, and the location of the applications, were measured. Because avalidated tool to measure adherence to the decolonization protocol, specifically, did not exist, aquestionnaire was designed to capture the above measures of adherence (see Appendix A). Thequestionnaire was administered to the enrolled patients by a registered nurse in the surgical daycare unit on the day of admission for surgery. The patients verbalized their responses to thequestionnaire and the registered nurse documented those responses. The questionnaire tookapproximately 5-10 minutes to complete.Sealable envelopes, with a code number labelled on the outside, were stored in the50surgical day care unit. Each envelope contained an adherence questionnaire for the nurses tocomplete with the patients. When the participants returned an envelope containing the self-efficacy survey and consent form to a nurse in the surgical day care unit on the day of surgery,the nurse took note of the code number on the envelope and then obtained the envelope with theadherence questionnaire and the same code number. In other words the envelope containing theconsent form and survey had the same code number as the envelope containing the adherencequestionnaire. The nurses were instructed to complete the adherence questionnaire with theparticipant, to place the completed questionnaire into the envelope, to seal the envelope, and toreturn the two envelopes to the research team. This process was in place to ensure that (a) theparticipants remained anonymous to the research team and (b) the self-efficacy questionnaire,consent form, and adherence questionnaire for each participant were kept together.Before the commencement of the study, the principal investigator conducted in-serviceeducation with all the nursing staff of the pre-surgical screening and surgical day care units toclarify the purpose of the study and the purpose and administration of the questionnaires. Theintention of the in-service education was to minimize information biases by ensuring that all thenurses provided the information package to the participants and documented the participants?responses to the questionnaire in a standardized manner. Furthermore, it was crucial that theywere aware of the numeric coding system used to ensure the anonymity of the participants.Data AnalysisTo determine whether there was a relationship between self-efficacy and adherence,correlation analysis was used to examine the relationship between self-efficacy scores andadherence scores. The self-efficacy survey consisted of 31 questions (see Appendix A). Twoquestions, specifically, assessed whether participants? had access to someone who could help51them with the decolonization protocol, should it be necessary. These 2 items were not includedin the calculation of the self-efficacy scores. The remaining 29 items in the survey measuredparticipants? levels of self-efficacy with respect to: 1) the application of chlorhexidine gluconatecloths (12 items), 2) the application of Mupirocin (15 items), and 3) participants? perceivedinfection risk with the use of both chlorhexidine gluconate cloths and Mupirocin (2 items). Eachitem on the questionnaire was scored on a 5-point scale with ?4? indicating that the participantwas ?extremely confident? and ?0? indicating that he or she was ?not at all confident.? For theapplication of chlorhexidine gluconate cloths, an averaged self-efficacy score, for eachparticipant, was calculated by summing the scores on each of the 12 items and then dividing by12. Similarly, an averaged self-efficacy score for the use of nasal Mupirocin was calculated bysumming the scores on each of the 15 items and then dividing by 15. Finally, an averaged self-efficacy score with respect to the participants? perceived infection risk was calculated bysumming the scores on the 2 items and then dividing by 2. Thus, the final scores ranged from 0.0to 4.0.Adherence scores for each participant were calculated based on the participants?responses to the adherence questionnaire. Adherence scores for the application of chlorhexidinegluconate cloths were calculated for all participants. For those participants who also requiredMupirocin treatment, an additional adherence score was calculated. Adherence to the use ofchlorhexidine gluconate cloths was based on a possible total score of 10 points with a score of 10indicating complete adherence to the timing and application of the chlorhexidine gluconatecloths. Similarly, adherence to the use of nasal Mupirocin was based on a total score of 10 pointswith a score of 10 indicating complete adherence to the timing, application, and use of nasalMupirocin. A breakdown of the score allocation is listed in Figure 2. When the adherence52questionnaire was verbally administered to the participants by the registered nurses in thesurgical day care unit, on the day of admission for surgery, the patients who had not adhered tothe chlorhexidine gluconate cloths or Mupirocin requirements was asked to explain why. Acontent analysis was conducted to examine some of the common themes for non-adherence.The self-efficacy and adherence scores were plotted with scatter plots and a Pearson?sproduct moment correlation coefficient (r) was calculated using SPSS software. Thiscorrelation coefficient was used to estimate the magnitude and direction of the relationshipbetween the variables.To examine the relationship between age and adherence, age and adherence scores wereplotted with scatter plots and a Pearson?s product moment correlation coefficient (r) wascalculated using SPSS software. The correlation coefficient (r) mathematically stated thedirection and strength of the relationships between age and adherence and self-efficacy andadherence, but it did not measure the amount of variance (r?) accounted for in the outcomevariable, adherence, given the explanatory variables, age and self-efficacy. To measure thisvariance (r?), multiple linear regression analysis with SPSS software was used. The regressionequation, y = a + b1x1 + b2x2 + e, where y equalled adherence, x1 and x2 corresponded to self-efficacy and age, respectively, b1 and b2 corresponded to the assigned weight given to self-efficacy and age, respectively, and e was the residual error, allowed us to predict the level ofadherence for a given age and self-efficacy score. Multiple regression analysis not only measuredthe significance of the correlation between the two independent variables and the outcomevariable, it also tested the significance of each of the b-weights (Munro, 2005). Testing each ofthe b-weights told us whether the explanatory variables associated with those b-weightssignificantly contributed to the total variance accounted for in the outcome variable, adherence,53and if so to what magnitude (Munro, 2005).Adherence to Chlorhexidine Gluconate ClothsName: Age:Did you apply the chlorhexidine clothsevening before sx.?Yes= 2 points No= 0 pointsIf yes, what time did you apply the cloths? After 5pm= 2 points Before 5pm= 0 pointsIf you did apply the cloths, how many didyou apply?If 6 applied= 2 pointsIf 5 applied= 0.5 pointsIf < 5 applied= 0pointsWhere did you apply the cloths? ?All 6 correct locations = 4points?Any combination of 5locations = 3 points?Any combination of 4locations = 2 points?Any combination of 3locations = 1 points?2 or fewer locations = 0Total Score /10Adherence to Nasal MupirocinName: Age:Did you fill the prescription for Mupirocin? Yes= 2 points No= 0 pointsWhen did you begin the application? 5 days beforesurgery= 2 points< 5 days before surgery= 0 pointsHow many times a day did you apply theMupirocin?Twice daily= 2 points < twice daily= 0 pointsWhere did you apply the Mupirocin? Nares= 2 points Any other area= 0 pointsDid you apply Mupirocin to each nostril? Yes= 2 points No= 0 pointsTotal Score /10Figure 2. Score Allocation for Adherence Questionnaire54Ethical ConsiderationsCaution was used and measures were put in place to design a sampling plan thatrespected participants? privacy and right to refuse to participate in the study. Approval from theUniversity of British Columbia, Behavioural Research Ethics Board, and Vancouver CoastalHealth was obtained before proceeding with this study.The study invitation explained the purpose of the study, the expectations of participation,the time investment required, the potential risks associated with the study, and the measures thatwere put in place to ensure the participants? privacy and confidentiality were respected. Theinclusion and exclusion criteria for the study were outlined in the invitation as was informationpertaining to the anonymization of participants to the research team. Eligible participants had aminimum of two weeks to review the invitation and if they had any questions or concerns, theyhad the opportunity to have their questions answered by contacting the research team via thecontact information indicated on the study invitation and consent form. This process ensured thatthe decision to participate in the study remained both voluntary and informed.Overcoming Potential ChallengesSeveral difficulties were anticipated with the current study plan. The first pertained to thein-service education of the nursing staff. Staff turnover was inevitable and ensuring that all of thenursing staff attended an in-service session was challenging. The study?s investigators workedclosely with the peri-operative clinician to devise a schedule that included all of the nursing staff.Staff attendance at the in-service sessions was monitored and documented. Similarly, there wasongoing communication between the study investigators and the peri-operative clinicianregarding staff turnover and new employees to ensure that all of the nursing staff were trained inthe study protocol.55The second difficulty pertained to eligible participants? possible concerns with theestablished protocol, especially the potential side effects associated with the use of chlorhexidinegluconate cloths and Mupirocin. Specifically, the participants could express concern aboutantibiotic resistance with the use of Mupirocin. Short term use of nasal Mupirocin has not beenshown to cause widespread resistance (Ammerlaan, Kluytmans, Wertheim, Nouwen, & Bonten,2009). These concerns were addressed by the registered nurses according to their usual practice.56Chapter 4: ResultsEligibility and Participants? Response RatesFrom May 1, 2013 to October 1, 2013, 187 total hip and knee replacement surgeries wereperformed at the study hospital. All 187 patients were assessed for eligibility. Eighty-fourpatients were found to be eligible and 103 patients were ineligible to participate based on theinclusion and exclusion criteria. Figure 3 below summarizes the eligibility rates of potentiallyeligible patients and the enrollment and response rates of participants.Figure 3. Eligibility and Response RatesHip and Knee ReplacementSurgeries Performed (n=187)Assessed for Eligibility(n=187; 100%)Eligible (n=84; 44.9%) Ineligible (n=103; 55.1%)Consented to participate(n=40; 47.6%)Did not consent toparticipate (n=44; 52.4%)Completed alldata requirements(n=34; 85.0%)Did not complete alldata requirements(n=6; 15.0%)EligibilityEnrollmentResponse RatesOnly self-efficacy survey completed (n=4)Only adherence questionnaire completed (n=2)57Sample DescriptionForty patients consented to participate. The sample included 13 (34.2%) men and 25(65.8%) women. In two cases, the gender of the participant was not reported. The participantsranged in age from 41 to 86 years with a mean age of 67.4 years (SD = 9.9 years). The medianage was 67.5 years. Two of the 31 items in the self-efficacy survey assessed whether theparticipants had someone available to help them with the decolonization protocol, should theyrequire it. Twenty-six (89.7%) respondents indicated that they had someone available to helpthem with the chlorhexidine gluconate wash cloths, if necessary. Only 3 (10.3%) of therespondents indicated that they had no one to help them. In 11 cases, this information was notprovided. Fourteen participants (82.4%) also indicated that they had someone available to helpthem with the application of Mupirocin, if necessary, while 3 (17.6%) of the respondentsindicated that they did not have someone available to help them. In 23 cases, this informationwas not provided.Self-Efficacy and Perceived Infection RiskAveraged self-efficacy scores reflecting the participants? levels of self-efficacy in relationto both the application of chlorhexidine gluconate and Mupirocin were calculated. The self-efficacy survey was divided into three sections consisting of a total of 29 items. Twelve itemsassessed the participants? levels of self-efficacy in relation to the application of chlorhexidinegluconate, 15 items assessed the participants? levels of self-efficacy in relation to the applicationof Mupirocin, and 2 items assessed the participants? perceived infection risk with the use ofchlorhexidine gluconate and Mupirocin. Each item on the questionnaire was scored on a 5-pointscale with ?4? indicating that the participant was ?extremely confident? and ?0? indicating thathe or she was ?not at all confident.? Averaged self-efficacy scores for the application of58chlorhexidine gluconate were calculated by summing the scores on each item and then dividingby 12. Similarly, average self-efficacy scores for the application of Mupirocin were calculated bysumming the scores on each item and then dividing by 15. Finally, scores for the participants?perceived infection risk were calculated by summing the scores on each item and then dividingby 2. Scores ranged from 0.0 to 4.0. Table 9 below summarizes the participants? self-efficacyand perceived infection risk scores.Table 9. Self-Efficacy and Perceived Infection Risk ScoresAveraged Scores Mean (X?) Median SD Range X? 95% CISelf-Efficacy - Chlorhexidine 3.35 3.50 0.56 1.7-4.0 3.16-3.53Self-Efficacy - Mupirocin 3.59 3.73 0.48 1.9-4.0 3.41-3.77Perceived Infection Risk 3.46 3.50 0.65 2.0-4.0 3.24-3.68Note. SD = Standard Deviation; CI = Confidence IntervalTable 10 below is a summary report of the participants? response rates with respect to each of thethree sections of the self-efficacy survey.Table 10. Self-Efficacy Survey Response RatesCasesSection of Self-EfficacySurveyNo. of respondents whocompleted the sectionNo. of respondentswho didn?t completethe section Totaln Percent n Percent n PercentChlorhexidine Application 37 92.5% 3 7.5% 40 100.0%Mupirocin Application 30 75.0% 10 25.0% 40 100.0%Perceived Infection Risk 37 92.5% 3 7.5% 40 100.0%59AdherenceAdherence scores were generated from the results of the adherence questionnaire. All ofthe participants were scored on their adherence to the use of the chlorhexidine gluconate cloths.For those participants who used both the chlorhexidine gluconate cloths and nasal Mupirocin,they were scored on each treatment separately. Adherence to the use of the chlorhexidinegluconate cloths was based on a possible total score of 10 points with a score of 10 indicatingcomplete adherence to the timing and application of the chlorhexidine gluconate cloths.Similarly, adherence to the use of nasal Mupirocin was based on a total score of 10 points with ascore of 10 indicating complete adherence to the timing, application, and use of nasal Mupirocin.Table 11 summarizes the participants? adherence scores.Table 11. Adherence ScoresAdherence Scores Mean (X?) Median SD Range X? 95% CIChlorhexidine Application 9.01 9.00 1.03 5.0-10.0 8.66-9.37Mupirocin Application 9.75 10.00 0.71 8.0-10.0 3.41-3.77Note. SD = Standard Deviation; CI = Confidence IntervalIn four cases, the adherence questionnaires were not completed by a registered nurse in thesurgical day care unit. In another case, items in the adherence questionnaire, measuringadherence to the use of chlorhexidine gluconate were not fully completed. Therefore, adherencescores reflecting the use of chlorhexidine gluconate cloths were based on 35 of the 40 enrolledcases. Only 8 of the 40 respondents required Mupirocin treatment, in addition to the use ofchlorhexidine gluconate. Adherence scores for the use of Mupirocin were based on the datacollected from all 8 respondents.60Correlations: Age, Self-Efficacy, and AdherencePearson?s product moment correlation coefficient (r) was calculated using SPSSsoftware. Table 12 below provides a statistical summary of the correlation coefficients and theircorresponding levels of significance for the relationships between the variables age, self-efficacy, and adherence.Table 12. Pearson?s Correlations: Age, Self-Efficacy, and AdherenceVariables Age ASE Score-ChlorhexidineASE Score-MupirocinPerceivedInfectionRisk ScoreAdherenceScore-MupirocinAdherenceScore-ChlorhexidineASE Score-Chlorhexidinerp valuen.081.63437-ASE Score-Mupirocinrp valuen.084.65830.701*.00030-PerceivedInfection RiskScorerp valuen.182.28037.467*.00436.437*.01630-AdherenceScore-Mupirocinrp valuen-.268.5228.679.0648.860*.0068.600.1168-AdherenceScore-Chlorhexidinerp valuen.334.05833.049.79232.249.23125-.173.34432-.114.7888-Note. ASE Score =Averaged Self-Efficacy Score* p < .05, 2-tailed.The correlation coefficient (r) presented in Table 12 mathematically stated the directionand strength of the relationships between age and adherence and self-efficacy and adherence, butit did not measure the amount of variance (r?) accounted for in the outcome variable, adherence,given the explanatory variables, age and self-efficacy. To measure this variance (r?), multiplelinear regression analysis with SPSS software was used. Only 8 of the 40 respondents requiredMupirocin treatment. With the exception of 1 case, there was no variance in the adherence61scores, for the use, timing, and application of Mupirocin among the remaining 7 respondents. All7 respondents received maximum scores of 10 points. One respondent received a score of 8points. Due to the lack of variance in adherence scores among those respondents who usedMupirocin, linear regression analysis was used only to examine the amount of variance in theoutcome variable, chlorhexidine gluconate cloth adherence, given the explanatory variables ageand self-efficacy. Table 13 below summarizes the results of the regression analysis that wasconducted to predict adherence to the use of chlorhexidine gluconate cloths given the variablesage, perceived risk for infection, and level of self-efficacy.Table 13. Regression Analysis Predicting Adherence to the Application of ChlorhexidineGluconate95% Confidence Interval for BVariables B ? Lower Bound Upper Bound Age .038* .378 .001 .075ASE Score-Chlorhexidine .264 .151 -.449 .977Perceived Infection Risk Score -.445 -.294 -1.067 .177Note. ASE Score = Averaged Self-Efficacy Score* p < .0562Chapter 5: Discussion, Implications, and ConclusionsDiscussionThis study examined the relationships between age, levels of self-efficacy, and patients?adherence to a Staphylococcus aureus decolonization protocol in the total joint arthroplastypopulation. The study was designed to test the hypothesis that there is a relationship betweenpatients? age and adherence and that there is a positive relationship between patients? level ofself-efficacy and adherence. The findings suggest that there may be a positive relationshipbetween age and adherence to the use of chlorhexidine gluconate cloths, and a negativerelationship between age and adherence to the use of nasal Mupirocin. These results, however,were not statistically significant. The relationship between age and adherence to therecommendation to apply chlorhexidine gluconate cloths was statistically significant once theself-efficacy and perceived risk of infection scores were controlled.With respect to the relationship between patients? level of self-efficacy and adherence,the results confirm that there is a statistically significant and strong positive relationship betweenpatients? level of self-efficacy in applying Mupirocin and their adherence to its use, timing, andapplication. Of interest, little if no relationship was found between the patients? level of self-efficacy to chlorhexidine gluconate cloths and their adherence to the use, timing, and applicationof the chlorhexidine gluconate. Other major findings include: (a) a statistically significant andstrong positive correlation between patients? level of self-efficacy to the use of chlorhexidinegluconate cloths and patients? level of self-efficacy to the use of Mupirocin, (b) statisticallysignificant and moderately positive correlations between perceived risk for infection, with theuse of both chlorhexidine gluconate and Mupirocin, and self-efficacy in using chlorhexidinegluconate cloths and Mupirocin, and (c) results of the multivariate regression analysis reveal that63age, but not self-efficacy or perceived infection risk, contributed significantly to the outcome,adherence.As previously discussed, Staphylococcus aureus has been identified as the mainperpetrator of surgical site infections in many tertiary health care centres (Anderson et al., 2007).Surgical site infections and their treatment, especially in the total joint arthroplasty population,have been shown to significantly affect patients? recovery from surgery and their quality of life,and generate considerable economic cost for the health care system (Rao et al., 2011). Van Rijenet al. (2008a) conducted a systematic review of randomized control studies that examined theeffects of a pre-operative screening and decolonization protocol on patients undergoing totaljoint arthroplasty. They found promising eradication rates of Staphylococcus aureus and areduced incidence rate of post-operative surgical site infections, although the results were notstatistically significant. In a review of the literature presented in Chapter two, a recognizable gaprelated to patient adherence to a decolonization protocol was identified as a possible missing linkor explanation for these findings. Given this literature gap, and given that the success andefficacy of such a protocol is heavily reliant on adherence, we can appreciate the importance ofenhancing our knowledge regarding the factors that promote and predict adherence. The resultsof this study have significant clinical implications for the design, implementation, and evaluationof patient education material relevant to the decolonization protocol. Constructs of self-efficacy(i.e., health perceptions, perceived risk, and health beliefs) related to the protocol, are guided andinfluenced by many factors including the type of education patients receive about the protocol,how that education is structured and designed, and how it is implemented and provided topatients by health professionals, especially nurses. One?s level of self-efficacy can in turn predictadherence. Having knowledge of this, appropriate changes and recommendations can be made, if64necessary, to the design, structure, and implementation of patient education. These changes andrecommendations would be aimed at enhancing or improving patients? level of self-efficacy andadherence to the protocol. In this study, there were opposing findings with respect to therelationship between self-efficacy and adherence. There was a strong positive relationshipbetween self-efficacy to Mupirocin and adherence to Mupirocin but little if no relationshipbetween self-efficacy to chlorhexidine gluconate washes and adherence to chlorhexidinegluconate washes. There are a number of possible explanations for these findings that arediscussed in the following sections; consideration must be given to the factors that influence andguide one?s level of self-efficacy, namely patient education related to the decolonizationprotocol.When considering the relationship between age and adherence, it is important tounderstand adherence as a function of age. In other words, it is important to comprehend howage-related differences might influence adherence depending on the patient population or healthdomain being considered. Similar to the relationship between self-efficacy and adherence,opposing outcomes were found with respect to the relationship between age and adherence. Agewas found to be positively correlated with chlorhexidine gluconate adherence but negativelycorrelated with Mupirocin adherence. These results point to age-related differences between theapplication of chlorhexidine gluconate and that of Mupirocin, and these age-related differences,if they truly exist, must be considered in the design, structure, and implementation of educationrelated to the decolonization protocol. Without a clear understanding of the relationship betweenage and adherence and the role and influence of self-efficacy on adherence, in the total jointarthroplasty population, a proper evaluation of the effectiveness of the decolonization protocolremains challenging as does the evaluation of patient education relevant to the protocol. Ongoing65assessment and evaluation of quality control measures in relation to physician-recommendedguidelines, such as the decolonization protocol, is necessary, to ensure that the needs of patientsare met and clinical practice remains evidence-based.Age and adherenceAs previously discussed, adherence to a decolonization protocol has been largelyunder studied in the total joint arthroplasty population making it necessary to consider otherhealth domains and patient populations where adherence has been examined: the areas ofmedication compliance, compliance to physician-recommended exercise programs, andcompliance to patient guidelines. Reflecting the current state of the literature and consistent withthe findings of a systematic review conducted by Taylor et al. (2011), which examined themedical, psychological, and socio-demographic factors associated with adherence to cardiacrehabilitation programs, the current study found equivocal results with respect to the direction ofthe relationship between age and adherence. This study found a moderately positive relationshipbetween age and adherence to the chlorhexidine gluconate washes but a moderately negativerelationship between age and adherence to Mupirocin. One would expect the relationshipbetween age and these two decolonization protocol measures to be similar. There are a numberof possible explanations for these findings, including lifestyle or contextual factors and age-related differences in cognition. It is uncertain whether these study outcomes can truly beattributed to age-related differences in cognition or lifestyle factors when there are severalsignificant methodological limitations to consider in this study, particularly the large samplingerror.The positive correlation found between age and adherence to the chlorhexidine gluconatecloths is similar to the outcomes from studies that have examined medication compliance and66adherence to patient screening guidelines. Older adults often display greater adherence rates thando younger adults (Barclay et al., 2007; Birkenfeld et al., 2011; Tarantino et al., 2010). Lifestylefactors, including family and work commitments and financial constraints, have been identifiedin the literature as possible explanations for this outcome (Barclay et al., 2007; Tarantino et al.,2010). Regimens or medications that require individuals to alter their lifestyles to accommodatepresent a great challenge for young adults who have family and work commitments and verybusy lifestyles. Changes in health perceptions and beliefs, such as greater perceivedsusceptibility to developing an illness or disease and heightened fear of mortality associated withnon-adherence, are often characteristic of older adults (Taylor et al., 2011). This may explainwhy older adults may be more likely to adhere to a prescribed treatment, at least in certainpatient populations or health domains (Taylor et al., 2011). The role and influence of these co-variables on the overall relationship between age and adherence is uncertain given that thesevariables were not examined nor controlled in the current study. Effort was made, however,when creating the adherence questionnaire, to include questions aimed at collecting informationabout the respondents? reasons for non-adherence. The hope was that these reasons would shedlight on some of the factors at play. The nurses, in the surgical day care unit, were directed to askthe respondents about their reasons for non-adherence when a respondent reported that she or hedid not fully adhere to one or more of the components of the decolonization measures. Thenurses however demonstrated poor compliance in addressing these questions with therespondents. As a result, a content analysis addressing and exploring the themes of non-adherence could not be executed as planned. It is possible that the nurses were not comfortableasking the respondents about their reasons for non-adherence. The nurses may have believed thatthe questions had an underlying notion of blame and they may have concluded that asking the67questions could have heightened the patients? anxiety levels before surgery. In only one case, anurse provided a reason for a participant having not washed with all six chlorhexidine gluconatecloths. It is therefore not known whether co-variables, such as lifestyle or contextual factors, orspecific health perceptions and beliefs, influenced the overall relationship between age andadherence.While there are many studies that have found greater adherence rates to medication andpatient screening guidelines in older adults when compared to their younger counterparts, there isa body of literature that reveals the converse, the vulnerability of the older adult population andthe multitude of factors that often put this population at risk for non-adherence (Liu & Gonzalez,2007; McDaniel & Einstein, 2007; Park & Meade, 2007). Cognitive factors such as memorydecline, particularly time-based memory decline, the memory of performing everyday tasks atthe time they are intended to be done, tends to diminish with age-related cognitive decline inadults older than 61 years of age (McDaniel & Einstein, 2007). The ability to consciouslyinterpret, comprehend, and translate instructions into action also diminishes with age, and this,more so than memory, is the greatest barrier to adherence in older adults (Park & Meade, 2007).Contextual factors, such as daily commitments and routines also change with age (Park &Meade, 2007). As one ages, these changes may not favour adherence, especially when theregimen to which they are adhering is complex (Park & Meade, 2007). It is possible that theolder respondents managed and coped better with the use, timing, and application of thechlorhexidine gluconate cloths rather than the Mupirocin. The older respondents may haveperceived the Mupirocin regimen to be more complex than the chlorhexidine gluconate washregimen, which only required them to wash with six chlorhexidine gluconate cloths once, theevening before surgery. Mupirocin, on the other hand, was required twice daily for five days68before surgery and it was necessary to have good vision and hand-eye coordination to effectivelyapply the nasal ointment. As one ages, vision and coordination can deteriorate making itchallenging for older adults to carry out particular tasks. Although there is a large body ofliterature that identifies cognitive decline as one of the major contributing factors of non-adherence in older adults, the extent to which this evidence can be used to explain the findings inthis study is questionable. The majority of respondents in this study were between 60 and 70years of age. Although adults most commonly begin displaying signs and symptoms of cognitivedecline after the age of 61 years, this cognitive decline is not likely to interfere with dailyactivities and tasks until greater than 70 years of age (Park & Meade, 2007). Cognitive factorsmay therefore be a weak explanation for these study outcomes. Furthermore, the size of thesampling error in this study was large and there was a lack of variability in the adherence scoresamong the eight respondents required to undergo treatment with Mupirocin. Seven of the eightrespondents received the maximum score of ten points for adherence. Again, thesemethodological limitations may be reasonable explanations of the negative relationship foundbetween age and the application of Mupirocin.Although the findings regarding the relationship between age and adherence wereequivocal in nature, the results from the multivariate linear regression analysis revealed that agecontributed significantly to the outcome variable, adherence. Despite these findings, however,age as a predictor of adherence remains a poorly understood phenomenon given the number ofco-variables at play, for which little is known or understood, and the limitations to thegeneralizability of study results from other patient populations to the orthopaedic population.Self-efficacy and adherenceLittle if no relationship was found between self-efficacy in the use of chlorhexidine69gluconate washes and adherence to the chlorhexidine gluconate washes, but a strong andstatistically significant positive relationship was found between self-efficacy to Mupirocin andadherence to Mupirocin. The latter results reflect the current state of the literature, which showsoverwhelming support for a positive association between overall levels of self-efficacy andadherence to medication and physician-recommended patient guidelines (Chia et al., 2006;Curtin et al., 2008; Ngamvitoj & Kang, 2007). In other words, adults who believe in their abilityto take medication or who believe in their ability to carry out a particular task related to atreatment regimen are more likely to adhere to the regimen. An examination of the relationshipbetween self-efficacy in applying the chlorhexidine gluconate cloths and adherence to washingwith the chlorhexidine gluconate cloths revealed results that were inconsistent with the overallfindings in the literature. One study conducted by Peddle et al. (2009) found that the influence ofsubjective norms (i.e., greater perceived social pressure or availability of social support), aconstruct derived from Bandura?s (1977) self-efficacy theory, and increased levels of self-efficacy, were strongly associated with greater adherence to a supervised exercise program forpatients awaiting surgical removal of malignant lung lesions. It may be that the studyrespondents who required Mupirocin treatment reported greater access to assistance with thedecolonization protocol than did the respondents who only required chlorhexidine gluconatewashes. Access to assistance with the protocol in turn resulted in higher levels of self-efficacyand greater adherence to Mupirocin. Although a possible explanation, it is weak in its ability toexplain the differences in adherence to the two decolonization measures. The respondents whorequired Mupirocin treatment reported similar results with respect to the availability of socialsupport with the decolonization protocol when compared with those who only required the use ofchlorhexidine gluconate cloths. Another possible explanation for these opposing findings relates70to the design of the chlorhexidine gluconate wash adherence questionnaire. As previouslydiscussed, the respondents were required to wash with six chlorhexidine gluconate cloths theevening before surgery and each cloth was assigned to one of six body areas: arms, legs, chest,back, buttocks, and genital area. The respondents were specifically told not to apply the cloth tothe genitals but rather to wash the skin folds within the groin area. On the adherencequestionnaire, the words genital area rather than groin were listed as one of the six bodylocations the participants were asked if they had washed. When the respondents were askedwhether they had washed their genital area, they may have thought that they were being asked ifthey had washed their genitals. The nurses may not have clarified that this in fact referred to thearea around the genitals (i.e., the groin area) and not the genitals themselves. Eleven of thirty-sixrespondents (30.6%) who were asked about their adherence to the chlorhexidine gluconate clothsresponded that they did not wash their genital area. These respondents were scored as if they didnot fully adhere to the recommended protocol. As a result, the overall self-efficacy scores mayhave been lower than predicted and this may have influenced the relationship between self-efficacy and adherence to the chlorhexidine gluconate washes. Despite these possibleexplanations, the results obtained from a correlation analysis conducted between the self-efficacyscores to chlorhexidine gluconate washes and the self-efficacy scores to Mupirocin, point to theinfluence of co-variables or methodological flaws within the study as more probableexplanations of the conflicting results. Results from the correlation analysis revealed that therewas a strong and statistically significant positive relationship between the self-efficacy scores tothe chlorhexidine gluconate washes and the self-efficacy scores to Mupirocin. Based on theseresults, one would expect similar correlations between self-efficacy and adherence to the twoseparate decolonization measures, but these expectations did not emerge in this study.71A construct of self-efficacy known as perceived risk was examined in this study. Therespondents were asked to rate how confident they were that chlorhexidine gluconate andMupirocin use would reduce their risk of developing an infection, post-operatively. Findingsfrom the literature suggest that greater perceived susceptibility to developing an illness or diseaseand stronger beliefs about the efficacy of a treatment in improving overall health and preventingdisease are likely to be associated with greater adherence (Boeka et al., 2010; Rogers & Prentice-Dunn, 1997; Rosenstock et al., 1988; Tarantino et al., 2010; Taylor et al., 2011). The currentstudy found a positive relationship between perceived infection risk and adherence to the use ofMupirocin. In other words, respondents who felt confident that the use of chlorhexidinegluconate washes and Mupirocin could prevent post-operative infection demonstrated greateradherence to Mupirocin, findings consistent with those of the published literature. These results,however, were not statistically significant. Of interest, opposing results were found for thecorrelation between perceived infection risk and adherence to the chlorhexidine gluconate cloths.Again these results were not statistically significant. The respondents who believed thatMupirocin and chlorhexidine gluconate could prevent post-operative infections demonstratedpoor adherence to the use of chlorhexidine gluconate cloths, findings that are inconsistent withthose in the literature. It is possible that the respondents perceived lesser risk of developing aninfection with the use of Mupirocin than with the chlorhexidine gluconate. One of the manyfactors that could not be controlled in this study was the patient education provided bypharmacists. All of the respondents requiring Mupirocin treatment were required to fill aprescription for the ointment at a community pharmacy of their choice. Consistent with theframework of professional practice produced by the College of Pharmacists of British Columbia(2006), pharmacists have a duty to their clients and are responsible for ensuring that clients are72educated about the potential side effects, dosing, and mode of administration of a prescription.Depending on the type of education the respondents received from their pharmacist, they mayhave perceived greater threat or risk of infection with non-adherence to Mupirocin than non-adherence to the use of the chlorhexidine gluconate cloths. The study hospital supplied therespondents with the chlorhexidine gluconate cloths and patient education related to the use ofthe cloths was provided solely by nurses in the pre-surgical screening clinic. This explanation,although reasonable, cannot be validated given that the respondents? perceived infection risk wasmeasured based on the two decolonization measures together rather than separate. On a differentnote, the perceived infection risk scores correlated significantly and positively with the self-efficacy scores to both the chlorhexidine gluconate washes and Mupirocin, which is strongevidence and support for the interconnectivity of the theoretical models used in the frameworkfor this study, specifically, Bandura?s (1977) self-efficacy theory, Theory of Planned Behavior(Ajzen, 1991), Protection Motivation Theory (Rogers & Prentice-Dunn, 1997), and the HealthBelief Model (Rosenstock et al., 1988).The results of the regression analysis revealed that neither self-efficacy nor perceivedinfection risk contributed significantly to the outcome, adherence. These results are consistentwith the findings of a study conducted by Ngamvitoj and Kang (2007), who examined the effectsof asthma self-efficacy, social support, and knowledge on adherence to peak expiratory flow rate(PEFR) monitoring in asthma patients. The current study, similar to that of Ngamvitoj andKang?s (2007), has methodological limitations, especially too small a sample size, as a majorcontributing factor for these results. Other contributing factors that Ngamvitoj and Kang (2007)noted in their study included decreased variability between the constructs of self-efficacy and theexistence of co-variables. Decreased variability between the constructs of self-efficacy in this73study may also have had an influence on the study outcomes. The concept of self-efficacy is oneof complexity. There is a great degree of overlap between the various constructs of self-efficacy(e.g., health beliefs, health perceptions, perceived risk, perceived threat) and multiple definitionsfor the concept exist. For example, self-efficacy can be defined in terms of medication ortreatment efficacy (i.e., perceived treatment utility), it can be defined in terms of perceivedsusceptibility to developing an illness or disease, or it can be defined in terms of the perceivedthreat to overall health associated with non-adherence. Depending on how self-efficacy is chosento be defined can generate different study outcomes. In this study, self-efficacy was defined asthe belief in the participants? ability to perform the decolonization measures. Items in the self-efficacy survey addressed many different constructs of self-efficacy (i.e., perceived threat,subjective norm, perceived risk) so it is possible, similar to Ngamvitoj and Kang?s (2007) study,that there was not enough variability between the constructs used to measure the respondents?levels of self-efficacy.A number of possible explanations for the study outcomes have been discussed but thereis another explanation, related to the education provided to the respondents about thedecolonization protocol, that is worthy of mention. In regard to patient education related to thedecolonization protocol, in one case, a respondent reported that she or he had not received adiagram showing the body areas to be washed with chlorhexidine gluconate. As a result, thisrespondent did not fully adhere to the decolonization measure before surgery. It may be possiblethat other respondents found themselves in similar situations and these cases were not reported.The nurses complied poorly with addressing the questions regarding non-adherence with therespondents, but it is also possible that the respondents did not report their lack of adherence forfear of having their surgeries cancelled. In two other cases, the respondents reported side effects74with the use of the chlorhexidine gluconate cloths, which was documented on the adherencequestionnaire by the nurses. One respondent reported burning of the skin and a rash with the useof the chlorhexidine gluconate while the other reported skin itchiness. Both of the respondentsdid not complete the chlorhexidine gluconate wash. It is uncertain as to whether there were othercases wherein side effects prevented the respondents from completing the chlorhexidinegluconate wash. Information regarding the side effects that the respondents may haveexperienced from the two decolonization measures was not collected and so it is difficult todecipher the degree of influence this may have had on the study outcomes.Clinical ImplicationsUnderstanding how age and age-associated predictors, namely self-efficacy, influenceadherence in adults has significant clinical implications. The state of the literature showspromising eradication rates with the use of a Staphylococcus aureus screening anddecolonization protocol even though the results have not been robust and lack statisticalsignificance (Van Rijen et al., 2008a). With the implementation of this protocol, a risk reductionof 1% has been shown to translate into cost savings in the millions of dollars for every tenthousand patients screened (Van Rijen et al., 2008a). These findings are of great clinicalrelevance not only to the health system but to patients as well when one considers the impact oftreating surgical site infections on patients? recovery and quality of life. Given the clinicalrelevance of the protocol, it is easy to appreciate the importance of obtaining a betterunderstanding of the factors that promote and predict patients? adherence to the protocol.Understanding adherence as a function of age, especially in the total joint arthroplastypopulation, is clinically significant given that the majority of patients requiring hip and kneereplacements are 60 years of age or older. Specifically, understanding how health perceptions75and health beliefs change with age and how age-related cognitive decline influences adherence isfundamental to the design and implementation of patient education materials related to adecolonization protocol. Similarly, the role and influence of self-efficacy, the factors thatpromote and prevent it, and how self-efficacy changes with age are all important concepts tokeep in mind when designing instructions for older adults and expecting that they will adhere tothem (Liu & Gonzalez, 2007). Furthermore, the complexity of the language used in instructionsor education material, the ease with which it is presented to patients and by whom, the setting inwhich it is presented or introduced, and how and when it is evaluated are all decisions thatrequire an understanding of the factors that promote and predict adherence. The findings fromthis study reveal inconclusive results with respect to the overall relationships between age, self-efficacy, and adherence. Despite these results, the study does highlight a number of age-relatedfactors and constructs of self-efficacy that are important for health professionals, especiallynurses, to consider when providing education to patients with the expectation that patients willadhere to what is being recommended.Methodological LimitationsA descriptive correlational study design was used to examine the relationships betweenage, self-efficacy, and adherence to the decolonization protocol. This study design, althoughappropriate for addressing the research questions, had several limitations. First, the study designwas weak in its ability to support a causal inference. Correlational studies work well forexamining relationships or associations between variables, but not for making causal inferences(Polit & Beck, 2012). Second, because the participants were not randomly sampled, selectionbias presented as a limitation. The participants may have chosen to participate in the study forreasons that could not be known or controlled and these reasons could have acted as confounding76variables. Such variables may have included income, level of education, health literacy,ethnicity, and social support. Selection bias can significantly limit the generalizability or externalvalidity of a study (Polit & Beck, 2012). Thirdly, the self-efficacy survey and adherencequestionnaire were not validated. The instruments were not tested for internal validity andreliability and caution must be used in the interpretation of the results. Fourthly, because nursesadministered the adherence questionnaire, the social desirability response set bias must beconsidered when interpreting the overall results. The participants may have altered theirresponses to the adherence questionnaire simply to save face and not be perceived as someonewho did not follow instructions (Polit & Beck, 2012). On the other hand, nurses demonstratedpoor compliance in addressing the reasons for non-adherence in instances where the respondentsreported that they had not fully adhered to the protocol. Hence, a content analysis aimed atidentifying potential co-variables could not be executed as planned. Lastly, this study issignificantly underpowered. In other words, the probability of detecting the relationshipsbetween age, self-efficacy, and adherence, if they in fact truly exist in the general population,was very low. It is very likely that a type II error has been made. Results of the power analysisinitially conducted to determine the sample size needed to detect relationships between patients?age, self-efficacy, and adherence revealed that 84 participants would be required. Because of anunanticipated reduction in the availability of surgical operating time at the hospital, during thedata collection period, fewer total hip and knee replacement surgeries were performed.Consequently, this reduced the number of eligible patients for the study. Furthermore, theexclusion criteria may have been too stringent. The hospital?s catchment area is home to a veryethnic population and so excluding patients from participating in the study based on theirproficiency in the English language, may have decreased the number of eligible patients. In order77to obtain a more precise and accurate estimate of the relationships between age, self-efficacy,and adherence, and improve the rigor and objectivity of the study, a larger sample size and morereliable and valid instruments to measure the variables would be needed.Recommendations for Future ResearchFuture research should be aimed at using a combination of both quantitative andqualitative study designs to examine the relationships between age, self-efficacy, and adherence(Yuhas et al., 2012). To obtain a better understanding of the constructs of self-efficacy (i.e.,individual health beliefs and personal health perceptions) and their influence on patient outcomes(i.e., adherence), the use of more qualitative study designs is needed. Qualitative study designswill provide researchers with a greater opportunity to explore patients? beliefs and concerns withrespect to a treatment, and they will also give researchers the ability to conduct more in depthevaluations of patients? understanding of physicians? instructions and guidelines (Yuhas et al.,2012).There are a limited number of reliable and valid instruments available to measure self-efficacy and adherence and there are a number of advantages and disadvantages associated withthe utilization of these tools. When possible and feasible to do so, researchers should beencouraged to use a combination of instruments (Osterberg & Blaschke, 2005). The use of bothelectronic monitoring devices and self-reported instruments to measure medication compliance,for example, could add precision and accuracy to the data collection processes and reduce theodds that the study results are due to chance alone (Osterberg & Blaschke, 2005). Furthermore,researchers should make greater effort to conduct studies that are designed to establish thereliability and validity of these instruments and report the statistical evidence. Without evidenceof reliable and valid instrumentations, it is difficult to ensure that accurate and precise data have78been collected, and without accurate and precise data, it is difficult to draw valid conclusions orto make valid predictions.Future studies should aim to use longitudinal study designs as opposed to cross-sectionaldesigns (Chia et al., 2006; Fawzi et al., 2012). As mentioned throughout this discussion, one ofthe disadvantages of using cross-sectional study designs is that researchers have limitations inthe control that can be introduced to minimize the effects of external factors that may be presentduring a particular time. When this design is used in conjunction with convenience sampling, thenumber of confounding variables at play significantly increases. Researchers should aim to uselongitudinal designs to minimize the number of confounding variables at play, and when possibleand feasible to do so, should use probability sampling as opposed to non-probability sampling.Furthermore, attrition in longitudinal studies may be an important aspect of adherence that mayoften get overlooked with the use of cross-sectional designs (Fawzi et al., 2012).This study demonstrated that there are many contextual factors in health care that cannotbe controlled but that can significantly affect the power of a study (e.g., unanticipated operatingroom closures, staff shortages, and bed closures). To be better prepared for these situations andto be able to effectively manage them when they do occur, consideration should be given toobtaining appropriate funding that would give researchers more resources to minimize the effectsof these situations. For instance, researchers could utilize these resources to lengthen their datacollection period to obtain a more representative and appropriately sized sample. These resourcescould also be utilized to hire research assistants to help with the recruitment of subjects and thedata collection rather than rely on health professionals who often have heavy workloads and arebusy providing patient care. Similarly, if health professionals, namely nurses, are involved andassisting with a research project in a clinical setting, it is crucial that the researchers assess how79much education the nurses might need to carry out their duties with respect to the project, howmany in-service hours they might require, whether they require close supervision in carrying outtheir research duties, the type and amount of resources they might need to complete their tasks,and whether incentives may be needed to compensate those who have chosen to assist with theresearch project. To increase the commitment level of health professionals and to improve thequality of work produced by those involved with the research, these factors must be consideredin the proposal of any research project.Finally, given the high level of interconnectivity between health beliefs and self-efficacy,it is important to acknowledge that health beliefs are highly influenced by culture and ethnicity(Fawzi et al., 2012). Constructs of self-efficacy such as beliefs or perceptions of illness anddisease are highly influenced by ethnicity and culture and they therefore must be examined ingreater depth across diverse patient populations (Chia et al., 2006; Fawzi et al., 2012).Researchers should be encouraged to collect larger and more ethnically diverse samples toexamine the effects of culture and ethnicity.ConclusionsIn conclusion, an examination of adherence to an established pre-operativeStaphylococcus aureus decolonization protocol in the total joint arthroplasty population waswarranted. Non-adherence to this protocol could have significant clinical implications forpatients? quality of life and the economic state of the health care system in light of the costsassociated with the treatment of surgical site infections (Rao et al., 2011). This study examinedthe relationships between age, self-efficacy, and adherence to a Staphylococcus aureusdecolonization protocol in the total joint arthroplasty population. The study found inconclusiveresults with respect to the relationships between age, self-efficacy, and adherence. In light of80these results, this study does highlight the many ways in which age and age-associatedpredictors, namely self-efficacy, can influence adherence in adults. It also identifies a number ofco-variables that warrant exploration in future studies given their potential influence on therelationships between age, self-efficacy, and adherence. Without a clear understanding of therelationship between age and adherence and the role and influence of self-efficacy on adherence,in the total joint arthroplasty population, a proper evaluation of the effectiveness of thedecolonization protocol remains challenging as does the evaluation of patient education relevantto the protocol. Ongoing assessment and evaluation of quality control measures in relation tophysician-recommended guidelines, such as the decolonization protocol, is necessary, to ensurethat the needs of patients are met and clinical practice remains evidence-based.81BibliographyAjzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human DecisionProcessing, 50, 179-211. doi:10.1016/0749-5978(91)90020-TAmmerlaan, H., Kluytmans, J., Wertheim, H., Nouwen, J., & Bonten, M. (2009). Eradication ofmethicillin-resistant staphylococcus aureus carriage: A systematic review. ClinicalInfectious Diseases, 48, 922-930. doi:10.1086/597291Anderson, D. J., Sexton, D. J., Kanafani, Z. A., Auten, G., & Kaye, K. S. (2007). Severe surgicalsite infection in community hospitals: Epidemiology, key procedures, and the changingprevalence of methicillin-resistant Staphylococcus aureus. Infection Control and HospitalEpidemiology, 28, 1047-1053. doi:10.1086/520731Bailey, R. R., Stuckey, D. R., Norman, B. A., Duggan, A. P., Bacon, K. M., Connor, D. L., . . .Lee, B., Y. (2011). Economic value of dispensing home-based preoperative chlorhexidinebathing cloths to prevent surgical site infection. Infection Control & HospitalEpidemiology, 32, 465-471. doi:10.1086/659763Bandura, A. (1977). A social learning theory. Upper Saddle River, NJ: Prentice-Hall.Barclay, T. R., Hinkin, C. H., Castellon, S. A., Mason, K. I., Reinhard, M. J., Marion, S. D., . . .Durvasula, R. S. (2007). Age-associated predictors of medication adherence in HIV-positive adults: Health beliefs, self-efficacy, and neurocognitive status. Health Psychology,26, 40-49. doi:10.1037/0278-6133.26.1.40Birkenfeld, S., Belfer, R. G., Chared, M., Vilkin, A., Barchana, M., Lifshitz, I., . . . Levi, Z.(2011). Factors affecting compliance in faecal occult blood testing: A cluster randomizedstudy of the faecal immunochemical test versus the guaiac faecal occult test. Journal ofMedical Screening, 18, 135-141. doi:10.1258/jms.2011.01014782Bissonnette, J. M. (2008). Adherence: A concept analysis. Journal of Advanced Nursing, 63,634-643. doi:10.1111/j.1365-2648.2008.04745.xBode, L. G., Kluytmans, J. A., Wertheim, H. F., Bogaers, D., Vandenbroucke-Grauls, C. M.,Roosendaal, R.,?Vos, M. C. (2010). Preventing surgical-site infections in nasal carriers ofstaphylococcus aureus. New England Journal of Medicine, 362, 9-17.doi:10.1056/NEJMoa0808939Boeka, A. G., Prentice-Dunn, S., & Lokken, K. L. (2010). Psychosocial predictors of intentionsto comply with bariatric surgery guidelines. Psychology, Health & Medicine, 15, 188-197.doi:10.1080/13548501003615282Brus, H., Van de Laar, M., Taal, E., Rasker, J., & Wiegman, O. (1999). Determinants ofcompliance with medication in patients with rheumatoid arthritis: the importance of self-efficacy expectations. Patient Education & Counseling, 36, 57-64. Retrieved from:http://www.ncbi.nlm.nih.gov/pubmed/10036560Caffrey, A. R., Woodmansee, S. B., Crandall, N., Tibert, C., Fielding, C., Mikolich, D., J., . . .Laplante, K. L. (2011). Low adherence to outpatient preoperative methicillin-resistantstaphylococcus aureus decolonization therapy. Infection Control & Hospital Epidemiology,32, 930-932. doi:10.1086/661787Chia, L. R., Schlenk, E. A., & Dunbar-Jacob, J. (2006). Effect of personal and cultural beliefs onmedication adherence in the elderly. Drugs & Aging, 23, 191-202. doi:10.2165/00002512-200623030-00002Christensen, A. J. (2004). Patient adherence to medical treatment regimens: Bridging the gapbetween behavioral science and biomedicine. New Haven, CT: Yale University Press.83Clark, N. M., & Dodge, J. A. (1999). Exploring self-efficacy as a predictor of diseasemanagement. Health Education & Behavior, 26, 72-89. doi:10.1177/109019819902600107Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155-159. Retrieved from:http://web.ebscohost.com.ezproxy.library.ubc.caCollege of Pharmacists of British Columbia. (2006). Framework of professional practice (FPP).Retrieved from: http://www.bcpharmacists.org/legislation_standards/provincial_legislationCurtin, R. B., Walters, B. A., Schatell, D., Pennell, P., Wise, M., & Klicko, K. (2008). Self-efficacy and self-management behaviors in patients with chronic kidney disease. Advancesin Chronic Kidney Disease, 15, 191-205. doi:10.1053/j.ackd.2008.01.006DiMatteo, M. R., Haskard, K. B., & Williams, S. (2007). Health beliefs, disease severity, andpatient adherence: A meta-analysis. Medical Care, 45, 521-528.doi:10.1097/MLR.0b013e318032937eDiMatteo, M. R., Hays, R. D., Gritz, E. R., Bastani, R., Crane, L., Elashoff, R.,?Marcus, A.(1993). Patient adherence to cancer control regimens: Scale development and initialvalidation. Psychological Assessment, 5, 102-112. doi:10.1037/1040-3590.5.1.102Fawzi, W., Abdel Mohsen, M. Y., Hashem, A. H., Moussa, S., Coker, E., & Wilson, K. C. M.(2012). Beliefs about medications predict adherence to antidepressants in older adults.International Psychogeriatrics, 24, 159-169. doi:10.1017/S1041610211001049GlaxoSmithKline. (2009, April). Prescribing information: Bactroban Nasal?. Research TrianglePark, NC. Retrieved from:https://www.gsksource.com/gskprm/htdocs/documents/BACTROBAN-NASAL.PDF84Granger, B. B., Moser, D., Harrell, J., Sandelowski, M., & Ekman, I. (2007). A practical use oftheory to study adherence. Progress in Cardiovascular Nursing, 22, 152-158.doi:10.1111/j.0889-7204.2007.888144.xHansen, D., Patzke, P., Werfel, U., Benner, D., Brauksiepe, A., & Popp, W. (2007). Success ofMRSA eradication in hospital routine: Depends on compliance. Infection, 35, 260-264. doi:10.1007/s15010-007-6273-yHaynes, R. B., Yao, X., Degani, A., Kripalani, S., Garg, A., & McDonald, H. P. (2005).Interventions for enhancing medication adherence. Cochrane Database of SystematicReviews 2005, Issue 4. Art. No.: CD000011. doi:10.1002/14651858.CD000011.pub2.Ingersoll, K. S., & Cohen, J. (2008). The impact of medication regimen factors on adherence tochronic treatment: A review of literature. Journal of Behavioral Medicine, 31, 213-224.doi:10.1007/s10865-007-9147-yJohnson, A. J., Daley, J. A., Zywiel, M. G., Delanois, R. E., & Mont, M. A. (2010). Preoperativechlorhexidine preparation and the incidence of surgical site infections after hiparthroplasty. Journal of Arthroplasty, 25(Suppl. 6), 98-102. doi:10.1016/j.arth.2010.04.012Kallen, A. J., Patel, P. R., & Hess, S. (2011). Intolerance of chlorhexidine as a skin antiseptic inpatients undergoing hemodialysis [Letter to the editor]. Infection Control and HospitalEpidemiology, 32, 1144-1146. doi:10.1086/662591Kim, D. H., Spencer, M., Davidson, S. M., Li, L., Shaw, J. D., Gulczynski, D.,?Richmond, J. C.(2010). Institutional prescreening for detection and eradication of methicillin-resistantStaphylococcus aureus in patients undergoing elective orthopaedic surgery. Journal ofBone and Joint Surgery. American Volume, 92, 1820-1826. doi:10.2106/JBJS.I.0105085Liang, S. Y., Yates, P., Edwards, H., & Tsay, S. L. (2008). Factors influencing opioid-takingself-efficacy and analgesic adherence in Taiwanese outpatients with cancer. Psycho-Oncology, 17, 1100-1107. doi:10.1002/pon.1326Liu, L. L., & Gonzalez, R. (2007). Judgment and decision processes in older adults' compliancewith medical regimens. In D. C. Park, & L. L. Liu (Eds.), Medical adherence and aging:Social and cognitive perspectives (pp. 201-232). Washington, DC: AmericanPsychological Association. doi:10.1037/11557-009McDaniel, M. A., & Einstein, G. O. (2007). Prospective memory components most at risk forolder adults and implications for medication adherence. In D. C. Park & L. L. Liu (Eds.),Medical adherence and aging: Social and cognitive perspectives (pp. 49-75). Washington,DC: American Psychological Association. doi:10.1037/11557-003McDonald-Miszczak, L., Maris, P., Fitzgibbon, T., & Ritchie, G. (2004). A pilot studyexamining older adults? beliefs related to medication adherence: the BERMA survey.Journal of Aging and Health, 16, 591-614. Retrieved from:http://www.ncbi.nlm.nih.gov/pubmed/15448274Mimoz, O., Villeminey, S., Ragot, S., Dahyot-Fizelier, C., Laksiri, L., Petitpas, F., & Debaene,B. (2007). Chlorhexidine-based antiseptic solution vs alcohol-based povidone-iodine forcentral venous catheter care. Archives of Internal Medicine, 167, 2066?2072.doi:10.1001/archinte.167.19.2066Morris-Morris, R., Weinman, J., Petrie, K. J., Horne, R., Cameron, L. D., & Buick, D. (2002).The revised illness perception questionnaire (IPQ-R). Psychology and Health, 17, 1-16.doi:10.1080/0887044029000149486Munro, B. H. (2005). Statistical methods for health care research (5th ed.). Philadelphia, PA:Lippincott Williams & Williams.Ngamvitroj, A., & Kang, D. (2007). Effects of self-efficacy, social support and knowledge onadherence to PEFR self-monitoring among adults with asthma: A prospective repeatedmeasures study. International Journal of Nursing Studies, 44, 882-892.doi:10.1016/j.ijnurstu.2006.03.001Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory (3rd ed.). New York: McGraw-Hill.Osterberg, L., & Blaschke, T. (2005). Adherence to medication. New England Journal ofMedicine, 353, 487-497. doi:10.1056/NEJMra050100Park, D. C., & Meade, M. L. (2007). A broad view of medical adherence: Integrating cognitive,social, and contextual factors. In D. C. Park & L. L. Liu (Eds.), Medical adherence andaging: Social and cognitive perspectives (pp. 3-21). Washington, DC: AmericanPsychological Association. doi:10.1037/11557-001Peddle, C. J., Jones, L. W., Eves, N. D., Reiman, T., Sellar, C. M., Winton, T., & Courneya, K.S. (2009). Correlates of adherence to supervised exercise in patients awaiting surgicalremoval of malignant lung lesions: Results of a pilot study. Oncology Nursing Forum, 36,287-295. doi:10.1188/09.ONF.287-295Perl, T. M., Cullen, J. J., Wenzel, R. P., Zimmerman, M. B., Pfaller, M. A., Sheppard, D.,?Herwaldt, L. A. (2002). Intranasal mupirocin to prevent postoperative staphylococcusaureus infections. New England Journal of Medicine, 346, 1871-1877. Retrieved from:http://www.nejm.org.ezproxy.library.ubc.ca/medical-index87Polit, D. F., & Beck, C. T. (2012). Nursing research: Generating and assessing evidence fornursing practice (9th ed.). Philadelphia: Lippincott Williams & Wilkins.Provincial Infection Control Network of British Columbia. (2008). Antibiotic resistant organism(ARO) guidelines. Retrieved from: http://www.picnetbc.ca/practice-guidelinesRao, N., Cannella, B. A., Crossett, L. S., Yates, A. J., McGough, R. L., & Hamilton, C. W.(2011). Preoperative screening/decolonization for staphylococcus aureus to preventorthopedic surgical site infection: Prospective cohort study with 2-year follow-up. Journalof Arthroplasty, 26, 1501-1507. doi:10.1016/j.arth.2011.03.014Rogers, R. W., & Prentice-Dunn, S. (1997). Protection motivation theory. In D. Gochman (Eds.),Handbook of health behavior research (pp. 113-132). New York: Plenum.Rosenstock, I., Strecher, V., & Becker, M. (1988). Social learning theory and the health beliefmodel. Health Education & Behavior, 15, 175-183. doi:10.1177/109019818801500203Shi, L., Liu, J., Koleva, Y., Fonseca, V., Kalsekar, A., & Pawaskar, M. (2010). Concordance ofadherence measurement using self-reported adherence questionnaires and medicationmonitoring devices. Pharmacoeconomics, 28, 1097-1107. doi:10.2165/11537400-000000000-00000Siegel, K., Karus, D., & Schrimshaw, E. W. (2000). Racial differences in attitudes towardprotease inhibitors among older HIV-infected men. Aids Care, 12, 423-434. Retrievedfrom: http://www.ncbi.nlm.nih.gov/pubmed/11091775Sirur, R., Richardson, J., Wishart, L., & Hanna, S. (2009). The role of theory in increasingadherence to prescribed practice. Physiotherapy Canada, 61, 68-77.doi:10.3138/physio.61.2.6888Sleath, B., Blalock, S., Robin, A., Hartnett, M., Covert, D., DeVellis, B., & Giangiacomo, A.(2010). Development of an instrument to measure glaucoma medication self-efficacy andoutcome expectations. Eye, 24, 624-631. doi:10.1038/eye.2009.174Tarantino, V., Cappellari, G., Cardaioli, C., Rumiati, R., Savadori, L., Barilli, E., & Bisiacchi, P.S. (2010). Sociocognitive factors associated with nonadherence to medication after hospitaldischarge. Behavioral Medicine, 36, 100-107. doi:10.1080/08964281003774935Taylor, G. H., Wilson, S. L., & Sharp, J. (2011). Medical, psychological, and sociodemographicfactors associated with adherence to cardiac rehabilitation programs: A systematic review.Journal of Cardiovascular Nursing, 26, 202-209. doi:10.1097/JCN.0b013e3181ef6b04Vall?s, J., Fern?ndez, I., Alcaraz, D., Chac?n, E., Cazorla, A., Canals, M.,?Mor?n, A. (2008).Prospective randomized trial of 3 antiseptic solutions for prevention of cathetercolonization in an intensive care unit for adult patients. Infection Control and HospitalEpidemiology, 29, 847-853. doi:10.1086/590259Van Rijen, M. M., Bonten, M., Wenzel, R. P., & Kluytmans, J. A. (2008a). Intranasal mupirocinfor reduction of staphylococcus aureus infections in surgical patients with nasal carriage: Asystematic review. Journal of Antimicrobial Therapy, 61, 254-261.doi:10.1093/jac/dkm480Van Rijen, M., Bonten, M., Wenzel, R., & Kluytmans, J. (2008b). Mupirocin ointment forpreventing staphylococcus aureus infections in nasal carriers. Cochrane Database ofSystematic Reviews 2008, Issue 4. Art. No.:CD006216. doi:10.1002/14651858.CD006216.pub289Vaughn-Cooke, M. (2009). Development of an integrated model for predicting patientcompliance. Human Factors and Ergonomics Society Annual Meeting Proceedings, 53,749-753. doi: 10.1518/107118109X12524442637066Wallston, K. A., Wallston, B. S., & DeVellis, R. (1978). Development of the multidimensionalhealth locus of control (MHLC) scales. Health Education & Behavior, 6, 160-170.doi:10.1177/109019817800600107Wenzel, R. P., & Perl, T. M. (1995). The significance of nasal carriage of staphylococcus aureusand the incidence of postoperative wound infection. Journal of Hospital Infection, 31, 13-24. Retrieved from:http://www.sciencedirect.com.ezproxy.library.ubc.ca/science/journal/01956701/31/1Yuhas, J., Mattocks, K., Gravelin, L., Remetz, M., Foley, J., Fazio, R., & Lampert, R. (2012).Patients' attitudes and perceptions of implantable cardioverter-defibrillators: Potentialbarriers to appropriate primary prophylaxis. Pacing & Clinical Electrophysiology, 35,1179-1187. doi:10.1111/j.1540-8159.2012.03497.xZywiel, M. G., Daley, J. A., Delanois, R. E., Naziri, Q., Johnson, A. J., & Mont, M. A. (2011).Advance pre-operative chlorhexidine reduces the incidence of surgical site infections inknee arthroplasty. International Orthopaedics, 35, 1001-100690Appendix ASelf-Efficacy SurveyThank you for participating in the survey.This survey will take you about 15 min to complete and consists of 3 sections. The answers yougive are strictly confidential and will not be shared with anyone other than the people conductingthis study.Please answer the following questions by circling the number that best describesyour level of confidence.0 = Not at all confident1 = Slightly confident2 = Somewhat confident3 = Moderately confident4 = Extremely confident1.0 How confident are you that you can apply the chlorhexidinecloths??..Not atallconfidentSlightlyconfidentSomewhatconfidentModeratelyconfidentExtremelyconfident1.1 When you are busy athome?0 1 2 3 41.2 When there is no one toremind you?0 1 2 3 41.3 When there is someoneto remind you?0 1 2 3 41.4 When they may causeside effects?0 1 2 3 41.5 When you are afraidthey might irritate orcause burning to yourskin?0 1 2 3 41.6 When they might smellunpleasant?0 1 2 3 4911.7 When you have othermedications to take?0 1 2 3 4How confident are you that you can carry out the followingtasks????..Not atallconfidentSlightlyconfidentSomewhatconfidentModeratelyconfidentExtremelyconfident1.8 Open the packages ofchlorhexidine cloths?0 1 2 3 41.9 Apply thechlorhexidine cloths tothe correct body areas?0 1 2 3 41.10 Wash the appropriatebody areas with thechlorhexidine cloths?0 1 2 3 41.11 Remember to washwith the chlorhexidinecloths the eveningbefore surgery?0 1 2 3 41.12 Use the chlorhexidinecloths without havingsomeone help you?0 1 2 3 41.13 If you are not confidentin using thechlorhexidine clothswithout someonehelping you, do youhave someone that canhelp you if needed?Please circle youranswer.Yes No NotApplicable2.0 How confident are you that you can apply nasalMupirocin???.Not atallconfidentSlightlyconfidentSomewhatconfidentModeratelyconfidentExtremelyconfident2.1 When you are busy athome?0 1 2 3 42.2 When there is no one toremind you?0 1 2 3 42.3 When there is someoneavailable to remindyou?0 1 2 3 42.4 When it may cause sideeffects?0 1 2 3 4922.5 When you are afraid itmight irritate or causeburning to yournostrils?0 1 2 3 42.6 When you have othermedications to take?0 1 2 3 4How confident are you that you can carry out the followingtasks????..Not atallconfidentSlightlyconfidentSomewhatconfidentModeratelyconfidentExtremelyconfident2.7 Fill your prescriptionfor nasal Mupirocin?0 1 2 3 42.8 Remember to apply thenasal Mupirocin twicedaily for 5 days beforesurgery?0 1 2 3 42.9 Open the tube of nasalMupirocin?0 1 2 3 42.10 Get the plastic seal offa new tube of nasalMupirocin?0 1 2 3 42.11 Squeeze the nasalMupirocin tube?0 1 2 3 42.12 Apply the Mupirocinointment to a Q-tip? 0 1 2 3 42.13 Correctly angle yourhead to accuratelyapply the Q-tip in eachnostril?0 1 2 3 42.14 Get the right amount ofointment onto the Q-tipeach time?0 1 2 3 42.15 Use the nasalMupirocin withouthaving someone helpyou?0 1 2 3 42.16 If you are not confidentin using the nasalMupirocin withoutsomeone helping you,do you have someonethat can help you ifneeded? Please circleyour answer.Yes No NotApplicable933.0 How confident are you that in using chlorhexidine cloths andnasal Mupirocin, if it is required??Not atallconfidentSlightlyconfidentSomewhatconfidentModeratelyconfidentExtremelyconfident3.1 You will prevent aninfection fromoccurring after yoursurgery?0 1 2 3 43.2 You will reduce yourrisk for developing aninfection?0 1 2 3 4What is your gender? Male Female Other, please specify_________________Please circle your answerIn what year were you born? _________________Please print the full year on the line above (example: 1980)What is the date of your surgery? ___________________Please write the date on the line above in day/month/year format (example: 25/Dec/1980)Please place your completed survey inside the envelope provided labeled ?consentform and survey.? Once you have placed this survey and your signed consent forminside the envelope, please seal the envelope by removing the white sticker at theback of the envelope. Please bring this envelope to the hospital on the day of yoursurgery and please give it to your nurse in the surgical day care unit when youcheck in.Thank you for completing this survey.94Adherence QuestionnaireAge: Sx. Date:Did you wash with the chlorhexidine cloths the evening before surgery? Please circle: Yes or NoIf no, why not?If yes, what time did you wash with the cloths? ___________am or pm (please indicate the time)If not the evening before surgery, why not?If you applied the cloths, how many did you apply? __________cloths (please provide a number)If not all 6 cloths, why not?Where did you apply the cloths? Please circle where the cloths were applied.One or both arms One or both legs Chest Back Genital Area ButtocksIf applicable?..Did you fill the prescription for Mupirocin? Please circle: Yes or NoIf no, why not?When did you begin the application? _________days before surgery. (Please provide a number)If not 5 days before surgery, why not?How many times a day did you apply the Mupirocin? Please circle: Once Twice NoneIf not twice daily, why not?Where did you apply the Mupirocin? Please circle: Nostrils or Other, please specify__________95If not the nostrils, why not?Did you apply Mupirocin to each nostril? Please circle: Yes or NoIf not to each nostril, why not?

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.24.1-0103388/manifest

Comment

Related Items