Skip Navigation


NDT Advance Access originally published online on April 18, 2007
Nephrology Dialysis Transplantation 2007 22(8):2339-2348; doi:10.1093/ndt/gfm149
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
22/8/2339    most recent
gfm149v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (3)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Chisholm, M. A.
Right arrow Articles by Mulloy, L. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Chisholm, M. A.
Right arrow Articles by Mulloy, L. L.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author [2007]. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

Predicting adherence to immunosuppressant therapy: a prospective analysis of the theory of planned behaviour

Marie A. Chisholm1, Gail M. Williamson2, Charles E. Lance3 and Laura L. Mulloy4

1University of Arizona College of Pharmacy, 2Life-Span Developmental Psychology Program, University of Georgia Department of Psychology, 3Applied Psychology Program, University of Georgia Department of Psychology and 4Section of Nephrology, Hypertension and Transplantation Medicine, Medical College of Georgia School of Medicine, USA

Correspondence and offprint requests to: M. A. Chisholm, Pharm.D., FCCP, FASHP, The University of Arizona College of Pharmacy, Pharmacy-Pulido Center, 1295 N. Martin Ave., Tucson, AZ 85721, USA. Email: Chisholm{at}pharmacy.arizona.edu



   Abstract
Top
 Abstract
Introduction
 Subjects and methods
 Results
 Discussion
 Appendix A. Theory of...
 Acknowledgements
 References
 
Background. The objective of this study was to increase the ability to predict renal transplant patients (RTPs) who are most likely to be non-adherent to their immunosuppressant therapy (IST).

Methods. One hundred and fifty-eight RTPs completed questionnaires assessing Theory of Planned Behaviour (TPB) variables (attitudes, subjective norms and perceived behavioural control) relevant to intentions to adhere to their IST, with the addition of a general measure of past adherence to medical advice. In the full sample, intentions to adhere to IST was the outcome variable. In a subsample of 70 RTPs, the primary outcome was IST adherence.

Results. TPB variables (attitudes, ß = 0.32, P < 0.01; perceived behavioural control, ß = 0.37, P < 0.01; but not subjective norms, ß = –0.001, ns) explained 41% of the variance in intentions to adhere to IST (P < 0.001). Past behaviour predicted perceived behavioural control (ß = 0.67, P < 0.001). Subsample analyses explained 33% (P < 0.001) of the variance in adherence, with intentions and past behaviour being the primary factors (P < 0.05).

Conclusions. RTPs particularly at risk may be those who have a history of non-adherence to medical advice, especially when they have negative attitudes about IST adherence and feel they have little control over their medication-taking behaviour. Interventions to improve attitudes about IST adherence and control of adherence behaviour are needed.

Keywords: adherence; immunosuppressant medications; renal transplant recipients



   Introduction
Top
 Abstract
 Introduction
Subjects and methods
 Results
 Discussion
 Appendix A. Theory of...
 Acknowledgements
 References
 
Non-adherence to recommended medical treatment has been declared a major public health problem [1]. Therapeutic regimens of renal transplant patients (RTPs) typically involve taking at least 10 different prescribed medications per day, including immunosuppressant therapy (IST) for as long as their grafts continue to function [2]. As a result, RTPs represent a population particularly prone to medication non-adherence. Research assessing both IST adherence and follow-up clinic visits has shown that 35–90% of non-adherent RTPs either lose their graft or die, compared with 18% of those who are adherent [3,4]. In fact, IST non-adherence is the leading avoidable cause of renal transplant failure [4]. A meta-analysis found that non-adherence contributed substantially to graft loss, with a median of 36% of graft losses associated with poor adherence. The odds of graft failure increased seven-fold in non-adherent compared with adherent RTPs [5]. Still, non-adherence is common, occurring in 20–50% of RTPs [5]. To implement effective strategies to increase adherence and decrease adverse outcomes, practitioners need to be able to identify those at greatest risk for non-adherence.

Several theories/models have been formulated to help predict and understand health-related behaviour. Each has its limitations, and few are well suited to identifying factors that contribute to non-adherence to a prescribed medical regimen as crucial as IST. For example, the Health Belief Model has demonstrated limited success in predicting health-related behaviour [6]. Other models, such as those based on ecological [7–9] or integrated [10,11] perspectives, are compelling because they posit multiple levels of influence and reciprocal causation. However, such models are limited in their utility for at least two reasons: (i) assessing all their variables requires a very large sample; and (ii) widely used measures for all of these factors are not yet available.

The Theory of Planned Behaviour (TPB), a psychosocial-cognitive model, has been used successfully in predicting a wide range of health-related behaviour and has several advantages compared with other health behaviour prediction models and theories [12]. First, the relatively small number of constructs in the TPB model (attitudes, subjective norms, perceived behavioural control, and intentions) make it possible to assess the full model in a sample of limited size (e.g. RTPs). Second, as an underlying assumption of the TPB is that other factors (e.g. culture environment) do not influence behaviour directly but operate through the model's constructs, it is not necessary to test a large number of demographic and medical factors. Third, ways to measure each of the TPB variables have been developed and repeatedly validated in a large number of different behavioural domains and subject populations [13,14]. Since the TPB concerns basic psychological process antecedents to actual behaviour, it can be used to predict any type of behaviour, so that the TPB instruments can be adapted to fit many populations including RTPs [13]. Finally, it has been proposed that the predictive power of the TPB could be enhanced by adding variables such as past behaviour to the model [14].

No previous research has utilized the TPB to determine either intentions to adhere or actual adherence to IST among RTPs. This study evaluated the ability of, (i) the fundamental TPB variables (attitudes, subjective norms and perceived behavioural control) plus past behaviour to predict intentions to adhere to IST and (ii) the ability of the full model (attitudes, subjective norms, perceived behavioural control, past behaviour and intentions) to predict actual IST adherence. The investigators hypothesized that, (i) the fundamental TPB variables (attitudes, subjective norms and perceived behavioural control) would predict intentions to adhere to IST; (ii) intentions to adhere to IST would predict actual adherence to IST; (iii) past behaviour would predict current behaviour (i.e. adherence); and (iv) associations between past behaviour and intentions would be mediated by perceived behavioural control (such that the extent to which past behaviour was related to intentions to take IST would depend, at least in part, on how much patients perceived they could control their IST-taking behaviour).



   Subjects and methods
Top
 Abstract
 Introduction
 Subjects and methods
Results
 Discussion
 Appendix A. Theory of...
 Acknowledgements
 References
 
Sample and procedure
The TPB questionnaire was mailed in March 2004 to 164 RTPs who were enrolled in the Medication Access Program (MAP). MAP is a statewide program that provides medication access to solid-organ transplant recipients who reside in the State of Georgia. A follow-up survey was mailed in April 2004 to those who did not complete and return the questionnaire in March. All patients in the study had a functioning graft, were taking cyclosporine or tacrolimus as their IST, and were 18 years of age or older. The portion of the model (attitudes, subjective norms, perceived behavioural control and past behaviour) predicting intentions was tested using the full sample of patients who completed and returned their questionnaires. The full model (Figure 1) predicting actual IST adherence was evaluated in a subsample of RTPs whose pharmacy refill records were available as the measure of adherence. The study was approved by the Human Assurance Committee at the Medical College of Georgia (MCG).


Figure 1
View larger version (15K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 1. Summary of regression results.

 
Measures
Sample characteristics data
Age, gender, race/ethnicity, marital status, donor type and time since transplant were collected from patients’ medical records. Annual income was available from income tax and social security benefit forms provided by patients.

TPB instruments
Evidence for the validity of the TPB model is supported by numerous studies [13,14], including meta-analyses indicating that components of the TPB, attitudes, subjective norms, perceived behavioural control and intentions, account for 40–50% of variance in targeted behaviour [12,15–17]. Intentions and perceived behavioural control explained 20–40% of variance in health-related behaviour when assessed prospectively [12,15,16]. Assessments of attitudes, subjective norms, perceived behavioural control and intentions were derived from previous TPB research and closely follow Ajzen's [13] recommendations for constructing a TPB questionnaire. See Appendix A for the TPB questionnaire used in the current study. All instruments have demonstrated internal reliability in past research as well as construct and predictive validity. Wording was changed to make instruments applicable to our population. Concurrent validity evidence was established using correlation coefficients between TPB variables and pharmacy refill records as a measure of the intended behaviour, IST adherence.

Attitudes
Attitudes are cognitive evaluations of the consequences (both positive and negative) of performing a target behaviour—in this case, taking IST as prescribed. Attitudes were measured by eight responses to the stem, ‘Taking my immunosuppressant medication(s) as prescribed is ...’ (e.g. 1 = Bad, 2 = Neither bad nor good, 3 = Good; 1 = Unpleasant, 2 = Neither unpleasant nor pleasant, 3 = Pleasant). Four questions assessed the extent to which patients felt that taking their IST: (i) ‘protects against rejecting (losing) my transplanted kidney’, (ii) ‘makes my life less enjoyable’, (iii) ‘makes my life difficult’ and (iv) ‘makes my life too complicated’ (1 = Strongly disagree to 5 = Strongly agree). Negatively worded items were reverse-scored so that higher values indicated more positive attitudes. In order to equate ranges and variances of these two subsets of items, the first eight items (anchored with 3-point scales) were weighted so their effective range was the same as the last four items (i.e. 1–5). Item scores were then summed up to create a composite measure that could range from 12 (extremely negative) to 60 (extremely favourable).

Subjective norms
The TPB defines subjective norms as perceptions of pressure/support from important others (e.g. family, healthcare providers) to perform the target behaviour (in this study the target behaviour is taking IST as prescribed). Two items asked whether ‘People who are important to me would (1 = Definitely advise me to take my immunosuppressant medication(s) as prescribed to 4 = Not advise me one way or the other)’ and ‘If they were in my place, the people who are important to me would (1 = Definitely take their immunosuppressant medication(s) as prescribed to 3 = Take their immunosuppressant medication(s) in whatever way they felt was best)’. Three additional items asked were whether: ‘My husband/wife/significant other thinks I should take my immunosuppressant medication(s) as prescribed’, ‘My friends and family think I should take my immunosuppressant medication(s) as prescribed and ‘My doctors (or other healthcare professionals) would approve if I take my immunosuppressant medication(s) as prescribed’ (1 = Definitely disagree to 5 = Definitely agree). The first two items were weighted to equate their scales to the last three items’ 5-point scales. Item scores were then summed up to create a composite measure of subjective norms that could range from 5 to 25, with higher values indicating greater perceptions that important others supported IST adherence.

Perceived behavioural control
Assessing perceived behavioural control involves measuring the extent to which people feel they are able to control the target behaviour. Specific to this study, RTPs reported how easy it was for them to take their IST as prescribed (1 = Very difficult to 5 = Very easy) and how much control they felt they had over taking their IST as prescribed (1 = No control to 4 = Complete control). The second item was weighted to equate to the first item's 5-point scale. Thus, composite scores could range from 2 (little or no control) to 10 (very much control).

Intentions
In the TPB, the immediate determinant of behaviour is the intention to act or not to act. In this study, RTPs responded to two items: ‘I have every intention of taking my IST as prescribed’ (1 = Strongly disagree to 5 = Strongly agree); and ‘How likely are you to take your IST as prescribed?’ (1 = Extremely unlikely to 5 = Extremely likely). The composite scores could range from 2 (little to no intention) to 10 (very high intention).

Past behavior
Because the strongest predictor of future behaviour is past behaviour in a similar situation [18], adding past behaviour to the TPB may enhance its predictive ability. However, rather than acting as a direct predictor of intentions, according to Ajzen [18], the effect of past behaviour on intentions should be mediated by perceived behavioural control, primarily because past behaviour can be an important source of information for perceptions of behavioural control. Past behaviour was conceptualized as long-term tendencies to comply with medical regimens in general. Two items asked RTPs how often (i) over their entire life and (ii) over the last 3 months they ‘did what doctors (or other healthcare professionals) recommended’(1 = Never to 4 = Always). The composite scores could range from 2 to 8 with higher scores indicating greater past adherent behaviour.

IST adherence
A subsample of RTPs received their medications from the MCG pharmacy, and their pharmacy refill data were used as the measure of IST adherence. No measure of adherence is without limitations, but refill records were used in this study as they provide an objective measure of medication access and presumably adherence [19–21] and were easily assessed as the subsample for patients who received their medications from the MCG pharmacy. A study examining the validity of a prescription claims database to estimate medication adherence found a high concordance rate between the database and pill counts, suggesting that the rate with which patients refill medications is generally consistent with the rate they consume them [22]. Furthermore, studies have indicated that RTPs who get their immunosuppressant refills filled appropriately are more likely to have desired immunosuppressant serum concentrations, indicating that refill records are a good measure of adherence [23].

Pharmacy refill records for cyclosporine (Gengraf®) or tacrolimus (Prograf®) were collected using the MCG pharmacy's computer system for a consecutive 3-month period beginning one month after patients completed our questionnaire. Three months is often adequate to observe patterns of adherence to chronic medications [24]. Medications were dispensed as a 30-day supply. Monthly refill records were compared with the prescribed regimen documented in medical records. Adherence rates were then calculated by using the number of days between IST refills. If the total number of days between refills was less than or equal to the total days’ supply of IST, adherence rate was 100%. If the number of days between refills was greater than the days’ supply, adherence rate was calculated according to the following formula:


Formula

Extra doses accumulated during the study period were assumed to be used as needed in order to adhere to prescribed therapy if medication refills were not obtained on time.

To make the study more practically applicable, the investigators utilized an adherence rate guideline to distinguish IST ‘adherent’ RTPs from ‘non-adherent’ RTPs. Since refill-based adherence rates of 80% or greater have been found to be correlated with other markers of IST adherence such as target serum IST concentrations and lower SCr concentrations (P < 0.05) [23,25,26], this percentage was used to develop an adherent vs non-adherent diagnostic index for IST. Thus, if the refill-based IST adherence rate was less than 80%, the patient was categorized as non-adherent; if the adherence rate was 80% or greater, the patient was considered adherent. Use of this diagnostic index (adherence categories) allows practitioners to identify those patients who are non-adherent based on adherence measures and patient-specific outcomes associated with adherence rates. Furthermore, it provides for targeting of RTPs who are at higher risk for non-adherence so that effective adherence interventions can be implemented. The 80% level has been widely used to categorize adherent and non-adherent patients in other patient populations [27–31].

Statistical analyses
Demographic characteristics have often been associated with IST adherence—patients at highest risk for IST non-adherence are more often male, single, non-Caucasian and have lower incomes [3]. However, these factors often account for small proportions of variance. The purpose of the present study was to investigate the ability of the TPB to predict IST adherence beyond the effects of patient demographic factors. Nevertheless, to control for the potential effects of demographic variables, we used bivariate correlations to examine associations between demographic factors and TPB constructs. If any demographic variable was correlated with any TPB variable(s), TPB variables were regressed onto that demographic factor, and the residuals were retained for further analyses. With these controls, the TPB model was tested using path analysis application of multiple linear regression with an alpha level of 0.05 [32].



   Results
Top
 Abstract
 Introduction
 Subjects and methods
 Results
Discussion
 Appendix A. Theory of...
 Acknowledgements
 References
 
Preliminary analyses
Initial data were obtained from 158 participants for a 96% participation rate. These participants were 60% male and had a mean age of 51 years (SD = 12.4, range 19–79). Data from a subsample of 70 individuals who were MCG RTPs were used in analyses of TPB and past behaviour constructs as predictors of adherence. This subsample (n = 70) was similar to the total sample (N = 158), with 58.6% male and a mean age of 51.6 years (SD = 12.5, range 28–79; Table 1). In fact, the MCG subsample did not differ significantly from the total sample on any of the variables shown in Table 1. Based on the 80% rate determining adherence, 27% of patients in the subsample were non-adherent, a percentage similar to findings in previous research [5]. Of the three RTPs who experienced a graft rejection, two were female. These patients were between 57–73 years of age (mean = 66 years, SD = 8.19) and had their grafts an average of 7.33 years (SD = 0.58; Table 1). All three rejections occurred within 14 weeks of completing the TPB questionnaire (two rejections occurred within 2 months after questionnaire completion; the third rejection occurred within 3 months).


View this table:
[in this window]
[in a new window]

  		Table 1. Characteristics of renal transplant patients (RTPs)

 
Of the patient demographic characteristics, only age was related to study variables (all other demographic characteristics were not statistically significant). Older patients had less favourable attitudes about IST adherence (r = –0.28, P < 0.01), felt they had less control over taking their IST (r = 0.31, P < 0.01), perceived weaker subjective norms (r = –0.35, P < 0.01), had lower intentions to adhere to IST regimen (r = –0.32, P < 0.01), reported less adherent past behaviour (r = –0.35, P < 0.01), and, in fact, were less adherent (r = –0.48, P < 0.01). Therefore, age was statistically controlled by regressing each of the study variables onto age, saving the regression residuals and performing the study's focal analyses on the residualized study variable scores.

Analyses of TPB measures
Reliability analyses were conducted for each component of the TPB model including past behaviour. Cronbach's alpha was 0.86 for the 12 attitudes items, 0.83 for the subjective norms items, 0.76 for the two-item perceived behavioural control composite, 0.87 for the intentions items and 0.82 for the two past behaviour items.

In order to provide validity evidence for the TPB measures as adapted here, the investigators correlated the measures with adherence data collected from pharmacy refill records obtained at approximately the same time as the TPB questionnaire data. Each of the TPB variables exhibited significant concurrent validity: attitudes (r = 0.67, P < 0.01), subjective norms (r = 0.63, P < .01), perceived behavioral control (r = 0.76, P < 0.01), intentions (r = 0.46, P < .01), and past behavior (r = 0.64, P < 0.01) were all related to IST adherence. As such, significant concurrent validity of the TPB variables supports their use in the present population.

Descriptive analyses were also conducted for each component of the TPB, and mean scores were as follows: attitudes was 53.4 (SD = 6.4, range 37–60); subjective norms was 23.4 (SD = 2.9, range 13–25); perceived behavioural control was 8.9 (SD = 1.6, range 4–10); and intentions was 9.6 (SD = 1.0, range 5–10). The average RTP also had a mean past behaviour (past medical recommendations) score of 6.7 (SD = 1.4, range 4–8).

Tests of models
Basic TPB model predicting IST adherence
In the full sample (N = 158), path analysis results indicated, as hypothesized, that more favourable attitudes (ß = 0.32, P < 0.01) and more perceived behavioural control (ß = 0.37, P < 0.01) predicted greater intentions to adhere to IST. Contrary to expectations, stronger subjective norms did not predict intentions (ß = –0.00, ns; see also standardized regression coefficients as presented in Figure 1). This model explained 41% of variance in intentions to adhere to IST (R2 = 0.41, P < 0.001). Subsample (n = 70) results supported the hypothesis that intentions to adhere to IST regimens would forecast actual adherence (ß = 0.26, P < 0.05).

Adding past behaviour to the TPB model
Based on previous empirical and theoretical work [12], the investigators hypothesized that past behaviour would directly predict perceived behavioural control and adherence. In analysis of the total sample (N = 158), past behaviour was a strong predictor of perceived behavioural control (ß = 0.67, P < 0.001), but not of intentions (ß = 0.01, ns; see Figure 1). As such, and as predicted, perceived behavioural control served to completely mediate the effects of past behaviour on intentions to adhere to IST. In the subsample (n = 70), past behaviour also independently predicted adherence (ß = 0.39, P < 0.05) after controlling for TPB constructs (i.e. attitudes, subjective norms, perceived behavioural control and intentions). Together, intentions and past behaviour explained 33% of variance in adherence (P < 0.001; intentions explained approximately 10% and past behaviour explained approximately 23% of the variance in adherence).



   Discussion
Top
 Abstract
 Introduction
 Subjects and methods
 Results
 Discussion
Appendix A. Theory of...
 Acknowledgements
 References
 
This study represents for the first time that TPB variables and patients’ past medically compliant behaviour have been incorporated into a single study of IST adherence among RTPs. From a clinician's perspective, our results may be particularly helpful because they provide insights about ways to forestall or possibly avoid non-adherence. Since the focus of this study was on basic psychosocial and cognitive variables that should theoretically apply to all members of a population, we excluded medically related factors (e.g. type of transplant donor, time since transplant and comorbid conditions). Indeed, even with these exclusions, TPB variables and past behaviour explained a large percentage of the variance in IST adherence. Specifically, patients with weaker intentions to adhere to their regimens were less likely to be adherent. We speculated that RTPs with a history of not following medical advice in general would be less likely to take their IST as recommended, and in fact, a history of non-adherence to medical advice was a significant factor in predicting both perceived behavioural control and adherence.

Important implications for intervention can be drawn from the results. First, it appears that patients who were, in general, non-adherent to medical advice in the past can be expected to be less adherent to their IST. This finding is consistent with evidence that pre-transplant non-compliance is related to graft loss in kidney transplantation [33]. A second implication is that the extent to which patients believe they are unable to control taking their IST as prescribed is a pivotal factor, one that is affected by past behavioural lapses. These findings are reminiscent of learned helplessness theory and research—i.e. the feeling of being unable to exercise control in one situation can generalize to other situations. Numerous psychological treatments afford ways to help people realize that they can control their behaviour. For example, behavioural modification techniques and stress and coping skills have been used successfully to increase patient adherence. In addition, we suspect that attitudes can become more favourable when patients are helped to realize that they can control their IST-taking behaviour. The overall message is that successful interventions can overcome past non-adherent behaviour and, consequently, increase intentions to be adherent. And, as shown in previous research, IST adherence is an important factor in graft maintenance [3,5].

With the exception of patient age, no demographic variables were related to our study variables. By contrast, age was associated with all of the study variables. That is, older RTPs reported less favourable attitudes, less behavioural control, weaker subjective norms, less adherent past behaviour, weaker intentions to adhere and less adherence to their IST. Findings in the literature regarding the relationship between age and IST non-adherence among RTPs have been inconsistent. Although some studies have found an association between younger age and IST non-adherence [34–36], other studies have found no relationship between age and non-adherence [25,37,38]. Furthermore, almost no studies of IST adherence have been conducted exclusively among elderly RTPs, who are the fastest growing segment of the RTP population and were approximately 13% of our sample. As our primary intent was to evaluate the TPB, a full examination of the meaning of age-related differences was beyond the scope of this study. Thus we controlled for age in our analyses and did not attempt to determine the mechanisms contributing to the age effect. However, we fully acknowledge the association between adherence and age, and believe this is an area in need of further research, particularly in light of the conflicting or inadequate evidence regarding the relationship between age and IST adherence among RTPs [25,34–38].

The study has several limitations. The first is that our data collection period of 3 months was not sufficient to assess the longer-term effects of IST non-adherence on graft function. This was not a goal of this particular study, but there is sufficient existing evidence demonstrating the importance of IST adherence to continued graft function [3,5]. Moreover, post hoc analyses of the three cases of biopsy-proven graft rejection revealed that adherence was negatively associated with graft failure (r = –0.35, P < 0.01). Having patients report their past adherence with medical recommendations may also be viewed as a limitation. Another source of this information could be clinicians’ ratings. However, in a study using electronic monitoring devices (e.g. MEMS caps) as the standard to measure adherence, the best measure of adherence was RTPs’ self-report and not clinicians’ ratings [19]. Since patient self-report is convenient, inexpensive, and achievable in the routine clinic setting, it appears that this method of evaluation should be relied on and studied further. Additional research on how best to facilitate disclosure in clinical settings may be the best way to develop adherence measures for use in routine practice [19]. Our study details a method that appears to be promising for predicting adherence.

Questions could be raised about whether completing the survey caused patients to change adherence behaviour and, thus influenced the results (i.e. a Hawthorne effect). To address this issue, we examined participants’ IST refill records for the three months prior to survey completion (the retrospective model), and found that age of RTPs related to study variables in the same manner as described in the results of the study's prospective model (involving IST adherence for a consecutive three-month period beginning one month after survey completion). In both the prospective and retrospective models, more favourable attitudes and greater perceived behavioural control were associated with intentions to adhere to IST, whereas subjective norms were not significant. Also, in both models, past behaviour was correlated with perceived behavioural control and independently predicted adherence. Similarly, subsample results in both models indicated that patient intentions to adhere to IST regimens were correlated with actual adherence [39]. Given these findings, we believe that possible changes in RTPs’ adherence because of participation in the study had limited (if any) effects.

An additional limitation of this study is that our results may not be generalizable to the entire population of RTPs. However, our sample was similar to the larger US renal transplant population in terms of gender (60.1% male in our sample vs 59.9% nationally), donor type (60.1% deceased donor in our sample vs 57.8% nationally), and age group (18–34 years—12.7% in our sample vs 17.5% nationally; 35–49 years—32.9% in our sample vs 32.4% nationally; 50–64 years—41.1% in our sample vs 38.3% nationally; 65+ years—13.3% in our sample vs 11.8% nationally). African-Americans were overrepresented in our sample (38.6%) compared with the larger US population of RTPs (23.4%); however, individuals in the ‘Other’ race category were underrepresented (5.1% in our sample vs 19.2% nationally). Perhaps an important difference between our sample and the larger US renal transplant population is that all participants in our sample were enrolled in MAP, which effectively removes some of the barriers to adherence that may result from low income that were not significant in our analyses. Further research is needed to validate our findings in larger, more representative samples.

Although our results provide strong support for the TPB, there was no evidence for an effect of subjective norms. Ajzen and Fishbein [14] have noted that, due to variations in the behaviour under investigation, it should not be expected that all TPB factors (attitudes, subjective norms and perceived behavioural control) will always be significant. Indeed, several studies have found stronger support for some components of the TPB than others, with subjective norms often emerging as the weakest predictor [41]. Among RTPs it may well be that social network pressure (subjective norms) to adhere to IST is routinely high because of its critical role in preventing rejection. Our results suggest this is the case, with high mean scores and little variability on the subjective norms factor. Still, we did not assess RTPs’ perceptions of social support, and this may have influenced the lack of results for subjective norms. Including measures of social support is an important direction for future research since, (i) social support has been shown to have both direct and indirect (mediated by subjective norms) associations with medical adherence [42]; and (ii) social support is related to adherence among RTPs [37].

Non-adherence remains the single most avoidable factor in graft rejection, and our study variables explained a large percentage of the variance in adherence—even without the demographic and medical factors posited in prior research. Based on our results, we suggest that our brief, easily administered instruments measuring attitudes about taking IST, perceived behavioural control, past adherence to medical recommendations, and intentions to adhere could increase practitioners’ ability to target for early intervention those individuals who are at increased risk for not taking their IST as prescribed. Interventions then can be targeted toward individual needs. For example, patients who have a history of non-adherence and those who have negative attitudes about taking medications might benefit from educational programs to increase understanding of why adherence is critical to their well-being, facilitating more positive attitudes. Similarly, for those who feel unable to comply due to complicated medical regimens or unpleasant taste, more simplified IST regimens or better palatable formulations of medications may also improve adherence [43]. See Table 2 for additional strategies to improve adherence [44–46]. Despite its limitations, results of this study are consistent with the TPB framework and provide insight into the factors underlying medication adherence.


View this table:
[in this window]
[in a new window]

  		Table 2. Strategies to promote medication adherence (44–46)

 


   Appendix A. Theory of planned behavior (TPB) instrument for IST adherence
Top
 Abstract
 Introduction
 Subjects and methods
 Results
 Discussion
 Appendix A. Theory of...
Acknowledgements
 References
 
Directions: For each question below, please check the option that best describes your feelings.
1. Taking my immunosuppressant medication(s) as prescribed is:
___Bad
___Good
___Neither bad nor good

2. Taking my immunosuppressant medication(s) as prescribed is:
___Harmful
___Good for me
___Neither harmful nor good for me

3. Taking my immunosuppressant medication(s) as prescribed is:
___Worrisome
___Reassuring
___Neither worrisome nor reassuring

4. Taking my immunosuppressant medication(s) as prescribed is:
___Unpleasant
___Pleasant
___Neither unpleasant nor pleasant

5. Taking my immunosuppressant medication(s) as prescribed is:
___Unsatisfying
___Satisfying
___Neither unsatisfying nor satisfying

6. Taking my immunosuppressant medication(s) as prescribed is:
___Negative
___Positive
___Neither negative nor positive

7. Taking my immunosuppressant medication(s) as prescribed is:
___Like punishment
___Rewarding
___Neither punishing nor rewarding

8. Taking my immunosuppressant medication(s) as prescribed is:
___Foolish
___Wise
___Neither foolish nor wise

Directions: Please circle the number of the response that best fits your feelings about each statement.

Formula

Directions: Please check the answer that best fits your feelings or opinions about each statement.

13. How easy is it for you to take your immunosuppressant medication(s) as prescribed?
___Very difficult
___Somewhat difficult
___Neither difficult nor easy
___Somewhat easy
___Very easy

14. How much control do you have over taking your immunosuppressant medication(s) as prescribed?
___No control
___Very little control
___Some control
___Complete control

15. People who are important to me would:
___Definitely advise me to take my immunosuppressant medication(s) as prescribed.
___Definitely advise me not to take my immunosuppressant medication(s) as prescribed.
___Advise me to take my immunosuppressant medication(s) in whatever way I feel is best.
___Not advise me one way or the other.

16. If they were in my place, the people who are important to me would:
___Definitely take their immunosuppressant medication(s) as prescribed.
___Definitely not take their immunosuppressant medication(s) as prescribed.
___Take their immunosuppressant medication(s) in whatever way they feel was best.

17. My husband/wife/significant other thinks I should take my immunosuppressant medication(s) as prescribed.
___Definitely disagree
___Somewhat disagree
___I don't know how they feel about it
___Somewhat agree
___Definitely agree

18. My friends and family think I should take my immunosuppressant medication(s) as prescribed.
___Definitely disagree
___Somewhat disagree
___I don't know how they feel about it
___Somewhat agree
___Definitely agree

19. My doctors (or other healthcare professionals) would approve if I take my immunosuppressant medication(s) as prescribed.
___Definitely disagree
___Somewhat disagree
___I don't know how they feel about it
___Somewhat agree
___Definitely agree

20. I have every intention of taking my immunosuppressant medication(s) as prescribed.
___Strongly disagree
___Somewhat disagree
___Neither agree nor disagree
___Somewhat agree
___Strongly agree

21. How likely are you to take your immunosuppressant medication(s) as prescribed?
___Extremely unlikely
___Somewhat unlikely
___I’m not sure
___Somewhat likely
___Extremely likely

22. Looking back over your entire life, how often did you do what your doctors (or other healthcare professionals) recommended?
___Never
___Sometimes
___Usually
___Always

23. Over the last three months, how often did you do what your doctors (or other healthcare professionals) recommended?
___Never
___Sometimes
___Usually
___Always



   Acknowledgements
Top
 Abstract
 Introduction
 Subjects and methods
 Results
 Discussion
 Appendix A. Theory of...
 Acknowledgements
References
 
This research was partially funded by the Carlos and Marguerite Mason Trust Fund (M.A. C., principal investigator). Participation in this research was facilitated by a grant from the National Institute on Aging (AG15321, G. M. W., principal investigator; C. E. L., co-investigator) and by a fellowship from the Institute for Behavioral Research at the University of Georgia (G. M. W.). All of the authors listed had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analyses.

Conflict of interest statement. None declared.



   References
Top
 Abstract
 Introduction
 Subjects and methods
 Results
 Discussion
 Appendix A. Theory of...
 Acknowledgements
 References
 

  1. Adherence to long-term therapies: evidence for action. In: World Health Organization [WHO] (2003) Accessed 1 November 2003. Available at http://www.who.int/chronic_conditions/en/adherence_report.pdf.
  2. DeGeest S, Borgermans L, Gemoets H, et al. Incidence, determinants, and consequences of subclinical noncompliance with immunosuppressive therapy in renal transplant recipients. Transplantation (1995) 59:340–347.[Web of Science][Medline]
  3. Rovelli M, Palmeri D, Vossler E, Bartus S, Hull S, Schweizer R. Noncompliance in organ transplant recipients. Transplant Proc (1989) 21:833–834.[Web of Science][Medline]
  4. Matas AJ, Humar A, Gillingham KJ, et al. Five preventable causes of kidney graft loss in the 1990s: a single-center analysis. Kidney Int (2002) 62:704.[CrossRef][Web of Science][Medline]
  5. Butler JA, Roderick P, Mullee M, Mason JC, Peveler RC. Frequency and impact of nonadherence to immunosuppressants after renal transplantation: a systematic review. Transplantation (2004) 77:769–776.[CrossRef][Web of Science][Medline]
  6. Becker MH, Rosenstock IM. Compliance with medical advice. In: Health Care and Human behavior.—Steptoe A, Mathews A, eds. (1984) London: Academic Press.
  7. McLeroy KR, Bibeau D, Steckler A, Glanz K. An ecological perspective on health promotion programs. Health Education Quarterly (1988) 15:351–377.[Web of Science][Medline]
  8. Stokols D. Establishing and maintaining healthy environments: toward a social ecology of health promotion. American Psychologist (1992) 47:6–22.[CrossRef][Medline]
  9. Stokols D. Translating social ecological theory into guidelines for health promotion. Am J Health Promotion (1996) 10:282–298.[Web of Science][Medline]
  10. Noar SM, Zimmerman RS. Health behavior theory and cumulative knowledge regarding health behaviors: are we moving in the right direction? Health Ed Res (2005) 20:275–290.[CrossRef]
  11. Fishbein M. The role of theory in HIV prevention. AIDS Care (2000) 12:273–278.[CrossRef][Web of Science][Medline]
  12. Hausenblas HA, Carron AV, Mack DE. Application of the theories of reasoned action and planned behavior to exercise behavior: A meta-analysis. J Sport and Exercise Psychology (1997) 19:36–51.
  13. Ajzen I. Constructing a TPB Questionnaire: Conceptual and Methodological Considerations (September 2002). Accessed 5 January 2005. Available at http://www.people.umass.edu/aizen/pdf/tpb.measurement.pdf.
  14. Ajzen I, Fishbein M. Questions raised by a reasoned action approach: Comment on Ogden (2003). Health Psychol (2004) 23:431–434.[CrossRef][Web of Science][Medline]
  15. Conner M, Sparks P. The theory of planned behaviour and health behaviours. In: Predicting Health Behavior: Research and Practice with Social Cognition Models.—Conner M, Norman P, eds. (1996) Buckingham, England: Open University Press. 121–162.
  16. Godin G, Kok G. The theory of planned behavior: a review of its applications to health-related behaviors. Am J Health Promotion (1996) 80:467–468.
  17. Sheeran P, Taylor S. Predicting intentions to use condoms: a meta-analysis and comparison of the theories of reasoned action and planned behavior. J Appl Soc Psychol (1999) 29:1624–1675.[CrossRef][Web of Science]
  18. Ajzen I. Attitudes, Personality, and behavior. (1988) Milton-Keynes, England: Open University Press and Chicago, IL, Dorsey Press.
  19. Butler JA, Peveler RC, Roderick P, Horne R, Mason JC. Measuring compliance with drug regimens after renal transplantation: comparison of self-report and clinician rating with electronic monitoring. Transplantation (2004) 77:786–789.[CrossRef][Web of Science][Medline]
  20. Morris PJ, Monaco AP. Detecting medication non-compliance: electronic devices or candid patients? Transplantation (2004) 77:767.[CrossRef][Web of Science]
  21. Osterberg L, Blaschke T. Adherence to medication. N Engl J Med (2005) 353:487–497.[Free Full Text]
  22. Grymonpre R, Cheang M, Matge C, Sitar DS. Validity of a prescription claims database to estimate medication adherence in older persons. Med Care (2006) 44:471–477.[CrossRef][Web of Science][Medline]
  23. Chisholm MA, Vollenweider LJ, Mulloy LL, Jagadeesan M, Wynn JJ, Rogers HE, et al. Renal transplant patient compliance with free immunosuppressive medications. Transplantation (2000) 70:1240–1244.[Web of Science][Medline]
  24. Stewart RB, Caranasos MD. Medication compliance in the elderly. Med Clin North Am (1989) 3:1551–1563.
  25. Hilbrands LB, Hoitsma AJ, Koene RAP. Medication compliance after renal transplantation. Transplantation (1995) 60:914–920.[Web of Science][Medline]
  26. Chisholm MA, Lance CE, Mulloy LL. Patient factors associated with adherence to immunosuppressant therapy in renal transplant recipients. Am J Health Syst Pharm (2005) 62:1775–1781.[Abstract/Free Full Text]
  27. Billups SJ, Malone DC, Carter BL. The relationship between drug therapy noncompliance and patient characteristics, health-related quality of life, and health care costs. Pharmacotherapy (2000) 20:941–949.[CrossRef][Web of Science][Medline]
  28. Mallion JM, Baguet JP, Siche JP, Tremel F, de Guademaris R. Compliance, electronic monitoring and antihypertensive drugs. J Hypertens Suppl (1998) 16:75–79.
  29. Benner JS, Glynn RJ, Mogun H, Neumann PJ, Weinstein MC, Avorn J. Long-term persistence in use of statin therapy in elderly patients. JAMA (2002) 288:455–461.[Abstract/Free Full Text]
  30. Psaty BM, Koepsell TD, Wagner EH, LoGerfo JP, Inui TS. The relative risk of incident coronary heart disease associated with recently stopping the use of beta-blockers. JAMA (1990) 263:1653–1657.[Abstract/Free Full Text]
  31. Monane M, Bohn R, Gurwitz JH, Glynn RJ, Levin R, Avorn J. Compliance with antihypertensive therapy among elderly Medicaid enrollees: the role of age, gender, and race. Am J Public Health (1996) 86:1805–1808.[Abstract/Free Full Text]
  32. James LR, Mulaik SA, Brett JM. Causal Analysis: Models, Assumptions, and Data. (1982) Beverly Hills, CA: Sage.
  33. Douglas S, Blixen C, Bartucci MR. Relationship between pretransplant noncompliance and posttransplant outcomes in renal transplant recipients. J Transpl Coord (1996) 6:53–58.[Medline]
  34. Frazier P, Davis-Ali S, Dahl K. Correlates of noncompliance among renal transplant recipients. Clin Transpl (1994) 8:550–557.[Web of Science]
  35. Sketris I, Waite N, Grober K, et al. Factors affecting compliance with cyclosporine in adult renal transplant patients. Transplant Proc (1994) 26:2538–541.[Web of Science][Medline]
  36. Greenstein S, Siegal B. Compliance and noncompliance in patients with a functioning renal transplant: a multicenter study. Transplantation (1998) 66:1718–1726.[CrossRef][Web of Science][Medline]
  37. Rosenbger J, Geckova AM, van Dijk JP, et al. Prevalence and characteristics of noncompliant behaviour and its risk factors in kidney transplant recipients. Transplant Int (2005) 18:1072–1078.[CrossRef][Web of Science][Medline]
  38. Yavuz A, Tuncer M, Erdogan O, et al. Is there any effect of compliance on clinical parameters of renal transplant recipients? Transplant Proc (2004) 36:120–121.[CrossRef][Web of Science][Medline]
  39. Chisholm MA, Williamson GM, Lance CE, Mulloy LL, McGinty HE. Predicting renal transplant patients’ adherence to immunosuppressant therapy and non-adherence associated graft rejection. In: American Transplant Congress Joint Annual Meeting. (2005) May 24. Seattle, Washington.
  40. Organ Procurement and Transplantation Network. Accessed 11 October, 2006. Available at http://www.optn.org/latestData/step2.asp.
  41. Armitage CJ, Connor M. Efficacy of the theoru of planned behavior: a meta-analytic review. British Journal of Social Psychology (2001) 40:471–499.[CrossRef][Web of Science][Medline]
  42. Honda K, Kagawa-Singer M. Cognitive mediators linking social support networks to colorectal cancer screening adherence. J Behav Med (2006) 29:449–460.[CrossRef][Web of Science][Medline]
  43. Schroeder K, Fahey T, Ebrahim S. How can we improve adherence to blood pressure-lowering medication in ambulatory care? A systematic review of randomized controlled trials. Arch Intern Med (2004) 164:722–732.[Abstract/Free Full Text]
  44. Chisholm MA. Issues of adherence to immunosuppressant therapy after solid-organ transplantation. Drugs (2002) 62:567–575.[CrossRef][Web of Science][Medline]
  45. Becker MH, Maiman LA. Strategies for enhancing patient compliance. J Community Health (1980) 6:113–135.[CrossRef][Medline]
  46. Inui TS, Yourtee EL, Williamson JW. Improved outcomes in hypertension after physician tutorials: a controlled trial. Ann Intern Med (1976) 84:646–651.[Abstract/Free Full Text]
Received for publication: 25. 8.06
Accepted in revised form: 27. 2.07


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
22/8/2339    most recent
gfm149v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (3)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Chisholm, M. A.
Right arrow Articles by Mulloy, L. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Chisholm, M. A.
Right arrow Articles by Mulloy, L. L.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?