Development and validation of an immunosuppressant therapy adherence barrier instrument

Transcription

1 Nephrol Dial Transplant (2005) 20: doi: /ndt/gfh576 Advance Access publication 30 November 2004 Original Article Development and validation of an immunosuppressant therapy adherence barrier instrument Marie A. Chisholm 1, Charles E. Lance 2, Gail M. Williamson 3 and Laura L. Mulloy 4 1 University of Georgia College of Pharmacy and 4 Hypertension and Transplantation Medicine, Medical College of Georgia School of Medicine, 2 Applied Psychology Program and 3 Life-Span Developmental Psychology Program, University of Georgia Department of Psychology, Athens, GA, USA Abstract Background. To decrease allograft rejection as a result of non-adherence to immunosuppressant therapy (IST), a valid and reliable instrument that measures solid organ transplant patients adherence barriers is needed. Methods. An immunosuppressant therapy barrier scale (ITBS) was developed to assess transplant patients perceived barriers to IST adherence and was completed by 222 transplant patients who lived in Georgia, USA. A renal transplant population subset was used to test the ITBS reliability and validity. Scale reliability was estimated using Cronbach s alpha coefficient of internal consistency; scale dimensionality was assessed using principal components analysis. The criterion-related validity of the scale was assessed by relating subscale scores to adherence measures and graft rejection. Nomological validity was assessed by relating barrier subscales to specific patient factors. Results. Two subscales that represented controllable and uncontrollable barriers were found. Cronbach s alpha coefficients demonstrated acceptable reliabilities of 0.93, 0.86 and 0.91 for the uncontrollable and controllable subscales, and for the entire ITBS, respectively. The ITBS subscales correlated negatively with a self-reported measure of IST adherence, IST serum concentrations and IST pharmacy refill adherence rate (P<0.01). The uncontrollable barrier subscale was positively correlated to kidney graft rejection (P<0.01), thus demonstrating the ITBS s validity. Males and older patients reported more adherence barriers (P<0.05). Correspondence and offprint requests to: Marie A. Chisholm, PharmD, FCCP, UGA Clinical Pharmacy Program, th Street, CJ-1020, Medical College of Georgia, Augusta, GA 30912, USA. Mchishol@mail.mcg.edu Conclusions. The ITBS is a reliable and valid instrument that can be used to measure patients perceived barriers to IST adherence. Keywords: adherence; compliance; immunosuppressant therapy; kidney transplantation Introduction Adherence, defined as the extent to which a person s behaviour conforms to medical or health advice, to immunosuppressant therapy (IST) is critical to graft maintenance in solid organ transplant patients [1 3]. Non-adherence to IST is the most common cause of preventable kidney rejection [4]. Despite the obvious importance of IST adherence, 5 70% of transplant patients are non-adherent to their IST [2 6]. Efforts to encourage IST adherence are an important aspect of post-transplant medical care [1 7]. Failure of transplant patients to take prescribed drugs properly may not only be a significant obstacle to graft maintenance, but it may also result in dialysis, decreased productivity, decreased quality of life, and increased health care cost, morbidity and mortality. Studies describing factors that may be associated with nonadherence are available, but little has been published concerning IST adherence barriers, which logically should influence adherence. A valid and reliable instrument assessing patients IST adherence barriers is critical to identifying causes of non-adherence, and to designing and evaluating intervention strategies to enhance IST adherence in transplant patients. The objectives of this study included developing a valid and reliable instrument that measures transplant patients perceptions of IST adherence barriers, testing the hypothesis that IST adherence barriers are associated with lower adherence and rejection, Nephrol Dial Transplant Vol. 20 No. 1 ß ERA EDTA 2004; all rights reserved

2 182 M. A. Chisholm et al. and discussing methods to decrease IST adherence barriers. Methods The health belief model (HBM) and the health decision model (HDM) support that a major impediment to medication adherence is the barriers that patients encounter [8]. Items believed to assess significant barriers of IST adherence in transplant patients were developed by a focus group utilizing: (i) a portion of the HBM and the HDM as a theoretical framework; and (ii) a list of the top 10 barriers preventing transplant patients from adhering to IST that was collected from a survey of 50 solid organ transplant recipients in 2002 (30 kidney, nine liver, eight heart, two kidney pancreas and one lung). Initially, 20 items were formulated by the focus group, which consisted of a nephrologist, a transplant surgeon, a renal transplant patient, transplant nurses, pharmacists and a social worker. Each item was anchored by a 5-point Likert-type scale (1 ¼ strongly disagree to 5 ¼ strongly agree ) assessing respondent agreement to the statement. In September 2003, the 20-item scale was pre-tested in 25 renal transplant patients for item understanding, and appropriate modifications were made, including the deletion of two items that respondents found confusing. The resulting 18-item scale was then mailed to 244 solid organ transplant patients in the State of Georgia in October Renal transplants represented the most common type of transplant of those who received the 18-item scale. The investigators felt that it was important to include patients with other solid organ transplants (e.g. liver, heart and lung) in addition to kidney transplants in the study to: (i) compare the results among transplant types; and (ii) preliminarily test the instrument s utility in other transplant populations. A follow-up survey was mailed in November 2003 to those who did not complete and return the scale in October. Each respondent was also asked to complete the immunosuppressant therapy adherence scale (ITAS). The ITAS is a valid and reliable instrument that was developed to measure transplant patients IST adherence by asking respondents to indicate how often they were non-adherent to IST (defined as not taking their cyclosporin or tacrolimus IST as advised/prescribed) over a 3 month time period given a particular circumstance. A time period of 3 months was selected given that patients recall of adherence is limited and decreases over time, and 3 months is often adequate to observe patterns of adherence to chronic medications such as IST [9]. Specifically, the ITAS is a 4-item self-report measure that was developed to measure transplant recipients IST adherence by asking respondents to indicate how often they were non-adherent to IST given a particular circumstance. Items asked respondents how often they: (i) forgot to take their IST medications; (ii) were careless about taking their IST medications; (iii) stopped taking their IST medications because they felt worse; and (iv) missed taking their IST medications for any reason. Response options were A ¼ 0% of the time (none), B ¼ 1 20% of the time, C ¼ 21 50% of the time and D ¼ >50% of the time. The ITAS had positive correlations with immunosuppressant refill record adherence rates and target immunosuppressant serum concentrations (P<0.01), and the ITAS item scores were shown to be negatively related to rejection occurrence and increased serum creatinine (P<0.05). Additionally, the ITAS demonstrates acceptable reliability (Cronbach s alpha of 0.81), and has documented construct, convergent and nomological validity among solid organ transplant recipients [10]. All patients participating in the study had a functioning graft, resided in Georgia, were members of the Medication Access Program (a statewide programme whose mission is to increase medication access to transplant recipients who reside in Georgia), were taking cyclosporin or tacrolimus as their IST and were 18 years of age or older. The study was approved by the Human Assurance Committee (Institutional Review Board) at the Medical College of Georgia (MCG) and the University of Georgia. A subset of renal transplant patients was selected as the validation subgroup since renal transplants are the most common solid organ transplant performed. Sixty-nine of the 244 transplant patients who were surveyed received their post-transplant care from MCG. MCG renal transplant patients were selected as the validation group because the principal investigator is located on the MCG campus, which allowed easier access to MCG transplant patients clinic and pharmacy data. Clinical and pharmacy data collected included: (i) IST medication refill records; (ii) IST serum concentrations; and (iii) occurrence of a rejection episode (biopsy-proven rejection). Since all participating MCG renal transplant patients received their medications from the MCG pharmacy and post-transplant follow-up care from the MCG renal transplant clinic, patients refill data and immunosuppressant serum concentrations for cyclosporin and tacrolimus were used as objective measures of IST adherence. Data on recipient date of birth, gender, race/ethnicity, 2002 income (via 2002 income tax forms and social security benefit forms), transplant donor type (living or cadaveric), type of transplant (kidney, liver, heart, lung and/or pancreas) and date of transplant were also collected. Data from IST refill records and IST serum concentrations, and occurrence of a rejection for each recipient were recorded for the time period of August October Each recipient s pharmacy refill records for cyclosporin and tacrolimus were collected for the months of August October 2003 from the MCG out-patient pharmacy by using the pharmacy s prescription computer system. Each recipient s medical and monthly pharmacy refill records were used to obtain the immunosuppressant (cyclosporin and tacrolimus) regimen details and to calculate the next expected refill. The IST refill adherence rate was estimated by comparing each recipient s monthly pharmacy refill records with the prescribed regimen documented in the medical records. Refill adherence rates were calculated by using the number of days between prescription immunosuppressant medication refills. If the total number of days between refills was less than or equal to the total days supply of the immunosuppressant medication, the recipient s refill adherence rate was 100%. An assumption was made that any extra doses accumulated during the study period were used as needed by the recipient in order to adhere to the prescribed therapy if medication refills were not obtained on time. If the number of days between refills was greater than the days supply, the IST refill adherence rate was calculated by using the formula below: Adherence Rate days between refills total days supply ¼ 1 100%: days between refills

3 Immunosuppressant therapy barrier instrument 183 Serum concentrations were classified as achieving target or not achieving target using concentration ranges of and 5 17 ng/ml for cyclosporin and tacrolimus, respectively. Statistical analyses Data were entered in a spreadsheet (Excel 2000, Microsoft Õ Inc., Redmond, WA) and downloaded into SPSS 12.0 (SPSS Inc., Chicago, IL). In order to ensure that all data were within valid ranges and that each item had sufficient variance to proceed with further analyses, frequencies and descriptive statistics were computed for each item. Next, scale dimensionality was assessed by using exploratory factor analysis that included: (i) extracting factors on the basis of the principal components model; (ii) determining the number of components to retain on the basis of a combination of inspecting the scree plot, parallel analysis and meaningfulness of the solution; and (iii) rotating the factors according to the oblimin criterion for interpretation of the solution. Scale reliability was estimated using Cronbach s alpha coefficient of internal consistency. Criterion-related validity was assessed by correlating the IST adherence barrier scale scores with outcomes that were thought to be related to the perceived barriers, including ITAS scores (self-report measure of IST adherence), IST pharmacy refill adherence rates, IST serum concentrations and rejection episodes. The diagnostic sensitivity of the immunosuppressant therapy barrier scale (ITBS) subscales and total scores in relation to adherence (as indicated by refill records since they are considered more objective than self-report measures of adherence) were assessed and receiver operating characteristic (ROC) curves were created. Finally, since patient factors may influence barriers to IST adherence, the nomological validity of the scale was assessed by relating barrier subscale scores to various patient factors including demographic data (race, age, gender, marital status and income), type of IST (cyclosporin or tacrolimus) and transplantrelated data [time since transplant, type of kidney donor (living or cadaveric) and type of transplant (heart, kidney, liver, lung, pancreas or kidney pancreas)]. Results Two hundred and twenty-two transplant patients completed and returned the scale for a 91% response rate. Since the focus of the study was to test the reliability and validity of the ITBS in renal transplant patients, most of the study population (n ¼ 137; 61.7%) consisted of renal transplant patients. Sixty-six of the participants were MCG renal transplant patients. The demographics of study participants are included in Table 1. Although responses to most barrier items were positively skewed, with means clustering toward the lower end of the scale (i.e. 1 ¼ strongly disagree to 2 ¼ disagree ), responses to each item covered the entire range of the 1 (¼ strongly disagree ) to 5 (¼ strongly agree ) scale, and all item variances were sufficient to continue with further analyses. Principal components analysis (PCA) of the 18 barrier items suggested a three-factor solution. The scree plot, however, clearly pointed to a two-factor solution. For both the two- and three-factor solutions, three items failed to have significant loadings of at least 0.40 on any factor, one item exhibited significant crossloadings and one other item exhibited an unexpected negative factor loading. With these five items deleted, the PCA and scree plot of the remaining 13 items produced a two-factor solution. The oblimin rotation resulted in an estimated correlation of between factors 1 and 2. The oblimin-rotated factor pattern loadings and descriptive statistics for the 13 barrier items that were eventually retained for the ITBS are shown in Table 2. In consultation with a focus group of four practising transplant health care professionals (a nephrologist, a transplant surgeon, a transplant clinical pharmacist and a transplant nurse), a kidney transplant recipient and three clinical psychologists, factor 1 was interpreted as including barriers that patients may perceive to be beyond their control, such as having Table 1. Characteristics of the study population Total study participants (n ¼ 222) MCG RTRs a (n ¼ 66) Age (years)±sd 52.2± ±14.3 Gender Male 149 (67.1%) 50 (75.8%) Female 73 (32.9%) 16 (24.2%) Race African-American 84 (37.8%) 32 (48.5%) Caucasian 133 (59.9%) 32 (48.5%) Asian 2 (0.9%) 1 (1.5%) Hispanic 3 (1.4%) 1 (1.5%) Transplant donor type Deceased (cadaveric) N/A b 44 (66.7%) Living N/A b 22 (33.3%) Marital status Married 93 (41.9%) 44 (66.7%) Single 82 (36.9%) 14 (21.2%) Divorced 36 (16.3%) 5 (7.6%) Widowed 11 (5.0%) 3 (4.5%) Income (US$)±SD ± ±7099 Time since transplant (years)±sd 6.52± ±5.56 Type of base immunosuppressant Cyclosporin N/A b 42 (63.6%) Tacrolimus N/A b 24 (36.4%) Total no. of drugs c ±SD N/A b 11.33±4.03 No. of immunosuppressants±sd 2.44±0.66 ITAS score d ±SD 11.04± ±1.52 Refill records adherence rate±sd N/A b 84.45±10.37 Target immunosuppressant levels Yes N/A b 46 (69.7%) No N/A b 20 (30.3%) Rejection Yes N/A b 6 (9.1%) No N/A b 60 (90.9%) a MCG RTRs ¼ Medical College of Georgia renal transplant recipients. b Data available for MCG RTRs only. c Total number ¼ total number of prescribed medications (drugs). d ITAS score ¼ ITAS scores range from 0 to 12; 0 is the lowest obtainable score and 12 is the highest.

4 184 M. A. Chisholm et al. Table 2. Descriptive statistics and oblimin-rotated factor loadings for immunosuppressant barrier items Item Descriptive statistics Factor loadings Mean SD Skew Factor 1 uncontrollable Factor 2 controllable 1. I have to take the immunosuppressant medication(s) too many times per day. 2. I have to take too many capsules (or tablets) of my immunosuppressant medication(s) at one time. 3. I cannot tell if my immunosuppressant medication(s) is (are) helping me I skip doses of my immunosuppressant medication(s) when I go out of town. 5. I miss doses of my immunosuppressant medication(s) when I feel depressed. 6. I get confused about how to take my immunosuppressant medication I do not understand when to take my immunosuppressant medication(s) I often run out (or do not have enough) of immunosuppressant medication(s). 9. It is hard for me to remember to take my immunosuppressant medication(s). 10. I miss a dose of my immunosuppressant medication(s) when I think there may be side effects. 11. I sometimes skip doses of my immunosuppressant medication(s) when I feel good (or better). 12. I miss doses of my immunosuppressant medication(s) when I get out of my daily routine. 13. I skip doses of my immunosuppressant medication(s) when I am short of money Scale: 1 ¼ strongly disagree ; 2 ¼ disagree ; 3 ¼ neutral ; 4 ¼ agree ; 5 ¼ strongly agree. Table 3. ITBS s Cronbach s alpha s and mean scale scores Total (n ¼ 222) MCG RTRs a (n ¼ 66) RTRs b (n ¼ 148) Liver transplant recipients (n ¼ 35) Heart transplant recipients (n ¼ 34) Lung transplant recipients (n ¼ 5) Cronbach s alpha coefficient for uncontrollable barriers Cronbach s alpha coefficient for controllable barriers Cronbach s alpha coefficient for entire ITBS Mean scores for uncontrollable barriers c 15.09± ± ± ± ± ±3.03 Mean scores for controllable barriers d 7.94± ± ±4 7.26±3.33 7± ±2.17 Mean scores for entire ITBS e 17.73± ± ± ± ± ±3.63 a MCG RTRs ¼ Medical College of Georgia renal transplant recipients. b RTRs ¼ renal transplant recipients. c 8 represents the lowest obtainable score and 40 represents the highest. d 5 represents the lowest obtainable score and 25 represents the highest. e 13 represents the lowest obtainable score and 65 represents the highest. to take too many medications, depression, confusion and lack of understanding regarding when to take IST. On the other hand, items having significant loadings on factor 2 referred to barriers that should be more within patients control, with non-adherence occurring because of carelessness or deliberate intention not to adhere (e.g. not remembering to take IST, or choosing not to take IST because the patient feels that there might be side effects, feels better or is short of money). These two were later labelled as uncontrollable barriers and controllable barriers, respectively, for factor 1 and factor 2. Subscale scores represent greater IST adherence barriers of each type, with 8 and 40 representing the lowest and highest scores for the uncontrollable barriers, 5 and 25 representing the lowest and highest scores for the controllable barriers, and 13 and 65 representing the lowest and highest scores for the entire 13-item ITBS. Using the total population of 222 recipients, the Cronbach s alpha coefficient was 0.93 for the items comprising the uncontrollable barriers, 0.86 for the controllable barriers and 0.91 for the 13-items of the ITBS. Mean scores for the 222 patients for factors 1 and 2, and the total ITBS were 15.09±8.55, 7.94±4.29 and 17.73± 8.24, respectively. See Table 3 for Cronbach s alpha coefficient and mean scale scores. Criterion-related validity coefficients for the uncontrollable barriers and controllable barriers factors are shown in Table 4. As predicted, the two ITBS subscales and the ITBS total scores correlated

5 Immunosuppressant therapy barrier instrument 185 Table 4. Criterion-related validity coefficients for ITBS Uncontrollable barriers Controllable barriers ITBS total score Adherence measures ITAS adherence score a 0.629** 0.334** 0.637** IS serum concentrations b 0.401** 0.491** 05** IS refill rate a 0.544** 0.589** 0.644** Clinical outcome Rejection episode b 0.418** ** ITAS ¼ immunosuppressant therapy adherence scale; IS ¼ immunosuppressant. **P<0.01. a Pearson product moment correlations. b Point biserial correlations [20]. Sensitivity 1 - Specificity ROC Curve for "Uncontrollable" Barriers Area under the curve is 0.80 Fig. 2. ROC curve for the uncontrollable barriers and refill record adherence measure. Sensitivity 1 - Specificity Area under the curve is 0.81 Fig. 1. ROC curve for the ITBS and refill record adherence measure. negatively with all three adherence measures, ITAS scores, IST serum concentrations and IST refill adherence rate (P<0.01). Figure 1 shows the ROC curve for the ITBS total score in relation to IST refill adherence (defined as 80% adherence rate ¼ adherent and <80% adherence rate ¼ non-adherent) [11]. Areas under the ROC curves were 0.80 for uncontrollable barriers (see Figure 2), 0.78 for controllable barriers (see Figure 3) and 0.81 for the ITBS total scale score. The areas under the curves produced by the ITBS and its subscales were interpreted as having good discriminability, as an area of 0.80 was selected a priori as indicating good discriminability [12,13]. Also, patient reports of barriers correlated positively with rejection episode (a clinical outcome that has been associated with IST non-adherence) [4]. Although the correlation between controllable barriers and rejection episode was non-significant with P ¼ 0.36, the correlation between uncontrollable barriers and rejection was Sensitivity 1 - Specificity Area under the curve is 0.78 Fig. 3. ROC curve for the controllable barriers and refill record adherence measure. statistically significant (P<0.01). The correlation between the entire ITBS and rejection was also statistically significant (P<0.01). Overall, the results (shown in Table 4) support the criterion-related validity of the subscales of the ITBS. The nomological validity of the ITBS subscales in terms of their relationship to various patient factors was also assessed. There were no significant differences in patient reports of IST adherence barriers by race,

6 186 M. A. Chisholm et al. income, time since receiving the transplanted organ, kidney donor type (living vs cadaveric) or the type of organ transplanted. However, male patients reported significantly more barriers, both uncontrollable and controllable, than did female patients (P<0.05). Also, older patients reported more uncontrollable barriers than did younger patients (P<0.05), and patients taking tacrolimus reported significantly more controllable barriers than patients taking cyclosporin (P<0.05). Discussion There is a great need to be able to identify patients who are non-adherent as well as determinants of non-adherence prior to patients experiencing adverse effects, such as graft loss [1,3,5]. Patients who have greater barriers to adhering to medical advice are often non-adherent [9]. Thus, a valid and reliable scale to measure transplant patients barriers to IST adherence is needed to help identify those patients who are at risk for being non-adherent and to design effective adherence intervention programmes. To be useful in practical settings, the scale must be reliable and valid based on sound measures of IST adherence (pharmacy refill records, IST serum concentration and self-reports) and clinical outcomes (e.g. kidney graft rejection). Furthermore, the scale should be relatively short, containing <20 items, so it can be completed within 10 min during a clinic or office visit. The investigators of this study developed the ITBS using these criteria and produced a reliable and valid 13-item scale. Although the ITBS includes some underlying principles of self-efficacy, as many other scales, it is unique since it was designed specifically to measure IST adherence barriers in transplant recipients. Two scales that have been used in transplant to measure self-efficacy (although, not specifically designed for IST) include the medication behaviour self-efficacy scale (MBSS) and the appraisal of self-care agency scale (ASA scale) [1,14]. The ITBS shares some similarities (e.g. ascertaining medications expense and adverse effects) and dissimilarities (e.g. the ITBS is more succinct and uses a 5-point rating scale instead of a 3-point rating scale) with the MBSS [14]. Similar to the ITBS, the ASA uses a 5-point Likert-type scale to rate items and has a Cronbach s alpha coefficient of at least The ASA scale, like the MBSS, is a longer instrument than the ITBS [1]. The ITBS consists of two subscales (factors). The uncontrollable barriers, which contain eight items, are described as those being beyond the patient s control (although many are within the control of health care practitioners, therefore strategies to reduce these barriers may increase patients IST adherence). For example, many patients may be confused as to how to take their IST, not intentionally or as a result of their own doing, but because of the regimen s complexity (e.g. multiple doses needed throughout the day) or because they were not told how to take the medication(s), thus facilitating non-adherence. The controllable barriers consist of items that the patient has more control over, and non-adherence is more likely to be due to carelessness or deliberate intention not to adhere, such as patients missing doses when they change their daily routine. These two subscales, uncontrollable barriers and controllable barriers, have a similar underlying foundation to the theory supporting the health locus of control (HLC). The HLC, a social learning theory, states that individuals develop a belief that certain outcomes are a result of their actions (internals) or a result of other forces independent of themselves (externals) [15,16]. However, items of the HLC are not specific enough to draw direct conclusions concerning IST adherence, whereas the ITBS is specific for IST adherence. Nonetheless, both the HLC and the ITBS support the concept of implementing interventions that facilitate increasing patients internal locus of control. This could be beneficial since it may empower patients to take responsibility and change their IST behaviour (empowering them to be more adherent). For example, educating patients as to how to properly take their IST, providing patients with devices to remind them to take their IST (such as watches with alarms indicating it is time to take their IST) and improving patients access to IST may improve adherence [5,7 9]. The results of the study support our hypothesis that those patients with greater barriers to IST adherence are less adherent to therapy, and that greater IST adherence barriers are associated with graft rejection. It is important to note that the correlations between patient-reported IST adherence barriers and adherence (measured by the ITAS scores and refill record adherence rates) tended to be higher than the correlations between patient-reported IST adherence barriers and graft rejection (see Table 4). From a practical perspective as experienced health care professionals and a conceptual perspective, this is to be expected, since barriers lower the probability of adherence to IST, and decreased IST adherence increases the likelihood of graft rejection. Due to advances in IST to prevent rejection and the short time frame of 3 months to monitor for rejection, only six individuals with graft rejections were found, thus limited study conclusions concerning graft rejections can be made. Nonetheless, since controllable barriers were not found to be significantly associated with graft rejection, response bias could be present. Perhaps patients may have rejected because of something that they could control (a controllable barrier ) but were ashamed to admit this since it is not socially well accepted. We know from the adherence literature that one of the biggest criticisms of using patient self-report is that it is dependent on the honesty of patients, and patients may be dishonest for many reasons (including dishonesty due to the truth potentially leading to interference with being socially accepted, as the likelihood of being reprimanded for not adhering to IST is high if a rejection was due to IST non-adherence) [9]. As patient self-reports tend to

7 Immunosuppressant therapy barrier instrument 187 overestimate medication adherence [9], patient dishonesty could possibly interfere with the validity of the ITBS (in particular the controllable barriers scale). Older patients reported more uncontrollable barriers than younger patients, and this was true with respect to each of the eight items of the uncontrollable barriers subscale. Generally, older patients tend to have more barriers. For example, there is a higher incidence of cognitive impairment, diseases, medications needed and depression with increasing age [9,17]. Although studies have indicated that females are more at risk for being non-adherent than males [18], studies have also indicated that male transplant patients are less adherent to IST than females, suggesting that men may have more IST adherence barriers [19]. Patients taking tacrolimus reported significantly more controllable barriers than patients taking cyclosporin (P ¼ 0.01). This was due mainly to tacrolimus users more often reporting that they find it hard to remember to take their IST, that they miss taking their IST when they change daily routines and that they skip their IST when they are short of money. These findings may have resulted because: (i) tacrolimus costs more than cyclosporin since no generic products are available for tacrolimus; and (ii) it is more difficult for patients to qualify to receive free medication from the tacrolimus assistance programme available through the pharmaceutical company than from the cyclosporin assistance programme. Thus, the higher cost of tacrolimus to the patient may be a significant barrier to adherence. Since kidney transplants are performed more frequently than any other solid organ transplant, renal transplant patients clinical and pharmacy data were used to validate the ITBS. Although the clinical and pharmacy data for only 66 renal transplant patients were used to confirm the ITBS s validity, the ITBS was found to have good validity in these patients. However, to make the claim of validity in solid organ transplant populations other than kidney transplant, further research using other populations and clinical data pertinent to that condition (transplant) should be part of the validation process (e.g. using liver function tests in liver transplant patients). Since solid organ transplant patients have many similar IST adherence barrier issues, and no statistical differences in ITBS scores were found between transplant types, the chance of the ITBS having validity (in terms of the association of the ITBS with clinical and pharmacy data) in solid organ transplant patients other than kidney transplant is promising. Furthermore, results of the PCA analysis and the respectable internal reliability found in this study when using the entire 222 solid organ transplant patients (representing kidney, liver, lung and heart transplants) add further validation to the ITBS for use in all solid organ transplants. Given the small number of transplant recipients used to confirm the validity of the ITBS and the promising study outcomes, this study serves as a springboard for future research using the ITBS. Conclusion To decrease graft rejection as a result of IST nonadherence, it is important to identify those patients who have IST adherence barriers. This study is the first of its kind to develop a reliable and valid scale to assess transplant patients adherence barriers. The ITBS, a 13-item scale with two subscales that measure uncontrollable barriers and controllable barriers, was developed and validated. The ITBS subscales correlated negatively with all three adherence measures (ITAS scores, IST serum concentrations and IST pharmacy refill adherence rate). Also, patient reports of barriers correlated positively with graft rejection, thus confirming its clinical validity as a measure of IST adherence barriers. Acknowledgements. This research was partially funded by the Carlos and Marguerite Mason Trust Fund. Conflict of interest statement. None declared. References 1. 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: Didlake RH, Dreyfus K, Kerman RH, Van Buren CT, Kahan BD. Patient noncompliance: a major cause of late graft failure in cyclosporine-treated renal transplants. Transplant Proc 1988; 20: Hilbrands LB, Hoitsma AJ, Koene RA. Medication compliance after renal transplantation. Transplantation 1995; 60: Gaston R, Hudson S, War M, Jones R, Macon R. Late renal allograft loss: noncompliance masquerading as chronic rejection. Transplant Proc 1999; 31: 21s 23s 5. Chisholm MA, Mulloy LL, Jagadeesan M, DiPiro JT. Impact of clinical pharmacy services on renal transplant patients compliance with immunosuppressive medications. Clin Transplant 2001; 15: 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: Chisholm MA. Issues of adherence to immunosuppressant therapy after solid-organ transplantation. Drugs 2002; 62: Eraker SA, Kirscht JP, Becker MH. Understanding and improving patient compliance. Ann Intern Med 1984; 100: Stewart RB, Caranasos MD. Medication compliance in the elderly. Med Clin North Am 1989; 3: Chisholm MA, Lance CE, Williamson GM, Mulloy LL. Development and validation of the immunosuppressant therapy adherence instrument (ITAS). Patient Educ Counsel 2004, in press 11. Chisholm MA, Vollenweider LJ, Mulloy LL et al. Renal transplant patient compliance with free immunosuppressive medications. Transplantation 2000; 70:

8 188 M. A. Chisholm et al. 12. Egan JP. Signal Detection Theory and ROC Analysis. Academic Press, New York, NY; Wickens TD. Elementary Signal Detection Theory. Oxford University Press, New York, NY; Baines LS, Joseph JT, Jindal RM. Compliance and late acute rejection after kidney transplantation: a psychomedical perspective. Clin Transplant 2002; 16: Lau RR. Origins of health locus of control beliefs. J Pers Soc Psychcol 1982; 42: Wallston BS, Wallston KA. Locus of control and health: a review of the literature. Health Educ Monogr 1978; Spring: Park HL, O Connell JE, Thomson RG. A systematic review of cognitive decline in the general elderly population. Int J Geriatr Psychiatry 2003; 18: Jindel RM, Joseph JT, Morris MC, Santella RN, Baines LS. Noncompliance after kidney transplantation: a systematic review. Transplant Proc 2003; 35: Siegal BR, Greenstein SM. Postrenal transplant compliance from the perspective of African-Americans, Hispanic-Americans, Anglo-Americans. Adv Ren Replace Ther 1997; 4: Nunnally J, Bernstein IH. Psychometric Theory, 3rd edn. McGraw-Hill, New York, NY; 1994 Received for publication: Accepted in revised form: