Patient factors associated with adherence to immunosuppressant ...

REPORTS

Immunosuppressant therapy

REPORTS

Patient factors associated with adherence to immunosuppressant therapy in renal transplant recipients MARIE A. CHISHOLM, CHARLES E. LANCE, AND LAURA L. MULLOY

M

any solid-organ-transplant recipients do not take their immunosuppressant therapy (IST) as prescribed. Nonadherence to IST after transplantation is one of the leading causes of allograft rejection, graft loss, and death.1,2 Despite the devastating consequences of IST nonadherence, which include decreased quality of life, increased health care costs, need for dialysis, morbidity, and mortality, reported nonadherence rates range from 5% to 68%.3,4 Therefore, adherence, defined as the extent to which a person’s behavior conforms to medical or health advice,5 is a critical issue in transplant medicine.6 As more potent and effective immunosuppressants become available to decrease acute-rejection episodes, immunosuppressant adherence emerges as an important factor determining outcomes of solid-organ transplantation.7 Medication compliance (adherence) is measured by a variety of methods, with no one method of adherence being superior in all aspects

Purpose. Factors associated with adherence to immunosuppressant therapy (IST) in renal transplant recipients were studied. Methods. The Immunosuppressant Therapy Adherence Scale (ITAS) was completed by adult renal transplant recipients in Georgia. Those completing the ITAS were classified as adherent to IST if their ITAS score were 12 and nonadherent if their score was less than 12. The relationship between the dichotomized ITAS scores and patient variables that are readily available to clinicians, such as sex, age, kidney donor type, income, marital status, race or ethnicity, and time since transplantation, was assessed. The relationship of ITAS scores to patients’ clinical and pharmacy data (e.g., graft rejection, serum IST concentrations, serum creatinine [SCr] concentrations, and pharmacy refillbased adherence rates) was also assessed. Results. One hundred thirty-seven patients completed the ITAS. Eighty-nine patients (65%) were adherent to IST, and the

to another.8-10 Although direct questioning of patients (self-reports) as a method of measuring adherence is desirable, since it is not invasive or expensive, self-reports depend on patient accuracy, and most scales

MARIE A. CHISHOLM, PHARM.D., FCCP, is Associate Professor of Pharmacy, College of Pharmacy, University of Georgia (UG), and Associate Clinical Professor of Medicine, School of Medicine, Medical College of Georgia (MCG), Augusta. CHARLES E. LANCE, PH.D., is Professor of Psychology, Applied Psychology Program, Department of Psychology, UG. LAURA L. MULLOY, D.O., FACP, is Professor of Medicine and Chief, Section of Nephrology, Hypertension and Transplantation Medicine, School of Medicine, MCG.

remaining 48 (35%) were nonadherent. Patient sex was unrelated to adherence. Compared with nonadherent patients, adherent patients tended to be younger, to take cyclosporine, to have lower incomes, to have received their transplant more recently, to have targeted immunosuppressant concentrations, to have greater refill-based adherence rates, and to be less likely to exhibit an increase in SCr concentration (p < 0.05). There was no significant difference in the number of rejections between adherent and nonadherent patients. Conclusion. Patient age, income, time since transplantation, and the immunosuppressant agent prescribed were associated with IST adherence. Index terms: Age; Blood levels; Compliance; Cyclosporine; Graft rejection; Immunosuppressive agents; Patients; Sex; Sociology; Transplantation Am J Health-Syst Pharm. 2005; 62:1775-81

used to measure adherence have not been tested for reliability and validity.8 Nonetheless, many studies have used patient self-reports to measure IST adherence and to relate IST adherence to patient-specific factors.

Address correspondence to Dr. Chisholm at the UGA Clinical Pharmacy Program, 1120 15th Street, CJ-1020, Medical College of Georgia, Augusta, GA 30912 ([email protected]). Supported by the Carlos and Marguerite Mason Trust Fund. Copyright © 2005, American Society of Health-System Pharmacists, Inc. All right reserved. 1079-2082/05/0901-1775$06.00. DOI 10.2146/ajhp040541

Am J Health-Syst Pharm—Vol 62 Sep 1, 2005

1775

REPORTS


Studies have found IST nonadherence to be more common in African Americans, Hispanics, younger adults, single adults, and those who have had their transplants for a longer time than in Caucasians, older adults, married adults, and recent transplantees.2,3,11-15 Patients with low social economic status have also been found to be more nonadherent than those in higher socioeconomic groups,11,16,17 although those with college educations were found to be less adherent than those without postsecondary education.15 In these studies, the method of defining and confirming recipients’ socioeconomic status was not detailed.11,16,17 To target IST adherence interventions at those who are at highest risk for IST nonadherence, it is beneficial to determine a profile of those who are more likely to be nonadherent. In March 2004, the journal Transplantation featured several articles highlighting the significance of IST adherence.6,7,10,18-20 Our study serves as a supplement to these articles, since it examined patient data (which are easily accessible to practitioners) and the association of the data (patientspecific factors) with IST adherence. Our purpose was to determine the prevalence of IST nonadherence in a statewide sample of Georgia renal transplant recipients (referred to as patients) and to assess relationships between self-reported IST adherence and patient sex, age, kidney donor type (deceased [cadaveric] or living), income, marital status, race or ethnicity, time since transplantation, rejection episode, serum creatinine (SCr) concentrations, adherence rate according to immunosuppressant refill records, and serum immunosuppressant concentrations. Methods In fall 2003, the Immunosuppressant Therapy Adherence Scale (ITAS) was mailed to 146 renal transplant patients in Georgia. The ITAS is a four-item, valid, reliable

1776

self-report measure that was developed to assess transplant recipients’ IST adherence by asking them to indicate how often they were nonadherent to IST, given a particular circumstance.21 Items ask respondents how often they (1) forgot to take their IST medications, (2) were careless about taking their IST medications, (3) stopped taking their IST medications because they felt worse, and (4) missed taking their IST medications for any reason. Response options are 0% of the time, 1–20%, 21– 50%, and greater than 50%. The ITAS has been found to have positive correlations with immunosuppressant-refill-based adherence rates and targeted serum immunosuppressant concentrations (p < 0.01). ITAS item scores were negatively related to rejection occurrence and increased SCr level (p < 0.05). The ITAS has acceptable reliability (Cronbach’s α of 0.81) and has construct, convergence, and nomologic validity among solidorgan-transplant recipients.21 To be included in the study, renal transplant recipients had to be members of the Medication Access Program (a statewide program whose mission is to increase medication access among transplant recipients who reside in Georgia), to have a functioning graft, to be taking cyclosporine or tacrolimus, and to be at least 18 years of age. Sixty-eight of the 146 patients surveyed were from the Medical College of Georgia (MCG); the remaining 78 received their posttransplant care from other transplant centers in Georgia. In addition to the data collected for the entire sample, data collected from the MCG subgroup included immunosuppressant refill records, blood (serum) immunosuppressant concentrations, and occurrence of a rejection episode (biopsy proven). Since all participating patients received their medications from the MCG pharmacy and received posttransplant care from the MCG renal transplant clinic, their refill data and


serum cyclosporine and tacrolimus levels were used as objective measures of IST adherence. Cyclosporine and tacrolimus were the only immunosuppressant agents examined, because (1) they are commonly used (most renal transplant recipients are prescribed either cyclosporine or tacrolimus as their base IST), and (2) blood concentrations of these agents are routinely measured in practice, thereby providing data to support adherence. Furthermore, cyclosporine and tacrolimus are not used in the same regimen, since they are both calcineurin inhibitors and inhibit the calcineurin pathway. Each patient’s date of birth, sex, race or ethnicity, 2002 income (all patients enrolled in the Medication Access Program during 2003 provided their 2002 income tax forms and Social Security benefit forms), transplant donor type (living or cadaveric), and date of transplantation were also collected. Immunosuppressant refill records, serum immunosuppressant concentrations, SCr concentrations, and occurrence of rejection for each recipient were collected for the period from mid-August through mid-November 2003. A period of three months was selected, because patients’ recall of adherence is limited and decreases over time, statements on the ITAS refer to a three-month period, and three months is often adequate for observing patterns of adherence to longterm medications.8 The study was approved by the human assurance committee (institutional review board) at MCG. Refill records. Each patient’s MCG pharmacy refill records for cyclosporine and tacrolimus were collected between mid-August and mid-November 2003 by using the pharmacy’s prescription computer system. Medical and monthly refill records were used to obtain the details of the cyclosporine or tacrolimus regimen and to calculate the next expected refill. Adherence was

REPORTS

estimated by comparing patients’ monthly refill records with the prescribed regimen documented in the medical records. Adherence rates were calculated from the number of days between refills. If the total number of days between refills was less than or equal to the total number of days of supply of the immunosuppressant, the adherence rate was 100%. If the number of days between refills was greater than the number of days of supply, the adherence rate was calculated as follows22: {1 – [(Days between refills – total days’ supply)/days between refills]} × 100%. We assumed that any extra doses accumulated during the study period were used as needed by the recipient to adhere to the prescribed therapy if refills were not obtained on time. For example, if the total number of days between cyclosporine refills was three days more than the number of days of supply for month 3, and the same patient refilled his or her month 2 supply five days early, we assumed that the patient used the extra doses obtained in month 2 to cover the three days of supply needed in month 3. Since refill-based adherence rates of 80% or greater have been found to be well correlated with other markers of IST adherence, such as serum IST concentrations,23 this percentage was used to develop an adherent-versusnonadherent diagnostic index for immunosuppressant refill records. If the refill-based adherence rate was less than 80%, the patient was characterized as nonadherent, and if the rate was 80% or greater, the patient was considered adherent.23 Serum cyclosporine and tacrolimus concentrations measured between August 2003 through November 2003 were recorded. Serum concentrations were classified as achieving or not achieving targeted minimum concentrations. Targeted blood concentrations of at least 100 ng/mL and at least 5 ng/mL were used for cyclosporine and tacrolimus, respective-

ly.24 The minimum targeted concentrations were used as a conservative method of assessing adherence and to allow for varying posttransplant time (lower concentrations are generally more acceptable as time since transplant surgery passes and among patients who are at lower risk for rejection, whereas higher concentrations are associated with more recent transplant surgery). Statistical analysis. Data were entered into a spreadsheet (Excel 2000, Microsoft Inc., Redmond, WA) and downloaded into SPSS, version 12.0 (SPSS Inc., Chicago, IL). Each ITAS item was coded 3 (0% of the time), 2 (1–20%), 1 (21–50%), and 0 (greater than 50%). A higher composite ITAS score would indicate greater adherence. Thus, the four-item ITAS composite score ranged from a low of 0, indicating very poor adherence, to a high of 12, indicating perfect adherence. Since there is abundant literature indicating that self-report measures of adherence tend to overestimate adherence,25-27 patients were dichotomized as being adherent by having a perfect adherence score (score of 12) on the ITAS to correct for response bias. Dichotomized ITAS scores of 1, representing adherent (composite ITAS score of 12), and 0, representing nonadherent (composite score of less than 12), were coded in the data set for each patient. Dichotomized refill-based adherence scores of 1, representing adherent (refill record greater than 80%), and 0, representing nonadherent (refill record less than 80%), were coded in the data set for each patient. The relationships between the dichotomized ITAS scores and patient factors were assessed by using chisquare and Student’s t tests where appropriate. The a priori level of significance was 0.05. Results One hundred thirty-seven patients completed and returned the ITAS, for a 94% response rate. The


study population had a mean ± S.D. age of 52.52 ± 14.02 years and a mean ± S.D. annual income of $16,691 ± $10,747 (Table 1). Eighty-nine (65%) of the patients had a composite ITAS score of 12 and therefore were considered adherent; the remaining 48 patients (35%) had a composite score of less than 12 and were considered nonadherent. Data for the MCG subsample (n = 65, response rate of 96%) indicated that 50 patients (77%) were adherent by composite ITAS scores and that 15 (23%) were nonadherent. Data for the MCG subsample also indicated that 41 patients (63%) were adherent by refill records and that 24 (37%) were nonadherent. Table 2 summarizes patients’ responses to the ITAS questions. Women were no more likely to be adherent than men (p = 0.237) (Table 3). Compared with nonadherent patients, adherent patients were significantly younger, more often took cyclosporine than tacrolimus, had lower average incomes, were more often within targeted range of immunosuppressant concentrations, had greater refill-record-based adherence rates, and less often exhibited a rise in SCr concentration (to greater than 0.3 mg/dL) (p < 0.01). Adherent patients had received their transplants more recently (p = 0.017). Patients who experienced a graft-rejection episode had a mean ± S.D. ITAS score of 10.00 ± 2.45. There was a nonsignificant tendency (p = 0.100) for nonadherent patients to more often experience a rejection episode. There were no significant differences between the groups in donor type or race. Discussion The percentages of patients who were classified as adherent to IST by ITAS composite scores and refill records were 77% (n = 50) and 23% (n = 15), respectively, and the percentages classified as nonadherent were 63% (n = 41) and 37% (n = 24),


1777

REPORTS


respectively. Thus, 23–37% of patients were nonadherent, depending on which adherence measure was used. These data support two important points. First, a high rate of IST nonadherence existed among renal transplant recipients, with as many as one patient out of three being nonadherent. Second, although dif-

ferent measures of adherence may yield different adherence scores, adherence behaviors assessed by valid adherence measures should be associated with each other and with clinical outcomes (Table 3). Although a recent study suggests that measures of adherence in current clinical use do not perform well when tested

Table 1.

Demographics of Sample

Characteristic Sex, no. (%) Male Female Race, no. (%) Caucasian African American Hispanic Other Mean ± S.D. age

All Patients (n = 137)

Non-MCGa Patients (n = 72)

MCG Patients (n = 65)

88 (64) 49 (36)

38 (53) 34 (47)

50 (77) 15 (23)

68 (50) 64 (47) 3 (2) 2 (1) 52.52 ± 14.02

36 (50) 33 (46) 2 (3) 1 (1) 50.17 ± 13.56

32 (49) 31 (48) 1 (2) 1 (2) 55.12 ± 14.17

aMCG = Medical College of Georgia.

against other adherence measures (e.g., electronic monitoring),6 our study indicates that patients’ refillbased adherence rates, serum immunosuppressant concentrations, and SCr concentrations are associated with each other. Specifically, lower ITAS scores were associated with less IST adherence, as indicated by lower refill-based adherence rates, not achieving targeted serum immunosuppressant concentrations, increased SCr concentrations, and graft rejections.21 Our study is unique in its use of three measures of adherence; the association of all three measures with one another supports their validity. The association of adherence with clinical outcomes is of vital importance in transplant medicine. Rovelli and colleagues16,17 conducted retrospective and prospective studies that assessed nonadherence in organ transplant recipients. Although the mean time since transplant surgery

Table 2.

Responses on Immunosuppressant Therapy Adherence Scale No. (%) Patients Giving Response

Question and Response

All Patients (n = 137)

Non-MCGa Patients (n = 72)

MCG Patients (n = 65)

MCG Patients with Rejection (n = 6)

98 (72) 37 (27) 2 (1) 0

49 (68) 22 (31) 1 (1) 0

49 (75) 15 (23) 1 (2) 0

3 (50) 2 (33) 1 (17) 0

107 (78) 29 (21) 0 1 (1)

55 (76) 16 (22) 0 1 (1)

52 (80) 13 (20) 0 0

4 (67) 2 (33) 0 0

127 (93) 10 (7) 0 0

68 (94) 4 (6) 0 0

59 (91) 6 (9) 0 0

4 (67) 2 (33) 0 0

89 (65) 43 (31) 4 (3) 1 (1)

40 (56) 29 (40) 2 (3) 1 (1)

49 (75) 14 (22) 2 (3) 0

3 (50) 2 (33) 1 (17) 0

In the last 3 months, how often did you forget to take your immunosuppressant medications? 0% of the time 1–20% 21–50% >50% In the last 3 months, how often were you careless about taking your immunosuppressant medications? 0% of the time 1–20% 21–50% >50% In the last 3 months, how often did you stop taking your immunosuppressant medications because you felt worse? 0% of the time 1–20% 21–50% >50% In the last 3 months, how often did you miss taking your immunosuppressant medications for any reason? 0% of the time 1–20% 21–50% >50% aMCG = Medical College of Georgia.

1778


REPORTS

Table 3.

Association of Patient Characteristics with Adherence Variable

Adherent Patients (ITASa Score, 12)

Nonadherent Patients (ITAS Score, 2 yr Refill records, no. (%)e,f Adherent Nonadherent Rejection episode, no. (%)e With rejection Without rejection Increase in serum creatinine conc. no. (%)e With increase Without increase

p 0.237c

54 (61) 35 (40) 49.67 ± 11.94 26–74

34 (71) 14 (29) 57.79 ± 16.07 21–84

16 (32) 34 (68)

5 (33) 10 (67)

38 (76) 12 (24)

4 (26) 11 (73)

14,387 ± 7,644

21,079 ± 14,081

0.001d 0.923c