Medication adherence and graft survival among heart transplant recipients

Abstract

Abstract Background: Though medication adherence is essential for graft survival, little is known about the impact of non-adherence on heart transplant survival. Aims: The objective of this study was to examine the association between graft survival and adherence in heart transplant recipients. Methods: This retrospective, observational cohort study used claims data from a single, large national pharmacy chain (claims data from 2013-2016) and post-transplant follow-up data from the OPTN database (data from post-transplant to 2016). The sample included adult, deceased-donor heart transplant recipients (most recent if more than one) who had >2 pharmacy claims for any immunosuppressant >150 days apart in the 12-months after their first fill in the study period (2013–2016). Proportion of days covered (PDC) by any immunosuppressant for 12-months after first fill was calculated as a measure of adherence (defined as PDC >80%). Graft survival was defined as having a surviving graft at the end of the study period. Logistic regression was used to estimate the association between adherence and graft survival controlling for covariates (age at transplant, time since transplant, gender, race/ethnicity, copay, number of prescriptions for chronic conditions, pharmacy insurance plan, brand medication usage, digital fills, filling at a transplant specialized pharmacy, and receiving financial assistance). Results: Of the 3,435 heart transplant recipients who were eligible for the study, 75% were adherent and 81% had a surviving graft (range = 6–10,012 days post-transplant; median = 1,409 days). After adjusting for covariates, the odds of having a surviving graft were almost double for adherent patients than for non-adherent patients (OR = 1.94 [95% CI = 1.58–2.37]; p < 0.001). Other notable factors associated with graft survival included having three or fewer post-index prescriptions for chronic conditions (OR = 4.33 [3.55–5.27]; p < 0.001) and filling immunosuppressants digitally (OR = 2.25 [1.13–4.48]; p < 0.001). A sensitivity analysis using a PDC >90% as the definition for adherence showed that the odds of having a surviving graft were 2.01 (95% CI [1.67–2.43]) times more likely for adherent patients. Conclusions: This analysis suggests adherent patients had greater odds of having a surviving graft than those who were not adherent to immunosuppressants. Future studies should aim to show which patient behaviors contribute to medication adherence and what PDC threshold should be used for transplant research.