Tu1241 Does Race Play a Role in Medication Adherence in Inflammatory Bowel Disease Patients?

Abstract

Background: High quality care in inflammatory bowel disease (IBD) includes complex medical therapies that require adherence to treatment plans and frequent monitoring. Studies of IBD patients in the United States have identified race-based differences in healthcare delivery and utilization. Prior investigations of race and medication adherence in IBD have relied on survey tools and self-reporting. Our aim was to compare medication adherence between African American (AA) and Caucasian (C) IBD patients using objective pharmacy refill data. Methods: After Emory IRB approval, we retrospectively reviewed the charts of IBD patients treated in a tertiary center's subspecialty IBD clinic between 10/2013 and 12/ 2013. Medication adherence was determined using electronic pharmacy refill data (including all pharmacies used by the patient) for the 6 months following a visit to the IBD clinic. Poor adherence was defined as a medication possession ratio of <0.8 during the periods that the patient was prescribed the medication. Data was collected on rates of disease flares over this 6-month period. Statistical analysis included Mann-Whitney U for continuous variables and Chi-square testing for categorical variables, as well as a logistic regression analysis. Results: One hundred and sixty IBD patient charts were reviewed (53% F, mean age 44.1 ± 1.3). In this cohort, 23% were AA, 69.4% C, and 6.6% other races (33% UC, 65% CD, and 2% indeterminate colitis). Twenty-five percent were new patients and 75% established patients. Among AA patients, 81.1% were adherent to medications (n=30/37), as compared to 88.3% of C patients (n=98/111). This differencewas not statistically significant (p=0.3). No difference was seen in anti-TNF therapy prescriptions, where 33.3% of C patients and 43.2% of AA were prescribed biologics (p=0.2). Patients who were non-adherent to medical therapy were more likely to be younger than adherent patients (age 33 vs.46, p= 0.001). In a logistic regression model adjusting for sex, race, prior surgery, and prior hospital admission, younger age was the only significant predictor of poor adherence to IBD therapies (OR= 0.9, 95% CI 0.9-0.98, p<0.01). Non-adherent patients had more flares (0.8 vs.0.4 flares/patient in 6 months, p=0.014), and were prescribed a larger number of IBDmedications (1.6 vs.1 medication, p<0.01). Conclusions: An objective evaluation of medication refills in AA and C IBD patients seen in the same clinic revealed no significant differences in medication adherence. The only significant predictor of poor adherence was younger age. Lower medication adherence was associated with increased IBD flares, as well as being prescribed a larger number of IBD medications. While prospective validation is needed, this suggests that race-based disparities in medication adherence are decreased with similar access to care.