Methods: We used a previously-validated algorithm and national Veterans Health Administration (VHA) administrative data to identify a cohort of Veterans initiating PrEP during 2012-2015 (i.e. TDF/FTC use, no other antiretroviral use, and no diagnosis codes for HIV, hepatitis B, or needle-stick injury). We used pharmacy refill data to calculate the proportion of days covered (PDC) by TDF/FTC in the 365 days after initiation, and multivariable logistic regression to identify patient characteristics associated with high adherence (i.e. PDC > 0.80).
Results: Most (96%) of the 706 Veterans initiating PrEP were men, reflecting the overall demographics of patients in VHA. The median age was 38. Reported race in administrative data was 141 (20.0%) black, 485 (68.7%) white, 42 (5.9%) other, and 38 (5.4%) missing. Only 30 PrEP users (4.2%) lived in rural areas. Co-existing diagnoses included substance use disorder in 264 (37.4%), hypertension in 250 (35.4%), and diabetes in 95 (13.5%). The median PDC for TDF/FTC in the first year was 0.78 (IQR 0.41-0.96), and a minority (N=120, 17%) had only a single TDF/FTC fill. Predictors of high adherence were older age (OR 1.88, 95% CI 1.20-2.84 for age 50-64 compared to age < 35); white compared to black race (OR 1.96, 1.30-2.94), Male sex (OR 4.17, 1.67-12.50), absence of a substance use diagnosis (OR 1.47, 1.05-2.04), and presence of diabetes (OR 1.66, 1.02-2.75).
Conclusion: Adherence to TDF/FTC in the first year of PrEP was overall high. Racial, gender, and substance-use-related differences in PrEP adherence mirrored those previously described for PrEP uptake. Interventions to promote equitable PrEP uptake should include strategies to support adherence.
B. Beck, None
B. Alexander, None
P. Van Epps, None
M. Maier, None
B. Lund, None
M. Vaughan-Sarrazin, None