Study Drug and Visit Adherence: Data from the BABY HUG Trial.

BABY HUG (NCT00006400) is an NHLBI-NICHD sponsored Phase III multi-center, randomized, double blind placebo controlled study of daily oral hydroxyurea in infants 9–17 months of age with sickle cell anemia. Subject retention and adherence are essential to maintain the power and validity of the trial. Initially, we hypothesized that families with higher socioeconomic status (SES) and advanced education would be better able to comply with the complex study procedures. We evaluated SES and other demographic factors which may influence adherence in the ongoing study. Both study drug adherence and visit adherence were examined. Study visits occurred every two weeks until a stable dose was reached and then every four weeks unless toxicity was encountered. Subjects remained on study for two years. Study drug adherence was monitored by measuring liquid in returned bottles at each visit. Bottles were included in the analysis if the subject was dispensed a bottle at a visit, returned it at the next scheduled visit with a non-zero volume, and did not have a drug stop order between the visits. From these bottles, each subject’s median drug adherence was calculated. Drug adherence (%) was defined as the amount consumed divided by the amount which should have been consumed. Visit adherence (%) was based on expected routine study visits. Subjects were withdrawn from the study upon family request or if they continually missed visits. We defined ‘good adherence’ as taking ≥80% study drug and attending ≥90% study visits. Adherence data was available on 186 of the 193 randomized infants. The median drug adherence was 101.7%, with 88.9% of subjects having drug adherence of ≥80%, and three infants having drug adherence >150%. The mean visit adherence was 97.3% ± 8.4 SD, with 82.3 % of subjects having no missed visits. Thirteen infants had a visit adherence of <90%; six were withdrawn from the study due to inactive follow-up status. Early withdrawal from the study was not significantly associated with drug adherence. However, infants on a stable dose who had visits every four weeks had a small but significantly worse drug adherence than infants whose dose was being titrated and had visits every two weeks (i.e. infants became less adherent over time in taking their study drug). Drug adherence and visit adherence were not significantly associated with the number of children or adults in the household, the primary caregiver’s employment or education level, or household income. To date, none of the patients have been nonadherent with both study visits and drug administration. A few subjects appeared to take more study drug than prescribed, which may be related to spilled doses. Although we hypothesized that higher SES would be associated with better adherence, we did not find this to be the case, suggesting that socioeconomic factors should not be used to prejudge the suitability of subjects for a randomized clinical trial. Our findings also suggest that special efforts are needed to maintain drug adherence, particularly when study visits become less frequent.