Subject retention and adherence in a randomized placebo-controlled trial of a disease-modifying osteoarthritis drug

Objective. To describe the methods by which remarkable levels of subject retention and adherence were achieved in a 30-month multicenter randomized placebo-controlled trial (RCT) of a disease-modifying osteoarthritis drug (DMOAD). Methods. Subjects were obese 45-64-year-old women with unilateral knee osteoarthritis. Before randomization, each volunteer completed a 4-week "faintness-of-heart" (FOH) test, during which she was required to demonstrate reliable appointment keeping and greater than or equal to80% adherence to the dosing regimen. Subjects who passed the FOH test were randomized to treatment with doxycycline or placebo for 30 months. The double-blind phase entailed 15 bimonthly followup visits; intervisit adherence data were downloaded from the dosing monitor and used to estimate therapeutic coverage and to identify correctable patterns of nonadherence. Subjects received token incentives and a small cash payment at each followup visit. Measures to prevent or treat side effects of doxycycline were dispensed free of charge. Study coordinators monitored safety and reinforced participation through between-visit telephone calls. Results. Of 463 eligible volunteers, 32 (7%) failed the FOH test and were excluded from the double-blind phase. Among the 431 subjects randomized to treatment groups, 307 (71%) completed the 30-month RCT and 124 discontinued the study drug prematurely. Nearly half of the dropouts returned for their 16- and 30-month radiographs, resulting in loss to followup of 14.8% of randomized subjects. The 2 treatment groups did not differ significantly with respect to rates of discontinuation or retention. Therapeutic coverage over 30 months was very high in both groups. Conclusion. The rate of discontinuation in this 30-month RCT (29%) was lower than that of any DMOAD trial of 2:2 years duration published to date. The proportion of subjects for whom 30-month radiographs were available (85%) and adherence to the dosing regimen (mean >80%) also were remarkably high.