Adherence with protocol medication use and mortality from unrelated causes in a prevention trial.

Several studies have shown that non-adherence to medication use is associated with lower use of preventive services and increased mortality. We aimed to study the relationship between initial adherence to medication use and mortality in the Prostate Cancer Prevention Trial (PCPT). The PCPT randomized men age 55 and over to a finasteride or placebo arm. Duration of treatment was seven years, followed by end-of-study prostate biopsy. Extended follow-up for mortality was performed by linkage to the National Death Index. Non-adherence was defined as taking under 80% of required pills during the first or second 6-month trial period. Proportional hazards models were used to assess the relationship between adherence and all-cause mortality (excluding prostate cancer deaths). Three models were developed as follows: Model I (controlling for demographics and trial arm), Model II (Model I factors plus specific medical conditions), Model III (Model II factors plus lifestyle factors). Of 18,667 men included in the analysis, 3082 (16.5%) were non-adherent. The most common reasons for non-adherence were side effects (33.9%) and forgetting to take pills (22%). Through 5 and 10 years of follow-up, 178 (5.9%) and 483 (15.7%) non-adherent men died versus 581 (3.7%) and 1887 (12.1%) adherent men. Hazard ratios (HRs) at 5 years were 1.62 (95% CI: 1.37-1.91), 1.55 (95% CI: 1.30-1.83) and 1.49 (95% CI: 1.25-1.76) for Models I-III. HRs at ten years were lower but still statistically significant. Non-adherence to taking protocol medications was associated with increased mortality from unrelated conditions.