Compliance assessment in drug trials: has there been improvement in two decades?

Abstract

Background: Compliance is a key element in the success of therapy, both in practice and research. A study from 1974 demonstrated that compliance in clinical trials was only determined in 19% of studies requiring it.

Objectives: The objective was to determine if there has been an improvement in compliance assessment in clinical trials.

Methods: All drug studies published in the British Medical Journal, Journal of Pediatrics, and Lancet from 1997 to 1999 were reviewed. Clinical trials were evaluated as to their measurement of compliance and the method of assessment.

Results: Of 303 studies in which the effects of drugs were reported, 165 required the incorporation of a measure of compliance, 86 did not, and in 52, compliance could not be measured. Of the studies requiring estimation of compliance, compliance was evaluated in 78 (47%). This rate did not vary between the journals examined or between trials in adults or children. The most common methods used to evaluate compliance were pill count (33%) and self report (25%). The use of drug assays (14%) and close supervision (9%) was less common. Electronic devices and other methods were uncommonly used (5%). In 16% of cases, a combination of methods was used.

Conclusion: Although the rate of evaluation of compliance in drug trials has improved over the past 25 years, it continues to be examined in less than half of the clinical studies of drug effects in which compliance assessment is required. This rate appears to be similar in paediatric and adult drug studies.