Unpredictability of deception in compliance with physician-prescribed bronchodilator inhaler use in a clinical trial.

OBJECTIVE To identify subject characteristics that may be predictive of intentional dumping of metered-dose inhalers (MDIs) during a clinical trial. DESIGN Nebulizer Chronologs (NCs; Medtrac Technologies; Lakewood, CO), which record the date and time of each MDI actuation, were attached to the MDIs of participants who were given a prescribed medication schedule to follow in a clinical trial. Participants were not informed of the function of the NC or that their medication use was being monitored. SETTING The Lung Health Study, a 5-year clinical trial to evaluate the effect of intensive smoking cessation counseling and regular use of an inhaled bronchodilator on the progression of COPD. PARTICIPANTS One hundred one smokers, 35 to 60 years of age, with mild to moderate airways obstruction enrolled in The Lung Health Study. MEASUREMENTS AND RESULTS Thirty of these 101 participants (30%) actuated their inhalers > 100 times within a 3-h interval on at least one occasion during the first year of this 5-year trial. Only 1 of an additional 135 participants who had full foreknowledge of the MDI monitoring capability of the NC did so. Most of these dumping episodes occurred shortly before a clinic follow-up visit, suggesting an active attempt to hide noncompliance from the clinic staff. Whereas self-reported inhaler usage and canister weights were similar for the "dumpers" and "nondumpers," NC data indicated significantly lower compliance rates for dumpers (chi(2); p < 0.05). When demographic variables, treatment and clinic assignments, smoking status, pulmonary function test results, respiratory symptoms, and disease history of dumpers and nondumpers were analyzed, no predictors of dumping could be found. CONCLUSIONS Deception among noncompliers occurs frequently in clinical trials, is often not revealed by the usual methods of monitoring, and cannot be predicted by data readily available in clinical trials.