An objective method of compliance assessment with metered-dose inhalers

Evaluation of compliance with metered-dose inhalers (MDIs) is especially difficult. We investigated the correlation between serial weight change of the MDI canister and the number of doses delivered. Canister weight strongly correlated with number of activations for three different bronchodilators (metaproterenol, r2 = 0.9965; albuterol, r2 = 0.9984; terbutaline, r2 = 0.9913). To validate these results, a one-month trial designed to mimic patient use was conducted. Nine aerosol bronchodilator MDIs (three each of terbutaline, albuterol, and metaproterenol) were evaluated by three volunteers, each given one of each type of MDI. Number of activations were recorded in a diary and canisters were weighed weekly. At the end of the four-week period the number of activations were determined from weekly canister weight. Predicted activation numbers, calculated from both regression line and baseline weight, were compared to actual activation number. Statistical analysis was done using analysis of variance. Predicted number of activations for all three drugs ranged from 77 to 125 percent of observed and did not significantly differ depending on the type of bronchodilator (p = 0.3340). The method of prediction, regression intercept or baseline weight, led to significantly different predictions (p = 0.0569). The interaction between the method of prediction and type of bronchodilator was significant (p = 0.0197). Using either method, the actual number of doses dispensed can be predicted within 25 percent.