Background: Asthma disease management programs typically use pharmacy data to identify high-risk individuals for outreach. Provider-directed pharmacy profiling seeks to identify physicians whose prescribing of recommended asthma medication is suboptimal. Both strategies require an accurate approach to counting prescribed asthma medication.
Objective: We compare two methods for counting the use of bronchodilators and inhaled anti-inflammatory medication. One approach uses simple counts of dispensed medication. An alternative, canister-equivalent method standardizes these medications on the basis of variation in both potency and medication-days supplied per prescription. We evaluate whether these alternative methods yield different population risk profiles when applied to managed care enrollees who have asthma and to the physicians treating them.
Methods: Retrospective cohort study of patterns of medication use by asthmatic patients receiving care within a group-model health maintenance organization and prescribing of asthma medications by the physicians treating them.
Results: Each method yields a different risk profile of the patient and physician populations, respectively. Relative to simple counts, the canister-equivalent method results in a 40% increase in the population identified as having high bronchodilator use and chronic anti-inflammatory medication use. On the physician-level, the mean anti-inflammatory:bronchodilator ratio (AIF:BD) was 1.50 by the canister-equivalent method compared with 1.08 by the simple-count method. When stratified by each method, 36% of physicians were assigned to different quartiles of anti-inflammatory:bronchodilator ratio.
Conclusions: A novel canister-equivalent method for counting dispensed asthma medications yields different risk profiles compared with simple counts of asthma medications. Asthma disease management programs should consider alternative approaches to improve the accuracy of risk profiling based on patterns of medication use.