Validation of a β-agonist long-term asthma control scale derived from computerized pharmacy data

Background: Asthma control has been defined clinically by using validated tools, but an asthma control scale using administrative data has not been reported. Objective: We sought to validate a beta-agonist asthma control scale derived from administrative data. Methods: Surveys that included validated asthma symptom and control tools were completed by a random sample of 2250 health maintenance organization members aged 18 to 56 years with persistent asthma. Linked computerized pharmacy data provided beta-agonist canister and oral corticosteroid dispensings. The proposed 4-level asthma control scale was based on the number of short-acting beta-agonist canisters dispensed in 12 months. Construct validity and predictive validity were assessed. Results: For construct validity, factor analysis showed significant loading of the beta-agonist scale on the symptom control factor, and the beta-agonist scale was significantly related to the validated asthma control and symptom scales (r = 0.31, P < .0001). For predictive validity, each progressive level of the proposed P-agonist control scale was associated with an increased risk of subsequent asthma hospitalizations or emergency department visits and oral corticosteroid use, independent of prior use. Conclusion: A scale based on the number of beta-agonists dispensed in a 1-year period and derived from administrative data reflects asthma symptom control over that period of time. This scale can help identify patients who are at risk for future acute asthma health care use. Clinical implications: This information can he used in population management and by clinicians to assess long-term asthma control and identify patients who need intervention to prevent future morbidity.