Order set to improve the care of patients hospitalized for an exacerbation of chronic obstructive pulmonary disease

Rationale: Physicians' adherence to prescribing evidence-based inpatient and outpatient therapies for chronic obstructive pulmonary disease (COPD) is low, and there is a paucity of information about the utility of admission order sets for patients with COPD exacerbations. Objectives: To determine if implementation of a locally designed, evidence-based, multidisciplinary computer physicianorderentryset in the electronic health record improves the quality of physician pharmacologic prescribing for patients hospitalized for COPD exacerbations. Methods: This study was performed before and after implementation of a computerized order set for patients hospitalized for COPD exacerbations. The primary outcome was the rate of zero prescribing errors by physicians for inpatient and discharge drugs for COPD over a 1-year period before implementation and for 6 months after implementation. Errors were defined as no therapy or inappropriate therapy in the following categories: antibiotic, systemic corticosteroid, short-acting bronchodilator, long-acting bronchodilator, and inhaled corticosteroid. Secondary outcomes included mean physician pharmaceutical prescribing error rate; types of errors; hospital lengths of stay; and unscheduled physician visits, emergency department visits, rehospitalizations, and deaths within 30 days from discharge. Measurements and Main Results: There were 194 COPD exacerbation admissions during the 1-year preimplementation period and 81 admissions during the 6-month postimplementation period. Compared with the preimplementation period, the percentage of patients receiving all recommended pharmacologic therapies for the 6 months after implementation increased from 18.6% to 54.3% (P<0.001). The mean number of errors decreased from 1.76 to 0.65 (P<0.001). Antibiotic and systemic corticosteroid errors decreased from 39% to 16% (P<0.001) and from 58% to 28% (P<0.001), respectively. Fewer patients were discharged without a short-acting bronchodilator (13.9% vs. 2.5%;P = 0.005), a long-acting bronchodilator (16.5% vs. 7.4%; P = 0.047), or inhaled corticosteroid (18% vs. 9.9%; P = 0.089). Improvements were sustained over the 6-month postimplementation period. Hospital length of stay decreased from 4 (±3) days preimplementation to 2.9 (±1.9) days postimplementation (P = 0.002). There were no significant differences in 30-day clinical outcomes, including the rates of unscheduled physician or emergency department visits, rehospitalizations, or deaths.