National adherence to evidence-based guidelines for the prescription of nonsteroidal anti-inflammatory drugs

Background & Aims: Our objective was to assess adherence to evidence-based guidelines by providers of the Department of Veterans Affairs nationwide. Methods: This was a cross-sectional study among veterans prescribed a nonsteroidal anti-inflammatory drug (NSAID) from January 1, 2002, to December 31, 2002. Prescription data were linked to inpatient and outpatient medical records and death files. The population was characterized as high risk based on the following: age 65 years or older, concurrent corticosteroid or anticoagulant use, history of peptic ulcer, and high average daily dose of NSAIDs. Adherence was defined as the prescription of a traditional NSAID with gastroprotection or a coxib in high-risk NSAID users. Univariate and multivariate analyses assessed the potential predictors of adherence. Results: Three hundred three thousand seven hundred eighty-seven met our definition of high risk. Most (97.3%) were male; 55.6% were white, 9.6% black, and 34.8% of other/unknown race. Age 65 years or older was the largest high-risk subset (87.1%). Overall, only 27.2% of high-risk veterans (n = 82,766) were prescribed an adherent strategy. Among veterans with at least 2 risk factors, adherence was 39.7%; among those with 3 risk factors, adherence was 41.8%. Predictors of adherence included history of upper gastrointestinal events, anticoagulant use, rheumatologic disease, high Deyo comorbidity index score, use of low-dose salicylates, and concurrent corticosteroid use. Predictors of nonadherence included prescriptions ≥90 days and high average daily dose of NSAIDs. Conclusions: Adherence to evidence-based guidelines for safe prescription of NSAIDs in the Department of Veterans Affairs is low (27.2%). The likelihood of adherence is further decreased if veterans are prescribed NSAIDs for ≥90 days.