Use of gastrointestinal preventive therapy among elderly persons receiving antiarthritic agents in Nova Scotia, Canada

Background: Two different strategies, referred to as gastrointestinal (GI) preventive therapy (GIPT), have been recommended for high-risk patients to prevent GI complications associated with antiarthritic therapy with nonsteroidal anti-inflammatory drugs (NSAIDs): (1) use of a gastroprotective agent (GPA) along with the NSAID or (2) use of a cyclooxygenase-2-selective inhibitor (COX-2SI). The COX-2SIs rofecoxib and celecoxib have been shown to be as effective as traditional NSAADs for pain relief, but with an improved GI safety profile. Objective: The purpose of this study was to examine the utilization of GIPT by elderly persons in Nova Scotia who were taking antiarthritic medications and to identify the factors associated with their use of GIPT. Methods: A retrospective, cross-sectional study was conducted using administrative data from the Nova Scotia Seniors' Pharmacare Program database. Study participants were aged ≥65 years and had filled a prescription for a COX-2SI, a traditional NSAID, or high-dose aspirin at some point between January 1, 2001 and August 31, 2002. Subjects with at least 1 risk factor (as defined by our study) who received GIPT were classified as receiving appropriate therapy. Subjects with risk factors who did not receive GIPT were classified as potential underutilizers of GIPT. Subjects without risk factors who received GIPT were classified as potential overutilizers of GIPT. Descriptive statistics were presented, and factors independently associated with receiving GIPT were assessed using logistic regression. Results: The study included 14,587 seniors: 3647 used COX-2SIs, 9412 used traditional NSAIDs alone, and 1528 used traditional NSAADs plus a GPA. Subjects were predominantly female (age range, 65-74 years). In subjects with at least 1 risk factor, 63% were classified as potential underutilizers of GIPT. Thirty-three percent of subjects with no risk factors were classified as potential overutilizers of GIPT. Factors significantly associated with receiving a GIPT included Female gender, annual income >$50,000, urban residence, age ≥75 years, GI complication in the previous year, and concomitant use of warfarin or corticosteroids. Conclusion: Potential under utilization of GIPT in this subject population was more prevalent than potential overutilization of GIPT. Although all hypothesized risk factors were significantly associated with receiving GIPT, physician education on GI risk factors might improve prescribing of GIPT for elderly persons in Nova Scotia.