Vanderbilt Center for Education/Research on Therapeutics

点击翻译按钮获取中文摘要

DESCRIPTION (PROVIDED BY APPLICANT): Chronic musculoskeletal problems and associated pain are common in clinical practice and affect Medicare enrollees disproportionately. The most effective treatment, non-steroidal anti-inflammatory drugs (NSAIDs) substantially increase the risk of serious peptic ulcer disease. The coxibs, a new class of anti inflammatory agents designed to avoid gastrointestinal complications, increase the risk of cardiovascular events. The remaining alternative, an NSAID with therapy to prevent gastrointestinal complications, is attractive, and although recommended in evidence-based guidelines, is not widely used. Endoscopic studies have shown that NSAID-induced erosions are decreased by proton pump inhibitors (PPI) and double-dose H2 receptor antagonist (H2RA), well tolerated drugs. However, there is no clinical trial showing that an NSAID with a PPI or H2RA decreases the risk of clinically important NSAID-induced gastrointestinal complications; nor is there an adequately powered trial comparing this combination therapy with a coxib. Such trials now are unlikely to be performed since many of the drugs are off patent and the hazards of both NSAID and coxib therapy are well known. Thus, robust clinical data to evaluate the comparative safety of NSAIDs with gastroprotective therapy is lacking. The absence of these comparative effectiveness data also will impede the development of Medicare policies that promote optimal pharmacotherapy for patients with arthritis and non-traumatic joint disorders while minimizing the risks of peptic ulcer disease, ischemic heart disease, and stroke. We propose to address these questions with pharmacoepidemiologic studies in the TennCare database, a resource we have used extensively. We will conduct a methodologically novel "new-user" design cohort study that will decrease channeling bias, to address two primary hypotheses: H1: The incidence of serious gastropathy in current users of NSAIDs with gastroprotective therapy is lower than that for NSAIDs absent gastroprotective therapy. H2: The incidence of serious gastropathy in current users of coxibs is lower than that for NSAIDs with gastroprotective therapy. The cohort will include 109,290 new NSAID/coxib users and will have a power (=.05, 1.-=.80) to detect protective rate-ratio of .74 for both H1 and H2.