Channelling bias and the incidence of gastrointestinal haemorrhage in users of meloxicam, coxibs, and older, non-specific non-steroidal anti-inflammatory drugs

Background: Although clinical trial results suggest that meloxicam has less gastrointestinal toxicity than most other non-steroidal anti-inflammatory drugs ( NSAIDs), in practice it has been associated with a large number of yellow card reports of gastrointestinal complications. Aims: To estimate whether meloxicam and the coxibs, rofecoxib and celecoxib, have been channelled towards high risk patients, and to estimate the risk of hospitalisation for gastrointestinal haemorrhage associated with the use of these drugs, allowing for the effects of channelling. Patients: Using the UK General Practice Research Database, this study included 7.1 thousand patient years (tpy) exposure to meloxicam, 1.6 tpy exposure to coxibs, and 628 tpy exposure to older nonspecific NSAIDs. Methods: Cohort study of patients who received a prescription for an NSAID between June 1987 and January 2001. Exposure to newer NSAIDs ( meloxicam, rofecoxib, celecoxib) and to older non-specific NSAIDs was identified. Channelling was assessed on factors including: demographic variables; diagnosis of arthritis; history of NSAID use or gastrointestinal events, including gastrointestinal haemorrhage; and use of ulcer healing drugs. Results: Most risk factors for gastrointestinal haemorrhage were more prevalent among patients prescribed the newer NSAIDs. Adjusting for these risk factors reduced the relative risks of gastrointestinal haemorrhage on meloxicam and coxibs versus older non-specific NSAIDs to 0.84 (95% confidence interval 0.60, 1.17) and 0.36 (0.14, 0.97), respectively. Conclusions: Channelling towards high risk gastrointestinal patients occurred in the prescribing of newer NSAIDs. After attempting to correct for channelling bias, coxib exposure, but not meloxicam exposure, was associated with a significantly lower risk of gastrointestinal haemorrhage than older non-specific NSAID exposure.