Aim : To compare the incidence of abdominal pain, dyspepsia and/or nausea associated with valdecoxib, nonspecific nonsteroidal anti-inflammatory drugs and placebo in patients with rheumatoid arthritis and osteoarthritis.

Methods : Data from five randomized, double-blind 12-week trials were pooled. Independent risk factors for abdominal pain, dyspepsia and/or nausea were also determined.

Results : The final analysis consisted of 4394 patients. Nonspecific nonsteroidal anti-inflammatory drug users (n = 1185) received naproxen 1000 mg/day (n = 766), ibuprofen 2400 mg/day (n = 207) or diclofenac sodium 150 mg/day (n = 212). Valdecoxib users received 10 mg/day (n = 955), 20 mg/day (n = 851) or 40 mg/day (n = 430). A total of 973 patients received placebo. The nonspecific nonsteroidal anti-inflammatory drug group was most likely to report abdominal pain or dyspepsia, while the placebo group reported the highest incidence of nausea. The most important risk factors for abdominal pain, dyspepsia and/or nausea were nonspecific nonsteroidal anti-inflammatory drug use, gastrointestinal history of nonspecific nonsteroidal anti-inflammatory drug-related intolerance or gastroduodenal ulcers, osteoarthritis diagnosis, female gender and age <65 years.

Conclusion : This pooled analysis demonstrates a clear decrease in dyspepsia and an improvement in upper gastrointestinal tolerability for patients with osteoarthritis and rheumatoid arthritis taking valdecoxib, even at supratherapeutic doses, compared with those taking nonspecific nonsteroidal anti-inflammatory drugs over 12 weeks.