Article Text


OP0137 Less healthcare resource utilisation for perforations, ulcers and bleeds with rofecoxib vs. naproxen in the vigour (vioxx gi outcomes research) study
  1. C Bombardier1,
  2. L Laine2,
  3. GW Carides3,
  4. JM Pellissier3,
  5. DR Ramey3,
  6. DJ Watson3,
  7. A Reicin3
  1. 1Clinical Epidemiology Division, Wellesley Hospital Research Institute, Toronto, Canada
  2. 2USC School of Medicine, Los Angeles, CA, USA
  3. 3Merck Research Laboratories, West Point, PA, USA


Background The VIGOUR Study was a randomised, double-blind, clinical trial assessing the GI safety of rofecoxib 50 mg. once daily (n = 4047) versus naproxen 500 mg twice daily (n = 4029) in the patients with rheumatoid arthritis. The mean age was 58 years; 80% of patients were female; 8% had a prior history of a GI ulcer or bleed; 82% reported NSAID use at baseline; 21% reported dyspepsia within 14 days prior to baseline; the treatment groups were not different in these respects. The median duration of follow-up 9.0 months, the maximum was 13 months.

Objectives To compare the rates of hospitalisation and mean number of outpatient healthcare contacts associated with upper GI perforation, ulcer, or bleeding (PUB) in patients treated with rofecoxib, a highly selective COX-2 inhibitor, vs. naproxen in the VIGOUR Study.

Methods The proportions of patients with hospitalizations for PUB in the two treatment groups were compared for all randomised patients. All PUBs were confirmed by an outside adjudication committee. The proportions of patients hospitalised for PUB, and the rates of PUB hospitalisation (per 100 patient-years, and based on the Kaplan-Meier method) were compared between treatment groups. Additional analyses compare proportions of PUB patients hospitalised and the mean number of outpatient contacts given a PUB event.

Results The incidence of confirmed PUBs, the primary endpoint in VIGOUR, was significantly lower with rofecoxib compared to naproxen (2.1 vs 4.5% per year; p < 0.001, relative risk 0.46). Patients treated with rofecoxib were hospitalised for PUB less often than patients treated with naproxen (crude rate: 0.4% vs 0.8%, 95% CI of difference -0.7% to -0.1%). Fewer patients treated with rofecoxib versus naproxen were hospitalised for complicated PUBs (i.e., perforation, obstruction, or clinically serious bleeding) (0.2% vs 0.6%, 95% CI of difference-0.6% to -0.1%); there was no difference between groups in the proportion of patients hospitalised with uncomplicated PUBs (0.2% both groups). PUB hospitalisation rates per 100 patient years were lower rofecoxib than for naproxen (0.63 vs. 1.26, risk ratio 0.50, 95% CI of risk ratio 0.28 to 0.89). These rates were higher in the US for both treatment groups (rofecoxib 0.91, naproxen 1.65) compared with ex-US (rofecoxib 0.44, naproxen 1.00), however the reduction in rates with rofecoxib was similar in the US and ex-US. Analyses of the one-year hospitalisation incidence rates for PUBs based on the Kaplan-Meier method produced analogous results. As expected, in those patients who experienced a PUB, rates of outpatient contacts and hospitalizations were not different between treatment groups.

Conclusion In the VIGOUR Study, RA patients treated with the highly selective COX-2 inhibitor rofecoxib for up to 13 months had significantly fewer PUBs and required fewer hospitalizations for PUBs than patients treated with naproxen.

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.