Outcome of upper gastro-intestinal bleeding and use of ibuprofen versus paracetamol

BACKGROUND: Nonsteroidal anti-inflammatory drugs (NSAIDs) are known to increase the risk of upper gastrointestinal bleeding (UGIB). Whether the severity of outcome of UGIB associated with NSAIDs differs from non-NSAID-related UGIB is less clear.

METHOD: Medical records of 228 patients hospitalized for UGIB in the Danish county of North Jutland were evaluated. Preadmission characteristics and clinical outcomes were compared between 112 patients who had been prescribed ibuprofen and 116 patients who had been prescribed paracetamol within 90 days of the hospitalization.

RESULTS: The baseline characteristics of UGIB patients prescribed ibuprofen tended to differ from those prescribed paracetamol. The ibuprofen group significantly less often had histories of ulcer (11% vs 36%) and dyspepsia (19% vs 44%), or had been prescribed medications for these or other conditions, and had lower co-morbidity indices. Ibuprofen users also were somewhat less likely (31% vs 37%) to report GI pain at admission, but among hospitalized patients with endoscopic examinations were more likely (75% vs 58%) to be diagnosed with ulcer or hematemesis vs normal or gastritis/dyspepsia/reflux. For the clinical outcomes, 30 days case fatality rates were 12% for both ibuprofen and paracetamol users. The ibuprofen-related cases of UGIB more often required surgery (11% vs 3%) or transfusions (66% vs 57%), and those prescribed ibuprofen averaged 11 days in hospital, 4 days longer than those prescribed paracetamol. Adjustment for baseline characteristics and underlying conditions, or analyses eliminating patients with unconfirmed diagnoses and prior ulcers or restricted to patients with current hospital diagnoses of ulcer or hematemesis, did not materially alter the ibuprofen vs paracetamol differences in outcome measures. Generally similar results were obtained when restricting the analyses to patients prescribed ibuprofen or paracetamol within 30 days of UGIB hospitalization, except for a reduction in the 30 days case fatality rate among those prescribed ibuprofen.

CONCLUSIONS: UGIB patients with antecedent ibuprofen prescriptions experienced about the same case fatality rates, but more surgery and longer hospital stays, than patients prescribed paracetamol. The differences appear in part due to differing characteristics among those prescribed ibuprofen compared with those prescribed paracetamol, but also raise the possibility of drug-related effects.