Ketorolac use in outpatients and gastrointestinal hospitalization: a comparison with other non-steroidal anti-inflammatory drugs in Italy

Objective: To compare the risk of hospitalization for gastroduodenal ulcer associated with the use of ketorolac and other non-steroidal anti-inflammatory drugs (NSAIDs). Methods: A cohort and a nested case-control study were carried out. All residents in the region of Umbria (Italy), aged 35-84 years, who had been given at least one NSAID prescription in 1993 and 1994 were identified. Exposure to drugs was ascertained through a drug prescription database. We estimated rate ratios of hospitalization for gastroduodenal ulcer with or without complications in the current, recent or past period according to exposure to different NSAIDs. Results: Rate ratio estimates, adjusted for age and sex, were 2.8 for any current NSAID and 1.4 for any recent NSAID. The highest rate ratios of lesions of any severity for current NSAID use were observed for piroxicam (RR: 4.6) and ketorolac (RR: 3.4). For gastrointestinal haemorrhage or perforation the highest rate ratios were those for ketorolac (RR: 5.9) and piroxicam (RR: 4.8). Rate ratio estimates did not change after adjustment for concomitant use of gastrotoxic drugs, use of gastroprotective agents not associated with NSAIDs and prior use of NSAIDs. Conclusion: Our study demonstrates the need to adhere to the restrictions relating to the indications and duration of use of ketorolac. At present piroxicam represents a greater public health concern since it is confirmed to be among the most gastrotoxic NSAIDs and is one of the most commonly prescribed NSAIDs in Italy.