Perception of the risk of gastrointestinal adverse drug reactions with non-steroidal anti-inflammatory drugs (including coxibs): differences among general practitioners, gastroenterologists and rheumatologists.

OBJECTIVE: To compare the perception of risk of gastrointestinal adverse drug reactions (ADRs) with non-steroidal anti-inflammatory drugs (NSAIDs) (including coxibs) in general practitioners, gastroenterologists and rheumatologists.
METHODS: Physicians (69 general practitioners, 45 gastroenterologists and 58 rheumatologists; 172 total) were interviewed using visual analogue scales in order to define a score of perceived risk (ranking from 0 to 10) of gastrointestinal ADRs associated with four classes of NSAIDs (indolics, oxicams, arylcarboxylic derivatives, coxibs).
RESULTS: When the three groups of physicians were taken as a whole, indolic derivatives were ranked as the most dangerous NSAIDs [median score: 5.3 (25-75th percentiles: 3.5-7)] followed by oxicams [median score: 5.0 (25-75th percentiles: 3.1-6.2)], arylcarboxylic derivatives [median score 4.7 (25-75th percentiles: 3.0-6.5)] and coxibs [median score 1.9 (25-75th percentiles: 1.2-3.0)]. Among physicians, general practitioners gave the most important mean score [median score 4.8 (25-75th percentiles: 3.4-5.7)] for NSAIDs as a whole followed by gastroenterologists [median score 4.6 (25-75th percentiles: 3.6-5.6)] and rheumatologists [median score 3.6 (25-75th percentiles: 2.3-4.7)]. There was no significant difference among the three groups of physicians in the median score of perceived digestive risk associated with the use of indolic derivatives, whereas significant differences were found for arylcarboxylic derivatives, oxicams and coxibs, with higher median scores given by gastroenterologists followed by general practitioners and rheumatologists.
CONCLUSION: This study shows differences in the perception of risks of gastrointestinal ADRs. Rheumatologists systematically considered NSAIDs as less harmful than general practitioners and gastroenterologists did. In contrast, the widely demonstrated difference in gastrointestinal risk among the different groups of NSAIDS (with, for example, a higher risk for oxicams) was not pointed out by the three studied groups of physicians. There are differences in the perception of gastrointestinal risks of NSAIDs (including coxibs) among physicians according to their medical education and medical specialisation.