Bridevaux Pierre-Olivier, Leuenberger Philippe
Service de pneumologie, CHUV,1011 Lausanne.
Rev Med Suisse. 2005 Nov 16;1(41):2673-4, 2677-9.
In Switzerland, high dose of inhaled corticosteroids are frequently prescribed to COPD patients. Observational studies have suggested a reduction in mortality, although several randomised controlled trials did not show such a beneficial effect. Two important exposure biases are highlighted by a careful review of the methodological aspects of these observational studies: the immortal time bias and the failure to perform a time dependent analysis lead to underestimate the risk of death among exposed patients therefore incorrectly suggesting a survival benefit of regular inhaled steroid use. Two recent pharmaco-epidemiological studies, controlling for these exposure biases did not find any reduction in mortality due to inhaled steroids among COPD patients.