RESUMO
OBJECTIVE: To assess whether antibiotic prescriptions in a department of pneumology are in accordance with the hospital conventions and whether documented prescriptions and probabilistic prescriptions are correctly adapted with regard to microbiological results and clinical progression of patients. METHOD: A 3-month prospective observational study was performed from november 1999 to january 2000 in a department of pneumology on all antibiotic prescriptions. The referential used for the conformity of probabilistic was the guide to pneumology antibiotic protocols regarding the most frequently encountered diseases. Treatment was considered adapted if the choice of the molecule and the dose administered were effective. Documented antibiotherapy was considered appropriate if the dose was correct and/or the micro-organisms were sensitive to at least one of the molecules administered. Probabilistic treatment was considered appropriated if progression was good without any change in the antibiotics. RESULTS: Among the 404 patients hospitalised in Pneumology, 132 (33%) received at least one antibiotherapy. There was a total of 163 treatments, 142 (87%) of which were probabilistic and 21 were documented. Seventy-two percent of the probabilistic prescriptions were in accordance with the protocols. The majority of those which did not conform (60.5%) was due to the dose rather than the choice of the antibiotic. More than three quarters of the probabilistic and documented treatments (79%) were clinically adapted. Treatment failures were more often due to a mistake in spectrum rather than an inappropriate dose. CONCLUSION: The use of antibiotics has become increasingly complex because of the need for therapeutic efficacy, limitation of the selection of microbial resistance, and the cost of treatment. Application of antibiotic protocols drawn-up by the clinicians, microbiologists and chemists concerned appears to be an efficient solution that is clearly acceptable to the various participants.
