[Prospective study of community-acquired pneumonias seen in a general hospital. Diagnostic errors].

Background: The patients with community acquired pneumonias seen in an emergency service of a basic general hospital during one year were evaluated to assess their etiological, clinical and radiological features, and also to investigate the initial and final diagnosis of the disease, its evolution and the parameters associated with each microbiological type.

Methods: A medical team investigated daily the clinical records. For etiologic diagnosis, blood cultures, serological studies, urine counterimmunoelectrophoresis (CIE) and, in individualized patients, pleural fluid culture, bronchoaspirate and bronchoalveolar lavage were performed. The discordance between the initial clinical and radiological diagnosis and the microbiological results, and also the initial and final diagnostic errors were analyzed.

Results: 311 cases of pneumonia (150 adults and 161 children) were diagnosed. 95 (30%) had microbiological confirmation. Streptococcus pneumoniae was the most commonly isolated organism. Serologic studies were the diagnostic method with the highest yield. Complications developed in 28% of the patients and the mortality rate was 2%. There was a relationship between mortality and a high respiratory rate on admission. The initial-final diagnostic discrepancy was 43% in atypical pneumonias and 40% in bacterial pneumonias.

Conclusions: Clinical parameters permitting the differentiation between atypical and bacterial pneumonia were not found. The initial diagnostic error was 12%, consisting of false positives in all instances, and the final diagnostic error included 15% false positives and 10% false negatives. Underlying diseases have a influence on the evolution of pneumonia. The mean respiratory rate on admission should be measured as a prognostic indicator. In the present study, urine CIE was a poorly sensitive method.