Abstract
Pneumonia is one of the main causes of intensive care unit admission and death in Brazil. There is a need to standardize the use of microbiological tests for the diagnosis of pneumonia. OBJECTIVE: To evaluate microbiological diagnostic data for severe pneumonia. METHOD: Comparative studies that evaluated the microbiological diagnosis of community pneumonia and hospital-acquired pneumonia were analyzed. The literature review was guided by two questions and used flowcharts prepared by experts. RESULTS AND DISCUSSION: Diagnostic tests for pneumonia should be requested based on the severity of the disease, the patient´s immune status, and the risk of infection by multidrug-resistant bacteria. Gram staining may aid in excluding S. aureus pneumonia and evaluating the quality of respiratory samples, while culture of these clinical samples may identify the infectious agent in up to half of the cases and allow antimicrobial susceptibility testing to be carried out. Blood cultures have a low sensitivity but may be useful for diagnosing extrapulmonary infections or situations involving sepsis or septic shock. Rapid molecular panels have high sensitivity and specificity for viral and bacterial targets for both types of pneumonia, and their use can reduce the length of hospital and intensive care stays and allow optimization of antimicrobial therapy. RT-PCR is highly accurate in diagnosing SARS-CoV-2 pneumonia. CONCLUSION: The rational use of molecular panels for the diagnosis of severe pneumonia can reduce the length of hospital and intensive care stays and 90-day mortality.