Invasive and non-invasive diagnostic approaches for microbiological diagnosis of hospital-acquired pneumonia

BACKGROUND: Data on the methods used for microbiological diagnosis of hospital-acquired pneumonia (HAP) are mainly extrapolated from ventilator-associated pneumonia. HAP poses additional challenges for respiratory sampling, and the utility of sputum or distal sampling in HAP has not been comprehensively evaluated, particularly in HAP admitted to the ICU. METHODS: We analyzed 200 patients with HAP from six ICUs in a teaching hospital in Barcelona, Spain. The respiratory sampling methods used were divided into non-invasive [sputum and endotracheal aspirate (EAT)] and invasive [fiberoptic-bronchoscopy aspirate (FBAS), and bronchoalveolar lavage (BAL)]. RESULTS: A median of three diagnostic methods were applied [range 2–4]. At least one respiratory sampling method was applied in 93% of patients, and two or more were applied in 40%. Microbiological diagnosis was achieved in 99 (50%) patients, 69 (70%) by only one method (42% FBAS, 23% EAT, 15% sputum, 9% BAL, 7% blood culture, and 4% urinary antigen). Seventy-eight (39%) patients underwent a fiberoptic-bronchoscopy when not receiving mechanical ventilation. Higher rates of microbiological diagnosis were observed in the invasive group (56 vs. 39%, p = 0.018). Patients with microbiological diagnosis more frequently presented changes in their empirical antibiotic scheme, mainly de-escalation. CONCLUSIONS: A comprehensive approach might be undertaken for microbiological diagnosis in critically ill nonventilated HAP. Sputum sampling determined one third of microbiological diagnosis in HAP patients who were not subsequently intubated. Invasive methods were associated with higher rates of microbiological diagnosis. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (10.1186/s13054-019-2348-2) contains supplementary material, which is available to authorized users.