Prediction of ventilator-associated pneumonia outcomes according to the early microbiological response: a retrospective observational study.

Background: Ventilator-associated pneumonia (VAP) is a leading infectious cause of morbidity in critically ill patients, yet current guidelines offer no indications for follow-up cultures. We aimed to evaluate the role of follow-up cultures and microbiological response 3 days after diagnosing VAP as predictors of short- and long-term outcomes.

Methods: We performed a retrospective analysis of a cohort prospectively collected from 2004 to 2017. VAP was diagnosed based on clinical, radiographical and microbiological criteria. For microbiological identification, a tracheobronchial aspirate was performed at diagnosis and repeated after 72 h. We defined three groups when comparing the two tracheobronchial aspirate results: persistence, superinfection and eradication of causative pathogens.

Results: 157 patients were enrolled in the study, among whom microbiological persistence, superinfection or eradication was present in 67 (48%), 25 (16%) and 65 (41%), respectively, after 72 h. Those with superinfection had the highest mortalities in the intensive care unit (p=0.015) and at 90 days (p=0.036), while also having the fewest ventilator-free days (p=0.019). Multivariable analysis revealed shock at VAP diagnosis (OR 3.43, 95% CI 1.25-9.40), isolation at VAP diagnosis (OR 2.87, 95% CI 1.06-7.75) and hypothermia at VAP diagnosis (OR 0.67, 95% CI 0.48-0.95, per +1°C) to be associated with superinfection.

Conclusions: Our retrospective analysis suggests that VAP short- and long-term outcomes may be associated with superinfection in follow-up cultures. Follow-up cultures may help guide antibiotic therapy and its duration. Further prospective studies are necessary to verify our findings.