Understanding antimicrobial prescribing in suspected ventilator-associated pneumonia: a prospective cohort study

Background Ventilator-associated pneumonia (VAP) is associated with significant healthcare cost, morbidity, and mortality, but can be difficult to identify, resulting in over-diagnosis and excessive use of broad-spectrum antimicrobial therapy. In addition, some organisms commonly cultured from the airways of critically ill patients, in particular Candida species, are of unknown clinical significance. Methods A prospective cohort study was conducted across five intensive care units in the North-West of England. Participants were enrolled within 24 hours of commencing antimicrobial therapy for suspected VAP. Laboratory-confirmed VAP was defined by quantitative culture of a known pneumonia-causing pathogen above predetermined growth thresholds. Univariate logistic regression was used to determine the impact of laboratory-confirmed VAP, APACHE II, culture of Staphylococcus aureus, and culture of Candida species on 30-day mortality. Results The prevalence of laboratory-confirmed VAP was 43/96 (44%), and the median number of antimicrobials prescribed for VAP was 1 (range: 1-4). Candida species were identified in 32/96 patients (33%). The overall 30-day mortality was 22/96 (26%). None of the variables analysed were associated with 30-day mortality, except for culture of Candida species, which was associated with survival (odds ratio 0.26, 95% CI 0.07 to 0.98; p= 0.047). Conclusions Ventilator-associated pneumonia was confirmed in under half the patients commenced on antimicrobial therapy for suspected VAP, which highlights the urgent need for improved diagnostic strategies. In our clinical practice, Candida species are not treated as pathogenic in VAP, and in this study, growth of Candida species was not associated with excess 30-day mortality.