Ventilator-associated events criteria in the assessment of Ventilator-Associated Pneumonia (IMPACTO MR-PAV): A prospective cohort

Background

Ventilator-Associated Pneumonia (VAP) is a critical healthcare-associated infection, but no universal surveillance standard exists. In 2013, the CDC revised its criteria, incorporating Ventilator-Associated Events (VAEs) with VAPs as a subset. In Brazil, however, the Health Regulatory Agency (ANVISA) chose to retain the traditional VAP criteria. This study aimed to evaluate the incidence of VAP using both the traditional and revised criteria.

Method

We conducted a prospective multicentric cohort of critically ill adult patients who required mechanical ventilation in 12 Brazilian Intensive Care Units (ICU) over six months. We evaluated the level of agreement between the two criteria considering frequency and kappa coefficient. This study was registered at ClinicalTrials.gov, NCT05589727.

Results

The study included 987 patients and revealed that 85.7 % of VAP reported by the centers according to ANVISA criteria were not confirmed by the adjudicators. Among the adjudicators, a 16.7 % disagreement (kappa = 0.32) suggested subjectivity in applying VAP criteria. Between the two sets of criteria, an 11% disagreement (kappa = 0.12) was observed. However, manual adjudication of automatically generated VAEs showed only a 4 % disagreement, indicating greater objectivity in the VAE criteria. Despite the high agreement in VAE adjudication, this did not necessarily translate to a more reliable exclusion of non-events, which is essential for accurate surveillance.

Conclusion

The findings highlight the challenges in identifying and classifying VAP, emphasizing the need for improved surveillance methods. The results could inform enhancements in VAP monitoring in Brazil and potentially impact other countries using similar criteria.