To explore the effects of bacterial filters (BFs) combined hot-water humidification on reducing the incidence of ventilator-associated pneumonia (VAP).

A prospective randomized controlled trial was performed in ICU of Nanjing Medical University from March 2017 to March 2020 that enrolled patients under mechanical ventilation for ≥48 hours. These patients were randomly assigned to the bacterial filter (BF) group (104 cases) or non-bacterial filter (NBF) group (106 cases). NBF group was treated with hot-water humidification (HH) while BF group was treated with HH combined BF. Univariate logistic analysis was applied to explore the risk factors for VAP. The Kaplan-Meier method was used to compare the cumulative incidence of VAP in group 2 patients. Cox proportional-hazards regression and post hoc subgroup analysis were used to determine the effects of the filter on the incidence of VAP.

The proportion of patients with tracheotomy was low and statistically significant (19.2% vs 38.7%, χ²=9.634, P=0.002), the prevalence of VAP in two groups had no difference without statistical significance (22.1% vs 34.0%, χ²=3.647, P=0.066). Logistic regression analysis revealed that the bacterial filter failed to effectively reduce VAP occurrence (OR=0.552, 95% CI=0.299-1.019, P=0.058). The Kaplan-Meier analysis did not show significant advantages of bacterial filters in preventing VAP (log-rank test P=0.060). Subgroup analysis and Cox survival regression analysis showed that bacterial filters significantly reduced VAP occurrence in patients with mechanical ventilation of＜25 d (HR=0.373, P=0.004).

Bacterial filters can significantly reduce the incidence of VAP in patients under mechanical ventilation for＜25 days with hot-water humidification, although this benefit became insignificant for prolonged mechanical ventilation.