Design: Pragmatic cluster-randomised crossover trial (AB/BA) in six ICUs
Population: 1035 adults ventilated for at least 48 hours in medical ICUs in Bhubaneswar, India
Primary outcome: VAP incidence per 1000 ventilator-days
Effect: 18.8 vs 16.9 per 1000 ventilator-days; adjusted IRR 1.08 (0.84–1.38)

A pragmatic cluster-randomised crossover trial, published in Critical Care on 29 October, compared a standardised oral-care bundle with and without 0.12% chlorhexidine in six adult medical ICUs at one academic hospital in Bhubaneswar. Each ICU used both strategies in turn, for nine months each.

Among 1035 patients ventilated for at least 48 hours, pneumonia occurred at 18.8 per 1000 ventilator-days with chlorhexidine and 16.9 without, with no meaningful difference. Mortality at 28 days was similar (26.4% vs 28.7%). ICU stay was about 0.9 days longer with chlorhexidine.

Chlorhexidine mouthwash has been a fixture of ventilator bundles for years, but earlier meta-analyses had already questioned its benefit outside cardiac surgery and raised concerns about harm. This Indian trial, in a setting close to many readers' units, adds direct evidence that the rest of the bundle does the work.

The practical step is to keep the oral-care bundle, brushing, suction, head-up positioning, and drop routine chlorhexidine in medical ICU patients.

Consider stopping routine chlorhexidine mouthwash in ventilated medical ICU patients.
Keep the rest of the oral-care bundle: toothbrushing, suction and moisturising.
Maintain head-of-bed elevation and daily sedation and weaning checks.
This trial did not study cardiac surgery patients; local policy for them may differ.

Why it matters

It removes a long-standing bundle element that costs time and may cause harm without benefit.

The statistics, in plain English

The adjusted incidence rate ratio of 1.08 (95% CI 0.84 to 1.38) crosses 1 and points, if anything, towards slightly more pneumonia with chlorhexidine. The extra 0.89 days in ICU (0.19 to 1.58) is a secondary outcome and should be read cautiously.