Etiology and antimicrobial resistance profiles of ventilator-associated pneumonia in adult intensive care units in India: implications for infection prevention and antimicrobial stewardship - a systematic review and meta-analysis

Ventilator-associated pneumonia (VAP) is a leading healthcare-associated infection in intensive care units (ICUs) and disproportionately affects low- and middle-income countries, where multidrug-resistant Gram-negative organisms predominate. No systematic review has pooled the extensive single-centre Indian literature to provide an India-specific synthesis of the available evidence to inform infection prevention and antimicrobial stewardship. This systematic review and meta-analysis followed PRISMA 2020 (PROSPERO CRD420261512214). MEDLINE (via PubMed), Embase, Scopus and Web of Science were searched to 15 September 2026 for studies (2016–2026) reporting VAP etiology and susceptibility in Indian ICUs, predominantly in adults (≥ 18 years); three further studies came from supplementary sources. Risk of bias was assessed with the Joanna Briggs Institute prevalence checklist. Proportions were pooled with a binomial–normal generalised linear mixed model (GLMM; primary) and a Freeman–Tukey double-arcsine model (sensitivity), with prediction intervals, subgroup, meta-regression, leave-one-out, trim-and-fill and age-restricted analyses; certainty was rated with a GRADE-informed approach. Twenty-six studies were included; 22 (4,386 isolates) contributed isolate counts. Pooled prevalence among VAP isolates (genus- and species-level reports combined) was 28.5% (95% CI 24.4–33.0) for Acinetobacter (21 studies, 2,285 isolates; I²=75%), 25.5% (21.5–29.9) for Klebsiella (22 studies, 4,386 isolates; I²=81%) and 19.9% (18.8–21.1) for Pseudomonas (22 studies, 4,386 isolates; I²=34%); prediction intervals were wide (Acinetobacter 14.2–49.1%; Klebsiella 11.5–47.4%). Klebsiella prevalence was higher in studies from 2020 onwards (34.6% vs. 21.7%; p = 0.001), a between-study comparison confounded by hospital type and region. Results were stable across model, leave-one-out, age-restricted and species-versus-genus analyses; the Klebsiella small-study signal (Egger p = 0.047) depended on one large study. Multidrug resistance was 74.4% (58.4–85.7) (6 studies; I²=95%) and carbapenem resistance in Acinetobacter 63.9% (34.8–85.3) (8 studies; I²=96%; range 52.6–78.8% across sensitivity analyses), both of very low certainty. Colistin resistance ranged from 0% to 15.3% across four studies (7.1% in the largest after excluding intrinsically resistant genera); pooled values were model-dependent and not interpretable as a national rate. Acinetobacter and Klebsiella are the leading and near-equally frequent causes of VAP in Indian ICUs, with high but highly heterogeneous multidrug and carbapenem resistance. Pooled figures summarise single-centre studies and are not uniform national rates; marked between-hospital variation supports unit-level antibiogram benchmarking and infection-prevention bundle compliance rather than uniform empirical antibiotic escalation.