Clinical impact of dual carbapenemase co-production on 28-day mortality in ICU patients with carbapenem-resistant Klebsiella pneumoniae: a prospective study from South India

Abstract

Abstract Background Carbapenem-resistant Klebsiella pneumoniae (CRKP) is a critical priority pathogen associated with substantial morbidity and mortality, particularly in intensive care units (ICUs). Although carbapenemase co-production is increasingly recognized, its clinical significance and relationship with mortality remain incompletely defined, particularly in South Asian ICU populations. Methods We conducted a prospective observational study among ICU patients with culture-confirmed CRKP infection in a tertiary care hospital in South India (January 2024-January 2025). Carbapenemase genes ( bla NDM , bla OXA−48−like , bla IMP , bla VIM , and bla KPC ) were detected using real-time polymerase chain reaction. Patients were stratified according to genotype ( bla NDM , bla OXA−48−like , co-producers, or non-producers) and the primary outcome was 28-day all-cause in-hospital mortality. Univariable and multivariable Firth penalized logistic regression were used to estimate unadjusted and adjusted odds ratios (OR), respectively. Because of the limited number of mortality events, the adjusted analysis was considered exploratory. Model diagnostics included assessment of multicollinearity and linearity in the logit. Results Among 106 patients with CRKP infection, 30 (28.3%) had carbapenemase co-producing isolates and 16 (15.1%) died within 28 days. Mortality differed significantly across genotypes, with the highest proportion observed among co-producer isolates (9/30, 30.0%). In univariable Firth regression analysis, mechanical ventilation (OR 12.47, 95% CI 3.83–52.03; p < 0.001), higher APACHE II score (OR 1.29, 95% CI 1.16–1.49; p < 0.001), septic shock (OR 25.11, 95% CI 7.19-100.65; p < 0.001), and co-producer status (OR 4.09, 95% CI 1.41–12.36; p = 0.010) were associated with 28-day all-cause in-hospital mortality. In the exploratory multivariable Firth penalized regression model, higher APACHE II score (aOR 1.24 95% CI 1.08–1.50; p = 0.001), septic shock (aOR 12.23 95% CI 2.18–93.77; p = 0.004), and carbapenemase co-production (aOR 7.50 95% CI 1.50-51.67; p = 0.014) were associated with 28-day all-cause in-hospital mortality. The adjusted OR for carbapenemase co-production was 7.50, but the CI was wide (1.50-51.67), indicating substantial uncertainty in the magnitude of the association. No problematic multicollinearity was identified (VIF range 1.12–1.28). The apparent AUC was 0.975; this estimate was not internally validated. Conclusion In this single-center ICU cohort, carbapenemase co-production was associated with higher odds of 28-day all-cause in-hospital mortality. The wide CI, limited number of deaths, possible residual confounding related to antimicrobial management, and lack of internal validation require cautious interpretation. Larger multicenter studies with detailed treatment and source-control data are needed to determine whether this association is reproducible. Clinical trial number Not applicable.