Cost-effectiveness of screening, decolonisation and isolation strategies for carbapenem-resistant Enterobacterales and methicillin-resistant Staphylococcus aureus infections in hospitals: a sex-stratified mathematical modelling study

Background: Methicillin-resistant Staphylococcus aureus (MRSA) and carbapenem-resistant Enterobacterales (CRE) impose the greatest burden among critical bacterial pathogens. Evidence for sex differences among antibiotic resistant bacterial infections is increasing but a focus on policy implications is needed. We assessed impact of CRE/MRSA on excess length of hospital stay, intensive care unit admission, and mortality by sex from a retrospective cohort study (n = 873) of patients in three Chilean hospitals, 2018–2021. Methods: We used inverse-probability weighting combined with descriptive, logistic, and competing-risks analyses. We developed a sex-stratified deterministic compartmental model to analyse hospital transmission dynamics and the cost-effectiveness of nine interventions. We compared interventions based on the incremental cost-effectiveness ratio (ICER) per quality-adjusted life year (QALY) gained and estimated net benefits. Findings: The adjusted odds of women acquiring CRE and MRSA were 0.44 (0.28–0.70; p = 0.0013) and 0.73 (95% CI = 0.48–1.01; p = 0.050), respectively. Competing-risk models indicated higher mortality rates among women, compared to men. Mathematical model projections showed that pre-emptive isolation across all newly admitted high-risk men was the most cost-effective intervention (ICER = $1366/QALY and $1083/QALY for CRE and MRSA, respectively). Chromogenic agar coupled with MRSA decolonisation was the second most cost-effective intervention ($2099/QALY), followed by screening plus isolation or pre-emptive isolation strategies (ICER ranged between $2411/QALY and $4216/QALY across CRE and MRSA models). Probabilistic sensitivity analysis showed that strategies were ICER < willingness-to-pay in 80% of simulations, except for testing plus digestive decolonisation for CRE. At a 20% national hospital coverage at least $12.2 million could be saved. Interpretation: Our model suggests that targeted infection control strategies would effectively address rising CRE and MRSA infections. Maximising health-economic gains may be achieved by focusing on control measures for men as primary drivers for transmission, thereby reducing the disproportionate disease burden borne by women. Funding: Agencia Nacional de Investigación y Desarrollo ANID, Chile.