OBJECTIVES: This study aims to estimate the excess annual mortality from nosocomial infection caused by multi-drug resistant (MDR) bacteria in Thailand.

METHODS: A retrospective case cohort study was conducted at Ramathibodi Hospital, Bangkok, one of the largest tertiary care hospitals in Thailand, between January 1^st, 2008 and December 31^st, 2012. A multivariate Cox proportional hazard regression model was used to estimate the excess mortality incidence rate of antimicrobial resistance (AMR) relative to those patients without resistance (non-AMR) after controlling for an array of potential confounders including the Charlson co-morbidity score.

RESULTS: The overwhelming majority of patients with nosocomial infection (73.80%) were MDR cases over the study period, 2008 – 2012. Patients in the AMR and non-AMR groups had similar comorbidities, severity of baseline diseases, and sites of infection. Compared to patients in the non-AMR group, the mortality incidence rate for patients in the AMR group (susceptibility group = 1) who had acquired A, E, P. aeruginosa, and Swere 92%, 25%, 60%, and 84% higher, respectively, but 16% lower for patient who had acquired Kpneumoniae CONCLUSIONS: In the complete absence of MDR bacteria, this study estimated that annually, in Thailand, there would be 111,295 fewer AMR cases, 48,258 fewer deaths and 123,000 fewer hospitalization days.