Clinical and economic burden of bloodstream infections caused by meticillin-resistant and meticillin-susceptible Staphylococcus aureus in Chinese tertiary hospitals from 2010 to 2017: a multicentre, retrospective cohort study

Abstract

BackgroundGlobally, antimicrobial resistance has become one of the most complex public health challenges. Accurate estimates of health and economic burden of antimicrobial resistance has not been available in China because of the absence of representative empirical data. We therefore aimed to estimate the effect of antimicrobial resistance attributable to Staphylococcus aureus bloodstream infections in Chinese hospitals, and to compare the effect of resistant strains with susceptible strains. MethodsWe did a multicentre, retrospective cohort study of 11 tertiary hospitals in China. The cohort consisted of all inpatients with S aureus bloodstream infections from the sample hospitals, and we randomly included individuals free of S aureus bloodstream infections. We stratified the patients into two categories of S aureus bloodstream infections: meticillin-susceptible S aureus (MSSA) or meticillin-resistant S aureus (MRSA). The primary outcome was the effect of antimicrobial resistance on excess length of stay in the hospital, hospital mortality, and cost of treatment. We adopted a multistate model to estimate the expected length of stay. Multi-level Cox proportional hazards models were adopted to compare the hazard of reaching hospital death, discharge alive, and the combined end-of-stay endpoint. Two bed-day values were adopted when estimating the cost of treatment. This study was approved by Peking University Health Science Center Ethics Committee, with a waiver for individual informed consent. FindingsFrom Jan 1, 2010, to June 31, 2017, we assessed 861 inpatients with S aureus bloodstream infections and 10 000 individuals free of S aureus bloodstream infections. Of the 861 admissions with an S aureus blood stream infection detected, 324 (38%) were due to MRSA and 537 (62%) were due to MSSA. Bloodstream infections caused by MRSA prolonged the length of stay in the hospital by 13·4 days (95% CI 11·1–16·3), and MSSA bloodstream infections prolonged the length of stay by 11·9 days (95% CI 9·7–14·1). However, bloodstream infections caused by MRSA did not increase the hazard of death (adjusted hazard ratio [aHR] 0·31, 95% CI 0·09–1·10). MSSA bloodstream infections also did not increase the hazard of death (aHR 0·59, 95% CI 0·14–2·54). Compared with MSSA, MRSA did not significantly increase the hazard of death (aHR 0·44, 95% CI 0·19–1·03), excess length of stay (1·5 days, 95% CI −0·4 to 2·8), and cost of treatment. Although the cost of MRSA treatment was higher per infection than MSSA treatment, the annual cost of MSSA was higher overall. InterpretationBy accounting for time-dependent bias and important confounders, this study added to the existing literature about the effect of antimicrobial resistance in China. Current results of the study implicate the importance of surveillance and infection control policies that target infections rather than resistance. FundingChina Medical Board (17-270).