BACKGROUND: Prevalence surveys in U.S. hospitals showed that on any given day, 1 of 25 patients had a health care-associated infection in 2011, as compared with 1 of 31 patients in 2015. We repeated the survey in 2023 to assess changes in the prevalence of such infections. METHODS: Using established methods, 10 Emerging Infections Program (EIP) sites recruited up to 25 hospitals each. Hospitals selected a survey day between May 1 and September 30, 2023. EIP staff reviewed medical records of randomly selected patients to identify health care-associated infections using National Healthcare Safety Network definitions. We described patient and hospital characteristics, compared the prevalence of health care-associated infections in 2023 and 2015, and estimated the national burden of health care-associated infections in 2023. RESULTS: In 2023, of 13,653 patients in 218 hospitals, 355 (2.6%; 95% confidence interval CI, 2.3 to 2.9) had at least one health care-associated infection, as compared with 394 of 12,299 patients (3.2%; 95% CI, 2.9 to 3.5) in 199 hospitals in 2015. Approximately 60% of the health care-associated infections were not associated with devices or procedures. Among 151 hospitals in both surveys, patients were less likely to have a health care-associated infection in 2023 than in 2015, after adjustment for other factors (risk ratio, 0.73; 95% CI, 0.63 to 0.85). We estimated that there were 518,000 health care-associated infections (95% CI, 494,500 to 542,000) in U.S. hospitals in 2023. CONCLUSIONS: The prevalence of health care-associated infections was lower in 2023 than in 2015, with 1 of 38 patients on any given day having such an infection; however, the burden of health care-associated infections in U.S. hospitals continued to be high. A majority of the health care-associated infections that were identified were not associated with a device or procedure. (Funded by the Emerging Infections Program cooperative agreement from the Centers for Disease Control and Prevention.).
Chea et al. (Wed,) studied this question.