Decreased Incidence of Methicillin-Resistant Staphylococcus aureus Bacteremia in Intensive Care Units: a 10-Year Clinical, Microbiological, and Genotypic Analysis in a Tertiary Hospital

Abstract

There are limited long-term data on the trends in incidence and characteristics of methicillin-resistant Staphylococcus aureus (MRSA) bacteremia (MRSAB) in intensive care units (ICUs) in which infection control measures have been adopted. We evaluated the trend of incidence and changes in characteristics of MRSA bacteremia in ICUs at a tertiary-care hospital over 10 years using prospective cohort data. ICU-acquired bacteremia was defined as S. aureus bacteremia (SAB) that occurred 48 h or more after ICU admission. MRSA isolates were collected and subjected to microbiological and genotypic analyses. A total of 529 SAB episodes were identified among 367,175 ICU patients. Of these episodes, 288 (54.4%) were ICU acquired, 238 (82.6%) of which were MRSAB. The incidence density of ICU-acquired MRSAB decreased from 1.32 per 1,000 patient-days to 0.19 per 1,000 patient-days (a decrease of 20% annually; P < 0.001 for trend), whereas that of non-ICU-acquired MRSAB fluctuated and did not decrease significantly. The decline in ICU-acquired MRSAB was due to lower catheter-related infection and less pneumonia. Rates of persistent bacteremia and 12-week mortality also fell significantly. A total of 183 isolates were collected from 238 ICU-acquired MRSAB cases. There were no significant changes in the geometric means of vancomycin MICs, vancomycin heteroresistance, or the sequence types of MRSA isolates over time. Chlorhexidine MICs decreased (P < 0.001 for trend) in association with a decline in frequency of the qacA or qacB gene that was related to reductions in specific spa types. The incidence of MRSAB in ICUs has decreased dramatically over time, but most of the microbiological and genotypic characteristics of MRSA isolates have not changed.

Keywords: bacteremia; genotype; incidence; intensive care unit; methicillin-resistant Staphylococcus aureus; vancomycin MIC.

FIG 1

Incidence density of S. aureus bacteremia. (A) ICU-acquired S. aureus bacteremia (bacteremia >48 h after ICU admission) and (B) non-ICU-acquired S. aureus bacteremia (bacteremia within 48 h of ICU admission).

FIG 2

Incidence density of ICU-acquired MRSA bacteremia according to site of infection.

FIG 3

Changes in microbiological and genotypic characteristics of MRSA blood isolates in ICUs. (A) Geometric means of vancomycin MICs (P = 0.62 for trend) and frequencies of hVISA phenotype (P = 0.21 for trend), (B) MLST (P = 0.07 for trend), and (C) geometric means of chlorhexidine MICs (P < 0.001 for trend) and frequencies of qacA or qacB (P < 0.001 for trend) and smr (P = 0.01 for trend).