Zero tolerance for healthcare-associated MRSA bacteraemia: is it realistic?

Abstract

Background: The term 'zero tolerance' has recently been applied to healthcare-associated infections, implying that such events are always preventable. This may not be the case for healthcare-associated infections such as methicillin-resistant Staphylococcus aureus (MRSA) bacteraemia.

Methods: We combined information from an epidemiological investigation and bacterial whole-genome sequencing to evaluate a cluster of five MRSA bacteraemia episodes in four patients in a specialist hepatology unit.

Results: The five MRSA bacteraemia isolates were highly related by multilocus sequence type (ST) (four isolates were ST22 and one isolate was a single-locus variant, ST2046). Whole-genome sequencing demonstrated unequivocally that the bacteraemia cases were unrelated. Placing the MRSA bacteraemia isolates within a local and global phylogenetic tree of MRSA ST22 genomes demonstrated that the five bacteraemia isolates were highly diverse. This was consistent with the acquisition and importation of MRSA from the wider referral network. Analysis of MRSA carriage and disease in patients within the hepatology service demonstrated a higher risk of both initial MRSA acquisition compared with the nephrology service and a higher risk of progression from MRSA carriage to bacteraemia, compared with patients in nephrology or geriatric services. A root cause analysis failed to reveal any mechanism by which three of five MRSA bacteraemia episodes could have been prevented.

Conclusions: This study illustrates the complex nature of MRSA carriage and bacteraemia in patients in a specialized hepatology unit. Despite numerous ongoing interventions to prevent MRSA bacteraemia in healthcare settings, these are unlikely to result in a zero incidence in referral centres that treat highly complex patients.

Keywords: methicillin-resistant Staphylococcus aureus; outbreak; whole-genome sequencing.

Figure 1.

Epidemiology and bacterial phylogeny of MRSA bacteraemia cases. (a) Epidemiological map of four patients (P1–P4) with MRSA bacteraemia. Each row represents a single patient and the coloured blocks for each represent the time spent as an inpatient on the hepatology ward, other wards at the CUH or at other hospitals. The length of each box denotes the duration of admission and the scale bar represents days after the date of the first bacteraemia. (b) Phylogenetic tree based on whole-genome sequencing of nine carriage and bacteraemia isolates of MRSA from the four patients shown in (a). The numbers shown represent SNP differences between isolates.

Figure 2.

Phylogenetic context of MRSA isolates from hepatology patients. Phylogenetic tree based on the whole-genome sequence of nine MRSA isolates from four bacteraemia patients (P1–P4, different colours) and 42 MRSA isolates from patients with recent MRSA acquisition admitted to the hepatology ward (P5–P47, coloured grey, left-hand tree). The same isolates were included in a phylogenetic tree with the 193 ST22 MRSA isolates from a global collection (right-hand tree). Grey lines linking the two trees indicate the position of the hepatology isolates in the global phylogeny. The three clusters (a, b and c) represent suspected MRSA transmission events, two of which were corroborated by epidemiological evidence.