Monitoring the spread of meticillin-resistant Staphylococcus aureus in The Netherlands from a reference laboratory perspective

Abstract

Background: In The Netherlands, efforts to control meticillin-resistant Staphylococcus aureus (MRSA) in hospitals have been largely successful due to stringent screening of patients on admission and isolation of those that fall into defined risk categories. However, Dutch hospitals are not free of MRSA, and a considerable number of cases are found that do not belong to any of the risk categories. Some of these may be due to undetected nosocomial transmission, whereas others may be introduced from unknown reservoirs.

Aim: Identifying multi-institutional clusters of MRSA isolates to estimate the contribution of potential unobserved reservoirs in The Netherlands.

Methods: We applied a clustering algorithm that combines time, place, and genetics to routine data available for all MRSA isolates submitted to the Dutch Staphylococcal Reference Laboratory between 2008 and 2011 in order to map the geo-temporal distribution of MRSA clonal lineages in The Netherlands.

Findings: Of the 2966 isolates lacking obvious risk factors, 579 were part of geo-temporal clusters, whereas 2387 were classified as MRSA of unknown origin (MUOs). We also observed marked differences in the proportion of isolates that belonged to geo-temporal clusters between specific multi-locus variable number of tandem repeat analysis (MLVA) clonal complexes, indicating lineage-specific transmissibility. The majority of clustered isolates (74%) were present in multi-institutional clusters.

Conclusion: The frequency of MRSA of unknown origin among patients lacking obvious risk factors is an indication of a largely undefined extra-institutional but genetically highly diverse reservoir. Efforts to understand the emergence and spread of high-risk clones require the pooling of routine epidemiological information and typing data into central databases.

Keywords: Algorithm; Epidemiology; Meticillin-resistant Staphylococcus aureus; Surveillance.

Figure 1

Isolates included in the study. Of the 14,042 isolates submitted to the Dutch National Institute for Public Health and the Environment, 5090 isolates were excluded (2864 duplicates and 2226 lacking information). A further 2657 MC398 isolates were excluded to reduce computational time, yielding a total of 6295 isolates used in the integrated algorithm.

Figure 2

Data sources for cluster detection in meticillin-resistant Staphylococcus aureus isolates in The Netherlands 2008 to 2010, MLVA (multi-locus variable number of tandem repeat analysis) complex 8. (A) The postal code of the patients' home addresses, (B) the MLVA complex, here shown as a minimum spanning tree, with each circle showing a single MLVA type (size increasing with number of isolates), and (C) time at which the sample was isolated, here shown in weekly aggregate numbers. Light grey denotes isolates outside clusters; blue, red, green, and black show the isolates present in the four largest clusters. The other clustered isolates are shown in dark grey. Whereas each of the sources shows weak clustering pattern, the three sources, taken together, reveal considerably stronger clustering pattern.

Figure 3

The proportion of isolates that is part of a cluster according to the algorithm. Dots show the point estimate; box plots (median, interquartile range, 95% interval) include jack-knife estimates for 90% of the isolates. Frequently occurring hospital-associated meticillin-resistant Staphylococcus aureus (MRSA) strains (MC5, MC45, MC22) show higher proportions of clustering than community (MC8)- or livestock (MC398)-associated MRSA.

Figure 4

The odds of belonging to a cluster of isolates differ considerably between epidemiological risk categories. The results of the clustering algorithm overlap with expectations based on risk group. The isolates lacking obvious risk factors more often fall outside the clusters, whereas isolates obtained as part of contact tracing during outbreaks were more often found within clusters. This is a healthcare worker (HCW)-specific risk group. OR, odds ratio; MRSA, meticillin-resistant Staphylococcus aureus.

Figure 5

The proportion of isolates positive for Panton–Valentine Leukocidin (PVL) and the proportion of community-acquired (CA) isolates are higher among cases outside the clusters. Black bars show clustered isolates; grey bars show non-clustered isolates. (A) The proportion of PVL-positive cases for the largest MLVA (multi-locus variable number of tandem repeat analysis) clonal complexes (MCs), and (B) for isolates obtained as part of contact tracing during outbreaks, those isolated within 48 h after admission (CA), and without apparent risk factors. (C) In absolute number of isolates, most PVL-positive isolates are found among isolates without apparent risk factors outside clusters. (D) The proportion of CA meticillin-resistant Staphylococcus aureus (MRSA) for the largest MCs, and (E) for isolates obtained as part of contact tracing during outbreaks and isolates without apparent risk factors. (F) In absolute number of isolates, most CA-MRSA isolates are found outside clusters.