An Investigation of the Prevalence, Population Structure and Relatedness of Environmental, Healthcare Worker and Patient Staphylococcus aureus Isolates Recovered from Multi-bed Patient Rooms in an Irish Teaching hospital by Whole-Genome Sequencing

Abstract

The putative role of the air in Staphylococcus aureus transmission in multi-bed ward areas of hospitals, where the majority of in-patients are cared for, has not yet been elucidated. The work described in this thesis included extensive environmental sampling and concurrent in-depth observation of the clinical activities in two differently ventilated multi-bed wards over four five-hour study days (with two days spent on each ward). Patients and healthcare workers (HCWs) occupying the study area were invited to volunteer to be screened for S. aureus (including methicillin-resistant S. aureus, or MRSA) carriage. Isolates recovered were investigated by whole-genome sequencing (WGS). This study aimed (1) to investigate the prevalence of environmental S. aureus under active clinical conditions in the two multi-bed hospital wards, (2) to observe the impact of routine activities and surface hygiene on S. aureus recovery and (3) to identify the population structure, molecular characteristics and relatedness of isolates recovered. This study was set in two surgical wards in an 820-bed tertiary referral hospital in Dublin, Ireland. The wards were chosen as they share similar patient caseloads and clinical activities, so are well-matched in most respects with the exception of how they are ventilated. Ward A has a HEPA-filtered ventilation system, and Ward B is ventilated naturally, via windows. The first part of this study involved environmental and participant sampling and surface hygiene assessments alongside observation of ward activity. Screening of patients included sampling of the nares and oropharynx. Surface sampling was undertaken by taking contact plate (ColorexTM Staph Aureus chromogenic medium (E&O Laboratories Ltd., Scotland)) samples of sites observed to be frequently touched by the hands of HCWs during the study. The air was sampled actively (using a surface air sampler) and passively (by use of settle plates). All samples were cultured using ColorexTM Staph Aureus chromogenic agar with putative S. aureus colonies definitively identified using the tube coagulase test and by latex agglutination with the Pastorex Staph Plus latex agglutination kit (Bio-Rad, France). All S. aureus isolates were tested for growth on chromogenic agar selective for MRSA (ColorexTM MRSA, E&O Laboratories Ltd., Scotland). Surface contamination was assessed on sampled sites by visual assessment (currently advised in Irish national guidance for the audit of cleaning in hospitals) and surface adenosine triphosphate (ATP) was quantified (using a SystemSure II luminometer, (Hygiena Int. Ltd., UK)). Variables that have previously been associated with recovery of airborne S. aureus were recorded, including bed-making, movement of privacy curtains, occupant density and clinical activities. A total of 375 samples were taken including 117 active air samples, 100 settle plates, 120 contact plates and nasal and oral samples from 19 participants (five patients and 14 HCWs). Six HCWs and one patient were found to be colonised with S. aureus. Overall, almost one third (32.2%) of air samples yielded S. aureus. Staphylococcus aureus was recovered from settle plates on six occasions (6%), representing contamination of surfaces via the air to a frequently touched site by patients- the overbed table and from active air samples on 38 occasions. Frequently touched surfaces were observed to be contaminated with S. aureus (11/120; 9.2%). All isolates (n=65) were initially identified as methicillin-susceptible S. aureus (MSSA) as none grew on MRSA selective agar. Lower ATP values were observed on surfaces that appeared visually clean, but this difference was not significant. Seven surfaces that passed a visual hygiene assessment harboured S. aureus. Privacy curtain movement was significantly associated with patients washing and dressing (p <000.1), and these actions were significantly associated with airborne S. aureus recovery (p <0.05). Antimicrobial susceptibility testing identified seven isolates that exhibited resistance to cefoxitin. These isolates failed to grow on MRSA selective agar but were confirmed as MRSA by the MRSA GeneXpert Assay (Cepheid, USA). All seven isolates were susceptible to oxacillin and were designated oxacillin-susceptible MRSA (OS-MRSA). Contamination of the near-patient environment did not appear to have resulted from shedding by the bed occupant, as on all occasions when S. aureus was recovered from a settle or contact plate, the individual occupying the bedspace was not known to be colonised with S. aureus. The detection of OS-MRSA isolates among the study collection was unexpected. These were recovered from environmental samples (six active air samples and one contact plate used to sample a nightstand). Irish guidelines for the control of MRSA in hospitals recommend the use of chromogenic agar for the culture of screening samples, which would not have identified these isolates as MRSA. In the second part of this study, 62 recovered S. aureus isolates were investigated by short-read WGS using the Illumina MiSeq (Illumina, the Netherlands) platform. Isolates were assigned to 13 multilocus sequence types (STs) within 11 clonal complexes (CCs). The most frequently identified CC was CC1 (22/62, 35.5%), comprising ST1 (n=19) and ST109 (n=3) isolates. Isolates were considered related if they exhibited &#8804; 24 allelic differences following whole genome (wg) MLST analysis. Using this criterion, 10 related isolate groups involving 46 isolates were identified. Each of these groups involved one or more isolates recovered from an active air sample. In-silico genotyping for virulence and resistance genes identified a wild type mecA gene in all seven OS-MRSA isolates identified in the first part of the study, and one additional OS-MRSA isolate originally identified as an MSSA, cultured from a settle plate and which was phenotypically susceptible to both oxacillin and cefoxitin and did not grow on MRSA chromogenic agar. All eight OS-MRSA isolates were assigned to ST8 and staphylococcal chromosome cassette mec (SCCmec) type V (5C2&5) and all were closely related by wgMLST, exhibiting 0-2 allelic differences. OS-MRSA have previously been associated with mutations in the femXAB operon, which encodes proteins critical for peptidoglycan synthesis in MRSA when native penicillin binding proteins are inhibited by antibiotic agents. Mutations in femA and femX were identified in the eight OS-MRSA isolates. When compared to an MRSA reference strain, these mutations resulted in two amino acid substitutions within FemA (Y195F and E234D) and three within FemX (N18H, 151V and E261K). Similar mutations have been reported previously in investigations of OS-MRSA isolates and are thought to inhibit synthesis of a pentaglycine cross-linking structure which is integral to cell wall function in S. aureus and may contribute to oxacillin susceptibility. This study investigated environmental contamination with S. aureus in two multi-bed surgical wards in a large Dublin teaching hospital. Whole-genome sequencing of isolates recovered revealed that the majority of recovered isolates were considered related to at least one other isolate. A number of isolates were unrelated to others, potentially reflecting spurious shedding by room occupants. A cluster of OS-MRSA was revealed which was not detected using chromogenic MRSA selective agar. Prevalence of OS-MRSA in Ireland is currently not known and should be established. Finally, the results of this study, and wgMLST identifying relatedness between the air and isolates obtained from different sources indicate that the air may play an underappreciated role in contaminating near-patient sites in multi-bed hospital rooms with pathogenic bacteria.