Undetected carriage explains apparent Staphylococcus aureus acquisition in a non­outbreak healthcare setting

Undetected carriage explains apparent Staphylococcus aureus acquisition in a non-outbreak healthcare setting. Study and Abstract Objectives: Previous studies have been unable to identify patient or staff reservoirs for the majority of the nosocomial S. aureus acquisitions which occur in the presence of good infection control practice. We set out to establish the extent to which undetected pre-existing carriage explains apparent nosocomial S. aureus acquisition. Methods: Over two years elective cardiothoracic admissions were screened for S. aureus carriage before and during hospital admission. Routine screening (nose/groin/wound sampling), was supplemented by sampling additional body sites (axilla/throat/rectum) and culture-based methods optimised to detect fastidious phenotypes (small colony variants, cell wall deficient variants) and molecular identification by PCR. Results: 35% of participants (53/151) were S. aureus carriers according to routine pre-healthcare screening; increasing to 42% (63/151) when additional body sites and enhanced cultures were employed. 71% (5/7) of apparent acquisitions were explained by pre-existing carriage using augmented measures. Enhanced culture identified a minority of colonised individuals (3/151 including 1 MRSA carrier) who were undetected by routine and additional screening cultures. 4/14 (29%) participants who became culture-negative during admission had S. aureus genomic material detected at discharge.` Conclusions: Conventional sampling under-estimates carriage of S. aureus and this explains the majority of apparent S. aureus acquisitions among elective cardiothoracic patients.


Public and patient involvement
The study was designed and developed with input from clinical stakeholders within the cardiothoracic surgery department at Brighton and Sussex University Hospital NHS Trust and without patient or public involvement.

Declaration of Interest
All authors have completed the Unified Competing Interest form (available on request from the corresponding author) and declare: no support from any organization for the submitted work; no authors received personal fees in the past three years outside the submitted work; no other relationships or activities that could appear to have influenced the submitted work.

Background
Staphylococcus aureus is a common commensal of the human nose and throat but also a major human pathogen. In the UK, healthcare-associated methicillin-resistant S. aureus (MRSA) has waned but community-onset S. aureus disease and nosocomial methicillinsusceptible S. aureus (MSSA) infections remain important in the UK and globally. S. aureus is the most commonly isolated single pathogen associated with nosocomial infection. 1  Carriage of S. aureus is common and person-to-person transmission occurs during close contact in hospitals and the community. Acquisition of new strains is associated with greatest risk of invasive disease. 4 In hospitals efforts are made to decolonise MRSA carriers to reduce infection risk and prevent person-to-person transmission. We have recently undertaken detailed studies of S. aureus transmission in critical care. Surprisingly, neither patients 5 nor healthcare workers 6 could be implicated as sources for the majority of acquisitions.
Conventional screening methods may fail to detect S. aureus carriage (i) at un-sampled sites such as the gastrointestinal tract 7 8 , (ii) present below the threshold for detection in particular if within tissues or cells 9 and (iii) as fastidious growth phenotypes (small colony variants (SCVs) and cell wall deficient variants (CWD)). 10 11 These concealed carriage states at admission may mean some patients falsely appear to acquire S. aureus while in hospital. In this study we aim to apply enhanced detection methods in order to describe patterns of S. aureus carriage in patients before and during healthcare admission and to investigate the possibility that undetected carriage might explain instances of apparent nosocomial acquisition.  Figure 1).

Sampling
Patient samples were taken before and during hospital stay, using augmented sampling and culture to maximise recovery of isolates and apply molecular techniques (16S rRNA PCR) to categorically identify new acquisition events in patients with pre-existing S. aureus colonisation and carriage loss with persisting concealed carriage. Table 1 depicts participant sampling for S. aureus. All elective cardiothoracic patients are routinely screened for MRSA carriage (nose and groin plus any wound) at pre-assessment and admission to the ward. All routine screens were retrieved and cultured for all types of S. aureus. In order to determine S. aureus carriage at additional sites enhanced screening was performed of throat, axilla, and rectum prior to healthcare exposure. To determine the extent of carrying cryptic forms such as SCV and CWD, which can reside outside and within human epithelial cells, additional nose and throat samples were taken to capture epithelial cells as well as extra-cellular bacteria 12 .

Ethics
Ethical approval for this study was gained from London-Harrow Research Ethics Committee (16/LO/2111).

Statistical analysis
Data were analysed using STATA 15.0. Associations between carriage profiles and the variables described were univariately assessed using the Fisher's exact test for categorical data and the Kruskal-Wallis test for nonparametric continuous and categorical variables.
Probability values of ≤0.05 were considered significant.
86/151 (58%) participants were electively admitted to hospital during the study period; 24 had a single further screen performed, 45 had two and 17 three further screens. 65 participants were not admitted as their procedures were cancelled or postponed outside the study period. No differences in patient demographics, co-morbidities or pre-healthcare sampling results were observed between those participants who were serially screened and those who only received pre-healthcare samples (Supplementary

Acquisitions
Seven acquisitions were identified where participants were culture-negative on pre-admission routine screening who subsequent sampling during admission was culture-positive for S. aureus. Of these, two yielded S. aureus from augmented screening of pre-healthcare samples; one through additional throat sampling (Patient 20 in Figure 3C) and one through enhanced culture (Patient 16 in Figure 3A). Culture-negative samples from the remaining five (Participants 1-5 in Figure 3C) underwent analysis by 16S RNA PCR; pre-admission samples from three participants had S. aureus detected. Overall, of seven acquisitions detected by routine sampling five had evidence of preceding undetected carriage.

Carriage Loss
14 participants were culture-positive on at least one routine pre-healthcare screening sample and were culture-negative on final sample taken during admission, which met our culturebased definition of carriage loss (Participants 6-19 in Figure 3C). Of these nine had S. aureus carriage detected by routine screening and five by augmented sampling (two by throat alone, one by throat and rectum, one by axilla and rectum, and one by enhanced culture). Culturenegative samples from these participants were analysed by 16S RNA PCR; four had S. aureus detected on discharge sample (Patients 6, 8, 18 & 19 in Figure 3C).

Antibiotic exposure
81/86 (94%) serially screened participants received antibiotics prior to recruitment. Patient characteristics between three carriage profiles identified by augmented screening did not significantly differ from those from routine screening (Supplementary Table 2). When comparing previous antibiotic exposure, patients exhibiting continuous detection (identified by augmented screening) tended to have received antibiotics in the 2 months preceding the study and those with no detected carriage received antibiotics over 6 months prior to the study. No trends with chronic skin wounds were observed.
Overall we identified:  seven apparent acquisitions among 86 serially screened participants among which five had S. aureus identified prior to healthcare exposure (one by additional screening, one by enhanced culture and three by 16S RNA PCR), indicating preexisting carriage.  10 instances of S. aureus carriage prior to admission which were missed by routine screening including one of MRSA carriage: eight by additional site sampling and two by enhanced culture methods.  14 of 89 participants screened throughout admission met our culture-based definition of S. aureus carriage loss and, of these, four had molecular evidence of S. aureus identified at discharge.

Discussion
The aims of our study were to describe S. aureus carriage patterns in patients before and during healthcare admission and evaluate whether undetected carriage explains apparent S.
aureus acquisitions during healthcare exposure. Our work reveals five key findings: The true burden of healthcare-acquired S. aureus in unknown. We previously report that the incidence of culture-based S. aureus acquisition in hospitalised patients in critical care is 6-8%. 5 6 Studies are confounded by differences between carriage and infection, focus on MRSA (in turn dependent on local prevalence rates), variation in definitions, sensitivity of screening tests, use of culture to determine presence or absence of organism, focus on extracellular organisms. It is plausible that we are underestimating true incidence of S. aureus acquisition. We conducted a study of patients in elective cardiothoracic surgery setting and identified 8% (7/86) patients who were  carriage. These are inherently difficult to culture organisms and that itself lies a limitation -just because we didn't find them doesn't mean they weren't there.

Impact of augmented screening
We employed methodologies specifically to aid identification of intracellular S. aureus detection (Supplementary Materials). Whilst there were three instances where S. aureus was cultured from samples optimised to detect intracellular cryptic variant and not by samples optimised to detect extracellular cryptic variants, there were no cases of carriage detected by intracellular carriage alone. Although swabbing the anterior nares with cotton-tip swabs has been shown to effectively capture nasal epithelial cells 12 it is plausible that cellular capture from deep or other sites may contribute towards undetected carriage. Further work is required.

Role of molecular testing
Selected samples assessed using 16S RNA PCR revealed seven instances of culture-negative samples containing genomic signatures of S. aureus (three refuting to apparent acquisitions and four refuting apparent losses). As a pilot study this work was not designed to interrogate all samples with molecular testing and in turn there will be selection bias. Our work does suggest that molecular testing may play a role in determining undetected carriage and further investigation is required.

Carriage profiles
We investigated in detail 86 patients who were serially screened during the course of a healthcare admission. Serial culture-based methods of nose and groin sampling revealed 20% were culture-positive on every sample, 58% culture-negative on every sample, and 22% intermittently culturing S. aureus. This is broadly consistent with traditional culture-based carriage profiles. 19   screens taken either pre-admission, post-operatively, weekly during admission or discharge. 5 Colours represent culture results: red = MRSA, blue = MSSA, white = culture-negative. 6 Samples positive for S. aureus DNA (*) are highlighted.  Table 1. Participant sampling during the study. Sampling (x) comprises swabs taken as part 15 of routine clinical practice (Routine) and swabs taken as part of the study including additional 16 site sampling (Additional) and enhanced culture (Enhanced) for small colony variants and