Introduction

Staphylococcus aureus is a leading pathogen causing hospital-acquired infections in children [1]. Approximately five percent of children admitted to pediatric intensive care units (PICUs) are colonized with methicillin-resistant Staphylococcus aureus (MRSA) [2]. Studies have proven that MRSA colonization increases the risk of subsequent invasive MRSA infection in both children [3] and adults [4–6]. Children who acquire MRSA after admission to the PICU are at a higher risk for invasive disease [3] and detection via PCR is sensitive to predict MRSA infection [7]. Colonization pressure within a critical care area is directly related to MRSA transmission in children. Popoola et al. showed that colonization pressure above 10% increases transmission of MRSA by 3-fold [8], leading to widespread screening and decolonization programs.

Coagulase-negative Staphylococci (CoNS), once thought to be largely commensal organisms, are now increasingly recognized as significant healthcare-associated pathogens, especially in patients with indwelling medical devices common in PICUs [9]. The rising prevalence of methicillin resistance among CoNS strains, which can be as high as 80%, limits the effectiveness of first-line antibiotics and complicates treatments [9]. This resistance contributes to increased morbidity, mortality, and healthcare costs [9], although CoNS has not yet been included in widespread screening programs. To understand the value of PICU infection control strategies, this study includes screening for both MRSA and methicillin resistant CoNS.

If the aim of PICU screening is to decrease transmission of hard-to-treat Staphylococcus within critical care area, then we hypothesize that there are two key components: a) determining if the prevalence of all methicillin-resistant Staphylococcus colonization pressure is high enough for transmission, and b) identifying individuals who may be at higher risk of colonization if exposed due to a lack of colonization resistance in their nasal microbiome. Risk factors for colonization include increased likelihood of exposures (contact with healthcare setting, high endemicity community) [10, 11], selective pressure to maintain (antibiotic use) [12], and colonization resistance from other microbes within the nasal community [13]. While exposure plays a major role in determining who will be colonized by multi-drug resistant organisms, host susceptibility to these organisms may be determined by other flora at the sites of colonization.

Colonization resistance, where the resident microbial community block an intruder’s establishment, plays a crucial role in nasal Staphylococcus ecology. The host’s microbial community can prevent colonization by outcompeting the intruder for resources and space; or through active mechanisms like producing antimicrobial compounds, or co-opting host immune responses. The interactions can become sophisticated evolutionary arms races, as seen when S. aureus develops catalase to neutralize H₂O₂ produced by S. pneumoniae or when S. epidermidis releases serine proteases to inhibit S. aureus [14]. While these specific antagonistic relationships are well-documented [15] the complex interactions that determine successful colonization remain incompletely understood. Understanding these mechanistic interactions between microbiota and methicillin-resistant Staphylococcus is essential for predicting colonization risk. Therefore, this study aims to investigate the prevalence of methicillin resistant Staphylococcus in the PICU of a high endemicity community and determine if nasal microbiome signatures can identify patients susceptible to transmission.

Methods

Results

Prevalence of methicillin-resistant Staphylococcus

Nasal MRSA colonization was found in 6.5% patients admitted to the pediatric intensive care unit from 2021–2021, lower than the 27–45% observed from 2000–2002 ²⁶. Despite the low MRSA prevalence, screening for MRCoNS revealed a 35% prevalence of nasal mecA+ colonization (MRSA and MRCoNS combined). Carriage of methicillin-resistant Staphylococcus did not differ between patients admitted to PICU for suspicion of infection and those admitted for post-operative observation (Table 1). In univariate analysis, patients with methicillin-resistant Staphylococcus were similar to those without in terms of age, sex, race, viral coinfection, asthma and antibiotic use. However, multivariate analysis showed carriage of methicillin-resistant Staphylococcus was independently associated with fever during PICU stay, with an adjusted odds ratio of 7.84 (95% confidence interval 1.95–31.45; p value of 0.004).

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Characteristics of patients admitted to LBCH PICU (Memphis, TN) and results of univariate analysis for risk factors for nasal colonization with either methicillin-resistant Staphylococcus aureus (MRSA) or methicillin-resistant coagulase-negative Staphylococci (MRCoNS).

https://doi.org/10.1371/journal.pone.0316460.t001

Nasal microbiome traits in colonized critically ill children

We characterized the nasal microbiota in critically ill children by analyzing admission nasal swabs using 16S ribosomal RNA gene sequencing. The composition of the nasal microbiota varied widely between individuals (Fig 1). Surprisingly, there was no link between low diversity of nasal microbiome and carriage of MRSA or MRCoNS (S1 Table). To identify unique taxa that influence colonization with methicillin-resistant Staphylococcus, we screened phylotypes that varied in distribution and abundance across multiple individuals [24] using a stringent criterion (FDR 0.01). These phylotypes selected to form ratios may originate from similar ecological niches within their communities.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Nasal microbiota diversity.

Diverse nasal microbiota on in 62 critically ill children with each microbiome a bar displayed as relative abundance; colors correspond to particular species or if not specified as other. Participants are ordered by age with heatmaps for methicillin-resistant Staphylococcus detection (mecA+, MRSA culture swab).

https://doi.org/10.1371/journal.pone.0316460.g001

After accounting for age, sex, race, and reason for PICU admission, we found that the ratio of anterior nasal mucosal microbes (Streptococcus agalactiae, Gleimia europaea and Moraxella sp.) or (Corynebacterium propinquum and Massilia consociata) to microbes more typical of deeper nasal mucosal sites (Haemophilus haemolyticus, Prevotella nanceiensis, and Streptococcus alactolyticus) could predict carriage with methicillin-resistant Staphylococcus (Table 2). The majority of identified nasal bacteria, including common commensals like Dolosigranulum pigrum and Neisseria cinerea and Pseudomonas sp., showed no significant association with methicillin-resistant Staphylococcus carriage (supplemental materials).

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Ratios of bacteria species found in the nasal microbiota of pediatric critically ill patients predicted carriage of methicillin-resistant Staphylococcus.

Ratios were transformed using log₁₀ before single and multivariate analysis to reduce skewness.

https://doi.org/10.1371/journal.pone.0316460.t002

Discussion

The primary aim of PICU screening is to mitigate transmission of methicillin-resistant Staphylococcus species, including MRSA and CoNS, which are challenging to treat in critical care units. We found that MRSA prevalence among PICU admits was 6.5%, lower than expected in a highly endemic community. However, when the overall prevalence for methicillin-resistant Staphylococcus (MRSA and CoNS) colonization was 35%, which may be sufficient to facilitate transmission. We hypothesized that the nasal microbiome could identify individuals at elevated risk of colonization who have insufficient colonization resistance to these Staphylococci. We discovered microbiome signatures predicting methicillin-resistant Staphylococcus colonization within this cohort, which differed from prior literature suggesting no predictive nasal microbiota patterns [26]. The predictive ratio included several notable species, such as Group B Streptococcus (S. agalactiae) which in most children is commensal but carries significant clinical implications due to antimicrobial resistance potential. Corynebacterium propinquum, typically a benign colonizer of anterior nares, can participate in bacterial interspecies interactions through production of bacteriocins. Moraxella species, while generally commensal, may enhance susceptibility to viral infections and can harbor beta-lactam resistance genes. The deeper nasal species identified, including Haemophilus haemolyticus and Prevotella nanceiensis, are typically non-pathogenic anaerobes that may contribute to colonization resistance through niche competition and metabolic interactions. These species roles in the nasal ecosystem could inform future strategies for preventing methicillin-resistant Staphylococcus colonization.

This study has several limitations. The sample size was restricted due to changes in PICU census and sample availability during the COVID-19 pandemic. Excluding patients with COVID-19 also reduced the number of eligible patients. Expecting higher MRSA prevalence, we combined colonization from MRSA and CoNS, acknowledging potential differences in colonization behaviors and risk factors in these genetically distinct species. Prophylactic antibiotics in post-operative patients and treatment antibiotics in patients with suspected infections could have altered their microbiomes. While this study focused on broad microbiome characterization, future work should include both antimicrobial susceptibility testing and multilocus sequence typing (MLST) of isolates. Susceptibility testing to commonly used PICU antibiotics would guide treatment decisions, while sequence typing could reveal whether particular Staphylococcus lineages associate with distinct microbial communities. Together, these approaches would enhance our understanding of methicillin-resistant Staphylococcus and improve prediction of colonization patterns in critically ill children.

Larger studies are needed to better understand methicillin-resistant Staphylococcus transmission in PICUs. Nonetheless, this study highlights the need to consider methicillin-resistant CoNS in screenings and highlights that risk is not uniform across PICU patients (as some may have colonization resistance from their nasal flora). This study was not sufficiently powered to determine if nasal colonization was associated with other outcomes, but it is intriguing that fever was significantly linked with methicillin-resistant Staphylococci colonization. Further evidence from larger studies would be influential in understanding what interventions are beneficial in the PICU setting to limit transmission.

Supporting information

Acknowledgments

The authors would like to thank the research group within the Margolis Lab at St. Jude Children’s Research Hospital for their help with this work.