WHAT IS ALREADY KNOWN ON THIS TOPIC

The prevalence of methicillin-resistant Staphylococcus aureus (MRSA) colonisation and infection was 2.8% (95% CI 1.6 to 4.0) among children (0–5 years old) between 2000 and 2010 reported in a meta-analysis and systematic review. The prevalence of MRSA among young children was higher than among healthy adults (0.8%; 95% CI 0.0 to 17.5), which may be due to their immature immune system and interaction with asymptomatic MRSA carriers.

WHAT THIS STUDY ADDS

Asymptomatic MRSA colonisation among young children may have increased in the last decade. Available studies were mostly regional, and more data on a global scale are needed. To the best of our knowledge, this study is the first systematic review and meta-analysis on the global prevalence and epidemiology of asymptomatic MRSA colonisation in young children.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

Public health policymakers should take asymptomatic MRSA transmission into consideration to implement antibiotic regulatory and infection control measures that will limit the spread, and protect children.

Introduction

Methicillin-resistant Staphylococcus aureus (MRSA) has been officially recognised as a significant pathogen with high antimicrobial resistance, as outlined in the 2017 global priority pathogens list published by the WHO.1 Community-associated MRSA (CA-MRSA) infection is often reported among healthy children under 6 years of age.2

Colonisation is defined as the presence of bacteria on the human body’s surface (eg, airway, skin and mouth) without displaying disease symptoms and is regarded as a prerequisite for infection.3 However, many publications confuse ‘colonisation’ with ‘infection’.3 Therefore, ‘asymptomatic colonisation’ is adopted to distinguish it from infection, as defined by Chisholm, Campbell.4

The respiratory tract of humans, particularly the anterior nares, serves as the primary ecological reservoir for S. aureus.5 S. aureus can adhere and multiply in the nose and transmit among the human nose via hands and air.6 Individuals who are persistently colonised by MRSA are at an increased risk for MRSA infection.7 The rate of invasive CA-MRSA infection cases among children rose from 1.1 cases per 100 000 children in 2005 to 1.7 cases per 100 000 children in 2010 according to a population-based active surveillance programme in the USA, with a modelled yearly increase of 10.2% (95% CI 2.7 to 18.2).8 Most studies were conducted nationally, while aggregate global data on asymptomatic MRSA colonisation among children are lacking. Hence, the role of ‘asymptomatic MRSA colonisation’ in MRSA transmission could be underestimated in surveillance programmes and in developing policies and programmes on infection control.4 Thus, understanding the prevalence and epidemiology of asymptomatic MRSA colonisation in the upper respiratory tract of young children could provide better evidence for developing infection control and prevention strategies locally and globally.

To the best of our knowledge, this is the first systematic review and meta-analysis focused on the global prevalence and epidemiology of asymptomatic MRSA colonisation in young children. The study aimed to identify the prevalence of S. aureus and MRSA asymptomatic colonisation and risk factors associated with asymptomatic MRSA colonisation in the upper respiratory tract of young children and to characterise the antibiotic resistance and genetic characteristics of relevant MRSA isolates.

Method

The systematic review and meta-analysis followed the guideline of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses 2020 statement9 and registered on PROSPERO.

Searching strategies and eligible criteria

All review authors collaboratively developed the search strategies (online supplemental appendix I) for publications. A comprehensive publication search was conducted on Medline, Embase, Web of Science and CINAHL on 3 March 2022. All the records were exported to Covidence (https://www.covidence.org/). Studies had to report prevalence data on both S. aureus and MRSA in young children to be eligible, and other eligibility criteria are listed in online supplemental appendix II.

Data extraction

A data collection form (online supplemental appendix III) was designed for data extraction. Missing data were extracted from figures if possible. The primary outcomes of interest were the prevalence of asymptomatic S. aureus and MRSA carriage among young children without infection. For cohort studies, if specimens for S. aureus and MRSA screening were collected from a subject more than once, only the first screening outcomes were extracted. As for the secondary outcomes, significant risk factors for asymptomatic colonisation, antibiotic resistance and genetic characteristics of MRSA in young children were extracted. Basic information about the studies was also extracted, including author names, publication year, country/continent, economic status of countries, screening setting, study design, sampling sites, total number of target subjects, age/age group, health status, S. aureus screening method, MRSA confirmation method and whether combining pre-enrichment (ie, broth enrichment to enhance the sensitivity of detection) before S. aureus detection.

Bias assessment and data analysis

The risk of bias in publications was assessed using Joanna Briggs Institution (JBI) standardised critical appraisal checklist for both cross-sectional and cohort studies. The checklist comprised nine questions to determine the possible bias in the design, conduct, and analysis experiment.10

The ‘meta’ package11 in R software (V.4.1.2) was employed for all the statistical analyses. As a binary primary outcome, the overall pooled prevalence and the 95% CI for asymptomatic S. aureus and MRSA colonisation were calculated with a random-effects model and visualised in forest plots. Statistical heterogeneity between studies was assessed by I² statistics and interpreted according to the Cochrane Collaboration.12 The possible causes of heterogeneity were examined using subgroup analysis based on certain potential independent factors. The estimated prevalence of each subgroup was reported. As part of the secondary outcomes, the available quantitative data on antibiotic resistance and genetic characteristics of MRSA isolates were analysed using a random effects model. A p-value less than 0.05 was considered statistically significant.

ResultsStudy selection and data extraction

A total of 12 974 publication records were initially collected (figure 1). Of these, 8843 unique records were screened by title and abstract, and 1190 were subjected to full-text screening. After excluding those not fulfilling the inclusion criteria, 35 articles were identified (34 peer-reviewed articles and one letter to the editor).

Figure 1

Flow diagram for publications screening. MRSA, methicillin-resistant Staphylococcus aureus.

The prevalence of S. aureus and MRSA was extracted from all included publications. Of the 35 studies,13–47 seven13 15 17 20 23 24 27 reported significant risk factors or predictors of asymptomatic MRSA colonisation in young children.

Study characterisation

The included studies were conducted from 2004 to 2019 in 22 countries on five continents (detailed in online supplemental appendix IV). Among them, non-duplicate subjects from three publications from Taiwan were eligible.44–46 Most included studies were cross-sectional studies (31/35), while the remainder were cohort studies. S. aureus and MRSA screening were conducted among children from children’s institutions (20/35), followed by healthcare centres and communities. A total of 18 190 subjects were young children (0–8 years old) without respiratory infections or other potential infections caused by S. aureus infections.

From the studies, only five had more than 1000 subjects, and the majority (20/35) had fewer than 500 subjects. Most screening specimens (23/35) were mainly collected from nares only and followed by nasopharynx only. Moreover, the risk of bias in all included studies assessed with the JBI standardised critical appraisal checklist for included studies was with an overall score (yes%) higher than 60%, which was with acceptable publication quality (detailed in online supplemental appendix V).

Overall pooled prevalence

All studies (n=35) were included in the pooled prevalence analysis using a random effects model. The pooled prevalence of asymptomatic S. aureus colonisation was 25.1% (95% CI 21.4 to 28.8) (figure 2). The prevalence of each study derived largely from 3.7% to 48.6% with a significant degree of statistical heterogeneity (I2: 97.9% (97.6 to 98.2); p≤0.0001). The highest prevalence was observed in studies conducted in Portugal (48.6%),39 followed by Brazil (48.0% and 47.3%).15 16 For MRSA, the pooled prevalence was 3.4% (95% CI 2.8 to 4.1) (figure 3). The prevalence of each study ranged from 0.0% to 29.4% with significant heterogeneity (I2: 96.7% (96.0 to 97.2); p≤0.0001). The highest prevalence of asymptomatic MRSA colonisation was reported in Taiwan (29.4%),44 followed by Iran (16.7%).42

Figure 2

Forest plot for the prevalence of S. aureus asymptomatic colonisation (n=35).

Figure 3

Forest plot for the prevalence of MRSA asymptomatic colonisation (n=35). MRSA, methicillin-resistant Staphylococcus aureus.

Subgroup analysis for the prevalence of MRSA

Subgroup analyses for the prevalence of asymptomatic MRSA colonisation were conducted using nine independent factors (online supplemental appendix VI). Among them, the pooled prevalence of each subgroup differed significantly when the subgroups were divided by continent (p<0.0001), screening setting (p<0.0001) and age group (p<0.0001). As shown in figure 4, MRSA prevalence was higher in Asia (5.8%; I2=97.6%) and South America (5.6%; I2=2.5%) compared with European countries (0.6%; I2=85.0%). Higher MRSA prevalence was reported by studies conducted in the community (5.7%; I2=96.7%) and healthcare centres (6.4%; I2=98.0%) when compared with children’s institutions (ie, nurseries, kindergartens, daycare centres and preschools) (1.5%; I2=89.9%). Although these significant factors can partially explain the heterogeneity, there is still relatively high heterogeneity within different subgroups.

Figure 4

Prevalence of asymptomatic colonisation of MRSA in the upper respiratory tracts of young children in five continents. MRSA, methicillin-resistant Staphylococcus aureus.

Significant risk factors

Seven of the 35 included observational studies reported the significant risk factors/predictors of asymptomatic MRSA colonisation for the target population. Specifically, previous antibiotic intake history and having an MRSA-colonised mother were significantly associated with asymptomatic MRSA colonisation in young children. Other significant risk factors are reported in table 1.

Table 1

Reported significant risk factors/predictors of the asymptomatic colonisation of MRSA in the upper respiratory tract in young children

Antibiotic resistance characterisation

Antibiotic resistance characteristics of MRSA isolates were extracted from a total of 14 studies out of 35 studies (detailed in online supplemental appendix VII). High pooled resistance rates (ie, higher than 60%) were observed for antibiotics, such as ampicillin (93.3%; I2=44.7%) and erythromycin (63.0%; I2=89.6%), while the resistance rates to some important antibiotics remained low, such as rifampin (0.3%; I2=0.0%) and fusidic acid (0.0%; I2=0.0). Only one isolate found by Sedighi et al 40 was resistant to vancomycin.

Molecular genetic characterisation

Only a few studies characterised the molecular genetic characteristics of MRSA isolates from their target population. Among 13 studies that tested for the mecA gene, all isolates were positive (539/539).20 21 23 25 27 28 32 33 36 39 43 46 Yet, the pooled positive rate of the Panton-Valentine leucocidin (PVL) gene was 14.7% (95% CI 4.3 to 25.0) with high heterogeneity (I2=89.1%) among 493 MRSA isolates from 13 studies.16 17 20–22 25 28 33 35 36 43 46 47 SCCmec type IV (319/487) was the predominant SCCmec type, followed by type VT (84/487), type I (20/487) and type V (14/487).16 21 22 25 28 32 33 35 36 43 46 47 Tsai et al 46 found that the PVL gene was predominantly associated with SCCmec type VT or IV isolates. Different MRSA variations were reported in different studies. In particular, ST59 was the most commonly observed isolate in Taiwan, and sporadic MRSA isolates (ie, single locus variants of ST9) belonging to livestock-associated MRSA (LA-MRSA) clones were found in two children.46 Scn-negative MRSA (a putative marker for LA-MRSA) was observed in a child from a family involved in industrial hog farming and processing23 and one ST398-MRSA-V was detected in a child without exposure to livestock.22

Discussion

To the best of our knowledge, this is the first systematic review and meta-analysis on the global prevalence and epidemiology of asymptomatic MRSA colonisation in young children. This study estimated the global asymptomatic MRSA colonisation in children using 35 studies published between 2010 and 2021. The pooled prevalence of asymptomatic S. aureus and MRSA colonisation was 25.1% (95% CI 21.4 to 28.8) and 3.4% (95% CI 2.8 to 4.1) in young children, respectively. Compared with pooled MRSA colonisation and infection prevalence of 2.8% (95% CI 1.6 to 4.0) among children (0–5 years old) in a systematic review using data published between 2000 and 2010,48 the overall asymptomatic colonisation prevalence increased among young children in the last 10–15 years. Furthermore, the prevalence was higher in young children than in healthy people, with a prevalence of 5.9% (95% CI 2.3 to 79.6) and 0.8% (95% CI 0.0 to 17.5) for S. aureus and MRSA colonisation, respectively, for the latter cohort.49 It is also supported by a meta-analysis conducted for the Asia-Pacific region reporting higher CA-MRSA prevalence in young children under 6 years old with less mature immune systems than older children or adults.50 This is a warning finding as a study conducted by von Eiff et al 51 presented evidence to support that S. aureus in the nasal mucosa could be an endogenous origin for bacteraemia. Compared with non-carriers, both low (Ct>24 cycles) burden and high burden (Ct≤24 cycles) of nasal colonisation was associated with subsequent MRSA infection.52

Considerable heterogeneity in estimates of prevalence is mainly related to the continent, age of subjects and screening setting. MRSA became more prevalent in some Asian and Latin American countries.53 54 A much higher MRSA prevalence in Asia (5.8%) and in South America (5.6%) was reported compared with Europe (0.6%). Studies conducted in mainland China and Taiwan34 44 46 reported a higher prevalence of asymptomatic MRSA colonisation than others, while studies conducted in Hungary reported a lower prevalence.31 32 It is important to note that a proportion of children recruited in a study conducted in the US was from households of industrial hog operation workers, which has been considered to be a risk factor for MRSA colonisation.23 Therefore, the prevalence of MRSA colonisation reported in this study was much higher than the prevalence reported in other publications from the North America.27 35 Furthermore, a higher prevalence of asymptomatic MRSA colonisation was observed among young children recruited from communities and healthcare centres (without infections caused by S. aureus) than in children’s institutions, although this classification may partially overlap due to the difficulty in clearly separating them based on the description in the original research publications. Nonetheless, children attending child care centres are usually with a higher risk to of MRSA infection55 and can be a potential reservoir for emerging MRSA genotypes.56 More importantly, a previous study suggested that CA-MRSA infections were more likely to occur in children under 3 years old than in older children.57 Similarly, in this study, the prevalence of asymptomatic MRSA colonisation in newborns (within the first month) was significantly higher than in the older children, which is consistent with a meta-analysis that suggested the prevalence of MRSA colonisation in newborns (within 28 days) was significantly higher than in young children (within 5 years old), regardless of the subject’s health status of subjects at the sampling time.48 This finding highlights the importance of monitoring the prevalence and persistence of asymptomatic MRSA colonisation in children, particularly in the younger age group. However, it should be noted that the subgroup analysis for age in our study was largely influenced by a longitudinal study conducted in Keelung City, Taiwan.44 In this cohort, 29.4% of the newborns were colonised by MRSA at the age of 1 month. Interestingly, the colonisation rate of both S. aureus and MRSA gradually decreased in the first year of their life, which may be due to the pneumococcal competition in the nasopharyngeal space.44

Genetically, SCCmec type IV or VT remained the predominant SCCmec type and was mainly associated with the presence of the PVL gene,46 consistent with the genetic pattern observed 10–15 years ago.48 The predominant SCCmec type IV, VT and V are associated with CA-MRSA, while type I is associated with hospital-acquired MRSA (HA-MRSA).58 59 In addition, LA-MRSA was reported in young children, though the prevalence was low.22 23 46 Therefore, it suggests that either CA-MRSA, HA-MRSA or LA-MRSA may be present in the upper respiratory tracts of young children without respiratory infection symptoms.

There are several limitations in this study. First, the languages of the included publications were restricted to English and Chinese, which may introduce bias to the prevalence estimates. Second, the substantial heterogeneity and publication bias across included studies may affect the estimation precision. However, since it is a small meta-analysis with 35 studies, it should be noted that the I2 is known to be inevitably less precise in practice.60 In addition, a large proportion of isolates (330 isolates) in our study was contributed by Tsai and colleagues,46 which may cause systematic errors in estimating the resistance rate to some antibiotics and genetic characterisation. Third, the prevalence of MRSA colonisation in the respiratory tract may be underestimated in this meta-analysis because all included studies used culture-based methods with lower sensitivity instead of the PCR method to detect S. aureus.61 Finally, considering that MRSA prevalence varies widely across different regions, and a higher number of studies may be conducted in regions with higher reported infection cases, more cross-regional routine surveillance studies are expected.

Conclusion

In conclusion, our study provides evidence from an important aspect for establishing MRSA infection control among young children. The prevalence of asymptomatic MRSA colonisation was associated with geography, screening setting, children with younger age, previous antibiotic intake history. The increasing colonisation rate and considerable heterogeneity in prevalence estimates highlighted the need for more cross-regional active surveillance on asymptomatic colonisation in early childhood care to avoid a wider spread of MRSA and subsequent infections.

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information.

Ethics statementsPatient consent for publication

Not applicable.