Staphylococcus aureus carriage among GPs in the Netherlands

Study population

At an annual conference of the Dutch College of General Practitioners (Nederlands Huisartsen Genootschap) in 2006, nasal swabs were taken from volunteering GPs.

Isolation of S. aureus

The swabs were analysed for the presence of S. aureus, using inoculation on colistin-nalidixic acid agar plates (BD Diagnostics, Breda, the Netherlands) and enrichment in Nutrient Broth No. 2 (Oxoid, the Netherlands) containing 6.5% sodium chloride. After overnight incubation, 10 μl of the broth was inoculated on oxacillin-resistant screening agar (Oxoid, the Netherlands). Putative S. aureus colonies were identified as S. aureus using Gram-stain, catalase, and coagulase testing as described previously.6

Antimicrobial susceptibility testing

The antimicrobial susceptibility pattern was determined using a microbroth dilution method with Mueller Hinton II Broth cation-adjusted (Becton, Dickinson and Company, Breda, the Netherlands), according to the guidelines of the Clinical Laboratory Standards Institute (CLSI).7 The microtitre plates containing freeze-dried antibiotics were provided by MCS Diagnostics BV, Swalmen, the Netherlands.

The antibiotics tested were (range in mg/l): cefaclor (0.06 to 128); cefuroxime (0.06 to 128); clindamycin (0.03 to 64); ciprofloxacin (0.128 to 32); clarithromycin (0.03 to 64); gentamicin (0.06 to 64); imipenem (0.03 to 64); linezolid (0.03 to 64); meropenem (0.06 to 64); moxifloxacin (0.12 to 4); oxacillin (0.03 to 64); penicillin (0.004 to 8); rifampicin (0.008 to 16); teicoplanin (0.06 to 128); tetracycline (0.03 to 64); trimethoprim–sulfamethoxazole (0.015 to 32); and vancomycin (0.06 to 128).

Susceptibility to fusidic acid (100 μg) and mupirocin (10 μg; NeoSensitabs, Rosco, Denmark) was determined using a disk-diffusion method on Mueller-Hinton agar plates (BD Diagnostics, 254081, Breda, the Netherlands).8,9 The MIC90 was defined as the minimum inhibitory concentration required to inhibit the growth of 90% of organisms. All isolates that were resistant to clarithromycin and susceptible to clindamycin were tested for inducible clindamycin resistance using the D-test according to CLSI guidelines.10

Genotypic determinations

The genetic background of the S. aureus isolates was determined using spa typing as previously described.11 The spa types were clustered into spa clonal complexes (spa-CC) using the algorithm ‘based upon repeat pattern’ (BURP) with Ridom StaphType software (version 1.5; http://www.ridom.de/staphtype/). The default settings recommended by the manufacturer were used. It has been shown that spa typing, together with BURP, yields results that are concordant with typing results obtained by multilocus sequence typing (MLST).11 The associated MLST-CCs were allocated through the Ridom SpaServer. (http://spaserver.ridom.de).

Statistical analysis

This study aims to asses the carrier rate of S. aureus in GPs at an annual conference of Dutch GPs. Assuming a 30% prevalence of S. aureus (between 25% and 35% as shown in earlier studies), with a power of 0.8 and α = 0.05 (two-sided), a sample size of 323 was required to demonstrate that the carriage rate of GPs did not differ from the rate in other populations.

Confidence intervals were determined with the exact method to guarantee at least 95% coverage, and were placed between brackets (95% confidence interval [CI] = lowest value to highest value). Statistical analyses were performed using SPSS (version 16.0).

Summary of main findings

This study showed that the prevalence of nasal carriage of S. aureus among GPs was 33% (95% CI = 28 to 37%). The low antibiotic resistance observed among the isolates suggests that the participating GPs are not an important reservoir for antibiotic-resistant S. aureus. The resistance to fusidic acid was 7% (95% CI = 3 to 13%) and has not been described before in the Netherlands. Although the nasal S. aureus carrier rate among GPs was within the normal range, 72% (95% CI = 67 to 83%) of the isolates had a MRSA-related genetic background, suggesting that GPs may have a potential high risk for future MRSA carriage.12–13

Strengths and limitations of the study

Statistical analysis showed that a sample size of 323 was sufficient for screening so the 395 tested provided sufficient power. The present study shows, to the best of the authors' knowledge, the first analysis of carriage of (antibiotic-resistant) S. aureus among GPs. Unfortunately, demographic details about the participants were not available, and a selection bias between GPs attending and not attending at the conference could have occurred. Furthermore, earlier studies showed that multiple-site screening will result in a higher yield.14,15 However, in practice, multiple-site screening is feasible in a hospital situation but not in a survey like the present one. The most practical solution was chosen.

Comparison with existing literature

Very few data on antimicrobial resistance among S. aureus isolates of GPs are available. One review, by Albrich and Harbarth, described the prevalence of S. aureus carriage among healthcare workers in hospitals as ranging from 0 to 40%, with a median of 24% in hospitals.5 An earlier study among hospital workers in the Netherlands showed a carrier rate of meticillin-sensitive S. aureus (MSSA) of 35%, and an MRSA prevalence of less than 5%.16 The carrier rate of MSSA in healthcare workers was comparable with the results found in the present study, but the MRSA prevalence among healthcare workers in the Netherlands was higher than that observed among the tested GPs.16

The only study among GPs was a study in west Ireland. This study found an MRSA prevalence of 7.7%.4 This was higher in comparison with that of the Irish community, but the same as the prevalence of MRSA among hospital staff (6.6%).4 In the present study, no MRSA was observed, which is in line with the low prevalence of MRSA among hospitalised patients (1 to 2%) and outpatients (0 to 2%) in the Netherlands.17 The observed S. aureus carriage among GPs was comparable with the S. aureus carriage prevalence in other populations in the Netherlands.18,19

Fusidic acid is used as a topical agent for local infections, especially for the eyes and the skin. Several studies have described that the topical use of this drug has led to an increase of resistance.20–22 Some suggest that a S. aureus clone responsible for the resistance to fusidic acid is spreading,22,23 and that this is resulting in increased resistance. An earlier study in the Netherlands from 2002 showed no resistance to fusidic acid among impetigo isolates.21 Therefore, according to the guidelines of the Dutch College of General Practitioners, fusidic acid remained the first choice for topical use in skin infections, especially impetigo.20,21

The 7% resistance to fusidic acid is, to the best of our knowledge, the first observation of such a high resistance percentage among commensal isolates in the Netherlands. The S. aureus of GPs are probably not exposed to fusidic acid through the GPs. Therefore, these results suggest that the GPs might become infected with S. aureus isolates by patients. There is no clear explanation for the relatively high resistance prevalence.

Further studies, investigating the prevalence of fusidic acid resistance in other populations in the Netherlands, especially among patients with impetigo infections, and the cause of increasing fusidic acid resistance, are warranted. Depending on these results, recommendations as to the first drug of choice in the case of impetigo infections in the Netherlands could be established.

MLST CC30, the main associated MLST CC found in this study, was also found in other studies among outpatients and hospitalised patients in the Netherlands and worldwide.24–27 Another common MLST CC worldwide is MLST CC45, which was also highly prevalent in the present study.25–27 As the Ridom SpaServer does not detect repeats longer than 24 bases, two isolates with a repeat of 25 base pairs were not type-able, in accordance with European rules for typing justified by SeqNet.org.28

In previous studies in the Netherlands, approximately 50% of the S. aureus isolated from several patient populations had a genetic background common to major MRSA clones.11,26 The high percentage (72%) of S. aureus with an MRSA genotype observed in this study was unexpected, and no explanation for this phenomenon could be found. Recently, Nubel et al showed that MRSA evolution occurs more often than one might expect.13 As Staphylococcal Cassette Chromose mec (SCCmec) is only stable in MSSA isolates with a MRSA related genotype,13 GPs could be a carrier of MRSA in the (near) future. One of the possible solutions to prevent that might be decolonisation of the nose of GPs.

Implications for future research and clinical practice

The role of GPs in transmission of MRSA to patients and community dwellers remains unclear and needs further study. In the Netherlands, the MRSA prevalence is still low, but is increasing.10 The higher percentage of S. aureus with a MRSA-related genetic background might result in a higher MRSA carriage rate among GPs, and in transmission of MRSA from GPs to patients.

Furthermore, the increased resistance in fusidic acid among commensal S. aureus of GPs is a point of concern and warrants further studies as to the prevalence of resistance among S. aureus from skin infections in GPs' patients.