Advertisement

Low prevalence of meticillin-resistant Staphylococcus aureus carriage at hospital admission: implications for risk-factor-based vs universal screening

  • J.A. Otter
    Correspondence
    Corresponding author. Address: Centre for Clinical Infection and Diagnostics Research, Department of Infectious Diseases, King's College London, Guy's and St. Thomas' Hospital NHS Foundation Trust, London SE1 9RT, UK. Tel.: +44 (0) 207 188 7188; fax: +44 (0) 1264 835917.
    Affiliations
    Centre for Clinical Infection and Diagnostics Research, Department of Infectious Diseases, King's College London, Guy's and St. Thomas' Hospital NHS Foundation Trust, London, UK
    Search for articles by this author
  • M.T. Herdman
    Affiliations
    Directorate of Infectious Diseases, Guy's and St. Thomas' Hospital NHS Foundation Trust, London, UK
    Search for articles by this author
  • B. Williams
    Affiliations
    Directorate of Infectious Diseases, Guy's and St. Thomas' Hospital NHS Foundation Trust, London, UK
    Search for articles by this author
  • O. Tosas
    Affiliations
    Centre for Clinical Infection and Diagnostics Research, Department of Infectious Diseases, King's College London, Guy's and St. Thomas' Hospital NHS Foundation Trust, London, UK
    Search for articles by this author
  • J.D. Edgeworth
    Affiliations
    Centre for Clinical Infection and Diagnostics Research, Department of Infectious Diseases, King's College London, Guy's and St. Thomas' Hospital NHS Foundation Trust, London, UK
    Search for articles by this author
  • G.L. French
    Affiliations
    Centre for Clinical Infection and Diagnostics Research, Department of Infectious Diseases, King's College London, Guy's and St. Thomas' Hospital NHS Foundation Trust, London, UK
    Search for articles by this author
Published:January 14, 2013DOI:https://doi.org/10.1016/j.jhin.2012.10.008

      Summary

      Background

      There is debate over the optimal policy for detecting meticillin-resistant Staphylococcus aureus (MRSA) colonization at hospital admission. The emergence of community-associated (CA)-MRSA may compromise targeted screening strategies based on risk factors for healthcare-associated (HA)-MRSA.

      Aim

      To determine the prevalence of MRSA colonization at admission, and the genotype and molecular epidemiology of the strains involved.

      Methods

      A 12-month observational study was performed at a 1200-bed London tertiary referral hospital from 1 April 2008 to 1 March 2009. All available MRSA isolates were genotyped by spa and staphylococcal cassette chromosome mec (SCCmec) typing.

      Findings

      The overall MRSA colonization rate was 2.0% of 28,892 admissions (range 6.6% in critical care to 0.8% in obstetrics/gynaecology/neonatology). The overall frequency of previously unknown carriage of MRSA on admission was 1.4%. Most colonizing strains were epidemic HA-MRSA-15 and -16. However, heterogeneous CA strains accounted for 18% of recovered isolates, including 37.5% of MRSA from accident and emergency and 23.1% of MRSA from surgery. The CA-MRSA strain types had significantly different epidemiological associations from the HA-MRSA strains, so risk factors used for the identification of HA-MRSA may not detect CA-MRSA reliably.

      Conclusion

      The low rate of HA-MRSA in the UK increases the relative proportion due to CA-MRSA, for which conventional risk-factor-based screening strategies may be less effective. Cost–benefit analyses of universal MRSA admission screening will need to take account of this new epidemiology.

      Keywords

      Introduction

      Asymptomatic colonization with meticillin-susceptible or meticillin-resistant Staphylococcus aureus (MRSA) is a risk factor for subsequent infection.
      • Lowy F.D.
      Staphylococcus aureus infections.
      The identification of MRSA carriers on admission can help to control the spread of MRSA in hospitals by facilitating targeted isolation and decolonization.
      • Dancer S.J.
      Considering the introduction of universal MRSA screening.
      • Coia J.E.
      • Duckworth G.J.
      • Edwards D.I.
      • et al.
      Guidelines for the control and prevention of meticillin-resistant Staphylococcus aureus (MRSA) in healthcare facilities.
      For many years, England and most other European countries have used a risk-based approach to MRSA screening.
      • Coia J.E.
      • Duckworth G.J.
      • Edwards D.I.
      • et al.
      Guidelines for the control and prevention of meticillin-resistant Staphylococcus aureus (MRSA) in healthcare facilities.
      • Cookson B.
      • Bonten M.J.
      • Mackenzie F.M.
      • Skov R.L.
      • Verbrugh H.A.
      • Tacconelli E.
      Meticillin-resistant Staphylococcus aureus (MRSA): screening and decolonisation.
      However, the English Government has mandated that National Health Service (NHS) acute hospitals perform universal MRSA screening for all elective and emergency admissions since December 2010.
      • Dancer S.J.
      Considering the introduction of universal MRSA screening.
      Several states in the USA have also mandated active MRSA surveillance cultures.
      • Weber S.G.
      • Huang S.S.
      • Oriola S.
      • et al.
      Legislative mandates for use of active surveillance cultures to screen for methicillin-resistant Staphylococcus aureus and vancomycin-resistant enterococci: position statement from the Joint SHEA and APIC Task Force.
      These legal directives have prompted considerable debate regarding the cost-effectiveness, ethics and practicalities of implementing universal screening programmes, with many experts maintaining that a targeted screening policy is most cost-effective.
      • Dancer S.J.
      Considering the introduction of universal MRSA screening.
      • Cookson B.
      • Bonten M.J.
      • Mackenzie F.M.
      • Skov R.L.
      • Verbrugh H.A.
      • Tacconelli E.
      Meticillin-resistant Staphylococcus aureus (MRSA): screening and decolonisation.
      • Weber S.G.
      • Huang S.S.
      • Oriola S.
      • et al.
      Legislative mandates for use of active surveillance cultures to screen for methicillin-resistant Staphylococcus aureus and vancomycin-resistant enterococci: position statement from the Joint SHEA and APIC Task Force.
      Risk factors for colonization with healthcare-associated (HA)-MRSA are well established.
      • Lowy F.D.
      Staphylococcus aureus infections.
      • Coia J.E.
      • Duckworth G.J.
      • Edwards D.I.
      • et al.
      Guidelines for the control and prevention of meticillin-resistant Staphylococcus aureus (MRSA) in healthcare facilities.
      • Harbarth S.
      • Sax H.
      • Fankhauser-Rodriguez C.
      • Schrenzel J.
      • Agostinho A.
      • Pittet D.
      Evaluating the probability of previously unknown carriage of MRSA at hospital admission.
      However, the same is not true for community-associated (CA)-MRSA strains that have emerged worldwide over the past decade and can affect otherwise healthy individuals of all ages in community settings.
      • Zetola N.
      • Francis J.S.
      • Nuermberger E.L.
      • Bishai W.R.
      Community-acquired meticillin-resistant Staphylococcus aureus: an emerging threat.
      • Otter J.A.
      • French G.L.
      Molecular epidemiology of community-associated meticillin-resistant Staphylococcus aureus in Europe.
      CA-MRSA have begun to transmit in hospitals, confounding epidemiological definitions and making a case for the definition and identification of CA-MRSA by their distinct genotypes.
      • Otter J.A.
      • French G.L.
      Community-associated meticillin-resistant Staphylococcus aureus: the case for a genotypic definition.
      The study hospital introduced a universal MRSA screening policy in April 2008. All MRSA isolates identified during the first year of universal screening were collected to determine the prevalence and molecular epidemiology of MRSA colonization among patients admitted to an acute hospital.

      Methods

      Setting, MRSA screening policy and culture methods

      Guy's and St. Thomas' NHS Foundation Trust comprises two hospitals in central London with about 1200 beds and approximately 120,000 admissions per annum, including day visits. From 1 April 2008, an admission MRSA screen was collected from all adult and paediatric elective surgical and medical patients during pre-admission clinics, and from all emergency cases within the first 48 h of admission. Repeated screens from the same patient were excluded. Cotton-tipped swabs (Sterilin Amies Transport, Sterilin Limited, Newport, UK) were used to sample nose, throat and perineal (groin in children) colonization sites. The three swabs were pooled and plated on to MRSA selective chromogenic agar (Brilliance™ MRSA, Oxoid, Basingstoke, UK). Swabs were also taken to detect rectal colonization in patients admitted to the adult intensive care unit (ICU) and high dependency unit and processed separately. Admission MRSA screens were taken at the same time from any clinical sites such as skin breaches and catheter urines.
      • Batra R.
      • Eziefula A.C.
      • Wyncoll D.
      • Edgeworth J.
      Throat and rectal swabs may have an important role in MRSA screening of critically ill patients.
      Presumptive MRSA isolates were confirmed by standard methods, and tested for antimicrobial susceptibility by automated broth microdilution (Vitek 2, bioMérieux, Basingstoke, UK). During the study period, one MRSA isolate per patient was collected prospectively and stored on a nutrient agar slope at room temperature.

      Identification and characterization of MRSA cases based on clinical and epidemiological factors

      Culture results of all MRSA admission screens collected between 1 April 2008 and 31 March 2009 were recorded prospectively. Patient age, gender, record of previous visits to the study hospitals, previous history of MRSA, admitting specialty and underlying medical conditions were obtained from patient electronic medical records. CA-MRSA and HA-MRSA strain types were defined genotypically (see below). Regardless of the strain types involved, cases were classified as healthcare-associated if: (1) their MRSA-positive screen during the study period was collected less than 12 months after a previous inpatient stay, or (2) the patient had (a) previous MRSA episodes, (b) regular day care (e.g. haematology and renal patients), (c) day surgery or (d) evidence of risk factors (e.g. long-term indwelling devices). All other cases were classified epidemiologically as community-associated.
      It was not possible to investigate MRSA acquisition because no discharge screening was performed. However, submission of an MRSA-positive specimen from a clinical site in the 12 months after admission was used as a proxy measure of MRSA infection.

      Genotypic definitions of HA- or CA-MRSA isolates

      DNA was extracted from MRSA using the ChargeSwitch™ gDNA mini-bacteria kit (Invitrogen Ltd., Paisley, UK). Isolates were characterized by staphylococcal cassette chromosome mec (SCCmec) type, spa type and carriage of Panton-Valentine leukocidin (PVL)-encoding genes; spa types were grouped into related clonal clusters (CCs) using the based upon repeat patterns (BURP) algorithm with a calculated cost between members of six or less, as described previously.
      • Otter J.A.
      • French G.L.
      The emergence of community-associated methicillin-resistant Staphylococcus aureus at a London teaching hospital, 2000–2006.
      CA-MRSA typically possess SCCmec types IV or V.
      • Otter J.A.
      • French G.L.
      Molecular epidemiology of community-associated meticillin-resistant Staphylococcus aureus in Europe.
      • Milheirico C.
      • Oliveira D.C.
      • de Lencastre H.
      Update to the multiplex PCR strategy for assignment of mec element types in Staphylococcus aureus.
      However, in the study setting, the most common HA-MRSA clone, multi-locus sequence type CC22-IV (EMRSA-15), is SCCmec IV.
      • Otter J.A.
      • French G.L.
      Utility of antimicrobial susceptibility-based algorithms for the presumptive identification of genotypically-defined community-associated methicillin-resistant Staphylococcus aureus at a London teaching hospital.
      Therefore, a combination of spa type and SCCmec type was used to define HA- and CA-MRSA. CA-MRSA strains were defined as isolates that were SCCmec IV or V that were not in the cluster relating to EMRSA-15.
      • Otter J.A.
      • French G.L.
      Utility of antimicrobial susceptibility-based algorithms for the presumptive identification of genotypically-defined community-associated methicillin-resistant Staphylococcus aureus at a London teaching hospital.
      Isolates with non-typeable SCCmec regions were defined as CA-MRSA because they were considered to be unlikely to represent common hospital lineages.
      • Otter J.A.
      • French G.L.
      Utility of antimicrobial susceptibility-based algorithms for the presumptive identification of genotypically-defined community-associated methicillin-resistant Staphylococcus aureus at a London teaching hospital.
      Two strains from a recently described CC22 CA-MRSA lineage were t005 and PVL positive,
      • Boakes E.
      • Kearns A.M.
      • Ganner M.
      • et al.
      Molecular diversity within clonal complex 22 methicillin-resistant Staphylococcus aureus encoding Panton-Valentine leukocidin in England and Wales.
      so were defined as CA-MRSA types despite clustering with EMRSA-15. All other isolates were classified as HA-MRSA.

      Statistical analysis

      Univariate and multiple logistic regression analyses were used to investigate risk factors for MRSA colonization on admission, and for submitting an MRSA-positive clinical specimen in the 12 months after admission (SPSS Inc., Chicago, IL, USA). Variables that were significant on univariate analysis (P < 0.05) were included in the multiple logistic regression analysis. Univariate analysis was used to compare the epidemiological associations of HA- and CA-MRSA strain types.

      Results

      In total, 583 (2.0%) of 28,892 admission screens were MRSA positive during the study period. Four hundred and seven (1.4%) of 27,727 patients without a history of MRSA had a positive admission screen. The highest prevalence of MRSA colonization was found in patients admitted to the ICU (6.6%), and the lowest prevalence was found in patients admitted to obstetrics/gynaecology/neonatology (0.8%) and paediatrics (1.2%); the prevalence of MRSA colonization was lower in surgical than medical specialties (1.6% vs 2.2%) (Table I).
      Table IPrevalence of meticillin-resistant Staphylococcus aureus-positive admission screens by specialty
      SpecialtyTotalPositive% positive% of all positives
      Surgery
      General surgery2361411.77.0
      Urology2250381.76.5
      Orthopaedics1822311.75.3
      Ear, nose and throat/oral surgery1561332.15.7
      Cardiothoracic surgery1317211.63.6
      Paediatric surgery1225141.12.4
      Plastic surgery1137111.01.9
      Vascular surgery480122.52.1
      Breast surgery38730.80.5
      Total surgery12,5402041.635.0
      Medicine
      General medicine46321423.124.4
      Cardiology2729321.25.5
      Paediatric medicine1732181.03.1
      Haematology/oncology1150171.52.9
      Renal medicine972192.03.3
      Respiratory medicine361174.72.9
      Elderly care21894.11.5
      Gastroenterology17142.30.7
      Ophthalmology13543.00.7
      Rheumatology12943.10.7
      Dermatology8578.21.2
      Total medicine12,3142732.246.8
      Accident and emergency
      Adult accident and emergency1280383.06.5
      Paediatric accident and emergency7211.40.2
      Total accident and emergency1352392.96.7
      Intensive care unit
      Adult intensive care unit624406.46.9
      Paediatric intensive care unit154117.11.9
      Total intensive care unit778516.68.7
      Obstetrics/gynaecology/neonatology
      Obstetrics/gynaecology1433161.12.7
      Neonatology47500.00.0
      Total obstetrics/gynaecology/neonatology1908160.82.7
      Grand total28,8925832.0
      Four percent of screens included non-standard sites: 1.6% were from incomplete nose, throat and perineum sets; 2.2% included rectal screens (standard in the ICU) and 0.3% included clinical sites. Compared with standard screening sets, MRSA-positive screens were more likely from clinical sites [P < 0.001, odds ratio (OR) 31.2, 95% confidence interval (CI) 20.6–47.3] and rectal screens (P < 0.001, OR 5.2, 95% CI 3.9–6.9).
      Significant associations with an MRSA-positive admission screen after adjusting for other factors were increasing age (OR 1.01, 95% CI 1.01–1.01; each year was associated with a 1% increase in the risk of carrying MRSA), having a history of MRSA (OR 16.8, 95% CI 13.7–20.5), admission to the ICU (OR 2.8, 95% CI 1.8–4.4), admission from a care home (OR 1.9, 95% CI 1.1–3.2) or hostel (OR 5.5, 95% CI 0.9–19.2), and overnight hospital stay in the 12 months prior to the screen (OR 1.8, 95% CI 1.4–2.3). Admission to obstetrics/gynaecology/neonatology was associated with a negative screen (OR 0.5, 95% CI 0.3–0.9) (Table II). Risk factors associated with previously unknown MRSA carriage were similar (Table III).
      Table IIRisk factors for meticillin-resistant Staphylococcus aureus (MRSA) colonization
      MRSA negative (N = 28,309)MRSA positive (N = 583)Unadjusted OR (95% CI)Univariate PAdjusted OR (95% CI)Multiple regression P
      Count(% of all negatives)Count(% positive)(% of all positives)
      Demographics
      Mean age (standard deviation)/years50 (25)58 (24)1.01 (1.01–1.02)<0.0011.01 (1.01–1.01)<0.001
      Median age (range)/years52 (0–109)64 (0–97)
      Gender (female)13,068(46.2)225(1.7)(43.7)0.9 (0.8–1.1)0.246
      Epidemiological data
      Previous hospitalization3676(13.0)136(3.6)(23.3)<0.001
       Overnight hospital stay in the past 12 months2028(7.2)96(4.5)(16.5)2.6 (2.1–3.3)<0.0011.8 (1.4–2.3)<0.001
       Day hospital visit in the past 12 months1648(5.8)40(2.4)(6.9)1.3 (1.0–1.9)0.0821.0 (0.7–1.4)0.931
      Previous positive for MRSA582(2.1)176(23.2)(30.2)20.6 (16.9–25.0)<0.00116.8 (13.7–20.5)<0.001
      Pre-admission screen8870(31.3)130(1.4)(22.3)0.6 (0.5–0.8)<0.0010.9 (0.7–1.2)0.530
      Admitted from a care home or hostel276(1.0)20(6.8)(3.4)<0.001
       Admission from care home254(0.9)18(6.6)(3.1)3.5 (2.2–5.7)<0.0011.9 (1.1–3.2)0.020
       Admission from homeless hostel22(0.1)2(8.3)(0.3)4.5 (1.1–19.3)0.0415.5 (0.9–19.2)0.022
      Specialty<0.001
      Surgery12,336(43.6)204(1.6)(35.0)0.6 (0.4–0.8)0.0010.8 (0.5–1.1)0.187
      Medicine12,041(42.5)273(2.2)(46.8)0.8 (0.5–1.1)0.120.9 (0.6–1.3)0.516
      Obstetrics/gynaecology/neonatology1892(6.7)16(0.8)(2.7)0.3 (0.2–0.5)<0.0010.5 (0.3–0.9)0.032
      Accident and emergency1313(4.6)39(2.9)(6.7)Reference groupReference group
      Intensive care unit727(2.6)51(6.6)(8.7)2.4 (1.5–3.6)<0.0012.8 (1.8–4.4)<0.001
      Total28,309(100.0)583(2.0)(100.0)
      OR, odds ratio; CI, confidence interval.
      Table IIIRisk factors for meticillin-resistant Staphylococcus aureus (MRSA) colonization excluding previously positive patients
      MRSA negative (N = 27,727)MRSA positive (N = 407)Unadjusted OR (95% CI)Univariate PAdjusted OR (95% CI)Multiple regression P
      Count(% of all negatives)Count(% positive)(% of all positives)
      Demographics
      Mean age (standard deviation /years49 (25)57 (25)1.01 (1.01–1.02)<0.0011.01 (1.01–1.01)<0.001
      Median age (range)/years52 (0–109)63 (0–97)
      Gender (female)12,805(46.2)188(1.4)(46.2)1.0 (0.8–1.2)1.000
      Epidemiological data
      Previous hospitalization3517(12.7)82(2.3)(20.1)<0.001
       Overnight hospital stay in the past 12 months1940(7.0)57(2.8)(14.0)1.8 (1.1–3.0)0.0111.9 (1.4–2.6)<0.001
       Day hospital visit in the past 12 months1577(5.7)25(1.6)(6.1)0.8 (0.6–1.3)0.4271.2 (0.8–1.7)0.489
      Pre-admission screen8760(31.6)90(1.0)(22.1)0.6 (0.5–0.8)<0.0010.8 (0.6–1.0)0.053
      Admitted from a care home or hostel256(0.9)15(5.5)(3.7)<0.001
       Admission from care home234(0.8)13(5.3)(3.2)3.9 (2.2–6.9)<0.0013.0 (1.7–5.3)<0.001
       Admission from homeless hostel23(0.1)2(8.0)(0.5)6.3 (1.5–27.2)0.0125.6 (1.3–24.1)0.022
      Specialty<0.001
      Surgery12,139(43.8)151(1.2)(37.1)0.7 (0.4–1.0)0.0580.8 (0.5–1.1)0.187
      Medicine11,739(42.3)182(1.5)(44.7)0.8 (0.5–1.3)0.3630.8 (0.5–1.3)0.461
      Obstetrics/gynaecology/neonatology1879(6.8)13(0.7)(3.2)0.4 (0.2–0.7)0.0040.5 (0.3–1.0)0.065
      Accident and emergency1268(4.6)24(1.9)(5.9)Reference groupReference group
      Intensive care unit702(2.5)37(5.0)(9.1)2.8 (1.6–4.7)<0.0013.0 (1.8–5.2)<0.001
      Total27,727(100.0)407(1.4)(100.0)
      OR, odds ratio; CI, confidence interval.
      Significant associations with an MRSA-positive clinical specimen in the 12 months after admission after adjusting for other factors were increasing age (OR 1.01, 95% CI 1.00–1.02; each year was associated with a 1% increase in the risk of carrying MRSA), an MRSA-positive admission screen (OR 48.7, 95% CI 35.4–66.9), having a history of MRSA (OR 4.8, 95% CI 3.4–6.9) and admission to the ICU (OR 4.5, 95% CI 2.0–10.1) (Table IV).
      Table IVRisk factors for a positive meticillin-resistant Staphylococcus aureus (MRSA) clinical specimen in the 12 months after hospital admission
      No clinical specimen (N = 28,660)Clinical specimen (N = 232)Unadjusted OR (95% CI)Univariate PAdjusted OR (95% CI)Multiple regression P
      Count(% of all negatives)Count(% positive)(% of all positives)
      Demographics
      Mean age (standard deviation)/years50 (25)61 (24)1.02 (1.01–1.03)<0.0011.01 (1.00–1.02)<0.001
      Median age (range)/years52 (0–109)64 (0–93)
      Gender (female)13,224(46.1)99(0.7)(42.7)1.2 (0.9–1.5)0.292
      Epidemiological data
      MRSA admission screen positive444(1.5)139(23.8)(59.9)95.0 (71.9–125.5)<0.00148.7 (35.4–66.9)<0.001
      Previous hospitalization5433(13.1)50(0.9)(21.5)<0.001
       Overnight hospital stay in the past 12 months2091(7.3)33(1.5)(14.2)2.2 (1.5–3.1)<0.0011.0 (0.6–1.5)0.873
       Day hospital visit in the past 12 months1671(5.8)17(1.0)(7.3)1.4 (0.8–2.3)0.1951.1 (0.6–1.9)0.822
      Previous positive for MRSA670(2.3)88(11.6)(37.9)25.5 (19.3–33.6)<0.0014.8 (3.4–6.9)<0.001
      Pre-admission screen8951(31.2)49(0.5)(21.1)0.6 (0.4–0.8)<0.0010.8 (0.5–1.1)0.166
      Admitted from a care home or hostel288(1.0)8(2.7)(3.4)0.001
       Admission from care home264(0.9)8(2.9)(3.4)3.8 (1.9–7.8)<0.0011.7 (0.7–4.0)0.237
       Admission from homeless hostel24(0.1)0(0.0)(0.0)0.9980.998
      Specialty<0.001
      Surgery12,451(43.4)89(0.7)(38.4)0.9 (0.5–1.6)0.6681.7 (0.8–3.6)0.187
      Medicine12,215(42.6)99(0.8)(42.7)1.0 (0.5–1.8)0.9701.2 (0.6–2.5)0.546
      Obstetrics/gynaecology/neonatology1904(6.6)4(0.2)(1.7)0.3 (0.1–0.8)0.0200.8 (0.2–2.6)0.657
      Accident and emergency1341(4.7)11(0.8)(4.7)Reference groupReference group
      Intensive care unit749(2.6)29(3.7)(12.5)4.7 (2.3–9.5)<0.0014.5 (2.0–10.1)<0.001
      Total28,660(100.0)232(0.8)(100.0)
      OR, odds ratio; CI, confidence interval.
      Four hundred and eighty-four (83%) of 583 MRSA admission isolates were recovered from storage. HA-MRSA strain types accounted for 82.4% of the recovered MRSA. The majority (80.2%) of the HA-MRSA isolates were in a cluster relating to ST22-IV (phage type EMRSA-15),
      • Otter J.A.
      • French G.L.
      Molecular epidemiology of community-associated meticillin-resistant Staphylococcus aureus in Europe.
      with t032 (49.1% of all HA-MRSA) predominating. A further 15.2% were in a cluster relating to ST36-II (EMRSA-16) and other HA-MRSA clones.
      CA-MRSA strain types accounted for 85 (17.6%) of the recovered isolates. Thus, assuming the same frequency of CA-MRSA strain types in the isolates that were not recovered, the prevalence of CA-MRSA strain types was 0.3%. CA-MRSA represented a substantial proportion of MRSA identified in certain specialties; for example, CA-MRSA strains accounted for 37.5% of MRSA from accident and emergency, 23.1% from surgery and 21.4% from obstetrics/gynaecology/neonatology. CA-MRSA isolates were more diverse than the HA-MRSA isolates: t127 (ST1, PVL negative) and t044 (ST80, PVL positive), both from spa CC5, accounted for 25.8% and 3.5% of the CA-MRSA strain types, respectively; 20.0% were spa CC3 (ST8) (41.2% of which were PVL positive) and a further 22.4% were either singleton lineages or had spa types that were too short to analyse by BURP clustering.
      Compared with HA-MRSA, CA-MRSA strains were associated with younger patients (OR 0.98, 95% CI 0.97–0.99; each year younger was associated with a 2% increase in the chance of carrying CA-MRSA), were less likely to be classified epidemiologically as healthcare-associated (OR 0.5, 95% CI 0.3–0.8), were less likely in medical specialties (OR 0.5, 95% CI 0.3–0.8) and the ICU (OR 0.2, 95% CI 0.1–0.8), were resistant to fewer classes of antimicrobial agents although more likely to be resistant to tetracycline (OR 3.5, 95% CI 1.7–7.3) and fusidic acid (OR 5.7, 95% CI 3.3–9.7), were more likely to be PVL positive (OR 1.3, 95% CI 1.2–1.4) and were more diverse (Table V). A high proportion (62.3%) of patients with CA-MRSA strain types were classified epidemiologically as healthcare-associated, and there was no significant difference in the likelihood of previous hospital contact in patients with HA- or CA-MRSA strains (Table V).
      Table VComparison of the characteristics of patients with healthcare-associated and community-associated meticillin-resistant Staphylococcus aureus (HA- and CA-MRSA) strain types
      HA-MRSA (N = 399)CA-MRSA (N = 85)OR (95% CI)P
      N(%)N(%)
      Demographics
      Mean age (standard deviation)/years61.2 (22.5)46.5 (25.1)0.98 (0.97–0.99)<0.001
      Median age (range)/years66 (0–97)50 (0–94)
      Gender (female)173(43.4)40(47.1)1.2 (0.7–1.9)0.533
      Epidemiological data
      Previous hospitalization100(25.1)15.0(17.6)0.318
      Overnight hospital stay in the past 12 months73(18.3)10(11.8)0.7 (0.2–2.4)0.139
      Day hospital visit in the past 12 months27(6.8)5(5.9)1.2 (0.5–3.4)0.642
      Previous positive for MRSA128(32.1)20(23.5)0.6 (0.4–1.1)0.122
      Pre-admission screen89(22.3)25(29.4)1.4 (0.9–2.4)0.163
      Admitted from a care home or hostel11(2.8)3(3.5)0.872
      Admission from care home10(2.5)3(3.5)1.4 (0.4–5.3)0.601
      Admission from homeless hostel1(0.2)0(0.0)1.000
      Healthcare-associated (epidemiologically defined)311(77.9)53(62.3)0.5 (0.3–0.8)0.003
      Specialty
      Surgery130(32.6)39(45.9)Reference
      Medicine196(49.1)28(32.9)0.5 (0.3–0.8)0.006
      Accident and emergency20(5.0)12(14.1)2.0 (0.9–4.4)0.090
      Intensive care unit42(10.5)3(3.5)0.2 (0.1–0.8)0.022
      Obstetrics/gynaecology/neonatology11(2.8)3(3.5)0.9 (0.2–3.4)0.888
      Antimicrobial resistance
      Ciprofloxacin380(95.2)28(32.9)0.02 (0.01–0.05)<0.001
      Erythromycin272(68.2)29(34.1)0.2 (0.1–0.4)<0.001
      Fusidic acid42(10.5)34(40.0)5.7 (3.3–9.7)<0.001
      Gentamicin67(16.8)7(8.2)0.4 (0.2–1.0)0.047
      Tetracycline21(5.3)14(16.5)3.5 (1.7–7.3)0.001
      Trimethoprim62(15.5)18(21.2)1.5 (0.8–2.6)0.206
      Mupirocin13(3.3)3(3.5)1.1 (0.3–3.9)0.899
      Number of non-beta lactam resistance classes
      Resistant to no non-beta lactam classes9(2.3)18(21.2)11.6 (5.0–27.0)<0.001
      Resistant to <2 non-beta lactam classes102(25.6)41(48.2)2.7 (1.7–4.4)<0.001
      PVL
      PVL0(0.0)19(22.3)1.3 (1.2–1.4)<0.001
      SCCmec type
      I1(0.2)0(0.0)
      II56(14.0)0(0.0)
      III10(2.5)0(0.0)
      IV312(78.2)63(74.1)
      V0(0.0)9(10.6)
      Non-typeable20(5.0)13(15.3)
      spa diversity
      Unique spa types78(19.5)35(41.2)<0.001
      Common spa types (inferred MLST CC)
      spa CC1 (CC22)320(80.2)2(2.3)
      t032196(49.1)
      t02219(4.8)
      spa CC2 (CC30)62(15.5)7(8.2)
      t01837(9.3)0(0.0)
      t0129(2.3)2(2.3)
      spa CC35(1.3)12(14.1)
      spa CC4 (CC8)12(3.0)17(20.0)
      t19010(2.5)(0.0)
      t0081(0.2)11(12.9)
      spa CC50(0.0)28(32.9)
      t12722(25.9)
      t0443(3.5)
      Other0(0.0)19(22.3)
      PVL, Panton-Valentine leukocidin; SCCmec, staphylococcal cassette chromosome mec; MLST, multi-locus sequence type; CC, clonal clusters; OR, odds ratio; CI, confidence interval.

      Discussion

      Only 2% of all patients presenting to a London acute hospital and 1.4% of patients without a history of MRSA were colonized with MRSA. Risk factors for MRSA carriage were similar to those reported in other studies.
      • Lowy F.D.
      Staphylococcus aureus infections.
      • Coia J.E.
      • Duckworth G.J.
      • Edwards D.I.
      • et al.
      Guidelines for the control and prevention of meticillin-resistant Staphylococcus aureus (MRSA) in healthcare facilities.
      • Harbarth S.
      • Sax H.
      • Fankhauser-Rodriguez C.
      • Schrenzel J.
      • Agostinho A.
      • Pittet D.
      Evaluating the probability of previously unknown carriage of MRSA at hospital admission.
      • Reilly J.S.
      • Stewart S.
      • Christie P.
      • et al.
      Universal screening for meticillin-resistant Staphylococcus aureus in acute care: risk factors and outcome from a multicentre study.
      Patients who had an MRSA-positive admission screen or a history of MRSA were considerably more likely to submit an MRSA-positive clinical specimen during the 12 months after their admission, in line with other studies.
      • Lowy F.D.
      Staphylococcus aureus infections.
      • Reilly J.S.
      • Stewart S.
      • Christie P.
      • et al.
      Universal screening for meticillin-resistant Staphylococcus aureus in acute care: risk factors and outcome from a multicentre study.
      Ninety-three (40%) patients who submitted a positive clinical specimen in the 12 months after admission had a negative admission screen, so it is likely that a proportion of these patients represent nosocomial acquisition.
      • Reilly J.S.
      • Stewart S.
      • Christie P.
      • et al.
      Universal screening for meticillin-resistant Staphylococcus aureus in acute care: risk factors and outcome from a multicentre study.
      Of the MRSA isolates tested, 82% were HA-MRSA strain types and 18% were CA-MRSA strain types. CA-MRSA strains were disproportionately represented among certain specialties, accounting for more than 20% of the recovered MRSA identified in surgical patients (mainly pre-admission screens) and 38% of those from accident and emergency; both groups present mainly from the community, with less prior hospital exposure. The CA-MRSA strain types had different epidemiological associations from the HA-MRSA strains, so risk factors used for the identification of HA-MRSA may not detect CA-MRSA reliably. However, no significant difference in the proportion of HA-MRSA and CA-MRSA strain types from patients with prior hospital admission or previous MRSA episodes was identified, reflecting the continuing breakdown of purely epidemiological definitions of CA-MRSA.
      • Otter J.A.
      • French G.L.
      Community-associated meticillin-resistant Staphylococcus aureus: the case for a genotypic definition.
      The finding of a 2% overall rate of colonization is comparable with hospitals performing universal screening in England between 2006 and 2009
      • Collins J.
      • Raza M.
      • Ford M.
      • Hall L.
      • Brydon S.
      • Gould F.K.
      Review of a three-year meticillin-resistant Staphylococcus aureus screening programme.
      and Scotland in 2010.
      • van Velzen E.V.
      • Reilly J.S.
      • Kavanagh K.
      • et al.
      A retrospective cohort study into acquisition of MRSA and associated risk factors after implementation of universal screening in Scottish hospitals.
      Rates from six Scottish hospitals performing universal screening in 2008–2009 were somewhat higher at 3.9%.
      • Reilly J.S.
      • Stewart S.
      • Christie P.
      • et al.
      Universal screening for meticillin-resistant Staphylococcus aureus in acute care: risk factors and outcome from a multicentre study.
      These rates are considerably lower than the rates reported from screening targeted specialities, including 6.7% of medical and surgical patients in the study hospital in 2006–2007,
      • Jeyaratnam D.
      • Whitty C.J.
      • Phillips K.
      • et al.
      Impact of rapid screening tests on acquisition of meticillin resistant Staphylococcus aureus: cluster randomised crossover trial.
      8.6% of emergency admissions to another London hospital in 2004–2005,
      • Gopal Rao G.
      • Michalczyk P.
      • Nayeem N.
      • Walker G.
      • Wigmore L.
      Prevalence and risk factors for meticillin-resistant Staphylococcus aureus in adult emergency admissions – a case for screening all patients?.
      and 5.1% of Swiss surgical patients in 2005–2006.
      • Harbarth S.
      • Fankhauser C.
      • Schrenzel J.
      • et al.
      Universal screening for methicillin-resistant Staphylococcus aureus at hospital admission and nosocomial infection in surgical patients.
      The present low rate of MRSA colonization at admission in England and Scotland may be the result of several factors. Firstly, universal screening includes a high proportion of patients at low risk of MRSA colonization and therefore inevitably reduces measured colonization rates compared with targeted screening. Secondly, as a result of national infection control improvement programmes, the rate of MRSA infection and transmission in hospitals in England and Scotland has fallen considerably in the last few years.
      • Edgeworth J.D.
      Has decolonization played a central role in the decline in UK methicillin-resistant Staphylococcus aureus transmission? A focus on evidence from intensive care.

      Health Protection Scotland, National Services Scotland, NHS Scotland MRSA Screening Pathfinder Programme; Final Report. Executive Summary 2011. Health Protection Scotland [Report].

      This has probably resulted in a smaller number of new patient carriers being discharged, and consequently a lower rate of ‘revolving door’ carriers being re-admitted.
      • Cooper B.S.
      • Medley G.F.
      • Stone S.P.
      • et al.
      Methicillin-resistant Staphylococcus aureus in hospitals and the community: stealth dynamics and control catastrophes.
      Indeed, in a recent study from Scotland evaluating the impact of the introduction of universal screening, the rate of MRSA colonization fell from 5.5% at the start of the study to 3.5% at the end of the study.
      • Reilly J.S.
      • Stewart S.
      • Christie P.
      • et al.
      Universal screening for meticillin-resistant Staphylococcus aureus in acute care: risk factors and outcome from a multicentre study.
      It is likely that this change in MRSA epidemiology is happening throughout the UK, and the cost–benefit analyses used previously to justify universal and economic analyses based on a higher rate of colonization (6–7%) will need to be reviewed.

      Jacqui Reilly, Sally Stewart, Traiani Stari, et al. Health Protection Scotland, National Services Scotland, NHS Scotland MRSA Screening; Update Report on Pathfinder project 2011. Health Protection Scotland [Report].

      • Robotham J.V.
      • Graves N.
      • Cookson B.D.
      • et al.
      Screening, isolation, and decolonisation strategies in the control of meticillin resistant Staphylococcus aureus in intensive care units: cost effectiveness evaluation.
      With a low overall admission prevalence of MRSA, it may be cost-beneficial to use targeted rather than universal screening if this could identify a significant proportion of carriers.
      • Dancer S.J.
      Considering the introduction of universal MRSA screening.
      • Coia J.E.
      • Duckworth G.J.
      • Edwards D.I.
      • et al.
      Guidelines for the control and prevention of meticillin-resistant Staphylococcus aureus (MRSA) in healthcare facilities.
      However, if, as seems likely, the prevalence of CA-MRSA strain types increases, targeted screening policies may require modification to accommodate the novel epidemiology of CA-MRSA strains. For example, in some parts of the USA, CA-MRSA now account for the majority of MRSA identified on hospital admission, so risk factors used to identify patients for targeted screening would require modification in these settings.
      • Otter J.A.
      • French G.L.
      Community-associated meticillin-resistant Staphylococcus aureus: the case for a genotypic definition.
      • Popovich K.J.
      • Weinstein R.A.
      • Hota B.
      Are community-associated methicillin-resistant Staphylococcus aureus (MRSA) strains replacing traditional nosocomial MRSA strains?.
      • Otter J.A.
      • French G.L.
      Community-associated meticillin-resistant Staphylococcus aureus strains as a cause of healthcare-associated infection.
      The strengths of this study include the large number of patients admitted, molecular typing on most of the strains involved, and clinical information on all patients. Limitations include possible underestimation of true CA-MRSA prevalence because these strains can colonize non-standard sites.
      • Yang E.S.
      • Tan J.
      • Eells S.
      • Rieg G.
      • Tagudar G.
      • Miller L.G.
      Body site colonization in patients with community-associated methicillin-resistant Staphylococcus aureus and other types of S. aureus skin infections.
      Screens from clinical sites were included, but these were collected at the discretion of clinical staff. Not all MRSA isolates were saved by the clinical laboratory, which may have underestimated the prevalence of certain genotypes. Relatively few CA-MRSA strains were identified, so the epidemiological associations of CA-MRSA strain types are less certain than for HA-MRSA strain types. Detailed epidemiological and demographic data were lacking for the MRSA-negative patients, such as antimicrobial use, ethnicity, socio-economic factors and health status, which would be required to make a thorough investigation of risk factors.
      At present, it is not clear what proportion of patients carrying MRSA at admission needs to be identified to prevent onward transmission. There is some evidence that both targeted screening and universal screening detect a sufficient amount of MRSA colonization to reduce transmission.
      • Dancer S.J.
      Considering the introduction of universal MRSA screening.
      • Coia J.E.
      • Duckworth G.J.
      • Edwards D.I.
      • et al.
      Guidelines for the control and prevention of meticillin-resistant Staphylococcus aureus (MRSA) in healthcare facilities.
      The situation is made more difficult by the changing epidemiology of MRSA in the UK. Overall MRSA colonization rates at admission are now low at approximately 2%. HA-MRSA colonization rates are low and apparently falling, and CA-MRSA rates are very low (approximately 0.3%) but represent a substantial proportion of MRSA in some specialties and may be increasing. Previous cost–benefit analyses of MRSA admission screening should be reviewed in light of the low overall colonization rate revealed by this study, and new screening and control strategies may need to be developed for the likely increasing prevalence of CA-MRSA.

      Acknowledgements

      The authors wish to acknowledge Dahir Mohamed, Guy's and St. Thomas' Hospitals for laboratory assistance.

      Conflict of interest statement

      JAO, JDE and GLF have support from Pfizer and the Guy's and St. Thomas' Charity for the submitted work. JAO is employed part-time by Bioquell and is supported by a grant from the Royal Commission for the Exhibition of 1851. MTH, BW and OT have no conflicts of interest to declare.

      Funding sources

      This study was supported by research grants from Pfizer and the Guy's & St Thomas' Charity .

      References

        • Lowy F.D.
        Staphylococcus aureus infections.
        N Engl J Med. 1998; 339: 520-532
        • Dancer S.J.
        Considering the introduction of universal MRSA screening.
        J Hosp Infect. 2008; 69: 315-320
        • Coia J.E.
        • Duckworth G.J.
        • Edwards D.I.
        • et al.
        Guidelines for the control and prevention of meticillin-resistant Staphylococcus aureus (MRSA) in healthcare facilities.
        J Hosp Infect. 2006; 63: 1-44
        • Cookson B.
        • Bonten M.J.
        • Mackenzie F.M.
        • Skov R.L.
        • Verbrugh H.A.
        • Tacconelli E.
        Meticillin-resistant Staphylococcus aureus (MRSA): screening and decolonisation.
        Int J Antimicrob Agents. 2011; 37: 195-201
        • Weber S.G.
        • Huang S.S.
        • Oriola S.
        • et al.
        Legislative mandates for use of active surveillance cultures to screen for methicillin-resistant Staphylococcus aureus and vancomycin-resistant enterococci: position statement from the Joint SHEA and APIC Task Force.
        Am J Infect Control. 2007; 35: 73-85
        • Harbarth S.
        • Sax H.
        • Fankhauser-Rodriguez C.
        • Schrenzel J.
        • Agostinho A.
        • Pittet D.
        Evaluating the probability of previously unknown carriage of MRSA at hospital admission.
        Am J Med. 2006; 119 (e15–23): 275
        • Zetola N.
        • Francis J.S.
        • Nuermberger E.L.
        • Bishai W.R.
        Community-acquired meticillin-resistant Staphylococcus aureus: an emerging threat.
        Lancet Infect Dis. 2005; 5: 275-286
        • Otter J.A.
        • French G.L.
        Molecular epidemiology of community-associated meticillin-resistant Staphylococcus aureus in Europe.
        Lancet Infect Dis. 2010; 10: 227-239
        • Otter J.A.
        • French G.L.
        Community-associated meticillin-resistant Staphylococcus aureus: the case for a genotypic definition.
        J Hosp Infect. 2012; 81: 143-148
        • Batra R.
        • Eziefula A.C.
        • Wyncoll D.
        • Edgeworth J.
        Throat and rectal swabs may have an important role in MRSA screening of critically ill patients.
        Intensive Care Med. 2008; 34: 1703-1706
        • Otter J.A.
        • French G.L.
        The emergence of community-associated methicillin-resistant Staphylococcus aureus at a London teaching hospital, 2000–2006.
        Clin Microbiol Infect. 2008; 14: 670-676
        • Milheirico C.
        • Oliveira D.C.
        • de Lencastre H.
        Update to the multiplex PCR strategy for assignment of mec element types in Staphylococcus aureus.
        Antimicrob Agents Chemother. 2007; 51: 3374-3377
        • Otter J.A.
        • French G.L.
        Utility of antimicrobial susceptibility-based algorithms for the presumptive identification of genotypically-defined community-associated methicillin-resistant Staphylococcus aureus at a London teaching hospital.
        Eur J Clin Microbiol Infect Dis. 2011; 30: 459-463
        • Boakes E.
        • Kearns A.M.
        • Ganner M.
        • et al.
        Molecular diversity within clonal complex 22 methicillin-resistant Staphylococcus aureus encoding Panton-Valentine leukocidin in England and Wales.
        Clin Microbiol Infect. 2011; 17: 140-145
        • Reilly J.S.
        • Stewart S.
        • Christie P.
        • et al.
        Universal screening for meticillin-resistant Staphylococcus aureus in acute care: risk factors and outcome from a multicentre study.
        J Hosp Infect. 2012; 80: 31-35
        • Collins J.
        • Raza M.
        • Ford M.
        • Hall L.
        • Brydon S.
        • Gould F.K.
        Review of a three-year meticillin-resistant Staphylococcus aureus screening programme.
        J Hosp Infect. 2011; 78: 81-85
        • van Velzen E.V.
        • Reilly J.S.
        • Kavanagh K.
        • et al.
        A retrospective cohort study into acquisition of MRSA and associated risk factors after implementation of universal screening in Scottish hospitals.
        Infect Control Hosp Epidemiol. 2011; 32: 889-896
        • Jeyaratnam D.
        • Whitty C.J.
        • Phillips K.
        • et al.
        Impact of rapid screening tests on acquisition of meticillin resistant Staphylococcus aureus: cluster randomised crossover trial.
        BMJ. 2008; 336: 927-930
        • Gopal Rao G.
        • Michalczyk P.
        • Nayeem N.
        • Walker G.
        • Wigmore L.
        Prevalence and risk factors for meticillin-resistant Staphylococcus aureus in adult emergency admissions – a case for screening all patients?.
        J Hosp Infect. 2007; 66: 15-21
        • Harbarth S.
        • Fankhauser C.
        • Schrenzel J.
        • et al.
        Universal screening for methicillin-resistant Staphylococcus aureus at hospital admission and nosocomial infection in surgical patients.
        JAMA. 2008; 299: 1149-1157
        • Edgeworth J.D.
        Has decolonization played a central role in the decline in UK methicillin-resistant Staphylococcus aureus transmission? A focus on evidence from intensive care.
        J Antimicrob Chemother. 2011; 66: ii41-ii47
      1. Health Protection Scotland, National Services Scotland, NHS Scotland MRSA Screening Pathfinder Programme; Final Report. Executive Summary 2011. Health Protection Scotland [Report].

        • Cooper B.S.
        • Medley G.F.
        • Stone S.P.
        • et al.
        Methicillin-resistant Staphylococcus aureus in hospitals and the community: stealth dynamics and control catastrophes.
        Proc Natl Acad Sci USA. 2004; 101: 10223-10228
      2. Jacqui Reilly, Sally Stewart, Traiani Stari, et al. Health Protection Scotland, National Services Scotland, NHS Scotland MRSA Screening; Update Report on Pathfinder project 2011. Health Protection Scotland [Report].

        • Robotham J.V.
        • Graves N.
        • Cookson B.D.
        • et al.
        Screening, isolation, and decolonisation strategies in the control of meticillin resistant Staphylococcus aureus in intensive care units: cost effectiveness evaluation.
        BMJ. 2011; 343: d5694
        • Popovich K.J.
        • Weinstein R.A.
        • Hota B.
        Are community-associated methicillin-resistant Staphylococcus aureus (MRSA) strains replacing traditional nosocomial MRSA strains?.
        Clin Infect Dis. 2008; 46: 787-794
        • Otter J.A.
        • French G.L.
        Community-associated meticillin-resistant Staphylococcus aureus strains as a cause of healthcare-associated infection.
        J Hosp Infect. 2011; 79: 189-193
        • Yang E.S.
        • Tan J.
        • Eells S.
        • Rieg G.
        • Tagudar G.
        • Miller L.G.
        Body site colonization in patients with community-associated methicillin-resistant Staphylococcus aureus and other types of S. aureus skin infections.
        Clin Microbiol Infect. 2009; 16: 425-431

      Linked Article

      • Screening for meticillin-resistant Staphylococcus aureus: targeted not universal screening
        Journal of Hospital InfectionVol. 85Issue 1
        • Preview
          Two recent publications in this journal on the carriage of meticillin-resistant Staphylococcus aureus (MRSA) pose interesting and important questions about strategies relating to screening as part of MRSA prevention and control.1,2 In a study of surgical wards in France, Greece, Italy and Spain, and excluding previous MRSA-positive patients, 111 of 2901 (3.8%) patients were identified as new carriers of MRSA on admission. While certain factors were identified as being associated with MRSA carriage (e.g.
        • Full-Text
        • PDF