Abstract. Introduction. ability to colonize and to survive and spread in humans,

ORIGINAL ARTICLE EPIDEMIOLOGY The search and destroy strategy prevents spread and long-term carriage of methicillin-resistant Staphylococcus aureus: results from the follow-up screening of a large ST22 (E-MRSA 15) outbreak in Denmark S. Böcher 1, R. L. Skov 1, M. A. Knudsen 2, L. Guardabassi 3, K. Mølbak 4, P. Schouenborg 2, M. Sørum 1 and H. Westh 5,6 1) National Center for Antimicrobials and Infection Control, Statens Serum Institut, Copenhagen, 2) Department of Clinical Microbiology, Vejle Hospital, Vejle, 3) Department of Disease Biology, Faculty of Life Sciences, University of Copenhagen, 4) Department of Epidemiology, Statens Serum Institut, 5) The Faculty of Health, University of Copenhagen, Copenhagen and 6) Department of Clinical Microbiology, Hvidovre Hospital, Hvidovre, Denmark Abstract In the aftermath of a methicillin-resistant Staphylococcus aureus (MRSA) ST22 hospital outbreak, we investigated the prevalence of longterm carriage, the efficacy of MRSA decolonization treatment (DT) and the spread of MRSA to households of patients and healthcare workers (HCWs). Furthermore, we evaluated the efficacy of repeated DT in long-term MRSA carriers. Of 250 index persons (58 HCWs and 192 patients), 102 persons (19 HCWs and 83 patients) and 67 household members agreed to participate. Samples from all 169 persons were taken from the nose, throat, wounds and devices/catheters, and urine samples were additionally taken from index persons. Samples from companion animals (n = 35) were taken from the nostrils and anus. Environmental sites (n = 490) screened were telephone, television remote control, toilet flush handle, favourite chair and skirting board beside the bed. Sixteen (19%) patients and two household members, but no HCWs, were ST22-positive. The throat was the most frequent site of colonization. In a multivariate analysis, chronic disease (p <0.001) and pharyngeal carriage (p <0.001) were associated with long-term MRSA carriage. MRSA was found in the environments of four long-term carriers. All animals tested were negative. MRSA-positive households were decolonized using nasal mupirocin TID and daily chlorhexidine body and hair wash for 5 days. Pharyngeal MRSA carriers also received fucidic acid (500 mg TID) combined with rifampicin (600 mg BID) or clindamycin (600 mg BID) for 7 days. The home environment was cleaned on days 2 and 5. At the end of follow-up, ten of 16 long-term carriers and the two household contacts were MRSA-negative. In conclusion, decolonization of MRSA carriers is possible, but should include treatment of household members and the environment. Keywords: Animals, E-MRSA15 outbreak, environment, household screening, MRSA decolonization treatment, ST22, pharyngeal carriage Original Submission: 11 March 2009; Revised Submission: 15 September 2009; Accepted: 20 November 2009 Editor: D. Raoult Article published online: 23 December 2009 Clin Microbiol Infect 2010; 16: 1427 1434 10.1111/j.1469-0691.2010.03137.x MRSA ST22, also known as EMRSA-15, has shown great Corresponding author: S. Böcher, Statens Serum Institut, National ability to colonize and to survive and spread in humans, Center for Antimicrobials and Infection Control, Artillerivej 5, DK-2300 Copenhagen S, Denmark companion animals (i.e. dogs and cats) and hospital environments [1,2]. In the UK, the MRSA prevalence was low until E-mail: sbc@ssi.dk the early 1990s but, as a result of the introduction of epidemic strains, including ST22 and EMRSA-16, the prevalence of MRSA in bactaeremias increased from 2% to 42% within Introduction 10 years [3]. The rise in MRSA infections resulted in increased admissions and mortality and extensive costs to For more than 30 years, Denmark has been a low-prevalence country for methicillin-resistant Staphylococcus aureus been tried, the epidemic clones in the UK are still a matter healthcare services and, although different approaches have (MRSA) with less than 1% of MRSA in Staphylococcus aureus of concern to the British healthcare system [4,5]. Pandemic bacteraemias (Ref: http://www.ssi.dk/sw3425.asp). This has spread of ST22 has also been reported in Portugal, Australia been ascribed to the restricted use of broad-spectrum antimicrobials and to a rigorous practice of isolation procedures Until 2002, ST22 was only observed sporadically in Den- and New Zealand [6 8]. concerning patients with MRSA in hospitals. mark. The small MRSA outbreaks have been contained by a Journal Compilation ª2010 European Society of Clinical Microbiology and Infectious Diseases

1428 Clinical Microbiology and Infection, Volume 16 Number 9, September 2010 CMI search and destroy strategy. In November and December 2002, five cases were seen in two hospitals in Vejle County and marked the onset of an outbreak, which, from November 2002 until December 2005, included 440 persons infected or colonized with ST22, spa-type t022. During the first 2 years of the outbreak, the only interventions implemented were isolation and campaigns for intensified hand hygiene. Decolonization therapy (DT) of patients was not systematically recommended. The outbreak increased during 2004 and, starting 1 January 2005, infection control strategies were extended, with screening of patients at admission and discharge and screening of all healthcare workers (HCWs). MRSA DT was free of charge for both HCWs and patients and finally resulted in curbing of the outbreak. The present study aimed to perform a follow-up assessment of long-term carriage and evaluate the efficacy of DT. Furthermore, we looked for spread of MRSA within households, to pets and to the environment, and we assessed the efficiency of repeated DT in persistent MRSA carriers. Materials and Methods Setting Vejle County, Denmark, has a population of approximately 355 000 inhabitants and is served by a network of six hospitals with 70 000 yearly admissions. Because of hospital specialization, patients are often transferred between hospitals. All microbiological specimens, collected in hospitals and by general practitioners and private specialists, are handled at one site: the Department of Clinical Microbiology, Vejle Hospital. MRSA decolonization treatment during the outbreak During the first 2 years of the ST22 outbreak in Vejle County, MRSA DT was not systematically recommended to patient carriers. From the beginning of 2004, the Department of Clinical Microbiology started recommending DT to patients, with mupirocin (2%, TID) and chlorhexidine (4%) hand wash, for 5 days, occasionally in combination with systemic antimicrobial treatment. Systemic antimicrobial treatment (Table 1) was mainly given to patients with DT failure. From 2005 onward, all patients and HCWs were offered DT free of charge (Table 1). Relatives were not offered screening and DT. Inclusion and exclusion criteria Patients, and their household members, living in Vejle County, who had been MRSA ST22-positive (infected or colonized) from November 2002 until December 2005 were eligible for inclusion. Patients living outside Vejle County and households of deceased patients were excluded. All HCWs who had worked at one of the hospitals in Vejle County and their households were eligible. The study was accepted by the ethics committee for Vejle and Funen (S-VF-20050181). All eligible households were contacted by mail with a letter providing information, an informed consent document and a questionnaire. Nonresponders were contacted again with a second letter after 6 weeks. Patients and HCWs were included after having provided their informed consent. Clinical and demographic data All persons positive for MRSA ST22 during the study period were identified in the laboratory database at the Department of Clinical Microbiology, Vejle Hospital. For all persons (both included and nonparticipants), the database was checked for DT control samples. Correctly taken DT control samples were defined as three samples, marked as MRSA control, from the nose, throat and perineum. These samples could not be taken earlier than 1 week after DT. TABLE 1. Treatment, cleaning and control recommendations for methicillin-resistant Staphylococcus aureus (MRSA)-positive persons identified during the household survey Topical treatment used for all carriers (National guidelines) Systemic decolonization treatment plus topical treatment for pharyngeal or urinary MRSA carriage (local guidelines) Cleaning recommendations for all positive households Control recommendations Mupirocin 2% ointment for nose, TID plus daily chlorhexidine (4%) body and hair wash for 5 days Pharyngeal carriers: fucidic acid 500 mg TID and rifampicin 600 mg BID or clindamycin 600 mg BID for 7 days plus mupirocin 2% ointment and chlorhexidine wash For children, same treatment but reduction of doses according to weight Urinary MRSA carriers: systemic treatment according to susceptibility profile, mostly trimethroprim and change of catheter during treatment plus topical treatment Private towels and facecloths. Linen washed at 90 C or as warm as possible Daily: Change underwear, towels and facecloths after the shower Air the rooms, pillows and duvets. Clean door handles, toilet and taps Day 2 and after treatment: Change the bed linen for all household members and Clean all horizontal surfaces. Vacuum clean the house Control at days 7, 14 and 21 after treatment, and after 6 12 months

CMI S. Böcher et al. The search and destroy strategy prevents spread and long-term carriage of MRSA 1429 Co-morbidity data and previous MRSA carriage and infection treatment were collected from the National Patient Registry, the hospital records and from the general practitioners of patients. Sampling during household visits Household screenings were carried out from December 2006 until March 2007 and were performed in 98 households of 102 persons (four households had two previously MRSA-positive persons) including 83 patients and 19 HCWs, and 39 household members of patients and 28 of HCWs. A total of 35 companion animals (20 dogs, 12 cats and three horses) from 13 patient households and nine HCW households were also screened. Samples from all household members were taken from the nose, throat and skin lesions using one flocked swab (Copan, Brescia, Italy) per anatomical site. Each swap was inoculated directly into 3 ml of tryptic soy broth containing 2.5% NaCl, 3.5 mg/l cefoxitin and 20 mg/l aztreonam (TSB SSI; SSI Diagnostica, Hillerød, Denmark). Urine samples were collected from the household index persons, considering the high frequency of urinary tract infections during the outbreak. A veterinarian collected samples from companion animals from the nose, anus and skin lesions if present. Environmental samples were taken from the household s telephone, the index person s favourite chair, the toilet flush handle, televison remote control and skirting board by the bed, using Biotrace dip-slides (3M, Glostrup, Denmark) containing a Baird Parker agar with ciprofloxacin (the ST22 strain is ciprofloxacin-resistant). Infection control precautions to prevent colonization of study personnel during study visits were performed according to national guidelines (National Board of Health: http:// www.sst.dk/sundhed3a/mrsa.aspx.), which specified an appropriate uniform, gloves when sampling and hand hygiene with alcohol disinfectants after each visit. The project nurse was screened for MRSA weekly. Human samples were processed at Vejle Hospital using the routine methods for MRSA identification. TSB SSI broths were incubated overnight at 35 C and PCR for MRSA was performed with the broth cultures obtained from all human swabs [9]. PCR-positive broths were sub-cultured on blood agar for final identification (ID) and antimicrobial susceptibility testing. Urine samples were spread onto Chromogenic MRSA agar (Oxoid, Hampshire, UK) and incubated overnight at 35 C. Veterinary samples were processed at the Faculty of Life Sciences in Copenhagen, using their standard methods for MRSA identification. The overnight incubated TSB SSI broths from companion animal samples were sub-cultured on MRSA ID agar (biomérieux, Marcy l Etoile, France) and blood (5%) agar. Environmental samples were processed at the Staphylococcus Laboratory, SSI. Dip-slides from environmental sampling were incubated for 42 48 h at 35 C in accordance with the manufacturer s instructions. Presumed staphylococcal colonies growing on the dip slides were inoculated into the TSB SSI broth and, after overnight incubation at 35 C, 20 ll were sub-cultured on MRSA ID agar. Susceptibility testing was performed on blood agar (SSI) using an inoculum giving semi confluent growth, NEO-Sensitabs (Rosco Diagnostica, Taastrup, Denmark) and overnight incubation. The antibiotics tested were: cefoxitin, cefuroxim, ciprofloxacin, clindamycin, erythromycin, fucidic acid, gentamicin, mupirocin, penicillin, polymyxin, rifampicin, tetracycline, trimethroprim and vancomycin. Interpretation was performed as specified by the manufacturer. (REF. Rosco s users guide: http://www.rosco.dk/default.aspx?id=198). At Statens Serum Institut, MRSA were confirmed by PCR for meca [10] and characterized by spa typing [11] and SCCmec typing [12]. Selected isolates were typed by multilocus sequence typing [13] and/or pulsed field gel electrophoresis [14]. Interventions and decolonization treatment Households were informed of the results by telephone. In households with MRSA-positive samples, all household members were offered DT and instructed to start DT at the same time (Table 1). Patients with wounds were given intensive wound care to ensure healing prior to DT. Cleaning instructions were given to all MRSA-positive households (Table 1). If patients needed assistance with DT or cleaning procedures, home care assistance and nursing was provided. All expenses were paid by the local municipalities in Vejle County. The efficacy of the DT was controlled by swabbing the nose, throat, perineum and other colonized sites on days 7, 14 and 21 after treatment. The laboratory database was used to follow up on MRSA positive samples for the next 18 months. Statistical analysis Univariate analyses were performed using Statistical Software 8 (Stata Corporation, College Station, TX, USA). Risk ratios and 95% CI were calculated for dichotomous variables. The multivariable model included previous infection, time since last positive sample, correct control after last positive sample and the number of sites previously colonized, chronic diseases and pharyngeal carriage. Logistic regression analyses were performed using LOGXACT, version 7.0, procedures for SAS 9.1.3 (SAS Institute Inc., Cary, NC, USA).

1430 Clinical Microbiology and Infection, Volume 16 Number 9, September 2010 CMI Results Four hundred and forty persons were positive for MRSA ST22 from November 2002 until December 2005. One hundred and sixty-nine were deceased and 21 lived in another county, leaving a total of 250 eligible persons of whom 102 (41%) agreed to participate. The participation rate was slightly higher among patients (n = 83, 43%) than among HCWs (n = 19, 34%). Among the 102 index persons, 74 (73%) had received DT, whereas 25 had never received DT. However, 16 of these 25 index persons had previously been treated with other anti-mrsa treatments. For three index persons, the previous use of DT could not be substantiated. Among the patients and HCWs participating in the study, 38 (46%) and 12 (63%), respectively, had three sets of negative MRSA control swabs after their last MRSA-positive test and five patients (6%) were known long-term MRSA carriers. For 37 patients and six HCWs, some negative control swabs existed, but not three full sets from the nose, throat and perineum (generally, the perineal samples were missing). Three patients had three negative sets, although sampling was performed immediately after DT. Finally, no control samples were taken from four patients and one HCW. For comparative purposes, we checked the MRSA status of the 148 persons who did not want to participate in the study; 42 (38%) patients and 30 (81%) HCWs had three sets of correctly taken negative control samples. In this group, we also found that 5% of patients were long-term carriers. Findings during the household survey and follow-up period At the household visits, we found 12 patients and two of their household contacts to be MRSA-positive. During follow-up, four additional patients tested positive: three at admission screening at the hospital and one at the general practitioner s office. The 16 MRSA ST22-positive patients had been intermittent or persistent carriers during 14 41 months; all had chronic underlying conditions and a history of pharyngeal carriage (Table 2). In the patients testing positive during the household survey, the carriage sites were throat (75%), nose (50%) and the urinary tract (19)% (all catheterized patients). Only three patients were multiple-site TABLE 2. Distribution, demographics, treatment and control data prior to household survey in 83 patients and 19 healthcare workers (HCW) with a history of methicillin-resistant Staphylococcus aureus (MRSA) ST22 infection or colonization and decolonization therapy Positive patients Negative patients HCW (all negative) Index persons 16 67 19 Households 15 64 19 Households with >1 previously positive: n (%) 3 (19) 6 (9) 1 (5) Sex (male/female) 10/6 38/29 1/18 Year of first positive sample: n (%) 2003 1 (6) 5 (8) 2 (11) 2004 5 (31) 17 (25) 4 (21) 2005 10 (63) 45 (67) 13 (68) Age at household visit: median (range) 73 (54 98) 70 (3 98) 42 (25 61) Households with pets: n (number of pets) 2 (2) 12 (13) 9 (20) Household size: mean (range) 1 (1 2) 1 (1 4) 2 (1 5) Sampled from all relevant sites at first MRSA screening: n (%) 4 (25) 20 (30) 7 (37) Previous symptomatic MRSA infection: n (%) 12 (75) 26 (39) 2 (11) Days from first positive sample to last positive sample: median (range) 784 (426 1248) 6 (1 1033) 1 (1 68) Days from last positive sample to the day of the household visit: median (range) 353 (57 714) 625 (40 1459) 689 (361 1173) Sites previously infected/colonized: median (range) 4 (1 8) 1 (1 6) 1 (1) Pharyngeal carriage: n (%) 16 (100) a 28 (42) 2 (11) Previous DT b Topical: n (%) 4 (25) 41 (61) 13 (68) Topical + systemic: n (%) 8 (50) 6 (9) 2 (11) No DT: n (%) c 4 (25) 17 (25) 4 (21) Unknown if DT: n (%) 3 (5) Three sets of negative controls prior to survey: n (%) 4 (25) 34 (51) 12 (63) Chronic disease: n (%) 16 (100) 46 (69) 0 Most frequent chronic diseases: n (%) Cardiovascular 7 (44) 29 (43) Chronic obstructive pulmonary disease 4 (25) 13 (19) Diabetes 2 (13) 10 (15) Cancer 4 (25) 10 (15) Patients are stratified according to MRSA status in the household survey. a One patient only tested positive in a pooled sample from nose and throat prior to the survey, but was positive in the throat sample in the survey; the others had previously tested positive in throat samples. b For decolonization treatment (DT), see Table 1. c Of the persons not having received DT; four of five positive patients, 12 of 17 negative patients and three of four HCW had received other anti-mrsa treatment. The HCW who had not received DT was unaware of MRSA carriage because the general practitioner had misinterpreted the laboratory result.

CMI S. Böcher et al. The search and destroy strategy prevents spread and long-term carriage of MRSA 1431 carriers (more than two positive sites) when tested during the survey (see Supporting Information, Table S1). The two positive household contacts were pharyngeal MRSA carriers only. MRSA ST22, was found in the environment of four households of long-term carriers: on the telephone (four households), skirting board near bed (two households), televison remote control (two households) and toilet flush handle (one household) (see Supporting Information, Table S1). These four patients had all been MRSA-positive for the last 2 3 years and none of the households had pets. In three of the four homes, one person had previously had an MRSA infection. Furthermore, in two of the four households, the index persons had previously tested MRSA-negative with three sets of control samples and MRSA in the environment may have been the cause of re-colonization in these two cases (see Supporting Information, Table S1). None of the HCWs or their households were MRSA-positive and none of the 35 pets were MRSA-positive. TABLE 3. Risk factors for long-term MRSA carriage in 83 patients. MRSA positive patients/all patients (%) Risk Ratio 95% CI P Sex Male 10/48 (21) 1.22 0.49 3.03 0.674 Female 6/35 (17) Age <70 years a 8/43 (19) 0.93 0.39 2.24 0.872 70 years 8/40 (20) Pets in the home Yes 2/14 (14) 0.70 0.18 2.76 0.604 No 14/69 (20) House hold with >1 previously MRSA positive Yes 3/9 (33) 1.90 0.67 5.41 0.258 No 13/74 (18) Previous MRSA infection Yes 12/38 (32) 3.55 1.25 10.11 0.009 No 4/45 (9) Throat carriage Yes 16 b /41 (37) ND d ND <0.001* No or unknown 0/42 (2) Number of sites previously colonized/infected >2 12/30 (40) 5.3 1.87 14.98 <0.001 2 4/53 (8) Chronic disease Yes 16/62 (26) ND ND <0.001* No 0/21 (0) Time since last positive sample <590 days c 13/41(32) 4.44 1.36 14.44 0.005 590 days 3/42 (7) Previous carrier treatment Yes 12/59 (20) 1.22 0.44 3.41 0.701 No or unknown 4/24 (17) Control after last treatment Yes 4/38 (11) 0.39 0.14 1.12 0.063 No 12/45 (27) a The median age of all patients was 70 years. b One patient had prior to the household survey only been tested positive in a pooled sample, this patient was throat positive at the household visit. All other positive patients had throat positive samples prior to the household survey. c The median number of days between last MRSA positive sample and household visit for all patients. d ND: Not defined. * Statistically significant in the logistic regression model. Clinical and demographic data of 83 patients and 19 healthcare workers: No difference was observed between MRSA-positive and - negative patients regarding DT and follow-up data (Table 2). In the univariate analyses (Table 3), the long-term MRSApositive patients had more recently had a positive sample (p 0.005), had previously had a symptomatic infection (p 0.009), had been pharyngeal carriers (p <0.001) and had more anatomical sites previously infected/colonized (p <0.001). All positive patients had one or more chronic diseases, most frequently cardiovascular disease and chronic obstructive pulmonary disease. In the multivariable model, only chronic diseases (p <0.001) and pharyngeal carriage (p <0.001) were found to be significant risk factors for long-term MRSA carriage (Table 3). Because none of the HCWs were MRSA-positive in our survey, they were not included in the statistical calculations. All HCWs were healthy and young and had no personal risk factors for MRSA. Compliance regarding DT and control after DT was higher among HCWs because they were suspended from work with salary until they had finished DT. Most HCWs were only colonized in the nose and were treated promptly; the time from first to last positive sample was thus much shorter. Intervention and decolonization treatment Twelve of the 18 cases (16 patients, two relatives) testing MRSA-positive in the present study received additional DT along with hygiene and cleaning instructions. Four patients were not treated because they had chronic conditions making successful MRSA DT unrealistic (e.g. fistulas from an abdominal prolene net or supra-pubic bladder catheter). One patient did not want DT and one patient died of pneumonia as a result of MRSA before anti-mrsa treatment was implemented. Five patients and two household contacts became MRSA-free after a single DT. Five additional patients were decolonized after two DTs. No pharyngeal carrier became MRSA-free without systemic antimicrobial treatment. Discussion Only few studies have followed nosocomially MRSA-infected or colonized patients after discharge from hospital [15 17] and, except for case reports, to our knowledge, the present study is the first to combine follow-up of household contacts with investigations of the home environments and companion animals.

1432 Clinical Microbiology and Infection, Volume 16 Number 9, September 2010 CMI Among patients, and in particular among HCWs, we observed a low participation rate, and from the reply letters from both groups, we received the impression that it was the emotional and social consequences of a possible MRSApositive test that made them reject participation. However, during the outbreak, 81% of HCWs (non-participants) had three correctly taken sets of samples that were negative after DT but, among the participating HCWs, only 63% had three correctly taken MRSA-negative control sets, which may have motivated them to participate. All HCWs had been successfully decolonized during the outbreak. Among patients, however, a long-term carriage rate of 19% was found in spite of DT as recommended by the local MRSA guidelines. During the outbreak only the index persons were offered MRSA DT, in disagreement with the protocol used in the present study, where the entire household was screened and treated simultaneously. As shown in Table 2, in only one-third of cases were samples taken from relevant sites when MRSA was first detected, which may have influenced the initial choice of the DT protocol. Furthermore, 25% of patients had not received DT and only half of the MRSA-positive patients had previously received treatment for pharyngeal carriage. That 25% of patients had not received DT during the outbreak may be a result of several factors: (i) during the first year of the outbreak, DT was only recommended occasionally and mostly for HCWs because it was believed that MRSA would disappear when the patient was discharged from hospital; (ii) MRSA DT was optional for patients, but not for HCWs; (iii) the majority of patients not receiving MRSA DT had previously had an MRSA infection and did not receive MRSA DT after treatment for their MRSA infection; (iv) some patients had personal risk factors, which made successful MRSA decolonization unlikely; and (v), in a few persons, colonization with ST22 disappeared spontaneously, without DT. It is noteworthy that the throat was the most frequent site of colonization in patients found to be positive at the household visit and that, in both household contacts identified, the throat was the colonized site. All the long-term MRSA-positive patients were or had at some stage been pharyngeal carriers. This type of carriage now appears to be more frequent than previously described, and the throat is a site not to be neglected when screening for MRSA [18,19]. Mertz et al. [20] have shown that pharyngeal carriage is agedependant and more frequent among young people and persons older than 75 years. The median age of the patients in the present study was 73 years. A higher detection rate in throat swabs compared to nose swabs might be a result of the screening method used. Solberg [21] showed that CFU counts from pharyngeal carriers are lower than from nasal and perineal carriers. We used a selective enrichment broth (TSB SSI), designed for use in the present study, which was found to allow detection of <20 CFU/mL and to increase the rate of detection of MRSA in throat samples [22]. In addition, we used flocked swabs that were placed directly in the enrichment broth. Despite the use of an enrichment step, four patients found to be MRSA-negative during household screening were found to be positive at a later date in the present study. Failure in detection of MRSA can be the result of several factors, such as incorrect sampling techniques, low MRSA numbers as a consequence of recent antibiotic treatment (which we inquired about but which patients did not always recall), intermittent carriage and intracellular MRSA carriage [23]. The treatment interventions made in the present study suggest that systemic treatment in addition to topical treatment may be necessary to remove pharyngeal carriage in long-term MRSA carriers. Failure in treating this carriage with topical treatment has been described in other studies [24,25]. Mupirocin resistance has been shown to be associated with treatment failure, although no mupirocin resistance was seen during this outbreak and in the follow-up study [26]. Similar to the hospital environment, the home environment is a possible reservoir for re-colonization [1,27]. Previous studies by Solberg [21] have shown that the amount of staphylococci dispersed to the environment vary according to the site of colonization, infection and co-morbidity. Nasal and perineal carriers disperse more staphylococci than pharyngeal carriers, and people with chronic illnesses disperse more than those without [21]. We took samples from five preselected household areas, although Kniehl et al. [27] have shown that MRSA can be found at many sites of the home. The present study confirms that, in cases of relapse after DT, the environment should be taken into consideration when planning for additional DT. We provided positive households with hygiene and cleaning instructions but other interventions (e.g. hydrogen peroxide vaporization) could have been used [28]. We did not detect MRSA in any of the 35 companion animals examined, two of which lived in households of long-term MRSA carriers. This result indicates that companion animals did not play a role in the Vejle outbreak. Several studies have shown that pet animals such as dogs and cats can be colonized or infected with MRSA ST22 [2], especially in countries where this MRSA lineage is widespread, such as in the UK. Apart from the recent emergence of MRSA ST398 in pig farming [29], only a few case reports have described humanto-animal or animal-to-human MRSA transmission [30,31]. In

CMI S. Böcher et al. The search and destroy strategy prevents spread and long-term carriage of MRSA 1433 most of these reports, the humans had an MRSA infection or the environment was heavily contaminated (e.g. veterinary hospitals contaminated by animal patients with MRSA). Because it is reasonable to assume that patients with MRSA are an important source of environmental contamination [21] and that pets are more likely to be colonized when their household environment is heavily contaminated, it should be noted that all animals included in the present study lived in households where no environmental MRSA contamination was detected. A limitation of the present study is that we used both prospective and retrospective data. Because the MRSApositive patients had not been screened systematically prior to our household visits, we could not with certainty determine whether they were intermittent or persistent carriers. In addition, the data on previous treatments for carriage were not complete, primarily because the wards in some cases had not stored records on decolonization treatment. The response rate from general practitioners was low and, although the patients filled in a questionnaire regarding treatment for carriage, there might have been recall bias. Also, the environmental Biotrace dip-slides often showed massive growth of other bacteria and fungi, which is one reason why MRSA may have been missed in some cases. In conclusion, we found that MRSA DT may have prevented long-term carriage in most patients and all HCWs; however, chronic disease and pharyngeal carriage were independent risk factors for long-term carriage. The prevalence of pharyngeal carriage was remarkably high and eradication of this carriage required systemic antibiotic therapy. The household environment was confirmed to be a possible reservoir for re-colonization with MRSA after DT. Companion animals were not involved in the spread of MRSA in the present setting. If early intervention in the form of MRSA DT and shortterm follow-up is thorough and supervised, further household spread, infection with MRSA and long-term carriage can be prevented. Long-term carriage increases the risk of spread to the environment whereby the eradication of MRSA becomes a much more complicated task. Acknowledgements We thank Emily Claire Nightingale from Faculty of Life Sciences, University of Copenhagen, for her assistance with the collection and analysis of animal samples; Tove Ibæk Kristiansen, Vejle Hospital, Denmark, for assistance with the data collection; and Copan, Italy, for providing us with flocked swabs for the present study. We thank Anders Mørup Jensen, Department of Biostatistics, Statens Serum Institut, for statistical assistance. The results from this study were presented at ECCMID 2008 in Barcelona, Spain (P654) and at SSAC 2008 in Copenhagen, Denmark. Transparency Declaration This project was sponsored by Statens Serum Institut and the former Vejle County, Denmark. Further Copan, Italy, sponsored the flocked swabs used for sampling. The authors declare that they have no conflicting interests. Supporting Information Additional Supporting Information may be found in the online version of this article: Table S1. Retrospective data on the 16 MRSA index persons and the results obtained from the household survey and follow up. Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article. References 1. Hardy KJ, Oppenheim BA, Gossain S, Gao F, Hawkey PM. A study of the relationship between environmental contamination with methicillin-resistant Staphylococcus aureus (MRSA) and patients acquisition of MRSA. Infect Control Hosp Epidemiol 2006; 27: 127 132. 2. Loeffler A, Boag AK, Sung J et al. Prevalence of methicillin-resistant Staphylococcus aureus among staff and pets in a small animal referral hospital in the UK. J Antimicrob Chemother 2005; 56: 692 697. 3. Johnson AP, Aucken HM, Cavendish S et al. Dominance of EMRSA-15 and -16 among MRSA causing nosocomial bacteraemia in the UK: analysis of isolates from the European Antimicrobial Resistance Surveillance System (EARSS). J Antimicrob Chemother 2001; 48: 143 144. 4. Griffiths C, Lamagni TL, Crowcroft NS, Duckworth G, Rooney C. Trends in MRSA in England and Wales: analysis of morbidity and mortality data for 1993 2002. Health Stat Q 2004; 21: 15 22. 5. Dancer SJ. Considering the introduction of universal MRSA screening. J Hosp Infect 2008; 69: 315 320. 6. Amorim ML, Faria NA, Oliveira DC et al. Changes in the clonal nature and antibiotic resistance profiles of methicillin-resistant Staphylococcus aureus isolates associated with spread of the EMRSA-15 clone in a tertiary care Portuguese hospital. J Clin Microbiol 2007; 45: 2881 2888. 7. Boyce JM, Cookson B, Christiansen K et al. Meticillin-resistant Staphylococcus aureus. Lancet Infect Dis 2005; 5: 653 663. 8. Smith JM, Cook GM. A decade of community MRSA in New Zealand. Epidemiol Infect 2005; 133: 899 904.

1434 Clinical Microbiology and Infection, Volume 16 Number 9, September 2010 CMI 9. Ornskov D, Kolmos B, Bendix HP, Nederby NJ, Brandslund I, Schouenborg P. Screening for methicillin-resistant Staphylococcus aureus in clinical swabs using a high-throughput real-time PCR-based method. Clin Microbiol Infect 2008; 14: 22 28. 10. Larsen AR, Stegger M, Sorum M. spa typing directly from a meca, spa and pvl multiplex PCR assay-a cost-effective improvement for methicillin-resistant Staphylococcus aureus surveillance. Clin Microbiol Infect 2008; 14: 611 614. 11. Harmsen D, Claus H, Witte W et al. Typing of methicillin-resistant Staphylococcus aureus in a university hospital setting by using novel software for spa repeat determination and database management. J Clin Microbiol 2003; 41: 5442 5448. 12. Kondo Y, Ito T, Ma XX et al. Combination of multiplex PCRs for SCCmec type assignment: rapid identification system for mec, ccr, and major differences in junkyard regions. Antimicrob Agents Chemother 2006; 51: 264 274. 13. Robinson DA, Enright MC. Multilocus sequence typing and the evolution of methicillin-resistant Staphylococcus aureus. Clin Microbiol Infect 2004; 10: 92 97. 14. Murchan S, Kaufmann ME, Deplano A et al. Harmonization of pulsedfield gel electrophoresis protocols for epidemiological typing of strains of methicillin-resistant Staphylococcus aureus: a single approach developed by consensus in 10 European laboratories and its application for tracing the spread of related strains. J Clin Microbiol 2003; 41: 1574 1585. 15. Vriens MR, Blok HE, Gigengack-Baars AC et al. Methicillin-resistant Staphylococcus aureus carriage among patients after hospital discharge. Infect Control Hosp Epidemiol 2005; 26: 629 633. 16. Beaujean DJ, Weersink AJ, Blok HE, Frenay HM, Verhoef J. Determining risk factors for methicillin-resistant Staphylococcus aureus carriage after discharge from hospital. J Hosp Infect 1999; 42: 213 218. 17. Datta R, Huang SS. Risk of infection and death due to methicillinresistant Staphylococcus aureus in long-term carriers. Clin Infect Dis 2008; 47: 176 181. 18. Nilsson P, Ripa T. Staphylococcus aureus throat colonization is more frequent than colonization in the anterior nares. J Clin Microbiol 2006; 44: 3334 3339. 19. Widmer AF, Mertz D, Frei R. Necessity of screening of both the nose and the throat to detect methicillin-resistant Staphylococcus aureus colonization in patients upon admission to an intensive care unit. J Clin Microbiol 2008; 46: 835. 20. Mertz D, Frei R, Periat N et al. Exclusive Staphylococcus aureus throat carriage: at-risk populations. Arch Intern Med 2009; 169: 172 178. 21. Solberg CO. Spread of Staphylococcus aureus in hospitals: causes and prevention. Scand J Infect Dis 2000; 32: 587 595. 22. Böcher S, Middendorf B, Westh H, Mellman A, Becker K, Skon R, Friedrich AW. Semi-selective broth improves screening for methicillin-resistant Staphylococcus aureus. J Antimicrob Chemother 2010; feb 3 [Epub ahead of print]. 23. Garzoni C, Kelley WL. Staphylococcus aureus: new evidence for intracellular persistence. Trends Microbiol 2009; 17: 59 65. 24. Wertheim HF, Verveer J, Boelens HA, van BA, Verbrugh HA, Vos MC. Effect of mupirocin treatment on nasal, pharyngeal, and perineal carriage of Staphylococcus aureus in healthy adults. Antimicrob Agents Chemother 2005; 49: 1465 1467. 25. Buehlmann M, Frei R, Fenner L, Dangel M, Fluckiger U, Widmer AF. Highly effective regimen for decolonization of methicillin-resistant Staphylococcus aureus carriers. Infect Control Hosp Epidemiol 2008; 29: 510 516. 26. Simor AE, Phillips E, McGeer A et al. Randomized controlled trial of chlorhexidine gluconate for washing, intranasal mupirocin, and rifampin and doxycycline versus no treatment for the eradication of methicillin-resistant Staphylococcus aureus colonization. Clin Infect Dis 2007; 44: 178 185. 27. Kniehl E, Becker A, Forster DH. Bed, bath and beyond: pitfalls in prompt eradication of methicillin-resistant Staphylococcus aureus carrier status in healthcare workers. J Hosp Infect 2005; 59: 180 187. 28. French GL, Otter JA, Shannon KP, Adams NM, Watling D, Parks MJ. Tackling contamination of the hospital environment by methicillinresistant Staphylococcus aureus (MRSA): a comparison between conventional terminal cleaning and hydrogen peroxide vapour decontamination. J Hosp Infect 2004; 57: 31 37. 29. Voss A, Loeffen F, Bakker J, Klaassen C, Wulf M. Methicillin-resistant Staphylococcus aureus in pig farming. Emerg Infect Dis 2005; 11: 1965 1966. 30. Weese JS, Caldwell F, Willey BM et al. An outbreak of methicillinresistant Staphylococcus aureus skin infections resulting from horse to human transmission in a veterinary hospital. Vet Microbiol 2006; 114: 160 164. 31. van Duijkeren E, Wolfhagen MJ, Box AT, Heck ME, Wannet WJ, Fluit AC. Human-to-dog transmission of methicillin-resistant Staphylococcus aureus. Emerg Infect Dis 2004; 10: 2235 2237.