Eur J Clin Microbiol Infect Dis (2000) 19 :101–107 Q Springer-Verlag 2000

Article

Molecular Epidemiology of Methicillin-ResistantStaphylococcus aureus in Finland

S. Salmenlinna, O. Lyytikäinen, P. Kotilainen, R. Scotford, E. Siren, J. Vuopio-Varkila

Abstract This study reports the recent trends in the occurrence of methicillin-resistant Staphylococcus aureus in Finland, with special focus on characterization ofthe strains linked to interhospital epidemics and local outbreaks. Between 1981 and1997, the annual number of methicillin-resistant Staphylococcus aureus isolationsranged from 89 to 272. Of all blood isolates of Staphylococcus aureus reported to theNational Infectious Disease Register during the period 1995–97 (np2049), only sixwere resistant to methicillin. Between 1992 and 1997, typing analysis by variousmethods (i.e., antibiogram, phage typing, ribotyping, and pulsed-field gel electro-phoresis) identified 18 different strains capable of causing intrahospital outbreaks orinterhospital epidemics. These 18 strains were separated into 13 different ribotypesand 14 major pulsed-field gel electrophoresis types. Multiresistance was investigatedas a possible marker for epidemicity. Eight of the ten interhospitally spread strainswere multiresistant compared to only three of the eight intrahospitally spreadoutbreak strains. More than one-third of the epidemic and local outbreak strainswere suspected to be of foreign origin. The majority (6 of 10) of the epidemics werelocalized in southern and western Finland, and the largest epidemic, which occurredin the Helsinki metropolitan area, involved over 200 persons. Thus far, the epidemicshave remained primarily intracity problems, and only two strains have becomeendemic.

S. Salmenlinna (Y), R. Scotford, E. Siren, J. Vuopio-VarkilaNational Public Health Institute, Department of Bacteriology,Mannerheimintie 166, 00300 Helsinki, Finlande-mail: Saara.Salmenlinna6ktl.fi

O. LyytikäinenNational Public Health Institute, Department of InfectiousDiseases Epidemiology, Mannerheimintie 172, 00300 Helsinki,Finland

P. KotilainenTurku University Central Hospital, Department of Medicine,Kiinamyllynkatu 4–8, 20520 Turku, Finland

Introduction

Methicillin-resistant Staphylococcus aureus (MRSA)has become a major problem in many hospitals world-wide [1, 2]. The prevalence of MRSA in hospitals in theUSA increased from 2.4% in 1975 to 35% in 1996 [3, 4].Similar trends have also been reported in Europeancountries; Italy reported an increase from 6% in 1981to 26% in 1986 and Germany an increase from 1.7% in1990 to 8.7% in 1995 [5, 6]. However, some other coun-

tries have succeeded in controlling MRSA. The preva-lence of MRSA in the Netherlands has leveled off [7],and Denmark has even reported a decline [8]. As awhole, the situation has been better in Scandinaviathan in southern Europe [2].

The increase in MRSA prevalence may result from thespread of individual strains. In some countries, MRSAstrains have spread from one hospital to another, whichhas led to a nationwide problem [5, 9, 10]. CertainMRSA strains have even been shown to spreadbetween continents [11]. Furthermore, the eradicationof endemic MRSA strains, the presence of whichresults from a long-term increase in MRSA numbers,has proved difficult [12, 13].

In recent years, several genotypic typing methods havebeen developed for the characterization of MRSAstrains. Of these, one of the best for epidemiologicalinvestigation is pulsed-field gel electrophoresis (PFGE)[14–22]. However, a combination of different typingmethods is often needed to obtain a reliable analysis of

102

strains that are suspected to be epidemiologicallyrelated.

In the present study, we report the recent trends of theMRSA occurrences in Finland and describe the molec-ular characteristics of MRSA strains linked to localoutbreaks in hospitals as well as to interhospitalepidemics from 1992 to 1997. The strains were charac-terized with the help of antimicrobial susceptibilitytesting, phage typing, ribotyping, and PFGE.

Materials and Methods

Surveillance of Methicillin-Resistant Staphylococcus aureus inFinland. Before 1992, data on annual numbers of MRSA isola-tions was collected voluntarily on a weekly basis and was avail-able only from the larger microbiology laboratories, whichincluded those of all central and university hospitals as well as thelarger private laboratories. Since 1992, all microbiology laborato-ries have been required to send MRSA isolates to the Staphylo-coccal Reference Laboratory at the National Public Health Insti-tute (KTL) for further study. Since 1995, the laboratories havealso been required to report MRSA isolations as well as all bloodand cerebrospinal fluid findings of Staphylococcus aureus to theNational Infectious Disease Register at the National PublicHealth Institute. In this register, the time interval within whichStaphylococcus aureus isolations from the same person are inter-preted as one case is 36 months. The following data concerningeach isolation are reported to the register: age and gender of thepatient, health care institution, date of isolation, and the source ofpositive culture.

Typing Scheme. All MRSA isolates submitted to the Staphylo-coccal Reference Laboratory were typed by phage and tested forantimicrobial susceptibility. Further studies using ribotyping andPFGE were performed if the isolates had any of the followingcriteria: (i) they were known to be linked to a suspected epidemicbased on the background information obtained from the submit-ting hospital, (ii) they possessed an antimicrobial susceptibilitypattern and a phage type similar to the strains identified earlier inthe laboratory, and/or (iii) they could not be typed by phages.

Antimicrobial Susceptibility Testing. The antimicrobial suscepti-bilities were tested after overnight growth at 37 7C on Mueller-Hinton agar plates by the disk diffusion method performedaccording to guidelines recommended by the National Committeefor Clinical Laboratory Standards [23]. The antibiotics testedwere mupirocin, erythromycin, clindamycin, gentamicin, tobra-mycin, chloramphenicol, trimethoprim-sulfamethoxazole, tetracy-cline, rifampin, ciprofloxacin, and vancomycin (Oxoid, Sweden).Resistance to oxacillin was tested by disk diffusion after incuba-tion at 30 7C for 24–48 h. Minimal inhibitory concentrations ofoxacillin were determined by the E test according to the manufac-turer’s instructions (AB Biodisk, Sweden). In strains with low-level resistance to oxacillin (1–6 mg/ml), the presence of the mecAgene was detected by mecA polymerase chain reaction (PCR) asdescribed previously [24].

Phage Typing. Phage typing was performed with the universalset of phages [25] at 1! and 100! routine test dilutions. Phagetyping was performed both with and without heat treatment ofthe bacteria [26]. The phage stock solutions were obtained fromthe Statens Serum Institut, Copenhagen, Denmark. A phagepattern was defined on the basis of the weakest test dilution thatproduced clear lytic reactions. The strains were divided into thefollowing four groups according to the phage reactions recorded:group III phages (6, 42E, 47, 53, 54, 75, 77, 83A, 84, or 85); group

III phages together with other phages; non-group III phages; andthe nontypeable group.

Ribotyping. Genomic DNA was extracted as describedpreviously [27] and digested separately with the restrictionenzymes EcoRI, HindIII, and ClaI (Boehringer Mannheim,Germany) as recommended by the manufacturer. DNA frag-ments were separated by electrophoresis in 0.7% agarose gels at20 V for 17–18 h, transferred to a nylon membrane (BoehringerMannheim) by vacuum blotting, and hybridized with plasmidpKK3535 containing the rrn operon of Escherichia coli [28]according to the manufacturer’s instructions. Digoxigeninlabeling of pKK3535 and detection (DIG labeling and detectionkit, Boehringer Mannheim) of the hybridized probe wereperformed according to the instructions. Digoxigenin-labeledDNA marker III (Boehringer Mannheim) and MluI-digestedCitrobacter koseri 32 [29] were used as molecular weight stand-ards. Ribopatterns were determined by visual comparison andgiven an arbitrary identification letter, an uppercase letter forHindIII, and lowercase letters for EcoRI and ClaI. A differenceof one band in the ribopattern was considered to represent a newpattern. A ribotype was a combination of the ribopatternsachieved by all three restriction enzymes individually.

Pulsed-Field Gel Electrophoresis. PFGE analysis was performedas described by Goering and Duensing [30] and Patel et al. [31]with slight modifications. A 0.5 ml volume of log-phase cells waswashed with TEN buffer (0.1 M Tris [pH 7.5], 0.15 M NaCl, 0.1 MEDTA) and suspended in 0.5 ml of TEN buffer. An equal volumeof 2% Sea Plaque agarose (FMC BioProducts, USA) in TENbuffer was added, and the mixture was cast in block molds. Solid-ification took place for 15 min on ice. For each strain, six blockswere incubated at 37 7C for 3–18 h in 3 ml of EC buffer (6 mMTris-HCl [pH 7.5], 1 M NaCl, 0.1 M EDTA, 0.5% Brij 58, 0.2%Na-deoxycholate, 0.5% sodium lauroyl sarcosine) containinglysostaphin (30 U/ml) (Sigma Chemical, Germany). The blockswere then transferred to 3 ml of ESP buffer (0.5 M EDTA [pH9.5], 1% sodium lauroyl sarcosine) containing proteinase K(250 mg/ml) and incubated overnight at 56 7C. Next, they werewashed at room temperature, once with 3 ml of TE buffercontaining 100 ml of 100 mM phenylmethylsulfonyl fluoride andthree times with plain TE buffer for 30–60 min each time. After30 min of incubation in 200 ml of restriction buffer, a singleagarose block was digested overnight with 10 U of SmaI (Boeh-ringer Mannheim) in 100 ml of restriction buffer. The chromo-somal fragments were separated with a Chef DR III (Bio-Rad,USA) for 24 h on 1% SeaKem agarose gel (FMC BioProducts) at6 V/cm; the pulse times were increased linearly, with the initialswitching time being 10 s and the final switching time being 60 s.A Lambda Ladder PFG marker (New England BioLabs, USA)was used as a molecular weight standard. The similarity of thePFGE patterns was assessed according to the guidelines ofTenover et al. [32].

Strain Definitions. A sporadic strain of MRSA was defined as astrain isolated from one person only and displaying an individualantimicrobial susceptibility pattern and/or phage type. For non-phage typeable strains, the strain relatedness was ascertained byribotyping and by PFGE. MRSA isolates sharing the same typingpatterns and isolated from two or more persons in the samehospital were defined as being representative of a locally spreadoutbreak strain (OMRSA). MRSA isolates sharing the sametyping patterns and isolated from two or more persons at two ormore hospitals were defined as representatives of an epidemicMRSA strain (EMRSA).

Infection Control Measures. In 1995 a group of experts producednational guidelines for the prevention of MRSA in Finland.These recommendations include protocols for rapid identificationof MRSA, prevention of MRSA transmission, and treatment ofMRSA infection and carriage, as well as for the actions to beundertaken when a new case of MRSA infection or carriage is

103

Figure 1 Number of MRSA isolations per year in Finland from1981 to 1997

recognized. For example, patients who have been hospitalizedoutside the Nordic countries are advised to be screened forMRSA carriage and treated in contact isolation until a negativeculture result is obtained. The patient records of patients knownto be colonized or infected with MRSA, should be labeledaccordingly. In case of patient transfers, it is also recommended toinform the new health care facility about the MRSA status of thepatient.

Results

Epidemiology of Methicillin-Resistant Staphylococcusaureus Isolations. On the basis of the laboratory noti-fications received at the National Infectious DiseaseRegister and data received from the major microbio-logy laboratories, the annual numbers of MRSA isola-tions range from 89 to 272 (Figure 1). Between 1995and 1997, which is the period from which more detailedinformation is available, a total of 317 MRSA isolationswere reported to the National Infectious DiseaseRegister. Forty-eight percent of the notifications werereceived from health care institutions located withinthe Helsinki metropolitan area. In addition, the healthcare institutions of three other cities reported morethan ten MRSA isolations each. The majority (51%) ofthe isolates were obtained from persons 60 years of ageor older; 10% of the isolates were from children (~16years). In most cases (95%) MRSA was isolated from asource other than blood, cerebrospinal fluid, or urine:only six of the isolates were recovered from blood.Staphylococcus aureus was isolated from blood a totalof 2049 times. Of these isolations, the number ofMRSA isolations was two in 1995, zero in 1996, andfour in 1997, with the percentage of methicillin resist-ance being 0.3%, 0%, and 0.5%, respectively.

Identification of Methicillin-Resistant Staphylococcusaureus Strains. During the period 1992–97, approxi-mately 900 isolates sent to the Staphylococcal Refer-ence Laboratory were confirmed as MRSA. Of these,202 were regarded as sporadic isolates based on theirunique antimicrobial susceptibility pattern and/orphage type. On the basis of further analysis by ribo-typing and PFGE as well as epidemiological informa-tion received from the hospitals, 143 additional strainsnontypeable by phage could be considered sporadic.Among the remaining 549 strains (61% of the total),the same phenotypic and molecular typing patternswere shared by at least two (2–200) other strains. Thesestrains were classified as EMRSA or OMRSA(Table 1) in accordance with the definitions describedabove.

Ribotyping of Outbreak Methicillin-Resistant Staphylo-coccus aureus Strains and Epidemic Methicillin-Resistant Staphylococcus aureus Strains. Ribotypingwith three different restriction enzymes divided theEMRSA and OMRSA strains into 13 differentcombined types (Table 1). Ribotypes B:b:d andAP:ad:c were found among both EMRSA and

OMRSA strains. Four EMRSA strains (E7, E10, E13,and E14) and one OMRSA strain (O8) were of ribo-type B:b:d. In addition, ribotype AP:ad:c was shared bystrains O11 and E12. The remaining ribotypes wereunique to either the EMRSA or the OMRSA group.

Pulsed-Field Gel Electrophoresis. PFGE analysisdifferentiated the EMRSA and OMRSA strains into 14main clones and eight subclones (Table 1 and Figure 2).Clone F could be further divided into subclones (F2,F3, and F4), three of which (F, F2, and F4) were foundamong the EMRSA strains and one of which (F3) wasfound among the OMRSA strains. Likewise, one of theEMRSA strains was identified as clone N, whereas thesubclone, N2, was an OMRSA strain. Other subtype-level PFGE type variants (A2, A3, E2, and J2; seeparentheses in Table 1) were found only among isolatesfrom some long-lasting epidemics (E1, E5, and E13).PFGE was able to differentiate MRSA strains withidentical ribotypes (Table 1). Strains E13 and E14 aswell as strains O11 and E12 were shown to be differentfrom each other by PFGE. Similarly, strains E7, O8,and E10 were divided into subclones. Interestingly, theribotypes of PFGE types F and N differed from thoseof F2, F3, F4 (ribopattern A:a:e vs. B:b:b), and N2(AG:aa:r vs. BB:g:r). The ribotypes of PFGE subtype-level variants observed during an epidemic remainedconstant.

Phage Typing. Among the EMRSA and OMRSAstrains, eight and seven different phage types, respec-tively, were found (data not shown). Of the tenEMRSA strains, three (E1, E6, and E13) reacted onlywith type III phages, three (E5, E10, and E12) reactedwith other phages besides type III phages, three (E2,E7, and E14) remained nontypeable even after heattreatment, and one (E17) showed only non-group IIIreactions (Table 1). Of the eight OMRSA strains, four(O3, O4, O8, and O18) showed group III reactions, two(O9 and O11) reacted with group III and other phages,and two (O15 and O16) showed non-group III reac-tions. When the heat treatment was used, three addi-tional strains (E13, O16, and O18) were typeable.

104

Table 1 Characteristics of EMRSA and OMRSA strains isolated in Finland

Strainno.

Timeperiod

No. ofcasesa

Geographicallocation

No. of healthcare facilitiesaffected

Ribotype PFGE pattern Phage groupb Multi-resistancec

E1 1992d 1200 Helsinkimetropolitan area

110 I :g :e A (cA2, A3) III yes

E2 1993–94 36 Helsinkimetropolitan area

4 A:a : j B NT yes

E5 1995–97 22 Helsinkimetropolitan area

8 AR:af :q E (cE2) IIIcother yes

E6 1991–92 35 western Finland 2 A:a :e F III yesE7 1992–93 57 western Finland 4 B:b :d F2 NT yesE10 1992–96 32 eastern Finland 6 B:b :d F4 IIIcother yesE12 1990d 43 central Finland 4 AP:ad :c I IIIcother noE13 1994–96 17 central Finland 6 B:b :d J (c J2) III yesE14 1992 5 western Finland 2 B:b :d K NT yesE17 1996 15 northern Finland 2 AG:aa : r N non-III noO3 1994 6 Helsinki

metropolitan area1 F:g :e C III yes

O4 1994–95 3 Helsinkimetropolitan area

1 A:a :z D III yes

O8 1992 2 western Finland 1 B:b :d F3 III yesO9 1993 14 western Finland 1 A:b :d G IIIcother noO11 1993–96 20 eastern Finland 1 AP:ad :c H IIIcother noO15 1993 5 western Finland 1 AF: j : y L non-III noO16 1993 3 western Finland 1 AN:a :e M non-III noO18 1997 7 northern Finland 1 BB:g : r N2 III no

a Number of different persons from whom the strain was isolatedbetween the beginning of the epidemic and the end of theepidemic, or by the end of 1997

b Each strain was categorized within a certain phage group basedon phage reactions recorded (see Methods)

c Resistance to more than three different antibiotic groups otherthan beta-lactam agents

d Epidemic incomplete. Strain is still isolated occasionally

Antimicrobial Susceptibility. Eight of the ten EMRSAstrains were multiresistant (resistant to 13 antibioticgroups other than beta-lactam agents). In contrast, onlythree (38%) of the OMRSA strains were multiresistant.All multiresistant EMRSA and OMRSA strainsshowed resistance to erythromycin, clindamycin,gentamicin, and tobramycin, but sensitivity to chloram-phenicol, tetracycline, rifampicin, and ciprofloxacinvaried. All but two of these strains were sensitive totrimethoprim-sulfamethoxazole. Slight variations in theresistance patterns of isolates from the same epidemicwere occasionally observed (data not shown).

Demographics of Epidemics Caused by Methicillin-Resistant Staphylococcus aureus. Most interhospitalepidemics and half of the local outbreaks occurred inthe southern part of Finland, where one-quarter of allFinns live. The largest epidemic, caused by strain E1,occurred in the Helsinki metropolitan area andinvolved more than 200 persons, including patients andhealth care workers in more than ten hospitals(Table 1). The outbreak started in 1992 and peaked in1994, though the strain is still isolated occasionally. Inaddition, two smaller epidemics (caused by strains E2and E5) occurred in the Helsinki metropolitan area,affecting 36 and 22 persons in four and eight hospitals,respectively. Two other major cities also encounteredepidemics of similar size. One city, in western Finland,had two epidemics (caused by strains E6 and E7) that

affected 35 and 57 persons in two and four health carefacilities, respectively. Another city, in central Finland,experienced two epidemics (caused by strains E12 andE13), with 43 and 17 persons involved in four and sixhealth care facilities, respectively.

Discussion

Our study shows no increase in the annual numbers ofMRSA isolated in Finland during the 9-year periodinvestigated. The proportion of methicillin resistanceamong Staphylococcus aureus strains causing invasiveinfections has also remained low. However, intensivetyping performed between 1992 and 1997 revealedseveral strains that had caused local intrahospitaloutbreaks or interhospital epidemics. On most occa-sions, the spread of strains was successfully controlledand the emergence of MRSA endemicity avoided.

According to the weekly reports collected up to 1995and the notifications submitted to the National Infec-tious Disease Register since 1995, the annual numberof MRSA isolations in Finland has been relatively low.However, a reliable comparison with other countries isdifficult, as proper denominator data, such as hospitaladmissions or patient-days, are rarely available andused in analyses. Some uncertainty also remains aboutthe data collected prior to 1995, since the diagnostic

105

Figure 2 (a) PFGE patterns of the epidemic MRSA (EMRSA)strains. Lanes 1 and 12 Lambda ladder; lanes 2–4 EMRSA strainsE1, E2, and E5, isolated in the Helsinki metropolitan area; lanes5, 6 and 10 strains E6, E7, and E14, from western Finland; lane 7strain E10, from eastern Finland; lanes 8 and 9 strains E12 andE13, from central Finland; and lane 11 strain E17, from northernFinland. (b) PFGE patterns of the outbreak MRSA (OMRSA)strains. Lanes 1 and 10 Lambda ladder; lanes 2 and 3 strains O3and O4, isolated in the Helsinki metropolitan area; lanes 4, 5, 7and 8 strains O8, O9, O15 and O16, from western Finland; lane 6strain O11, from eastern Finland; and lane 9 strain O18, fromnorthern Finland

criteria for detecting methicillin resistance and theguidelines for weekly reporting of MRSA were not welldescribed. Thus, it remains uncertain whether all of thestrains reported were indeed resistant to methicillin.

In Finland, the annual proportion of invasive MRSAisolates was extremely low (0–0.5%) compared to ratesin other countries. In England and Wales, the propor-

tion of invasive MRSA increased dramatically from1.5% in 1989 to up to 31.7% in 1997 [33]. Compared toFinland, other countries also have higher figures: theproportion of invasive MRSA was 3.5% in three Cana-dian tertiary care hospitals during a 6-year period [34],46% in a hospital in Hong Kong during a period of 22months [35], 46% in a Spanish tertiary care universityhospital during 4 years of an extensive outbreak ofMRSA [36], and 36% in another Spanish universityhospital during a 23-month period [37].

Despite the reasonably good situation on a nationallevel in Finland, hospital outbreaks and epidemicscaused by several different MRSA strains haveoccurred. More than half of the strains studied werecapable of spreading from one person to another. Bymeans of phage typing, antimicrobial susceptibilitytesting, ribotyping, and PFGE analysis, 18 differentMRSA strains were identified as causative strains. Tenof these strains were categorized by their capability tocause epidemic interhospital spread (EMRSA) andeight by their capability to cause local outbreaks(OMRSA) in various geographic locations in Finland.

The dissemination of MRSA strains in Finland has,thus far, been mostly an intrahospital or an intracityproblem. Unlike the situation in many other countries[9, 10, 21], there is no single clone (or few clones) thathas spread across the country. Although we were ableto identify several clones capable of spreading, thestrains that spread interhospitally usually remainedwithin the same city. This may be attributable to theless frequent patient transfers between the citieslocated outside the Helsinki metropolitan area.Frequent patient transfers between hospitals have beensuspected to provoke the spread of E1 MRSA inHelsinki. An analogous situation has been observedpreviously in Germany, where several epidemic clonescould be traced to distinct regions of Germany. Subse-quently, these clones disseminated throughout thecountry, mainly due to patient transfers between hospi-tals [6].

To minimize the dissemination of MRSA in Finnishhospitals, national guidelines for MRSA preventionwere established in 1995. In practice, the infectioncontrol methods differ slightly between various hospi-tals, and some hospitals may have been more successfulin controlling MRSA than others. In seven epidemics/outbreaks, the causative MRSA strains have beensuspected to be of foreign origin, as the strain identifiedwas first isolated from a patient who had recently beentransferred from a hospital outside the Nordic coun-tries. However, the actual figure may be higher, sinceinformation on patient transfers has not been collectedsystematically. Currently, it is recommended that allpatients who have been hospitalized outside the Nordiccountries be screened for MRSA upon hospital admis-sion.

106

Although MRSA spread is likely to be higher amonghigh-risk patients in high-risk units with frequentpatient transfers, it has been suggested that someMRSA strains possess a pronounced intrinsic capabilityto spread [D. Blanc et al., 8th European Congress ofClinical Microbiology and Infectious Diseases, 1997,Poster no. 443]. The molecular basis for this epidemicnature, however, remains to be elucidated. One impor-tant feature could be multiresistance, as 80% of allepidemic strains identified in the present study weremultiresistant. No single phage type could be linked toepidemicity, as the majority of the strains exhibitedreactions belonging to phage group III.

According to the present study, endemic MRSA strainsare not yet a major problem in Finland. However, ofthe ten Finnish MRSA strains showing epidemicspread, two are still isolated occasionally. These strainshave persisted since the beginning of the 1990s,although their numbers are decreasing. Both of thesestrains have been found predominantly within theparticular city affected. Periodic existence of dominantMRSA clones has been reported due to the overalldecline in MRSA numbers [38] or due to substitutionof the original clone by others [39]. In the UK, forinstance, the strain that was prevalent in the 1980s,EMRSA-1, was followed by a series of EMRSA strains(EMRSA strains 2–14) by 1986. The EMRSA strainscurrently prevalent in the UK are EMRSA 3, 15, and16 [40–43].

Genomic alterations among the epidemic MRSAisolates analyzed in this study were occasionallydetected by the appearance of PFGE subtype variantsof isolates belonging to the same epidemics. This mayreflect the evolution of epidemic strains during thestudy period. Even more variation within the EMRSAstrains could have been detected if all isolates had beenroutinely typed by PFGE. However, after the charac-teristics of a specific epidemic strain were establishedby all typing methods at the beginning of the epidemic,the subsequent isolates were usually identified on thebasis of phage type and antibiogram alone. The currentguidelines for interpreting PFGE patterns are restrictedto outbreak investigations within a limited time frame[32]. In cases of long-lasting epidemics during whichsubtype variation of the causative strain is expected tooccur, interpretation based on these guidelines may notbe recommendable [44].

In conclusion, the overall number of MRSA hasremained relatively stable in Finland, and, thus far, nosingle clone has dominated or replaced others. Thelimited spread of MRSA is most probably due tosuccessful preventive measures taken both nationallyand locally. A crucial part of infection control is theexchange of information between hospitals. The emer-gence of endemic strains and the increasing numbers ofpatient transfers may, however, challenge the infectioncontrol at both local and national levels in the future.

Acknowledgements This data was presented in part at the 4th

Conference of the Hospital Infection Society, 13–17 September1998, Edinburgh, Scotland, Poster 3.1.2.9.

References

1. Ayliffe GAJ: The progressive intercontinental spread ofmethicillin-resistant Staphylococcus aureus. Clinical Infec-tious Diseases (1997) 24, Supplement 1 :74–79

2. Voss A, Milatovic D, Wallrauch-Schwarz C, Rosdahl VT,Braveny I: Methicillin-resistant Staphylococcus aureus inEurope. European Journal of Clinical Microbiology & Infec-tious Diseases (1994) 13 :50–55

3. Panlilio A, Gulver D, Gaynes R, Banerjee S, Henderson T,Tolson JS, Martone WJ: Methicillin-resistant Staphylococcusaureus in US hospitals, 1975–1991. Infection Control andHospital Epidemiology (1992) 13 :582–586

4. National Nosocomial Infections Surveillance (NNIS) report,data summary from October 1986-April 1996. AmericanJournal of Infection Control (1996) 24 :380–388

5. Schito G, Varaldo P: Trends in the epidemiology and anti-biotic resistance of clinical Staphylococcus strains in Italy – areview. Journal of Antimicrobial Chemotherapy (1988) 21,Supplement C:67–81

6. Witte W, Cuny C, Braulke C, Heuck D, Klare I: Widespreaddissemination of epidemic MRSA in German hospitals. Euro-surveillance (1997) 2 :25–28

7. Frenay H, Schot C, Van Leewen W, Rost J, De Neeling A,Van Klingeren B: Methicillin-resistant staphylococcal infec-tion. Infection spreads between hospitals. British MedicalJournal (1994) 308 :58

8. Rosdahl V, Knudsen A: The decline in methicillin resistanceamong Danish Staphylococcus strains. Infection Control andHospital Epidemiology (1991) 12 :83–88

9. de Lencastre H, Severina E, Milch H, Konkoly Thege M,Tomasz A: Wide geographic distribution of a unique methi-cillin-resistant Staphylococcus aureus clone in Hungarianhospitals. Clinical Microbiology and Infection (1997)3 :289–296

10. Santos Sanches I, Ramirez M, Troni H, Abecassis M, PaduaM, Tomasz A, de Lencastre H: Evidence for the geographicspread of a methicillin-resistant Staphylococcus aureus clonebetween Portugal and Spain. Journal of Clinical Microbiology(1995) 33 :1243–1246

11. Aires de Sousa M, Santos Sanches I, Ferro M, Vaz M, SaraivaZ, Tendeiro T, Serra J, de Lencastre H: Intercontinentalspread of a multidrug-resistant methicillin-resistant Staphylo-coccus aureus clone. Journal of Clinical Microbiology (1998)36 :2590–2596

12. Kauffman C, Terpenning M, He X, Zarins L, Ramsey M,Jorgensen K, Sottile W: Attempts to eradicate methicillin-resistant Staphylococcus aureus from a long-term-care facilitywith use of mupirocin ointment. American Journal of Medi-cine (1993) 94 :371–378

13. Adeyemi-Doro FA, Scheel O, Lyon DJ, Cheng AF: Livingwith methicillin-resistant Staphylococcus aureus: a 7-yearexperience with endemic MRSA in a university hospital.Infection Control and Hospital Epidemiology (1997)18 :765–767

14. Struelens M, Bax R, Deplano A, Quint W, van Belkum A:Concordant clonal delineation of methicillin-resistant Staphy-lococcus aureus by macrorestriction analysis and polymerasechain reaction genome fingerprinting. Journal of ClinicalMicrobiology (1993) 31 :1964–1970

15. van Belkum A, Bax R, Prevost G: Comparison of four geno-typing assays for epidemiological study of methicillin-resistant Staphylococcus aureus strains. European Journal ofClinical Microbiology & Infectious Diseases. (1994)13 :420–424

107

16. Tenover F, Arbeit R, Archer G, Biddle J, Byrne S, GoeringR, Hancock G, Hébert A, Hill B, Hollis R, Jarvis W, Kreis-wirth G, Eisner W, Maslow J, McDougal L, Miller M,Mulligan M, Pfaller M: Comparison of traditional and molec-ular methods of typing isolates of Staphylococcus aureus.Journal of Clinical Microbiology (1994) 32 :407–415

17. van Belkum A, Kluytmans J, van Leeuwen W, Bax R, QuintW, Peters E, Fluit A, Vandenbroucke-Grauls C, van denBrule A, Koeleman H, Melchers W, Meis J, Elaichouni A,Vaneechouette M, Moonens F, Maes N, Struelens M,Tenover F, Verbrugh H: Multicenter evaluation of arbitraryprimed PCR for typing of Staphylococcus aureus strains.European Journal of Clinical Microbiology & InfectiousDiseases (1995) 33 :1537–1547

18. Cuny C, Witte W: Typing of Staphylococcus aureus by PCRfor DNA sequences flanked by transposon Tn 916 targetregion and ribosomal binding site. Journal of Clinical Micro-biology (1996) 34 :1502–1505

19. de Lencastre H, de Lencastre A, Tomasz A: Methicillin-resistant Staphylococcus aureus recovered from New YorkCity Hospital: analysis by molecular fingerprinting tech-niques. Journal of Clinical Microbiology (1996)34 :2121–2124

20. Kumari P, Keer V, Hawkey P, Parnell P, Joseph N, Rich-ardson J, Cookson B: Comparison and application of ribo-some spacer DNA amplicon polymorphisms and pulsed-fieldgel electrophoresis for differentiation of Staphylococcusaureus strains. Journal of Clinical Microbiology (1997)35 :881–885

21. Trzcinski K, van Leeuwen W, van Belkum A, Grzesiowski P,Klytmans J, Sijmons M, Verbrugh H, Witte W, HryniewiczW: Two clones of methicillin-resistant Staphylococcus aureusin Poland. Clinical Microbiology and Infection (1997)3 :198–207

22. Van Belkum A, van Leeuwen W, Kaufman ME, Cookson B,Forey F, Etienne J, Goering R, Tenover F, Steward C,O’Brien F, Grubb W, Tassios P, Legakis N, Morvan A, ElSolh N, de Ryck R, Struelens M, Salmenlinna S, Vuopio-Varkila J, Kooistra M, Talens A, Witte W, Verbrugh H:Assessment of resolution and intercenter reproducibility ofStaphylococcus aureus genotyping by pulsed-field gel electro-phoresis of SmaI macrorestriction fragments: a multi-centerstudy. Journal of Clinical Microbiology (1998) 36 :1653–1659

23. National Committee for Clinical Laboratory Standards:Performance standards for antimicrobial disk susceptibilitytests. Approved standard M2-A6. NCCLS, Villanova, PA(1997)

24. Murakami K, Minamide W, Wada K, Nakamura E, TeraokaH, Watanabe S: Identification of methicillin-resistant strainsof staphylococci by polymerase chain reaction. Journal ofClinical Microbiology (1991) 29 :2240–2244

25. Parker TM: Phage typing of Staphylococcus aureus. In: NorrisJR, Ribbons DW (eds): Methods in microbiology. AcademicPress, London (1972) pp 1–28

26. Davies H, Martin D: Heat shocking as a useful adjunct toroutine phage typing of Staphylococcus aureus. Journal ofHospital Infection (1987) 10 :4–9

27. Pitcher D, Saunders N, Owen R: Rapid extraction of bacterialgenomic DNA with guanidium thiocyanate. Letters inApplied Microbiology (1989) 8 :151–156

28. Brosius J, Ullrich A, Raker MA, Gray A, Dull T, Gutell R,Noller H: Construction and fine mapping of recombinantplasmids containing the rrn B ribosomal RNA operon of E.coli. Plasmid (1981) 6 :112–118

29. Grimont F, Grimont PAD: Determination of rRNA generestriction patterns. In: Howard J, Whitecombe DM (eds):Diagnostic bacteriology protocols. Humana Press, Totowa,NJ (1995) pp 181–200

30. Goering R, Duensing T: Rapid field inversion gel electro-phoresis in combination with a rRNA gene probe in theepidemiological evaluation of staphylococci. Journal of Clin-ical Microbiology (1990) 28 :426–429

31. Patel A, Foster T, Pattee P: Physical and genetic mapping ofthe protein A gene in the chromosome of Staphylococcusaureus 8325-4. Journal of General Microbiology (1989)135 :1799–1807

32. Tenover F, Arbeit R, Goering R, Mickelsen P, Murray B,Pershing D, Swaminathan B: Interpreting chromosomalDNA-restriction patterns produced by pulsed-field gel elec-trophoresis: criteria for bacterial strain typing. Journal ofClinical Microbiology (1995) 33 :2233–2239

33. Johnson AP, James D, Livermore DM: Increasing prevalenceof methicillin-resistant Staphylococcus aureus blood cultureisolates. Journal of Antimicrobial Chemotherapy (1999)43 :160

34. Suh K, Toye B, Jessamine P, Chan F, Ramotar K: Epidemio-logy of methicillin-resistant Staphylococcus aureus in threeCanadian tertiary care centers. Infection Control andHospital Epidemiology (1998) 19 :395–400

35. Cheng F, French G: Methicillin-resistant Staphylococcusaureus bacteremia in Hong Kong. Journal of Hospital Infec-tion (1988) 12 :91–101

36. Romero-Vivas J, Rubio M, Fernandez C, Picazo J: Mortalityassociated with nosocomial bacteremia due to methicillin-resistant Staphylococcus aureus. Clinical Infectious Diseases(1995) 21 :1417–1423

37. Pujol M, Pena C, Pallares R, Ayats J, Ariza J, Gudiol F: Riskfactors for nosocomial bacteremia due to methicillin-resistantStaphylococcus aureus. European Journal of Clinical Micro-biology & Infectious Diseases (1994) 13 :96–102

38. Frimodt-Moller N, Espersen F, Skinhrj P, Rosdahl V:Epidemiology of Staphylococcus aureus bacteremia inDenmark from 1957 to 1990. Clinical Microbiology and Infec-tion (1997) 3 :297–305

39. Aires de Sousa M, Santos Sanches I, van Belkum A, vanLeeuwen W, Verbrugh H, de Lencastre H: Characterizationof methicillin-resistant Staphylococcus aureus isolates fromPortuguese hospitals by multiple genotyping methods. Micro-bial Drug Resistance (1996) 2 :331–341

40. Marples R, Cooke M: Current problems with methicillin-resistant Staphylococcus aureus. Journal of Hospital Infection(1988) 11 :381–392

41. Richardson J, Reith S: Characterization of a strain of methi-cillin-resistant Staphylococcus aureus (EMRSA-15) byconventional and molecular methods. Journal of HospitalInfection (1993) 25 :45–52

42. Cox R, Conquest C, Mallaghan C, Marples R: A majoroutbreak of methicillin-resistant Staphylococcus aureuscaused by a new phage type (EMRSA-16). Journal ofHospital Infection (1995) 29 :87–106

43. Hookey J, Richardson J, Cookson B: Molecular typing ofStaphylococcus aureus based on PCR restriction fragmentlength polymorphism and DNA sequence analysis of thecoagulase gene. Journal of Clinical Microbiology (1998)36 :1083–1089

44. Van Leeuwen W, van Belkum A, Kreiswirth B, Verbrugh H:Genetic diversification of methicillin-resistant Staphylococcusaureus as function of prolonged geographic dissemination andas measured by binary typing and other genotyping methods.Research in Microbiology (1998) 149 :497–507