lable at ScienceDirect

American Journal of Infection Control 42 (2014) 1285-90

Contents lists avai

American Journal of Infection Control

journal homepage: www.aj ic journal .org

American Journal of Infection Control

Major article

Methicillin-resistant Staphylococcus aureus in public transportationvehicles (buses): Another piece to the epidemiologic puzzle

Jonathan K. Lutz PhD, CIH a, Joany van Balen DVMb, John Mac Crawford PhD a,John R. Wilkins III DrPH c, Jiyoung Lee PhD a,d, Rocio C. Nava-Hoet DVM, MSc b,Armando E. Hoet DVM, PhD b,c,*aDivision of Environmental Health Sciences, College of Public Health, The Ohio State University, Columbus, OHbDepartment of Veterinary Preventive Medicine, College of Veterinary Medicine, The Ohio State University, Columbus, OHcDivision of Epidemiology, College of Public Health, The Ohio State University, Columbus, OHdDepartment of Food Science and Technology, The Ohio State University, Columbus, OH

Key Words:Methicillin-resistant Staphylococcus aureusPublic transportationSurface contamination

* Address correspondence to Armando E. Hoet, D1920 Coffey Rd, Columbus, OH 43221.

E-mail address: hoet.1@osu.edu (A.E. Hoet).Conflicts of interest: None to report.

0196-6553/$36.00 - Copyright � 2014 by the Associahttp://dx.doi.org/10.1016/j.ajic.2014.08.016

Background: Little is known about the occurrence and epidemiology of methicillin-resistant Staphylo-coccus aureus (MRSA) in public transportation in the United States. This research sought to determine thebackground prevalence and phenotypic and genotypic characteristics of MRSA strains circulating onbuses from a large, metropolitan transportation agency.Methods: Electrostatic wipes were used to collect 237 surface samples from 40 buses randomly selectedfrom July-October 2010. Six samples were collected from each bus immediately postservice and beforeany cleaning and disinfection. Positive isolates were analyzed for antibiotic resistance, staphylococcalcassette chromosome mec (SCCmec) type, and pulsed-field gel electrophoresis; and potential epidemi-ologic factors were examined.Results: Of the buses, 68% (27/40) were contaminated with S aureus, and 63% (25/40) were contaminatedwith MRSA. Seats and seat rails were the surfaces most frequently contaminated, followed by the backdoor and stanchions. Most (62.9%) of the MRSA isolates were classified as community-associated MRSAclones (SCCmec type IV), and 22.9% were health careeassociated MRSA clones (SCCmec type II). Of theMRSA strains, 65% (5/20) were multidrug resistant.Conclusion: MRSA was frequently isolated from commonly touched surfaces in buses serving bothhospital and community routes. Phenotypic and genotypic analysis demonstrated that buses may beeffective mixing vessels for MRSA strains of both community and health careeassociated origin.

Copyright � 2014 by the Association for Professionals in Infection Control and Epidemiology, Inc.Published by Elsevier Inc. All rights reserved.

Public transportation is perhaps one of the most importantservices in urban life. This critical infrastructure is widely used,with approximately 10.5 billion trips per year in the United States.1

However, little data exist regarding the prevalence of pathogens onpublic transportation vehicle surfaces in the United States, and nostudies have isolated methicillin-resistant Staphylococcus aureus(MRSA) from public transportation vehicles in this country.2 Pre-vious work has reported the isolation of MRSA on public trans-portation vehicles, but such research has only been conducted inEurope and Asia.3-7 In prior reports from Europe, MRSA was

VM, PhD, A100Q Sisson Hall,

tion for Professionals in Infection C

isolated in 2 studies (both in Portugal), with 26% (22/85) and 36%(72/199) of sampled buses testing positive for MRSA contamina-tion.3-6 MRSA contamination was also found in Japanese trains,with 2.3% of vehicles found as positive.7

Thereare several factors thatmakepublic transportationvehiclesan ideal setting for the movement and spread of MRSA. First, thereare undoubtedly colonized and infected individuals using publictransportation because the U.S. colonization rate is 0.8%-1.5%.8,9

Second, hand-to-fomite contact is expected in public trans-portation vehicles. This type of contact has been previously impli-cated in community-associated (CA) MRSA transmission.10 Ridersroutinely touch stanchions, seat rails, doors, and seats, especiallyduring high-volume usage times (as vehicles become crowded, thishand-to-fomite contact increases). Finally, there is little to no op-portunity for hand hygiene during and immediately after

ontrol and Epidemiology, Inc. Published by Elsevier Inc. All rights reserved.

Table 1Descriptive statistics and comparisons for the presence of Staphylococcus aureus and MRSA on surfaces sampled in buses from a top urban regional transit authority in a large,Midwestern U.S. city

VariableS aureus,n (%)

Negative S aureus,n (%) OR (95% CI)

MRSA,n (%)

Negative MRSA,*n (%) OR (95% CI)

Surface samples 41 (17.3) 196 (82.7) NA 35 (14.8) 202 (85.2) NASampling date c2 and Fisher exact tests (P ¼ .134) c2 and Fisher exact tests (P ¼ .273)July 12y 14 (24.6) 43 (75.4) 2.5 (0.9-6.7) 13 (22.8) 44 (77.2) 2.2 (0.8-6.1)August 23 13 (21.7) 47 (78.3) 2.1 (0.8-5.7) 8 (13.3) 52 (86.7) 1.2 (0.4-3.4)October 11 7 (11.7) 53 (88.3) 1.0 (0.3-3.1) 7 (11.7) 53 (88.3) 1.0 (0.3-3.1)October 25 7 (11.7) 53 (88.3) Referent 7 (11.7) 53 (88.3) Referent

Facility c2 and Fisher exact tests (P ¼ .864) c2 and Fisher exact tests (P ¼ .363)1 21 (17.9) 96 (82.1) 1.1 (0.6-2.1) 20 (17.1) 97 (82.9) 1.4 (0.7-2.9)2 20 (16.7) 100 (83.3) 15 (12.5) 105 (87.5)

Sample location c2 and Fisher exact tests (P < .001)z c2 and Fisher exact tests (P < .01)z

Seats 13 (32.5) 27 (67.5) 5.9 (1.5-22.9) 13 (32.5) 27 (67.5) 18.8 (2.3-152.2)Seat rails 13 (35.1) 24 (64.9) 6.7 (1.7-25.9) 11 (29.7) 26 (70.3) 16.5 (2.0-135.6)Back door 7 (17.5) 33 (82.5) 2.6 (0.6-10.9) 6 (15.0) 34 (85.0) 6.9 (0.8-60.1)Stanchions 5 (12.5) 35 (87.5) 1.8 (0.4-7.9) 4 (10.0) 36 (90.0) 4.3 (0.5-40.6)Operator’s area 3 (7.5) 37 (92.5) Referent 1 (2.5) 39 (97.5) ReferentHVAC return vent 0 (0.0) 40 (100) NA 0 (0.0) 40 (100.0) NA

Multiple routes c2 and Fisher exact tests (P ¼ .605) c2 and Fisher exact tests (P ¼ .712)Multiple routes 26 (18.4) 115 (81.6) 1.2 (0.6-2.5) 22 (15.6) 119 (84.4) 1.2 (0.6-2.5)Same route for day 15 (15.6) 81 (84.4) 13 (13.5) 83 (86.5)

Hospital or nonhospital route c2 and Fisher exact tests (P ¼ 1.00) c2 and Fisher exact tests (P ¼ 1.00)Hospital 27 (17.4) 128 (82.6) 1.0 (0.5-2.1) 23 (14.8) 132 (85.2) 1.0 (0.5-2.2)Nonhospital 14 (17.1) 68 (82.9) 12 (14.6) 70 (85.4)

Ridership c2 and Fisher exact tests (P ¼ .173) c2 and Fisher exact tests (P ¼ .258)�200/d 24 (20.0) 96 (80.0) 1.5 (0.7-2.9) 20 (16.7) 100 (83.3) 1.4 (0.7-2.8)0-199/d 17 (14.5) 100 (85.5) 15 (12.8) 102 (87.2)

CI, confidence interval; HVAC, heating, ventilating, and air conditioning; MRSA, methicillin-resistant Staphylococcus aureus; NA, not applicable; OR, odds ratio.*Includes methicillin-susceptible S aureus and all negative results.yOn July 12, 57 samples were collected because 3 buses did not have seat rails; 60 samples were collected for each of the remaining sampling dates.zStatistically significant c2 and Fisher exact tests at P � .05 level.

J.K. Lutz et al. / American Journal of Infection Control 42 (2014) 1285-901286

transportationusage. Previous researchhasdemonstrated thathandhygiene is the primary method for the prevention of MRSAtransmission.11,12

Little is known about the background prevalence or variation ofMRSA strains circulating in the community, especially on fomitesurfaces in public transportation vehicles. In this research, the aimwas to determine the background prevalence and phenotypic andgenotypic characteristics of MRSA strains contaminating publictransportation vehicle surfaces in a large, metropolitan setting.Here, the results of a cross-sectional study conducted on 40 busesfrom a Midwestern United States transportation agency aredescribed. The results of this work provide critical data on theepidemiologic characteristics and clonal distribution of MRSAstrains contaminating a public environment in this importantcommunity setting.

METHODS

Transportation agency

The studied transportation agency is a top 50 (U.S.) urban-anchored regional transit authority from a large, metropolitanarea. The research teamsampled17% (40/239) of the busfleet (basedon a daily average for each facility), including the selection of anequal number of buses from each depot facility (facilities 1 and 2).Specific routes were targeted for sampling using the transportationagency’s system map. Such routes were chosen based on thefollowing criteria: (1) those that served major hospitals andnonhospital-related routes (to evaluate the possible variation inhospital-associated [HA] vs CA strains), (2) routes with high rider-ship and low ridership (to determine potential effect of crowdingand human density), and (3) those that served 1 route or multipleroutes (to assess possible differences in the number of strainsisolated).

Surfaces sampled and methodology

Sampling was conducted over 4 sampling days, starting in July2010 and ending in October 2010 (Table 1). The sampling dateswere chosen based on project goals and the transportation agency’savailability; each facility was visited twice. In each sampling date,10 buses were randomly selected from those available to besampled at the time of visit. Six predetermined sample locations(stanchions; seats; seat rails; back door; driver’s area; heating,ventilating, and air conditioning return vent) were collected fromeach bus (total of 240 surfaces). Sample locations inside the buseswere chosen as those having potentially high levels of skin-to-surface contact.

Sampling was conducted using commercially available elec-trostatic wipes (Swiffer, Procter & Gamble, Cincinnati, OH), aspreviously described.13 A wipe was unfolded and placed over thesurface to be sampled in a manner which maximized surface areacontact. Once the sample was collected, the wipe was folded andplaced into a sterile, labeled stomacher bag (Nasco, Fort Atkinson,WI). To enhance sampling efficiency, some samples werecollected as a pool. These pooled (composite) samples werecollected from the stanchions, seats, seat rails, and vehicle op-erator’s area (which included the steering wheel, arm rests,knobs, and headrest). For pooled samples, the same electrostaticcloth was used to wipe all the units of the same surface type (eg,9-10 seat rails were sampled with the same cloth), and anattempt was made to sample the same number of seats, seat rails,and similar linear footage of stanchions per bus. Sampling wasalways conducted by the same researcher, focusing only on thesurfaces on one side (driver’s side) of the bus (to control for anypotential differences between the 2 sides, which were of differentconfiguration). All sampling was conducted while vehicles wereimmediately postservice and before any cleaning anddisinfection.

J.K. Lutz et al. / American Journal of Infection Control 42 (2014) 1285-90 1287

Strict quality control standards were assured throughout thesampling process (such as, but not limited to, technician screeningfor MRSA colonization, use of N-95 masks, and glove changes andapplication of alcohol-based hand sanitizer between each samplecollected). In addition, for each collection date, 2 negative controlwipes (field blanks) and a positive control (MRSA ATCC 43300) wereheld for processing. All collected samples were transported andprocessed at the Diagnostic and Research Laboratory for InfectiousDiseases at The Ohio State University College of VeterinaryMedicine.

Isolation, identification, and phenotyping

All samples were transported to the laboratory for processingwithin 1 hour after each sampling event. The details of processingare described elsewhere.13 Briefly, all samples were placed in apreenrichment media (Trypticase Soy Broth, BD Diagnostic Systems,Sparks, MD) at 35�C for 24 hours. After pre-enrichment sampleswere transferred to BD Mannitol Salt Agar with Oxacillin (2 mG/mL)plates (BD Diagnostic Systems, Sparks, MD) for 24-48 hours incu-bation. Then,1-3 morphologically uniquemannitol-positive colonieswere selected per sample for purification, isolation, and furtheranalysis. S aureus was identified through morphology, color, size,hemolysis, mannitol fermentation, and the following biochemicaltests: catalase reaction (Catalase, BD BBL; Becton, Dickinson andCompany, Franklin Lakes, NJ); tube coagulase testing (Coagulaseplasma rabbit with EDTA, BD BBL Becton, Dickinson and Company);latex agglutination testing (Sure-Vue Color Staph ID, Biokit USA Inc,Lexington, MA); aniline reaction (ORSAB, Oxoid LTD, Basingstonke,Hampshire, England); Voges-Proskauer test (Vogues-Proskauer re-agents A and B, BD BBL Becton, Dickinson and Company); andpolymyxin B susceptibility (BD Diagnostic Systems). Coagulase-negative isolates were not further processed in this study. AllS aureus isolates underwent susceptibility testing to determinemethicillin phenotype per the Clinical and Laboratory StandardsInstitute procedure and as described by Van Balen et al.14,15 Fresh,24-hour colonies were inoculated onto oxacillin resistance agarplates containing 6 mG/mL oxacillin (Oxacillin Screen Agar [OSA], BDDiagnostic Systems). All media plates were incubated at 35�C for24 hours. Positive controls were used for all reactions (MRSA: ATCC43300; methicillin-susceptible S aureus [MSSA]: ATCC 29213;methicillin-resistant S pseudintermedius: in-house isolate). Isolateswere classified as MRSA when growth was confirmed on the OSAplates. If no growth was observed on the OSA plates, the isolate wasconsidered MSSA.

Resistance testing was conducted using the Kirby-Bauer diskdiffusion technique (disk concentrations listed accordingly),following the Clinical and Laboratory Standards Institute standardmethods.15 All MSSA and MRSA isolates underwent phenotypinganalysis for antibiotic resistance to 16 antibiotic agents from 9antibiotic classes: aminoglycosides (gentamicin 1 mG, amikacin30 mG); b-lactams (ampicillin 10 mG, amoxicillin-clavulanate 30 mG,oxacillin 1 mG, cephalothin 30 mG, cefpodoxime 10 mG); glycopep-tide (vancomycin); lincosamide (clindamycin 2 mG); tetracyclines(tetracycline 30 mG, doxycycline 30 mG); macrolide (erythromycin15 mG); quinolones (ciprofloxacin 2 mG, enrofloxacin 5 mG); phe-nicol (chloramphenicol 30 mG); and potentiated sulfonamide (sul-famethoxazole-trimethoprim 25 mG). To test for vancomycinresistance, vancomycin screening agar plates (BD BBL VancomycinScreen Agar, BD Diagnostic Systems, Sparks, MD) were inoculatedwith 10 mL of the bacterial solution (0.5 McFarland standard). The Dtest was used to detect inducible clindamycin resistance.16 Controlstrains were used throughout the phenotyping process, includingEnterococcus faecalis (ATCC 29212, ATCC 51299), S aureus (ATCC29253, ATCC 43300, ATCC 29213), Pseudomonas aeruginosa (ATCC27853), and Escherichia coli (ATCC 25922).

Confirmation of mecA, staphylococcal cassette chromosome mectyping, and pulsed-field gel electrophoresis

All isolates phenotypically determined to be MRSAwere furthertested to confirm the presence of the mecA gene. A multiplexpolymerase chain reaction technique was used (as described byOliveira and de Lencastre17 and Milheiriço et al,18 with modifica-tions described by Van Balen et al14) to confirm and type thestaphylococcal cassette chromosome mec (SCCmec) gene. Resultsare reported as types I-VI.

Pulsed-field gel electrophoresis (PFGE) was conducted to furthersubtype isolates, which permitted the determination of potentialepidemiologic origin (HA or CA) and clonal relatedness. S aureusgenomic DNA was digested using a SmaI restriction enzyme, ac-cording to standardized methods (Centers for Disease Control andPrevention-PulseNet). Salmonella serotype Branderup strain H9812was used as a molecular size marker and digested with XbaI. Elec-trophoresiswas conducted using a CHEFMapper XA system (Bio-RadLaboratories, Nazareth, Belgium). All bandpatternswere analyzed forclonal relatedness both visually and with BioNumerics version 6.6software (Applied Maths, Ghent, Belgium). Clonal relatedness wasassigned for all bands using the Dice coefficient and unweighted pairgroup method, using arithmetic averages to achieve dendrogramswith a 1% band position tolerance.19 Dendrogram analysis, includingcluster and pulsotype interpretation, was performed, as describedelsewhere,14 with cutoff points of >80% and >98% similarity,respectively. Designation of USA types was performed by comparingour study isolates to a database obtained from the U.S. Centers forDisease Control and Prevention that contains 100 S aureus strains(representing themost commonbandpatterns fromeachUSA type inthe United States), using �80% similarity as the cutoff point.

Statistical analysis of data

Statistical analysis was conducted using PASW Statistics 18.0(SPSS Inc, Chicago, IL). Data were analyzed using descriptivestatistics and c2 and Fisher exact tests (2-sided c2 and Fisherexact analysis was used for all variables except ridership, for whicha 1-sided analysis was conducted), and simple logistic regressionwas used to determine odds ratios. Total S aureus (MSSA, MRSA)and MRSA prevalence on public transportation vehicles wereexamined on a per vehicle level and overall (per sample level). Forall statistical analyses, P � .05 was considered statistically signifi-cant (P � .10 was considered notable). For assessment of ridership,continuous ridership data were obtained and dichotomized into2 levels (based on a 50% data cut-point) (Table 1): low ridership(0-199 riders/d) and high ridership (�200 riders/d).

RESULTS

A total of 237 surface samples were collected from 40 buses (39unique buses with 1 bus randomly repeated), which were equallydivided between the 2 facilities managed by the transportationagency. All preselected surfaces were collected from each busincluded in the study,with the exceptionof 3 buses that did nothaveseat rails. Approximately 16% of the total active fleet was sampledover the course of the study, covering 35 of the transportationsystem’s 68 routes (51%).

At the bus level, S aureus was isolated in 68% (27/40) of buses,and 63% (25/40) of all buses had at least 1 MRSA-positive surfacesample. At the surface level, S aureus was found in 17% (41/237) ofall surface samples, and MRSA was found in 15% (35/237) of allsurface samples. The S aureus and MRSA results (at bus and surfacelevels) are summarized in Table 1 and Table 2. The small differenceobserved between S aureus andMRSA contamination rates is highlylikely because during the early stages of the isolation process,

Table 2Proportions and ORs of Staphylococcus aureusepositive and MRSA-positive buses from a top urban regional transit authority in a large, Midwestern U.S. city

VariableS aureus,n (%)

Negative S aureus,n (%) OR (95% CI)

MRSA,n (%)

Negative MRSA,*n (%) OR (95% CI)

Buses 27 (67.5) 13 (32.5) NA 25 (62.5) 15 (37.5) NASampling date c2 and Fisher exact tests (P ¼ .134) c2 and Fisher exact tests (P ¼ .233)July 12 9 (90) 1 (10) 9.0 (0.8-100.1) 9 (90) 1 (10) 9.0 (0.8-100.1)August 23 8 (80) 2 (20) 4.0 (0.6-29.0) 6 (60) 4 (40) 1.5 (0.3-8.8)October 11 5 (50) 5 (50) 1.0 (0.8-5.8) 5 (50) 5 (50) 1.0 (0.2-5.8)October 25 5 (50) 5 (50) Referent 5 (50) 5 (50) Referent

Facility c2 and Fisher exact tests (P ¼ 1.00) c2 and Fisher exact tests (P ¼ .514)1 14 (70) 6 (30) 1.3 (0.3-4.7) 14 (70) 6 (30) 1.9 (0.5-7.0)2 13 (65) 7 (35) 11 (55) 9 (45)

Multiple routes c2 and Fisher exact tests (P ¼ .733) c2 and Fisher exact tests (P ¼ .527)Multiple routes 17 (70.8) 7 (29.2) 1.5 (0.4-5.6) 16 (66.7) 8 (33.3) 1.6 (0.4-5.7)Same route for day 10 (62.5) 6 (37.5) 9 (56.3) 7 (43.8)

Hospital or nonhospital route c2 and Fisher exact tests (P ¼ .316) c2 and Fisher exact tests (P ¼ .502)Hospital 16 (61.5) 10 (38.5) 0.4 (0.1-2.0) 15 (57.7) 11 (42.3) 05 (0.1-2.2)Nonhospital 11 (78.6) 3 (21.4) 10 (71.4) 4 (28.6)

Ridership c2 and Fisher exact tests (P ¼ .088) c2 and Fisher exact tests (P ¼ .257)�200/d 16 (80) 4 (20) 3.3 (0.8-13.3) 14 (70) 6 (30) 1.9 (0.5-7.0)0-199/d 11 (55) 9 (45) 11 (55) 9 (45)

CI, confidence interval; HVAC, heating, ventilating, and air conditioning; MRSA, methicillin-resistant Staphylococcus aureus; NA, not applicable; OR, odds ratio.*Includes methicillin-susceptible S aureus and all negative results.

J.K. Lutz et al. / American Journal of Infection Control 42 (2014) 1285-901288

a small amount of oxacillin was used. Therefore, a portion of sus-ceptible S aureus colonies was probably lost during the isolationprocess. Nonetheless, all results obtained (S aureus, MRSA) arementioned and discussed in this article.

The overall number of positive S aureus and MRSA samples andbuses decreased as the study progressed, with the highest contami-nation being found in July (S aureus: 14/57; MRSA: 13/57) and thelowest in October (S aureus: 7/60; MRSA: 7/60). Also, the degree towhich S aureus and MRSA could be isolated varied considerably be-tween surface sample locations on a bus, with the highest levels ofcontamination found on the seat rails (S aureus: 13/40;MRSA: 11/40),followed by the seats (S aureus: 13/40; MRSA: 13/40), back door (Saureus: 7/40; MRSA: 6/40), stanchions (S aureus: 5/40; MRSA: 4/40),andoperator’s area (Saureus: 3/40;MRSA:1/40).No S aureusorMRSAwas found on the heating, ventilating, and air conditioning returnvent (0/40). At the sample and bus levels, multiple variables wereconsidered in the analysis, such as the facility (1, 2), routes covered(single, multiple), if hospitals were served (hospital, nonhospital),and ridership. None of these factors demonstrated a significantdifference in their association with S aureus or MRSA surfacecontamination (Table 1 and Table 2). However, high ridership busesdemonstrated a notable difference in S aureus isolation comparedwith low ridership buses (16/20 vs 11/20, respectively).

Antibiotic resistance profiles, SCCmec typing, and PFGE

From the 35 MRSA-positive surfaces sampled, a total of 76 MRSAisolates were obtained and phenotypically characterized. Based ontheir morphologic characteristics and antimicrobial susceptibilitypatterns, only 1 unique isolate from each positive surface sampledwas selected to be genotypically characterized (35 isolates). Detailedinformation of these 35 isolates (locations, surface description,phenotypic profile, SCCmec type, USA type) can be found in Figure 1.Most of the isolates (62.9%, 22/35) were classified as SCCmec type IVand eitherUSA300orUSA400, characteristics present inMRSAclonesknown as CA MRSA. Moreover, 22.9% (8/35) of the isolates wereSCCmec type II and USA100, characteristics associatedwith HAMRSAclones. The results of the dendrogram analysis indicated 15 uniquepulsotypes,with 3 clusters identified. Over 45% (16/35) of the isolateswere grouped in cluster 2, and 26% (9/35) were in cluster 1 (Fig 1).

Based on the unique combination of their antimicrobial sus-ceptibility pattern (SCCmec type and PFGE pulsotype) 20 MRSAstrainswere identified. Besides b-lactam resistance, 80.0% (16/20) of

the strains were resistant to erythromycin, 50.0% (10/20) wereresistant to ciprofloxacin, and 30.0% (6/20) were resistant to enro-floxacin. Moreover, clindamycin resistance was observed in 45.0%(9/20) of the strains, where 33.3% (3/9) were considered inducible.Of the strains, 65% (13/20) weremultidrug resistant (resistant to�3classes, as defined in Magiorakos et al).20 Of the strains, 25% (5/20)were resistant to 4 classes, and 1 strain was resistant to 5 antibioticclasses. All strains were susceptible to amikacin, gentamicin,doxycycline, sulfamethoxazole-trimethoprim, and vancomycin.

DISCUSSION

Little is known about the overall occurrence and characteristics ofMRSAonpublic transportation vehicle surfaces in theUnited Sates. Inthis work, themain objective was to determine if buses could harborMRSA contamination. Further, if MRSA was isolated, the researchsought to elucidate the phenotypic and genotypic characteristics ofthis pathogen circulating in the vehicles. S aureus and MRSA strainswere isolated from 68% and 63% of the buses sampled, respectively.The rate of susceptible S aureus contamination reported in this articlecould be underestimated because of the isolation process used.Nevertheless, these findings indicate that public transportation ve-hicles could be a potential source of both S aureus and MRSA to thepeople using this service. Using these data, infection control pro-fessionals and epidemiologists can better understand the dynamicsand epidemiology of MRSA in the community.

To our knowledge, this is the first study to report MRSA on publictransportation vehicles in the United States. MRSA has been previ-ously isolated inpublic transportation systems inEuropeandAsia, andadditional work has resulted in the isolation of susceptible S aureus,but not MRSA.2-5,7 However, considerably higher MRSA prevalencewas noted in the current study than inpreviouswork (MRSA-positivevehicles: 63% vs 2.3%-36%).5-7 Low sample number, limited samplinglocation, less sensitive sampling technique, inhibitory processing,cleaning practices applied to the buses, and other factors may havecontributed to considerable variation between the studies.3-5,7,21,22

Further, different regional MRSA colonization rates in the commu-nities using these public transportation servicesmayhave also greatlyaffected the degree to which MRSA could be isolated on buses.23,24

Finding MRSA in a nonclinical, community setting (eg, publictransportation vehicles) has several important implications. First,these findings increase understanding about the degree to whichMRSA can be found outside of the hospital. Given that 0.8%-1.5% of

Fig 1. Clonal dendrogram of methicillin-resistant Staphylococcus aureus isolates from surfaces on buses from a top 50 urban-anchored regional transit authority in a large Mid-western city. The percentage similarity was calculated with Dice coefficients from the pulsed-field gel electrophoresis data. Band position tolerance and optimization were set at 1%.Amc, amoxicillin with clavulanic acid; Amp, ampicillin; Cep, cephalothin; Chl, chloramphenicol; Cip, ciprofloxacin; Cli, clindamycin; Cpd, cefpodoxime; Eno, enrofloxacin; Ery,erythromycin; Ind, inducible; Oxa, oxacillin; SCC, staphylococcal cassette chromosome; Tet, tetracycline.

J.K. Lutz et al. / American Journal of Infection Control 42 (2014) 1285-90 1289

people (in the United States) are colonized with MRSA, it is nosurprise that MRSA has been isolated on a variety of surfaces in thecommunity.8,9 However, many community studies that resulted inMRSA isolation have been conducted in settings that are health carerelated (eg, fire stations and emergency medical services, prisonclinics), after an outbreak (sauna), or limited to household trans-mission.25-27 Not only is the current study one of a few to findMRSAin public areas of the community, but public transportation alsorepresents a potentially important setting for transmission. Inaddition, despite the fact that previous studies have isolated MRSAfrom the community environment, the bacterium does not appearto be ubiquitous. Several published studies have failed to isolateMRSA in the community, and one may suspect that many moreunpublished studies have also failed to isolate MRSA.3,4,28,29

There were multiple factors that were associated with isolatingS aureus and MRSA from a vehicle. Vehicle sample location was animportant indicator, with seats being the most frequently MRSA-positive surface (33%). These results are consistent with priorwork regarding seat contamination. In a study of fire station sur-faces, couches were found to harbor the highest concentration ofMRSA.27 Further, Felkner et al30 found MRSA contamination on aprison van seat and clinic chair, and Baggett et al25 isolated MRSAfrom a sauna bench. Seat rails, which have a roughened surface thatmay enhance removal of bacteria from hands, were also frequentlyMRSA positive (30%).

Back doors also demonstrated high levels ofMRSA contamination(15%). These surfaces are a model of hand-to-fomite and fomite-to-hand transmission (although actual transmission has not beendemonstrated in this study). To operate the back door, the ridermust

either touch the back door surface (flat palmed) or grasp 2 verticaltubes and push on the door. Therefore, throughout a route, there arecountless occasions where riders’ hands contact the back door sur-face. Inprior researchof community surfaces, S aureus contaminationhas been found on vehicle rails, stair rails, door and faucet handles,remote controls, and key padsdall surfaces with similar hand-to-fomite dynamics as both the bus doors and seat rails.3,4,26,27,30

The sampling date was also associated with the degree to whichS aureus and MRSA could be isolated, with 90% and 80% Saureusepositive buses from the first 2 dates (90% and 60% MRSApositive, respectively), but only 50% positive buses from the second2 dates (50% both for MRSA). Little is known about any seasonaltrends in community MRSA prevalence, but some past research hasnoted increases in community MRSA infections during summermonths.31 This may be a reflection of increased skin-to-skin andskin-to-fomite contact because of increased temperature and moreexposed skin (wearing shorts and short sleeves).

There remains a paucity of data regarding the underlyingphenotypic and molecular characteristics of MRSA circulating in thecommunity. Most published work has either not provided moleculardataorhascomeas thedirect result of a specificoutbreak.3,25-29Giventhat outbreak strainsmay have increased fitness or occur in scenarioswhere enhanced transmission may be occurring, they may fail toprovide a true reflection of the underlying dynamics of MRSA in thecommunity.32 Therefore, the current research contributes insight intowhichMRSAstrains are circulating ina community,which is essentialto our understanding of pathogen spread and prevention efforts.

Dendrogram analysis showed low diversity among the MRSAisolates, with most isolates (54%, 19/35) classified as SCCmec type IV

J.K. Lutz et al. / American Journal of Infection Control 42 (2014) 1285-901290

USA300. These characteristics correspond to the most common CAMRSA strain found in the United States, demonstrating a potentialcommunity-acquiredorigin.33-36 In addition, 23% (8/35)of the isolatesexhibit characteristicspresent inHAMRSAclones,whichmay indicatea spillover of MRSA strains from health care facilities into the com-munity; these results are consistent with previous work.5,7 The mostpredominant strain in U.S. communities (SCCmec type IV USA300)was found in different buses across 3 different sampling dates.34,37

The results of this environmental study demonstrate that MRSAcontamination is consistently present onmultiple surfaces in publictransportation vehicles. Some surfaces (ie, seats, seat rails) weremore frequently contaminated with this pathogen; therefore, theymust be actively targeted for routine cleaning and disinfection. It isalso important to highlight that both CA and HA strains were iso-lated from these vehicles, in some cases from the same unit(vehicle). This was true regardless if they served hospital routes ornot. This finding underscores the conclusion that buses may act aspotential mixing vessels for MRSA strains in the community.

Acknowledgments

The authors wish to thank Shasha Bai, Jade Braman, MelissaMikolaj, Wannasawat Ratphitagsanti, PhD; Chongtao Ge, SophiaDailey, Colleen Shockling-Dent, and Debbie Lutz for their laboratoryand field assistance during this project. Also, Duncan MacCannell,from the Center for Disease Control and Prevention (CDC), forfacilitating the database containing S. aureus strains with the mosttypical band patterns for each USA type for PFGE characterization.We also want to thank Herminia de Lencastre, PhD from the Uni-versidade Nova de Lisboa in Portugal, for providing MRSA controlsisolates for the standardization of the SCCmec type multiplex PCR.Finally, we will like to thank the Network on Antimicrobial Resis-tance in Staphylococcus aureus (NARSA) program for providingseveral control strains. NARSA is supported under NIAID, NIHcontract number HHSN272200700055C.

References

1. American Public Transportation Association. Public transportation fact book.Available from: http://www.apta.com/resources/statistics/Pages/transitstats.aspx; 2012. Accessed September 26, 2014.

2. Yeh PJ, Simon DM, Millar JA, Alexander HF, Franklin D. A diversity of antibiotic-resistant Staphylococcus spp. in a public transportation system. Osong PublicHealth Res Perspect 2011;2:202-9.

3. Otter JA, French GL. Bacterial contamination on touch surfaces in the publictransport system and in public areas of a hospital in London. Lett ApplMicrobiol 2009;49:803-5.

4. Stepanovi�c S, Cirkovi�c I, Djuki�c S, Vukovi�c D, Svabi�c-Vlahovi�c M. Public trans-port as a reservoir of methicillin-resistant staphylococci. Lett Appl Microbiol2008;47:339-41.

5. Simões RR, Aires-de-Sousa M, Conceição T, Antunes F, da Costa PM, deLencastre H. High prevalence of EMRSA-15 in Portuguese public buses: aworrisome finding. PLoS One 2011;6:e17630.

6. Conceição T, Diamantino F, Coelho C, de Lencastre H, Aires-de-Sousa M. Contami-nation of public buseswithMRSA in Lisbon, Portugal: a possible transmission routeof major MRSA clones within the community. PLoS One 2013;8:e77812.

7. Iwao Y, Yabe S, Takano T, Higuchi W, Nishiyama A, Yamamoto T. Isolation andmolecular characterization of methicillin-resistant Staphylococcus aureus frompublic transport. Microbiol Immunol 2012;56:76-82.

8. Kuehnert MJ, Kruszon-Moran D, Hill HA, McQuillan G, McAllister SK,Fosheim G, et al. Prevalence of Staphylococcus aureus nasal colonization in theUnited States, 2001-2002. J Infect Dis 2006;193:172-9.

9. Gorwitz RJ, Kruszon-Moran D, McAllister SK, McQuillan G, McDougal LK,Fosheim GE, et al. Changes in the prevalence of nasal colonization with Staph-ylococcus aureus in theUnited States, 2001-2004. J Infect Dis 2008;197:1226-34.

10. Miller LG, Diep BA. Clinical practice: colonization, fomites, and virulence:rethinking the pathogenesis of community-associated methicillin-resistantStaphylococcus aureus infection. Clin Infect Dis 2008;46:752-60.

11. Boyce JM, Pittet D, Healthcare Infection Control Practices Advisory Committee.Society for Healthcare Epidemiology of America. Association for Professionalsin Infection Control. Infectious Diseases Society of America. Hand Hygiene TaskForce. Guideline for Hand Hygiene in Health-Care Settings: recommendationsof the Healthcare Infection Control Practices Advisory Committee and the

HICPAC/SHEA/APIC/IDSA Hand Hygiene Task Force. Infect Control Hosp Epi-demiol 2002;23(Suppl):S3-40.

12. Maltezou HC, Giamarellou H. Community-acquired methicillin-resistantStaphylococcus aureus infections. Int J Antimicrob Agents 2006;27:87-96.

13. HoetAE, JohnsonA,Nava-HoetRC,BatemanS,HillierA,Dyce J, et al. Environmentalmethicillin-resistant Staphylococcus aureus in a veterinary teaching hospitalduring a nonoutbreak period. Vector Borne Zoonotic Dis 2011;11:609-15.

14. Van Balen J, Kelley C, Nava-Hoet RC, Bateman S, Hillier A, Dyce J, et al. Presence,distribution, andmolecular epidemiology of methicillin-resistant Staphylococcusaureus in a small animal teaching hospital: a year-long active surveillance tar-geting dogs and their environment. Vector Borne Zoonotic Dis 2013;13:299-311.

15. Clinical and Laboratory Standards Institute. Performance standards for antimi-crobial susceptibility testing; nineteenth informational supplement. CLSI docu-mentM100eS19.Wayne (PA): Clinical and Laboratory Standards Institute; 2009.

16. Fiebelkorn KR, Crawford SA, McElmeel ML, Jorgensen JH. Practical disk diffu-sion method for detection of inducible clindamycin resistance in Staphylo-coccus aureus and coagulase-negative staphylococci. J Clin Microbiol 2003;41:4740-4.

17. Oliveira DC, de Lencastre H. Multiplex PCR strategy for rapid identification ofstructural types and variants of the mec element in methicillin-resistantStaphylococcus aureus. Antimicrob Agents Chemother 2002;46:2155-61.

18. Milheiriço C, Oliveira DC, de Lencastre H. Update to the multiplex PCR strategyfor assignment of mec element types in Staphylococcus aureus. AntimicrobAgents Chemother 2007;51:3374-7.

19. Murchan S, Kaufmann ME, Deplano A, de Ryck R, Struelens M, Zinn CE, et al.Harmonization of pulsed-field gel electrophoresis protocols for epidemiolog-ical typing of strains of methicillin-resistant Staphylococcus aureus: a singleapproach developed by consensus in 10 European laboratories and its appli-cation for tracing the Spre. J Clin Microbiol 2003;41:1574-85.

20. Magiorakos A, Srinivasan A, Carey RB, Carmeli Y, Falagas ME, Giske CG, et al.Multidrug-resistant, extensively drug-resistant and pandrug-resistant bacteria:an international expert proposal for interim standard definitions for acquiredresistance. Clin Microbiol Infect 2012;18:268-81.

21. Kassem II. Concerning public transport as a reservoir of methicillin-resistantstaphylococci. Lett Appl Microbiol 2009;48:268.

22. Lutz JK, Crawford J, Hoet AE, Wilkins JR, Lee J. Comparative performance ofcontact plates, electrostatic wipes, swabs and a novel sampling device for thedetection of Staphylococcus aureus on environmental surfaces. J Appl Micro-biol 2013;115:171-8.

23. Marschall J, Dürig P, Mühlemann K. Low prevalence of methicillin-resistantStaphylococcus aureus (MRSA) carriage in women from former Yugoslavialiving in Switzerland. Eur J Clin Microbiol Infect Dis 2006;25:535-6.

24. Wall EC, Ghazy A, Bulley S, Ahmad F, Aali A, Lynn W. Medical staff at a Londondistrict hospital are not chronic vectors of MRSA or Clostridium difficile. J Infect2009;58:469-71.

25. Baggett HC, Hennessy TW, Rudolph K, Bruden D, Reasonover A, Parkinson A,et al. Community-onset methicillin-resistant Staphylococcus aureus associatedwith antibiotic use and the cytotoxin Panton-Valentine leukocidin during afurunculosis outbreak in rural Alaska. J Infect Dis 2004;189:1565-73.

26. Scott E, Duty S, McCue K. A critical evaluation of methicillin-resistant Staph-ylococcus aureus and other bacteria of medical interest on commonly touchedhousehold surfaces in relation to household demographics. Am J Infect Control2009;37:447-53.

27. Sexton JD, ReynoldsKA. Exposureof emergencymedical responders tomethicillin-resistant Staphylococcus aureus. Am J Infect Control 2010;38:368-73.

28. Begier EM, Frenette K, Barrett NL, Mshar P, Petit S, Boxrud DJ, et al. A high-morbidity outbreak of methicillin-resistant Staphylococcus aureus amongplayers on a college football team, facilitated by cosmetic body shaving and turfburns. Clin Infect Dis 2004;39:1446-53.

29. Kazakova SV, Hageman JC, Matava M, Srinivasan A, Phelan L, Garfinkel B, et al.A clone of methicillin-resistant Staphylococcus aureus among professionalfootball players. N Engl J Med 2005;352:468-75.

30. Felkner M, Andrews K, Field LH, Taylor JP, Baldwin T, Valle-Rivera AM, et al.Detection of Staphylococcus aureus including MRSA on environmental surfacesin a jail setting. J Correct Health Care 2009;15:310-7.

31. Kaplan SL, Hulten KG, Gonzalez BE, Hammerman WA, Lamberth L, Versalovic J,et al. Three-year surveillance of community-acquired Staphylococcus aureusinfections in children. Clin Infect Dis 2005;40:1785-91.

32. Wagenvoort JH, Sluijsmans W, Penders RJ. Better environmental survival ofoutbreak vs. sporadic MRSA isolates. J Hosp Infect 2000;45:231-4.

33. KennedyAD,OttoM,BraughtonKR,WhitneyAR,ChenL,MathemaB, etal. Epidemiccommunity-associated methicillin-resistant Staphylococcus aureus: recent clonalexpansion and diversification. Proc Natl Acad Sci U S A 2008;105:1327-32.

34. Deleo FR, Otto M, Kreiswirth BN, Chambers HF. Community-associatedmethicillin-resistant Staphylococcus aureus. Lancet 2010;375:1557-68.

35. Tenover FC, Tickler IA, Goering RV, Kreiswirth BN, Mediavilla JR, Persing DH.Characterization of nasal and blood culture isolates of methicillin-resistantStaphylococcus aureus from patients in United States hospitals. AntimicrobAgents Chemother 2012;56:1324-30.

36. Talan DA, Krishnadasan A, Gorwitz RJ, Fosheim GE, Limbago B, Albrecht V, et al.Comparison of Staphylococcus aureus from skin and soft-tissue infections in USemergency department patients, 2004 and 2008. Clin Infect Dis 2011;53:144-9.

37. Mediavilla JR, Chen L, Mathema B, Kreiswirth BN. Global epidemiology ofcommunity-associated methicillin resistant Staphylococcus aureus (CA-MRSA).Curr Opin Microbiol 2012;15:588-95.