An abstract of this work was presented at the American College of Veterinary Internal Medicine (ACVIM) Forum 2008, San Antonio, TX. The study was performed at the University of Missouri (Kottler, Middleton, Cohn, Perry) and the University of Guelph (Weese). Dr Kottler is currently at 1 Intervale Road, Concord, NH 03301.
Prevalence of Staphylococcus aureus and Methicillin-Resistant Staphylococcus aureus Carriage in Three Populations
Article first published online: 30 NOV 2009
Copyright © 2009 by the American College of Veterinary Internal Medicine
Journal of Veterinary Internal Medicine
Volume 24, Issue 1, pages 132–139, January/February 2010
How to Cite
Kottler, S., Middleton, J.R., Perry, J., Weese, J.S. and Cohn, L.A. (2010), Prevalence of Staphylococcus aureus and Methicillin-Resistant Staphylococcus aureus Carriage in Three Populations. Journal of Veterinary Internal Medicine, 24: 132–139. doi: 10.1111/j.1939-1676.2009.0424.x
- Issue published online: 4 JAN 2010
- Article first published online: 30 NOV 2009
- Submitted April 22, 2009; Revised September 9, 2009; Accepted October 2, 2009.
- Antimicrobial resistance;
Background: A higher prevalence of methicillin-resistant Staphylococcus aureus (MRSA) colonization is reported in healthcare workers compared with nonhealthcare workers.
Hypothesis: The prevalence of MRSA colonization differed in people and pets in households with healthcare workers as compared with households without healthcare workers.
Subjects: A person and 1 dog or cat from 586 households defined as either a nonhealthcare (n = 213), veterinary healthcare (n = 211), or human healthcare (n = 162) worker household.
Methods: Prospective cross-sectional study. Samples from humans and pets were cultured in vitro. Staphylococcus aureus was identified as methicillin sensitive (MSSA) or MRSA with mecA polymerase chain reaction. Pulsed-field gel electrophoresis and spa-typing were used to characterize relatedness of S. aureus and MRSA and assign USA types.
Results: The prevalence of MSSA and MRSA in humans was 21.5% (126/586) and 5.63% (33/586), respectively, and 7.85% (46/586) and 3.41% (20/586), respectively, in pets. There were no differences in prevalences of either MSSA or MRSA between household types. The proportion of MRSA among all S. aureus isolates in humans and pets was 20.8% (33/159) and 30.3% (20/66), respectively. In <1.0% (4/586) of households, the same strain of MRSA was found in both a person and a pet.
Conclusions and Clinical Importance: There were no differences in the prevalences of MSSA or MRSA between healthcare worker and nonhealthcare worker households. Pets and people colonized with S. aureus were as likely to be colonized with MRSA. Colonization of a person and their pet with the same strain of MRSA was rare.
methicillin-resistant Staphylococcus aureus
methicillin-sensitive Staphylococcus aureus
penicillin binding protein 2a
polymerase chain reaction
pulsed-field gel electrophoresis
Approximately 30% of healthy people are colonized with Staphylococcus aureus, a Gram-positive, facultative anaerobic, catalase-positive coccus.1–4 Given the appropriate opportunity, S. aureus is capable of causing morbidity and mortality as a result of infection.5,6 Although pet dogs and cats are more likely to be colonized with Staphylococcus pseudintermedius, pets can become contaminated, colonized, or infected with S. aureus.1,3,7–9 Methicillin-resistant Staphylococcus aureus (MRSA) is S. aureus that has acquired and expressed the mecA gene. This gene encodes for an altered protein called penicillin-binding protein 2a (PBP2a), which has a low affinity for β-lactam antimicrobials including penicillins and cephalosporins.10 MRSA is therefore resistant to the aforementioned antimicrobials and MRSA strains frequently carry other genes that confer resistance to a variety of other antimicrobials.10 Colonization with MRSA may increase the risk of developing a MRSA infection.11 Although some animal species have been implicated as potential reservoirs for MRSA colonization or infection of people, interspecies transmission of MRSA is thought to occur primarily from humans to animals.3,12,13
The prevalence of S. aureus on healthy cats and dogs is thought to be lower than the approximately 30% prevalence of human colonization with S. aureus.2,8,14–18 Adherence of bacteria is an important factor in the process of colonization and infection. Staphylococcus pseudintermedius (formerly S. intermedius) preferentially adheres to the corneocytes of canine skin, whereas S. aureus preferentially adheres to the corneocytes of human skin, thus accounting for the reported low prevalence of S. aureus colonization in dogs and cats.19–21 In cats, adhesion of staphylococci to corneocytes may be even less effective than in dogs.19
Previous studies evaluating MRSA colonization in dogs and cats have not distinguished between pets living in households with or without a healthcare worker. The prevalence of S. aureus and MRSA in people is reported to vary with occupation and likelihood of exposure. Carriage of S. aureus in the nasal passages of humans in the general population is between 29 and 38%,3 whereas carriage of MRSA in the same population ranges from 0.8 to 3.5%.4,22 An increased prevalence of colonization with S. aureus (44.4 versus 29.6%) and MRSA (7.6–11.1 versus 3.5–5.4%) has been described in healthcare workers as compared with the general population.4,23–28 Additionally, although no studies have directly compared veterinary healthcare workers (VH) with nonhealthcare workers, carriage of MRSA by people involved in veterinary medicine has been reported to be higher than carriage in the general population.29,30
Numerous studies suggest that S. aureus, including MRSA, can be passed between people and pets.12,13,31–33 Because people can transmit MRSA to their pets, and because those pets might become a reservoir for MRSA, it is important to know if the pets of healthcare workers have a higher prevalence of MRSA colonization than do the pets of nonhealthcare workers. We hypothesized that the prevalence of MRSA would differ in people and pets who reside in households with veterinary or human healthcare workers (HH) from the prevalence in people and pets who reside in households with nonhealthcare workers. The present study compared the prevalence of S. aureus and MRSA colonization of people and pets living in households with people who are involved in veterinary healthcare, human healthcare, or non-healthcare–related occupations. When S. aureus was found in a person and pet in the same household, strain relatedness was investigated.
Materials and Methods
Owners were required to provide informed consent before participation in the study. As an incentive for participation, pet owners were given free heartworm prophylaxis,a microchips,b or pet food coupons.c Study participants were recruited via email and local media releases, and sample collection took place at the University of Missouri, community events, and in participant households or businesses.
Study Population and Inclusion Criteria
The inclusion criteria of the study were broad so as to generate a representative sample of each population. Pets without clinical signs and people without symptoms of a MRSA infection who resided in 1 of 3 different types of households were enrolled in 1 of 3 groups: (1) households with no healthcare workers (NH), (2) households with VH, and (3) households with HH. One person and 1 pet (dog or cat) were sampled per household. Only people over the age of 16 years were eligible for sampling, but pets of any age were sampled. Participants were asked to fill out a questionnaire detailing household demographic information, including number of people and pets and the presence of a veterinary or HH. Households in which there were both a human and a VH or in which a group could not be assigned due to an incomplete questionnaire were excluded from analysis. Additional information provided included a history of recent antimicrobial usage or documented infections in any member of the household. Households with VH included at least 1 individual who worked as a veterinarian, veterinary technician, veterinary assistant, veterinary student with animal contact, or kennel worker. HH households included at least 1 individual who worked as a doctor, nurse, medical student, physical therapist, technician, emergency medical technician, nursing home employee, or paramedic. No member of a nonhealthcare household worked in any medical field. Samples were collected under the approval of the Institutional Animal Care and Use Committee and the Institutional Review Board.
Prestudy sample size calculations were performed to estimate the numbers of households within each group required to detect a difference in occurrence of S. aureus and MRSA colonization among groups with a statistical power of 0.80 at an α of 0.05.d Sample size calculations were based on reported prevalences of MRSA among HHs (7.63 and 11.1%),4,24 VH (17.9%),30 and members of the general population (3.5 and 5.3%).4,24 No preliminary data were available for the prevalence of MRSA among pets housed in each of the target populations. The mean group size required based on the data above was 244 with a median of 257 and range of 84–390.
Sampling and Laboratory Methods
Sampling of the human in each household was performed by a veterinarian affiliated with the study, whereas sampling of the pet was performed by a veterinarian, veterinary technician, or veterinary student. Cotton-tipped culturette swabse moistened with sterile saline were used to collect samples using previously described procedures.30,34 The moistened swab was inserted in the anterior nares of both nostrils and gently rubbed against the mucosal surface for approximately 5–10 seconds. Additional swab samples were taken from pets by inserting a moistened swab just inside the anus and rubbing the swab against the mucosal surface for approximately 5–10 seconds.35–38 Swabs were placed in modified Stuart's mediume and refrigerated for up to 48 hours. Swabs were placed in 2 mL of enrichment broth (10 g/L Tryptone T, 75 g/L sodium chloride, 10 g/L mannitol, and 2.5 g/L yeast extract) and incubated at 37°C for 24 hours.38 Enriched samples were frozen in glass tubes at −20°C for no longer than 1 month before further analysis.
Enrichment broth tubes were thawed at room temperature (22°C) and initial screening for S. aureus was performed by streaking 5–10 μL of enrichment broth on either mannitol salt agarf or phenylethyl alcohol (PEA) agar.g The 1st 106 samples were plated onto mannitol salt agar, but because of frequent overgrowth of Gram-negative organisms, the procedure was modified. The remaining samples were plated onto PEA agar. Plates were incubated at 37°C for 24 hours and if growth was not observed, incubation was continued for an additional 24 hours.
S. aureus was presumptively identified based on Gram-staining, colony morphology, catalase reaction, the tube coagulase test, and fermentation of mannitol.h Coagulase-positive isolates were confirmed to be S. aureus with a latex agglutination test according to the manufacturer's directions.i Isolates that were borderline positive with this test were subjected to polymyxin B susceptibility to differentiate between S. aureus (resistant to polymyxin B) and other coagulase-positive staphylococci (susceptible to polymxyin B).
S. aureus isolates were stored at −80°C in phosphate-buffered glycerol until further analyses were performed. Isolates were grown from storage media and identified as MRSA by detection of the mecA gene by polymerase chain reaction (PCR).39
Relatedness of MSSA or MRSA isolates was determined using pulsed-field gel electrophoresis (PFGE) according to the method described by Middleton et al.40SmaI digestsj of staphylococcal chromosomal DNA were separated by PFGE in a 1% agarose gelk immersed in 0.5% Tris-boric acid-EDTA buffer at 14 °C for 20 hours at 6 V/cm with a 5–50-second pulse time on a CHEF DRIII PFGE machine.l A lambda ladder molecular weight size standardm,n was run in the 1st and last lanes of the gels. Isolates with identical banding patterns within households were considered the same.40
MRSA isolates were classified according to USA types primarily using PFGE. Isolate PFGE patterns were compared with established USA clone PFGE patterns previously described by McDougal et al.41 The USA type also was inferred based on spa typing and by detection of the panton-valentine leukocidin (PVL) gene by real-time PCR amplification of lukF-lukS.42–44 Real-time PCR for PVL with primers, PVLSC-F, GCTCAGGAGATACAAG and PVLSC-R, GGATAGCAAAAGCAATG was performed as described previously.43,44
For spa typing, the SSR region of the spa gene was amplified and sequenced according to a method described by Shopsin et al.45 Sequences were analyzed by the eGenomics software (http://tools.egenomics.com) and were reported using a numerical system that corresponded to a spa type sequence. Ridom database equivalents were identified by the Ridom Spaserver website (http://www.spaserver.ridom.de) and were reported using a numerical coding system.22
Pets were classified as colonized with MSSA or MRSA if they had a positive nasal sample, a positive rectal sample, or both. Humans were classified as colonized with MSSA or MRSA if they had a positive nasal sample. Persons or pets colonized with >1 strain of S. aureus were counted as a single colonized individual for purpose of data analysis. The proportions of individuals colonized with MSSA and MRSA were compared among groups using the Chi-square test or Fisher's exact test, as appropriate (P < .05).
Samples were collected from 586 households, which included 213 NH, 211 VH, and 162 HH households. In 144/162 (88.8%) HH households and in 195/211 (92.4%) VH households the person sampled was the healthcare worker. One member of the HH group and 16 members of the VH group were in their 1st or 2nd year of medical or veterinary school, respectively. The mean ± standard deviation (SD) number of people (including children) within sampled households was 2.28 ± 1.17. The mean ± SD number of pets per household was 4.80 ± 6.02. With regard to distribution of pets, 82.8% (485/586) of samples were obtained from households in which a dog was sampled and 17.2% (101/586) were obtained from households in which a cat was sampled.
Antimicrobial drugs had been used in either a person or a pet within the prior 2 months in 92 included households. Methicillin-sensitive S. aureus was isolated in 19/92 (20.7%) of these households (15 people, 4 pets). MRSA was isolated in 11/92 (12.0%) of these households (8 people, 5 pets). There was no statistically significant difference in the rate of MSSA or MRSA colonization between households with previous antimicrobial usage and households without antimicrobial usage (P= .158 and .130, respectively).
MSSA was isolated from 26.6% (156/586) households and MRSA was detected in 8.02% (47/586) households. The prevalences of MSSA and MRSA, and the proportion of all S. aureus isolates that were MRSA in people are shown in Table 1. There were 6 people colonized with both a MRSA and a MSSA. There were no statistically significant differences in the proportion of people colonized with MSSA or MRSA among household types (P= .22 and .76, respectively). Likewise, there was no statistically significant difference in the proportion of S. aureus isolates that were methicillin-resistant isolated from people among groups (P= .42).
|Group||S. aureus||MSSA||MRSA||MRSA/ S. aureus|
|Non-healthcare (n = 213)||30.0%||25.4%||4.69%||15.6%|
|Veterinary healthcare (n = 211)||25.6%||19.4%||6.16%||24.1%|
|Human healthcare (n = 162)||25.3%||19.1%||6.17%||24.4%|
|Total (n = 586)||27.1%||21.5%||5.63%||20.8%|
Among dogs, 8.25% (40/485) were colonized with MSSA and 3.3% (16/485) were colonized with MRSA. Among cats, 5.94% (6/101) were colonized with MSSA and 3.96% (4/101) were colonized with MRSA. The prevalence of MSSA and MRSA, and the proportion of all S. aureus isolates that were MRSA in pets are shown in Table 2. There were no statistically significant differences in the prevalences of MSSA, MRSA, or the proportion of all S. aureus isolates that were MRSA among groups (P= .20, .40, and .13, respectively). Animals were statistically less likely to be colonized with MSSA than people (P < .001). However, there was no statistically significant difference in the proportion of MRSA among S. aureus isolated from people or pets (P= .07).
|Group||S. aureus||MSSA||MRSA||MRSA/ S. aureus|
|Non-healthcare (n = 213)||10.3%||7.04%||3.29%||31.8%|
|Veterinary healthcare (n = 211)||12.8%||10.4%||2.37%||18.5%|
|Human healthcare (n = 162)||10.5%||5.56%||4.94%||47.1%|
|Total (n = 586)||11.3%||7.85%||3.41%||30.3%|
For pets, animal species and site of colonization were compared (Fig 1). No statistically significant difference was detected in the proportion of dogs versus cats colonized with either MSSA (40/46 dogs versus 6/46 cats, P= .54) or MRSA (16/20 dogs versus 4/20 cats, P= .98). Few animals were positive on both nasal and rectal cultures. Concurrent nasal and rectal colonization with MSSA was recognized in 5 animals and 1 animal was colonized with MRSA in both the nasal cavity and rectum. One dog was colonized with 3 different S. aureus (2 MRSA and 1 MSSA) strains and another dog was colonized with 1 MSSA and 1 MRSA. A single cat was colonized with 3 strains of S. aureus (1 MSSA and 2 MRSA).
Using a positive culture at any site (nasal or rectal) as the gold standard, the sensitivities for detection in the feline nasal cavity of MSSA and MRSA were 1.00 and 0.75, respectively. In dogs, the sensitivities for detection of MSSA and MRSA in the nasal cavity were 0.73 and 0.63, respectively. The sensitivities for detection of MSSA and MRSA from the rectum of dogs were 0.35 and 0.44, respectively. The sensitivities for detection of MSSA and MRSA from the rectum of cats were 0.33 and 0.25, respectively.
Characterization of MSSA Isolates
In 16 of 586 households (2.73%), MSSA were isolated from both the sampled person and pet. Of these, 8/16 were from NH households, 6/16 were from VH households, and 2/16 were from HH households. MSSA isolation was not statistically different among household types (P= .37). In 9 of these 16 households, the MSSA isolated from the person and pet had the same PFGE banding pattern (5 NH, 3 VH, and 1 HH) whereas in 7 of the 16 households the MSSA isolated from the person and pet differed by ≥3 PFGE bands (3 NH, 3 VH, and 1 HH). There was no statistically significant difference among groups with regard to the proportion of households that had a person and pet colonized with the same strain of MSSA (P= .40).
Characterization of MRSA Isolates
There were 58 MRSA isolates cultured from 53 humans or pets (3 humans, 1 cat, and 1 dog had 2 phenotypically distinct MRSA isolates). PFGE was successful in producing a banding pattern for 52 isolates from 48 humans or pets. Six isolates did not yield an interpretable PFGE result. For the 6 isolates without a PFGE type available, a USA type was inferred from the spa type and detection of the PVL gene.42 Overall, a USA type was assigned to 50 isolates from 47 individuals based on PFGE or spa-typing and detection of the PVL gene. For the remaining 8 isolates from 6 individuals, PFGE or spa-typing did not identify a USA clone. The most frequently identified USA type among the 47 individuals was USA 100 (27/47, 57.4%), followed by USA 500 (13/47, 27.7%), USA 200 (3/47, 6.38%), USA 700 (3/47, 6.38%), and USA 400 (1/47, 2.13%). The distribution of USA types among groups for both people and pets is presented in Table 3. There were no statistically significant differences between the proportions of people and pets colonized within USA type (P≥ .171), and hence the data for people and pets were combined for comparison among USA types. There were statistically significant differences between the proportion of MRSA isolates that were USA 100 versus USA 200, USA 400, USA 500, and USA 700 (P≤ .007); between USA 200 and USA 500 (P < .014); between USA 400 and USA 500 (P= .001); and between USA 500 and USA 700 (P= .014).
|VH||5 (30%)||1 (50%)||1 (100%)||5 (83%)||1 (100%)|
|HH||6 (35%)||1 (50%)||0||1 (17%)||0|
|NH||4 (40%)||0||0||3 (43%)||0|
|VH||3 (30%)||0||0||1 (14%)||1 (50%)|
|HH||3 (30%)||1 (100%)||0||3 (43%)||1 (50%)|
MRSA was found in both a person and a pet in 6 of 586 (1.0%) households representing 2 households in each group. Of the 6 households in which a person and a pet were colonized with MRSA, 4 person-pet pairs (2 NH, 1 VH, and 1 HH) had identical PFGE banding patterns. Of the remaining 2 households, the strain cultured from the person had a different spa-type and PFGE pattern from the strain cultured from the pet. There were no statistically significant differences among groups in the proportion of households with the same strain of MRSA in the person and the pet (P= .84).
In this study, no statistically significant difference was found in the occurrence of person or pet colonization with MSSA or MRSA between households with or without healthcare workers. Similarly, we found no statistically significant difference in the proportion of MRSA among all S. aureus in the NH, VH, or HH households. These results were surprising given previous reports of a higher prevalence of S. aureus colonization in people employed in either veterinary or human healthcare.4,24,29,30
Several studies have evaluated the prevalence of S. aureus, MRSA colonization, or both in people employed in either human or veterinary healthcare.4,23,24,29 In a study by Saxena et al,24 44.4% of healthy human healthcare providers were colonized with S. aureus, whereas 29.6% of nonhealthcare workers were found to be colonized with S. aureus. This investigation documented MRSA colonization of 11.1% of healthcare providers but only 5.4% of nonhealthcare workers. Similarly, colonization rates with MRSA for veterinarians and veterinary technicians (17.9%) have been reported to be higher than UK community surveys documenting MRSA colonization rates of 1.5 and 0.78%.30 Hanselman et al29 documented that 6.5% of attendees at an international veterinary conference were colonized with MRSA (7.0% of the veterinarians and 12.0% of the technicians). However, not all studies of healthcare professionals have documented an increased prevalence of colonization with S. aureus among healthcare providers. Lu et al4 found that 19.1 and 25.2% of human healthcare providers and members of a community, respectively, were colonized with S. aureus, but this study did document a greater prevalence of MRSA colonization among healthcare providers as compared with community subjects (7.63 and 3.5%, respectively). Although it is impossible to directly compare these diverse studies, the results of the present study show that all groups (HH, VH, and NH) had similar prevalences of S. aureus and MRSA colonization to those previously reported for nonhealthcare workers.3,4,22,24
As with people, no differences in prevalence of colonization with MSSA or MRSA were detected in pets from HH, VH, or NH households in the present study. Although a direct comparison cannot be made with previous veterinary studies documenting MRSA colonization in dogs and cats in contact with VH,30,46–48 it was hypothesized, based on previous studies, that there would be a statistically significant difference in the proportion of pets colonized with MRSA among household groups. Recently, Boost et al17 demonstrated that colonization of dogs with S. aureus was positively associated with having an owner involved in a healthcare profession. Given that in our study there was no statistically significant difference in MSSA or MRSA colonization among groups of people, it is not surprising that no statistically significant differences in pet colonization rates were identified among groups. When all groups were combined, we documented a 7.85% MSSA prevalence in pets, similar to previous studies17,49,50 and far below the prevalence of MSSA in people either in this study or in other reports.2,51,52 However, previous studies of MRSA colonization in pets have reported prevalences of 0–2%o15,17,35,38,53 whereas our study documented MRSA in 3.4% of 586 dogs or cats. This figure exceeds reported prevalence of MRSA in dogs or cats in all but 1 published study of 57 animals that reported a 7.0% prevalence.30
A similar proportion of S. aureus isolates that were methicillin-resistant was documented from persons and pets. Among people, the proportion of S. aureus isolates that were MRSA in our study was 20.8%, similar to that found in 1 report,24 but higher than that found in other reports where the prevalence ranged from 1.25 to 9.0%.2,18,54,55 Veterinary studies have reported the prevalence of MRSA in S. aureus-colonized individuals to range between 0 and 20%.8,15,17,38 However, only a single study of 50 healthy cats recognized a prevalence above 8.2%.8,17 The proportion of MRSA among S. aureus isolated from pet animals in the present study was 30.3%. If our study is reflective of the general pet population, this finding could be important in that approximately one third of pets colonized with S. aureus may be colonized with MRSA.
It is not unexpected that USA 100 was the most common USA type identified in this study overall in both people and pets based on findings in previous studies.8,56 The prevalence of USA 500 in people and pets was surprisingly high. Although USA 500 has been associated with colonized or infected horses57 and 5/6 of the people in this category were from VH households, information regarding exposure to horses was not available. It is possible that people and pets with USA 500 in this study had regular contact with horses. Alternatively, the prevalence of USA 500 may be increasing in the human and pet population.
S. aureus has been shown not to adhere well to canine and feline corneocytes,19–21 providing 1 explanation for the lower prevalence of S. aureus colonization in pets relative to humans. However, this does not preclude S. aureus adhering to other cells types, and several reports implicate animals as a source of persistent re-infection or colonization of people or the environment.12,13,31,32 In several such cases, people in close contact with a pet animal could not be successfully treated for infection or decolonized until colonization in the pet was addressed.12,13,31–33,58–60 Our study demonstrated that the number of households in which both a person and pet were colonized with MRSA was very low (<1.0% of the total households). In the uncommon situation in which MRSA was isolated from a person and their pet, isolates were identified by PFGE as the same in 4 of 6 households. For those 4 households, it is uncertain whether the MRSA originated from the person or pet. Nonetheless, the results from these 4 households suggest that it is possible for a healthy household pet to serve as a source of colonization for people in the household. Alternatively, transmission to both a person and pet in the same household could come from a point source.
Veterinarians may be requested on occasion to screen pet animals for MRSA carriage, especially in cases where a human is persistently infected or becomes frequently reinfected with MRSA. Objective study of optimal sampling sites currently is lacking. Griffeth et al15 cultured 5 anatomical sites from normal dogs and dogs affected with inflammatory skin disease. No site examined was statistically more likely to carry methicillin-resistant organisms. Likewise, Abraham et al8 did not find a statistically significant difference in MRSA carriage among multiple body sites in cats. In our study, there was no statistically significant difference between nasal and rectal carriage of MRSA in dogs or cats. Sampling of the nasal passages alone would have provided a false-negative result in 11 animals for the detection of MSSA and in 7 animals for the detection of MRSA. Ideally, both nasal and rectal samples should be obtained for surveillance cultures, although further study of different sites and sampling techniques is required.
Our findings may reflect the possibility that in the locale where we obtained our samples there was no difference in the likelihood of colonization based on occupation. Alternatively, our study population may have been inadequately sized to detect a subtle difference. Although the projected sample size of 244 households per group was approximated in all but 1 group, it is unlikely that we would have detected a difference among groups had we reached our target because of the small difference in occurrence of MRSA among groups. For example, in order to detect a difference in the prevalence between humans in NH households and HH households (4.69 versus 6.17%) with a power of 0.80, a sample size of 3,678 subjects per group would have been required. Furthermore, our study may have underestimated the carriage of S. aureus as a result of sampling only 1 human and 1 animal member of each household. The sampling design used in this investigation facilitated collection from a large number of households. Although ideally all members of the household would have been sampled, this was not practical owing to the fact that many samples were collected at public events or in areas where only a limited number of household members were present. Failure to sample the healthcare workers themselves may have resulted in a falsely low prevalence of MSSA and MRSA in VH or HH households. However, in almost 90% of cases a healthcare worker in the household was sampled. By classifying veterinary and medical students in their preclinical years as veterinary or HH, respectively, the potential for misclassification also exists because these individuals would have limited patient contact. However, these professional degree students accounted for only 1 member of the human healthcare group and 16 members of the veterinary healthcare group. Because only 1 animal in each household was sampled, this study likewise may have underestimated colonization of pets in the household. Failure to sample every member of each household might have led to failure to detect person-pet pairs with shared bacterial strains. Finally, we did not exclude animals or people who had received recent antimicrobial therapy. Although this may have led to a decreased likelihood of culturing S. aureus, we found no difference in rates of S. aureus colonization when antibiotics had been used.
We documented that MRSA colonization of both a healthy person and their healthy pet is rare, but possible. However, this study did not address the occurrence of colonization among pets in households where an animal or human MRSA infection had been documented. Additional studies should address the likelihood of animal colonization in such households, the duration of colonization in pet animals, and protocols for the prevention of MRSA transmission among species.
aTri-heart plus, Schering Plough, Kirkland, QC, Canada
bHome Again Pet Identification Microchip, Schering Plough
cPurina, St Louis, MO
dSigmaStat v. 3.1, SYSTAT Software Inc, San Rafael, CA
eBacti-swab, Remel, Lenexa, KS
fMannitol salt agar, Remel
gPEA agar, Remel
hCoagulase plasma, Remel
iPastorex Staph Plus, Bio-Rad Laboratories Ltd, Mississauga, ON, Canada
jSmaI, Invitrogen, Carlsbad, CA
kPulsed-field certified agarose, Bio-Rad, Hercules, CA
lChef DRIII PFGE, Bio-Rad
mLambda Ladder, Bio-Rad
nMasterPure DNA extraction kit, Epicentre Technologies, Madison, WI
oMurphy C, Reid-Smith R, Prescott J, et al. Occurrence of antimicrobial resistance in selected bacteria in healthy dogs and cats presented to private veterinary clinics in southern Ontario. J Vet Intern Med 2004;19:478 (abstract)
The authors acknowledge Dr Thomas Reilly and Dr William Fales for their intellectual contributions. Additionally, we would like to thank Dr Rebecca Johnson for her help with participant recruitment and Dr Michael Estrin for his help with sample collection.
The study was funded by the American College of Veterinary Internal Medicine Foundation, the Waltham Foundation, and the University of Missouri Committee on Research. Incentives for participation were provided by Purina and Schering Plough Animal Health.
- 3Methicillin-resistant Staphylococcus aureus: An emerging pathogen in small animals. J Am Anim Hosp Assoc 2005;41:150–157.
- 8Surveillance of healthy cats and cats with inflammatory skin disease for colonization of the skin by methicillin-resistant coagulase-positive staphylococci and Staphylococcus schleiferi ssp. schleiferi. Vet Dermatol 2007;18:252–259., , , et al.
- 42A preliminary guideline for the assignment of methicillin-resistant Staphylococcus aureus to a Canadian pulsed-field gel electrophoresis epidemic type using spa typing. Can J Infect Dis Med Microbiol 2008;19:273–281., , , et al.
- 54Survey of methicillin-resistant Staphylococcus aureus (MRSA) carriage in healthy college students, Hawaii. Hawaii Med J 2007;66:213–215., , , et al.