Surveillance of Bacterial Colonization in Operating Rooms

Abstract

Background:

Surgical site infections (SSIs) continue to occur at an unacceptably high rate, incurring direct costs of up to $10 billion per year in the United States and far more than that in total economic costs. There is a continued need to identify potential sources of microbial contamination that lead to surgical infections.

Methods:

Using contact culture plates, we randomly took a total of 517 samples of various surfaces in 33 operating rooms (ORs) over a 6-mo period.

Results:

Flat surfaces treated with decontamination techniques (floors, anesthesia carts, operating tables, and other flat surfaces) grew small numbers of bacterial colonies, as did other surfaces that were decontaminated less often. Personal items, especially the tops of shoes and personal hats, had much higher contamination than the surfaces just described, which in the case of these two items averaged 50–60 CFU/20 cm². The outsides of face masks contained slightly more organisms than did floors, but the insides had almost 100 times more organisms, which was of concern because of high leakage rates at the mask–face interface.

Conclusions:

The culture system used in our study can be used as a simplified and cost-effective method of identifying the comparative densities of organisms on different surfaces for surveillance of microbial contamination in the OR. To reduce bacterial contamination, shoe covers and disposable hair coverings should be worn at every operation in which there are substantial risks of SSI. Masks that reduce leakage at the mask–face interface should be worn and discarded after each operation.

The incidence of surgical site infection (SSI) continues to be unacceptably high, imposing a severe economic burden on healthcare systems worldwide. In the United States alone, it has been estimated that more than 500,000 SSIs occur yearly, at a direct cost of up to $10 billion per year [1]. Every 1% increase in the incidence of SSI creates $900 million annually in national in-hospital cost alone, and $1.6 billion in excess direct costs overall [2]. However, this constitutes only about 10% of the total economic cost of SSIs when factors such as time away from work, lost productivity, legal fees, and death benefits, among other factors, are taken into consideration. Additionally, the effect of SSIs on patient morbidity and mortality cannot be understated. A study by DiPiro et al. [3] found that the in-hospital mortality for patients with SSIs was eight-fold that of uninfected patients.

Extensive efforts have recently been made to improve the prevention of SSIs by mandating compliance with practices known to reduce the risk of such infections (e.g., the measures specified in the Surgical Care Improvement Project) [4]. Unfortunately, this has resulted in a surprisingly small reduction in SSIs [4 –9]. The reasons for this small effect are somewhat unclear because each of the specified measures has been shown independently to reduce the frequency of infections when causal relationships behind SSI are examined. However, the following two points are clear: (1) SSIs will not occur if microbial pathogens do not enter a surgical incision; and (2) an unacceptable majority of surgical incisions are contaminated by the ends of surgical operations [10]. Collectively, organisms are prevented from entering surgical incisions by an institutional commitment to a process that adheres to the principles of asepsis. Various studies have shown that during an operation, organisms can enter a surgical incision from a variety of sources, including the patient, the operating room (OR) personnel, and the air.

Surveillance for bacterial colonization for the purpose of identifying potential problem areas within modern ORs has been infrequent in recent years, partly because of inconvenience, partly because of cost, and partly because of perceived irrelevance to modern practice. The present study was done to determine whether surveillance in ORs can be done at a reasonable cost and whether it may be useful in identifying areas that are especially susceptible to bacterial colonization.

Materials and Methods

Contact culture plates (Petrifilm Aerobic Count Plate and Petrifilm Staph Express Count Plate; 3M, St. Paul, MN) were used to assess microbial growth from various surfaces in ORs. These culture plates were designed for efficient and low-cost surveillance for bacterial colonization and have been used extensively in the food industry. Each plate contains a dual layer of prepared culture medium containing standard nutrients, a gelling agent soluble in cold water, and an indicator that facilitates the enumeration of bacterial colonies. The plates were prepared according to the manufacturer's instructions. Each aerobic plate measures the spectrum of aerobic bacteria that will grow within 24 h, whereas each plate for staphylococcal colonization primarily indicates colonization by Staphylococcus aureus, although S. epidermidis and several other species, including Enterococcus and Bacillus spp., may grow on the plate. Growth of other organisms, including other staphylococcal species and anaerobic organisms, is inhibited. To utilize the plates, the entire surface of the culture medium is pressed against the surface of the object being cultured, thereby providing a culture surface area of approximately 20 or 30 cm² (depending on the size of the surface area that has been rehydrated) for the growth of organisms. Visible colonies are counted after incubation of a plate at 37°C for 24 h. All preparations, procedures, and colony counts were done according to the manufacturer's instructions for each type of plate used in the study. Counts were adjusted, when appropriate, to reflect colony-forming units (CFUs) per 20 cm². None of the cultures taken was used to identify specific organisms, but instead served only to compare the numbers of organisms on different surfaces.

Cultures were taken randomly from various flat surfaces, personnel attire, and equipment in 33 ORs at University Hospital in Cincinnati, OH, over a 6-mo period. Most of the samples from flat surfaces were obtained from clean ORs in the morning before surgery began. Samples from OR personnel were largely obtained immediately post-operatively. The study was approved by the local institutional review board as a surveillance study rather than a clinical study. Statistical analyses were done with the Student t-test.

Results

A total of 517 samples were taken during the period of observation. The aerobic counts (CFU/20 cm²) for flat surfaces undergoing regular decontamination are shown in Figure 1. Cultures of the OR floor were taken at sites at which the operating surgeon might stand and yielded very low bacterial counts overall, averaging 7 CFU/20 cm². Other flat surfaces that were sampled included tables or work surfaces on which charting might be done. No significant differences were found in ORs with standard air filtration systems and those using high-efficiency particulate air (HEPA) filtration (data not shown). Counts of aerobic organisms were also obtained from other surfaces in the OR (Fig. 2). All cultured surfaces that are disinfected routinely had low overall bacterial counts. Vertical surfaces at a variety of sites had very few bacteria (data not shown). Stationary telephones and telephone cords in the operating room are subject to intermittent decontamination, as is the Bair Hugger surgical blanket (Arizant Healthcare, 3M, Eden Prairie, MN) used to maintain a patient's body temperature during surgery. However, personal badges and computer mice are not decontaminated regularly and had high bacterial counts because of human contact. Samples of air from the Bair Hugger were taken from air blowing onto an aerobic plate with the Bair Hugger at an intermediate setting for 1 min.

FIG. 1.

Bacterial colony-forming units (CFU)/20 cm² of surface area on flat decontaminated surfaces in operating rooms. OR=operating room; NOS=not otherwise specified.

FIG. 2.

Bacterial colony-forming units (CFU)/20 cm² on various surfaces in operating rooms.

Microbial contamination was detected at high levels both on the tops of uncovered shoes and on the external surfaces of personal hats (Fig. 3). Disposable shoe covers and disposable hats had less than one-third the number of bacterial colonies found on personal shoes and hats. Samples were usually taken after the completion of surgeries. Unused masks generally had only a few bacteria on their surfaces. However, aerobic cultures of the outsides of masks showed an average of 20 CFU/20 cm². Twenty-eight OR personnel had cultures taken from the insides of their masks on both aerobic and staphylococcal plates (Fig. 4). The counts of aerobic organisms were approximately 100-fold greater than what was found on OR floors. The counts of organisms that grew on staphylococcal plates were approximately 50% lower than those of aerobic organisms at the same site.

FIG. 3.

Microbial contamination of the attire of operating-room personnel in colony-forming units (CFU) 20 cm².

FIG. 4.

Microbial contamination in colony-forming units (CFU)/ 20 cm² of face masks worn by operating-room personnel. Staph=Staphylococci.

Staphylococci were present at almost all sites from which aerobic cultures were taken. It must be appreciated that the numbers of counts reported represent only relative densities among different sites of sampling. The types and true numbers of organisms and any relevant pathogenicity cannot be assumed. The tests used in the study reveal only the relative microbial cleanliness of different sites in the OR, and no standards exist against which to compare such cleanliness.

Discussion

Our study shows that routinely disinfected flat and other surfaces in the OR have low levels of bacterial colonization. These are hard non-porous surfaces that are easier to clean than personal surfaces, including shoes and personal hats, which have a higher level of bacterial colonization (by about 10-fold). The study also found that surfaces routinely in contact with OR staff members and personnel had greater overall bacterial colonization than surfaces not in such contact. Masks showed even higher levels of bacterial colonization. Although not unexpected, these results raise the question of what measures can be taken to minimize the release of microbial organisms from the respiratory tract into the air during surgical procedures.

Suzuki et al., in 1984, conducted a study similar to ours [11]. In their survey, done with contact plates, OR floors cleaned with disinfectants had only 3.3 CFU/10 cm² (as compared with those in our study, which had 3.5 CFU/10 cm²). However, their survey found that the colony counts of floors cleaned with detergents ranged from 44.8 to 71.4 CFU/10 cm². When they cultured the floors of dressing rooms covered by carpets, the counts rose to 487.4 CFU/10 cm². Other surfaces in the OR, when cleaned with disinfectants, had an average of 2.8 CFU/10 cm² as compared with an average of 253.2 CFU/10 cm² for surfaces not cleaned with disinfectants. These findings are consistent with our data showing that disinfected flat surfaces in the OR harbor very few bacteria.

Hambraeus et al. [12] calculated that about 15% of the bacteria found in the air of ORs were disbursed from floors, and that higher levels were associated with walking. A later study found that many of the bacteria in OR air are brought in from the outside, including being brought in on shoes [13], with outdoor shoes being more heavily contaminated than those reserved only for the OR, in accord with our observations. Our study found that the colony counts on shoes covered with disposable shoe covers are lower than on uncovered shoes. Not surprisingly, SSIs increase progressively with the number of people in the OR [14]. According to a study by Lynch et al., the number of people in different specialties who entered and exited ORs varied from 19 to 50 per hour across different specialties [15]. Such entries and exits progressively increase the opportunity for the airborne dispersal of microbial organisms and potential microbial colonization of surgical wounds.

Both mobile and mounted telephones have consistently been shown to be contaminated, and can serve as reservoirs for the transmission of resistant bacteria [16 –19]. The use of both headlamps/loupes and operative microscopes has been shown to be associated with bacterial shedding in simulated spine-surgery procedures [20]. More than one-half of all sampled uses of these devices yielded shed pathogens. Another potential source of contamination are forced-air heating blankets, which could emit contaminated air into the OR, as shown in studies by Albrecht et al. [21,22]. The bacterial counts in our study did not show high levels of contaminants from these various sources, but small numbers of samples were taken. The mice used with computers in the OR had higher colony counts than did other surfaces. Fukada et al. [23] reported significant contamination of computer keyboards by anesthetists in ORs. All surfaces that were in ready contact with personnel showed higher levels of contamination than other surfaces, which is logical because all humans are colonized on the skin and in the respiratory tract. We could find no studies of the microbial densities on different types of hats worn in the OR, but it is notable that the densities of bacteria on personal hats were about 10-fold greater than those on OR floors.

Whyte et al. [24] performed a set of experiments that showed the importance of airborne bacterial contamination of wounds associated with hip and knee-joint replacement. They demonstrated that the bacterial count in the air during such arthroplastic surgery was 413 CFU/m³ in a conventionally ventilated OR, as compared with 4 CFU/m³ in an OR ventilated with laminar airflow. A washing out of incisions before skin closure yielded 105 CFU/m³ versus 3 CFU/m³ with conventionally ventilated rooms and those ventilated with laminar airflow, respectively. On the basis of these data, the authors concluded that in a conventionally ventilated OR, 98% of the bacteria in the patient's incisions at the time of wound closure after joint replacement came from the air. Of the organisms coming from the air, Whyte et al. felt that the minority (30%) came directly from the air and that the remainder came indirectly from other sources, such as surgical drapes contaminated from the air. In a subsequent study, Whyte et al. [25] concluded that it was the surgical team that was responsible for bringing and disbursing most of the airborne bacteria that were ultimately deposited into surgical incisions in clean procedures, and that the primary source of these bacterial organisms was the respiratory tract. The respiratory tract is colonized with multiple organisms in healthy individuals. Activities such as speech, coughing, or sneezing have been shown to expel particles into the air. For example, about 10⁶ particles are expelled during a sneeze, 5,000 particles by a cough, and 250 particles when 100 words are spoken loudly. A large percentage of these particles contain bacteria. Whyte et al. also showed that bacterial particles could be dispersed by clothing worn in the OR [26].

Other studies have shown that contamination of a variety of surfaces can occur relatively quickly from the air in an OR. Dalstrom et al. [27] studied the rate of contamination of sterile trays that were opened in an OR with positive air flow. Three of 30 uncovered trays were found to be contaminated immediately after opening. Of the remaining 27 uncovered trays, 4% were contaminated after 30 min, 15% after 1 h, 22% after 2 h, 26% after 3 h, and 30% after 4 h. Covering of the surgical trays with a sterile towel reduced significantly the risk of contamination, which was believed to be attributable most readily to airborne bacteria.

Face masks have been used by medical personnel for more than a century to prevent the transmission of disease, and have evolved with time [28,29]. Meleney and Stevens [30] first reported in 1926 that surgical site infection could be prevented by the use of face masks worn by a surgical team. Since that time, the use of masks by surgical teams has been routine practice. However, several studies have suggested that the wearing of face masks provides no significant benefit for the prevention of SSI [31 –35]. These findings, are somewhat puzzling, in that previous research [36] showed that during periods of talking, unmasked subjects expelled more than 5,000 bacterial contaminants per 5 ft³, whereas masked subjects expelled an average of 19 contaminants per 5 ft³.

Perhaps masks are not as effective as we believe. Mask design may be an important feature for the prevention of SSI. Quesnel [37] showed that the most effective masks contained more fabric, were softer, and were pleated. Efficiency is also linked to properties of the material from which a mask is made, including its mean pore size [38]. The performance of face masks as a question of filtration efficiency has been shown to be related primarily to two aspects of particle penetration: (1) Efficiency of filtration, and (2) leakage at the locations of contact of skin with the mask [29,39 –44]. Grinshpun et al. [42] demonstrated that the lack of a good face seal permitted from 4.8 to 5.8 more particles to enter the air than the number of particles passing through the filter component of the mask. Indeed, it has been estimated that 5%–40% of expelled bacteria are not removed by filtration through a face mask itself but exit through inadequate sealing of the mask to the face [28].

The efficiency of face masks may be especially important because the nares can contain many pathogens that cause surgical site infections. The pathogen most frequently responsible for surgical site infections is S. aureus. Nasal carriage of S. aureus has been studied extensively and appears to occur in approximately 30%–50% of individuals [45]. Eriksen et al. [46] conducted a study in which 104 healthy subjects had repeated cultures taken for S. aureus. It was detected in the noses of 83.7% of all of the subjects over a 19-mo period of observation, and 14.4% were persistent carriers of the organism, whereas only 16.3% were non-carriers. In their study, males were more frequent carriers than females.

Munoz et al. [47] conducted a 1-y observational study of the nasal carriage of S. aureus in patients undergoing major cardiac surgery. Twenty-seven percent of the patients were found to be nasal carriers of S. aureus, and 9.4% had methicillin-resistant S. aureus (MRSA). Patients who were nasal carriers of S. aureus had significantly higher rates of SSI than did non-carriers (12.5% vs. 5%, respectively), and the incidence of SSI in carriers of MRSA was 33%. Staphylococci were responsible for 64% of SSIs.

Coagulase-negative S. epidermidis are also frequent inhabitants of the nares. Bitkover et al. found that S. epidermidis causes almost as many SSIs in cardiac patients as does S. aureus (33%–62.5%) [5]. This group took environmental cultures from patients, OR personnel, and the air during elective coronary artery bypass grafting (CABG) or valve replacement in 20 patients, and demonstrated incisional contamination in 13 of the 20 procedures. Organisms from incisions could be traced to the patients themselves in three cases, to the surgeon's arm and forehead in one case, and to the surgeon's or assistant's nose in two cases. Thus, in this small study, 15% of the bacteria identified as pathogens came from the nares of the surgical team.

In our study, the aerobic microbial densities on the inside surfaces of face masks were approximately 32 times greater than those on the outer surfaces of the masks, but more than 150 times greater than the microbial densities measured on the OR floor. Many of the bacteria on cultured surfaces appeared to be staphylococcal species. With the poor filtration characteristics of face masks and the large numbers of bacteria involved, it is easy to see why the inefficiency of face masks in trapping and retaining bacteria could be a major factor in the contamination of surgical wounds. It is also easy to see how the failure to remove a used mask will contaminate other areas of the hospital environment.

Recommendations

The contact culture plates used in this study can be used to identify comparative densities of organisms for the surveillance of bacterial contamination in the OR. To reduce bacterial contamination in the OR, new disposable shoe covers and hair coverings should be worn for each operation. Only masks that reduce leakage at the mask–face interface should be worn, and should be discarded routinely after every operation.

It is understood that these recommendations are unlikely to be subjected to meaningful prospective randomized clinical trials to show that they reduce the incidence of SSI. However, such trials are neither necessary nor indicated. It has been known since the work of Semmelweis 165 years ago that the principle of exposing incisions to the smallest possible bacterial load (asepsis) is sound and the foundation of infection control. As Atul Gawande recently noted in his book Better [48], a knowledge of sound principles is not adequate without an institutional will and commitment to follow such principles. However, both of these elements are consistent with the concept that all measures that can potentially reduce bacterial contamination of surgical wounds should be routinely implemented as components of optimal patient care.

Footnotes

Author Disclosure Statement

There was no financial support for this research apart from departmental funding. All supplies for the study described herein were purchased.

References

Thompson

, Oldenburg

, Deschamps

et al. Chasing zero: The drive to eliminate surgical site infections. Ann Surg, 2011; 254:430–436.

de Sissovay

, Fraeman

, Hutchins

et al. Surgical site infection: Incidence of impact on hospital utilization and treatment costs. Am J Infect Control, 2009; 37:387–397.

DiPiro

, Martindale

, Bakst

et al. Infection in surgical patients: Effects on mortality, hospitalization, and post-discharge care. Am J Health Syst Pharm, 1998; 55:777–781.

Hawn

, Vick

, Richman

et al. Surgical site infection prevention: Time to move beyond the surgical care improvement program. Ann Surg, 2011; 254:494–501.

Awad

. Adherence to surgical care improvement project measures and post-operative surgical site infections. Surg Infect, 2012; 13:234–237.

Davis

, Kuo

, Ahmed

et al. Report card on Surgical Care Improvement Project (SCIP): Nationwide inpatient sample infection data 2001–2006. Surg Infect, 2011; 12:429–434.

Barie

. SCIP to the Loo? Surg Infect, 2011; 12:161–162.

Rosenberger

, Politano

, Sawyer

. The surgical care improvement project and prevention of post-operative infection, including surgical site infection. Surg Infect, 2011; 12:163–168.

Edmiston

, Spencer

, Lewis

et al.

Reducing the risk of surgical site infections: Did we really think SCIP was going to lead us to the promised land?

Surg Infect, 2011; 12:169–177.

10.

Bitkover

, Marcusson

, Ransjo

. Spread of coagulase-negative staphyloccoci during cardiac operations in a modern operating room. Ann Thorac Surg, 2000; 69:110–115.

11.

Suzuki

, Yoshimichi

, Matsuura

, Horisawa

. Bacterial contamination of floors and other surfaces in operating rooms: A five-year survey. J Hyg Camb, 1984; 93:559–566.

12.

Hambreus

, Bengtsson

, Laurell

. Bacterial contamination in a modern operating suite. 3. Importance of floor contamination as a source of airborne bacteria. J Hyg (Lond), 1978; 80:169–174.

13.

Amirfeyz

, Tasker

, Ali

et al.

Theater shoes—a link in the common pathway of postoperative wound infection?

Ann R Coll Surg Engl, 2007; 89:605–608.

14.

Pryor

, Messmer

. The Effect of traffic patterns in the OR on surgical site infections. AORN J, 1998; 68:649–660.

15.

Lynch

, Englesbe

, Sturm

et al. Measurement of foot traffic in the operating room: Implications for infection control. Am J Med Q, 2009; 24:45–52.

16.

Ulger

, Esen

, Dilek

et al.

Are we aware how contaminated our mobile phones with nosocomial pathogens?

Ann Clin Microbiol Antimicrob, 2009; 6:7.

17.

Brady

, Verran

, Damani

, Gibb

. Review of mobile communication devices as potential reservoirs of nosocomial pathogens. J Hosp Infect, 2009; 71:295–300.

18.

Jeske

, Tiefenthaler

, Hohlrieder

et al. Bacterial contamination of anaesthetists' hands by personal mobile phone and fixed phone use in the operating theater. Anaesthesia, 2007; 62:904–906.

19.

Nelson

, Bivens

, Shinn

et al. Microbial flora on operating room telephones. AORN J, 2006; 83:607–626.

20.

Weiner

, Kilgore

. Bacterial shedding in common spine procedures: Headlamp/loupes and the operative microscope. Spine, 2007; 32:918–920.

21.

Albrecht

, Gauthier

, Leaper

. Forced-air warming: A source of airborne contamination in the operating room? Orthop Rev (Pavia), 2009; 1:e28.

22.

Albrecht

, Gauthier

, Belani

et al. Forced-air warming blowers: An evaluation of filtration adequacy and airborne contamination emissions in the operating room. Am J Infect Control, 2011; 39:321–328.

23.

Fukada

, Iwakiri

, Ozaki

. Anaesthetists' role in computer keyboard contamination in an operating room. J Hosp Infect, 2008; 70:148–153.

24.

Whyte

, Hodgson

, Tinkler

. The importance of airborne bacterial contamination of wounds. J Hosp Infect, 1982; 3:123–135.

25.

Whyte

. The role of clothing and drapes in the operating room. J Hosp Infect, 1988; 11,Suppl C:2–17.

26.

Whyte

, Vesley

, Hodgson

. Bacterial dispersion in relation to operating room clothing. J Hyg (Lond), 1976; 76:367–378.

27.

Dalstrom

, Venkatarayappa

, Manternach

et al. Time-dependent contamination of opened sterile operating-room trays. J Bone Joint Surg Am, 2008; 90:1022–1025.

28.

Belkin

. The surgical mask: Are new tests relevant for OR practice? AORN J, 2009; 89:883–891.

29.

Huang

, Huang

. Evaluation of the efficiency of medical masks and the creation of new medical masks. J Int Med Res, 2007; 35:213–223.

30.

Meleney

, Stevens

. Post-operative hemolytic streptococcus wound infections and their relation to hemolytic Streptococcus carriers among operating personnel. Surg Gynecol Obstet, 1926; 43:338–342.

31.

Lipp

, Edwards

. Disposable surgical face masks for preventing surgical wound infection in clean surgery. Cochrane Database Syst Rev, 2002; CD002929.

32.

Webster

, Croger

, Lister

et al. Use of face masks by non-scrubbed operating room staff: A randomized controlled trial. ANZ J Surg, 2010; 80:169–173.

33.

Jepsen

, Pers

, Lester

, Jensen

. Importance of surgical masks for preoperative asepsis. Ugeskr Laeger, 1993; 155:1940–1942.

34.

Bahil

. Does evidence based medicine support the effectiveness of surgical facemasks in preventing postoperative wound infections in elective surgery? J Ayub Med Coll Abbottabad, 2009; 21:166–170.

35.

McCluskey

. Does wearing a facemask reduce bacterial wound infection? A literature review. Br J Theatre Nurs, 1996; 6:18–2029.

36.

Greene

, Vesley

. Method for evaluating effectiveness of surgical masks. J Bacteriol, 1962; 83:663–667.

37.

Quesnel

. The efficiency of surgical masks of varying design and composition. Br J Surg, 1975; 62:936–940.

38.

Leonas

, Jones

. The relationship of fabric properties and bacterial filtration efficiency for selected surgical face masks. JTATM, 2003; 3:1–8.

39.

Sanchez

. Filtration efficiency of surgical masks. 2010. Theses and DissertationsPaper 1760http://scholarcommons.usf.edu/etd/1760.

40.

Romney

. Surgical face masks in the operating theater: Re-examining the evidence. J Hosp Infect, 2001; 47:251–256.

41.

Chen

, Willeke

. Aerosol penetration through surgical masks. Am J Infect Control, 1992; 20:177–184.

42.

Grinshpun

, Haruta

, Eninger

et al. Performance of an N95 filtering facepiece particulate respirator and a surgical mask during human breathing: Two pathways for particle penetration. JOEH, 2009; 6:593–603.

43.

Balazy

, Toivola

, Adhikari

et al.

D95 respirators provide 95% protection level against airborne viruses, and how adequate are surgical masks?

Am J Infect Control, 2006; 34:51–57.

44.

Cho

, Reponen

, McKay

et al. Large particle penetration through N95 respirator filters and facepiece leaks with cyclic flow. Ann Occup Hyg, 2010; 54:68–77.

45.

Kluytmans

, van Belkum

, Verbrugh

. Nasal carriage of Staphyloccocus aureus: Epidemiology, underlying mechanisms, and associated risks. Clin Microbiol Rev, 1997; 10:505–520.

46.

Eriksen

, Espersen

, Rosdahl

, Jensen

. Carriage of Staphyloccocus aureus among 104 healthy persons during a 19-month period. Epidemiol Infect, 1996; 115:51–60.

47.

Munoz

, Hortal

, Giannella

et al. Nasal carriage of S. aureus increases the risk of surgical site infection after major heart surgery. J Hosp Infect, 2008; 68:25–31.

48.

Gawande

. Better. A Surgeon's Notes on Performance. New York: Metropolitan Books, Henry Holt, 2007.