BOWERS, ANDREA L.; HUFFMAN, G. RUSSELL; SENNETT, BRIAN J.
Cutaneous lesions including cellulitis, folliculitis, furunculitis, boil, and abscess are not uncommon among competitive athletes, presumably due to the direct physical contact and the risk for disruption of the skin barrier with rough play. As in the general population, such lesions in the athlete typically are suspected to be caused by Staphylococcus and Streptococcus bacterial species.
Traditionally, staphylococcal infections have been treated with synthetic penicillins, such as oxacillin, methicillin, or nafcillin; however, methicillin-resistant Staphylococcus aureus (MRSA) strains have grown more prevalent since they were first identified (1,5). Failure to recognize a resistant strain and continued treatment with an inappropriate antibiotic can contribute to patient morbidity and further resistance patterns among the community.
Within the last decade, culture-positive community-acquired MRSA (ca-MRSA) infections have been recognized in the athletic population. Although evidence in the literature is sparse, ca-MRSA infections have been reported across athletic disciplines including rugby (11), wrestling (7), weightlifting (4), and fencing (3). Infections among American football players have been reported at the high school (9), college (2,8,10), and professional (6) levels. Reports in the literature, although limited, speak to a growing problem with potentially disastrous implications. ca-MRSA infection is a disease process in which the incidence likely outweighs the prevalence, and the feasibility of level 1 or 2 studies in this area is limited. The highest level of evidence available currently to examine disease demographics in this case remains case series reporting. Nonetheless, consolidation of the existing reports might serve to generate a greater awareness for health care practitioners caring for the athletic population, thereby facilitating treatment and prevention of transmission of disease.
A review of the literature was performed to identify case reports of ca-MRSA among collegiate football players to better characterize the demographics of affected population. An electronic search of PubMed was performed using Boolean operators "and" or "or" and the terms "MRSA," "methicillin-resistant Staphylococcus aureus," "skin infection," "cutaneous infection," and "college football." The articles returned were obtained, and a review of the references for each paper was performed for additional appropriate reports. Only English-language papers were included. Reports involving high school or professional athletes, and those involving "football," but pertaining to soccer rather than "American football," were excluded. Reports were cross-referenced to eliminate duplicate entries. An outbreak involving athletes at our home institution is also reported here for the first time. Written informed consent was obtained for chart review from the involved subjects, and approval was obtained from the institutional review board.
Subject data were consolidated from reports meeting inclusion criteria. Each case series was scrutinized for data involving only culture-positive, proven MRSA infections. For each incidence, "day of infection" was calculated on the basis of the time that had elapsed since the start of training camp for that season. Univariate statistical analysis was performed using chi-square testing to evaluate probability of infection on the basis of timing of season, body part involved, and player position (offense vs defense; lineman vs nonlineman). Reverse Kaplan-Meier analysis was performed to evaluate cumulative risk for infection through the season.
Five reports specific to MRSA skin infection in the American collegiate football player were identified. On cross-referencing, two were identified as duplicate descriptions of the same outbreak. When appropriate, information was combined from the duplicate entries to provide a broader description of the outbreaks. Of these series, only cases in which MRSA was confirmed by culture were included for data consolidation and analysis.
The first documented report among collegiate football players was issued by the Centers for Disease Control and Prevention in 2003 regarding an outbreak in September 2000 among 10 athletes in Pennsylvania (3). Seven of these patients were hospitalized for management of their infection. Possible risk factors identified were skin trauma from turf burns and shaving and sharing of unwashed bath towels.
Another outbreak in August to October 2003, reported a year later as a retrospective cohort study, involved 10 cases among a Connecticut-based team of 100 athletes (2). Nine required incision and drainage to evacuate pus, and another underwent arthrocentesis. Two players received intravenous antibiotics; one for a recurrent infection after completing an oral regimen, and the other for presumed septic arthritis in the setting of confirmed MRSA abscess at a remote location (chin abscess; Table 1). All infections were verified as MRSA by culture and were further subtyped as the USA300 strain by pulsed-field gel electrophoresis. Polymerase chain reaction analysis identified the Panton-Valentine leukocidin toxin gene common among all 10 specimens. Nasal swabs were performed in all team members; zero revealed MRSA, although 43 of 100 samples grew methicillin-sensitive Staphylococcus aureus. Such cultures were consistent with colonization, and nasal colonization was concluded to play no role in the MRSA outbreak. Significant relative risk by position was reported to be 17.5 for cornerbacks and 11.7 for wide receivers. The presence of turf burn yielded a relative risk of 7.2, and the act of body shaving, thought to generate microabrasions vulnerable to contamination, was found to have a relative risk of 6.1. The authors postulated that direct contact was the likely mode of transmission.
In 2006, Romano et al. (10) reported 14 cases in 13 athletes spanning three seasons. A retrospective analysis was performed on 2 cases presenting in 2002, including an upper extremity necrotizing fasciitis, 11 in 2003, and 1 recurrent infection in 2004. All infections were confirmed to be MRSA by culture; another six were suspected by clinical diagnosis in the 2003 season. Fourteen infections during the 3-yr period involved abscesses that necessitated incision and drainage, two of which were performed in the operating room; one of these patients required multiple surgical debridements for reported necrotizing fasciitis. Six received intravenous antibiotics; these included athletes who underwent formal debridement in the operating room and others who "did not respond to [outpatient, oral] antibiotic treatment." Nasal cultures were obtained in the 2003 season after five infections were reported within a week. Of 106 samples, 7 (6.6%) yielded MRSA and 19 (17.9%) methicillin-sensitive Staphylococcus aureus. The MRSA carriers were decolonized with topical nasal mupirocin ointment and oral rifampin. Two weeks later, new samples were taken from all players. The seven previously MRSA-positive players were then found to be negative; however, four (3.7%) new players were identified and subsequently decolonized for MRSA. At the start of the 2004 season, two returning and one new player had positive nasal cultures and underwent decolonization. An analysis performed by infectious disease specialists (8) determined that sharing bars of soap and having preexisting cuts or abrasions were associated with infection. Nasal carriage was associated with having a locker near a teammate with skin infection, sharing towels, and living on campus.
THE PENN EXPERIENCE
During a 9-wk period from September 10 to November 13, 2006, 10 cultures in wounds of nine varsity collegiate football players were found positive for MRSA (Table 1). Four other team members presented with five wounds consistent with ca-MRSA by clinical examination that were not cultured. The initial four players presented skin lesions on day 19, and five MRSA-positive cultures were obtained from these players during the subsequent 36 h. Cultures of lesions were obtained from subsequent players on days 24, 25, 26, 27, 28, 36, 51, and 83 of the season. The last positive culture on record was obtained 2 months 3 d after the index case. The standard of care in managing abscess formations is surgical incision and drainage. In following this criteria, seven wounds required incision and drainage performed in the training room, or in one case, in the emergency room, because of the presence of pus under pressure. Two athletes were treated with intravenous antibiotics on the basis of the rapidity of onset and exacerbation of symptoms, extent of pain, and concomitant systemic symptoms including malaise and fever.
The culture-positive athletes ranged in age from 18 to 24 yr, with a mean of 21 yr. Anthropometric analysis reveals that player height averaged 73 inches (range = 71-76 inches), weight averaged 231 lb (range = 185-290 lb), with body mass index (BMI) averaging 30.0 (range = 25.1-35.6), not statistically significantly different from the team average of 73 inches, 222 lb, and BMI of 29.0. By position, the infections occurred in four defensive linemen (including defensive ends), two linebackers, one defensive back, one offensive lineman, and one wide receiver (Fig. 1).
FIGURE 1-Incidence o...Image Tools
The sites of infection were predominantly on the extremities. Five infections involved the lower extremities: one thigh, two around the knee joint, and two on the lower leg. Two upper-extremity infections involved the skin around the elbow and forearm. Three nonextremity sites included one on each of the buttocks, nipple, and chin (Fig. 2).
FIGURE 2-Anatomic lo...Image Tools
All cultured strains were sensitive to trimethoprim-sulfamethoxazole, vancomycin, clindamycin, and tetracycline. In each case, bacteria were resistant to ampicillin-sulbactam, cefazolin, erythromycin, levofloxacin, oxacillin, and penicillin. This ca-MRSA strain infected otherwise healthy hosts without immunosuppression, malnutrition, or other identifiable factors that would render them easily susceptible to infection.
Of the summarized reports, descriptions of three outbreaks among Connecticut (2), California (8,10), and Pennsylvania provided adequate demographic information for consolidated analysis (Table 1). Thirty-three among 491 collegiate football players are described, yielding an overall attack rate, or cumulative risk per season, of 6.7%.
By position, infections occurred in eight defensive linemen (including defensive ends), seven defensive backs (including cornerbacks and safety), five linebackers, five wide receivers, two offensive linemen, two quarterbacks, two "other" (described as "not cornerback or wide receiver"), one tight end, and one running back (Fig. 1). Because of the low numbers of total infections, meaningful statistical analysis could not be performed on the basis of individual position played. However, univariate statistical analysis revealed no difference in probability of infection between offense and defense or between lineman and nonlineman.
A majority (n = 22; 60%) of the skin infections were in the form of an abscess. Cellulitis was the second most frequent manifestation (n = 26; 16%). The remaining lesions included folliculitis, insect bites, impetigo, and necrotizing fasciitis. Incision and drainage was performed 90% of the time (n = 30); most were in the form of beside incision and drainage for abscess or impetigo, although in one extreme circumstance, repeated surgical debridements were performed for necrotizing fasciitis. Intravenous antibiotics were administered to 30% (n = 10), usually in the setting of operative debridement, systemic symptoms, or recurrent infection. All but seven lesions occurred on the extremities. The most common location of lesion was the elbow (n = 11), followed by the knee (n = 6), lower leg (n = 4), and forearm (n = 4; Fig. 2).
The first infections occurred within the first 2 wk of training camp. Most outbreaks occurred within the first third of the season, although isolated infections were observed toward the end of the second and third months of the season (Fig. 3). Chi-square testing revealed a significant probability of infection occurring in the first third of the season compared with either the second or the last thirds (P < 0.001 for each), and the probability of infection occurring during the middle third versus the last third of the season nearly approached statistical significance (P = 0.06). The cumulative risk for MRSA infection over the course ofa collegiate football season was 0.067 (Fig. 4).
Here, we present 10 new cases of cutaneous MRSA infection in nine collegiate football players. Forty-one such cases had been reported previously since the year 2000. Altogether, 37 cases in 33 athletes were described with enough detail for consolidated analysis.
Whereas a prior report noted a preponderance of infection among cornerbacks and wide receivers (2), the consolidated data do not support this trend, and in fact, the widespread distribution suggests that risk for infection may be independent of the position. Although the limited numbers prevented statistical analysis by individual position, we did not detect a difference in the probability between offense and defense or between lineman and backfield.
In each case, infections did not occur until the second week of training camp. This is typically the time that hitting drills begin and the players are increasingly subjected to direct physical contact with their teammates. Anatomically, MRSA infections most commonly affected the extremities. The elbow was the most frequently affected body part. Given that many infections occurred early in the season, the elbow may not have been as commonly covered, potentially increasing its exposure. We suspect that the skin, therefore, would be exposed and at risk for minor trauma, such as insect bites, abrasions, or "turf burn." The compromised cutaneous barrier could also directly contact skin and sweat of other players and place the athlete at risk for transmission of MRSA and subsequent soft-tissue infection. Notably, infection rates dropped as the season progressed, possibly because athletes competed with less exposed skin and less perspiration when playing in colder weather. Although this study did not encounter enough data to examine the use and effect of protective clothing, we predict that wearing protective clothing or gear might minimize the risk for cutaneous MRSA infection.
Prior reports (2,8,10) identified MRSA outbreaks that were isolated to the first third of the season. Again, exposed skin may have facilitated transmission during the early part of the season. Nonetheless, the Penn experience demonstrates that although infections are indeed more common early in the season, they can in fact occur later in the season as well. A high index of suspicion should be maintained throughout the football season.
Although the true incidence of such infections is not known, we suspect that ca-MRSA infections among athletes may actually be underreported and therefore underappreciated. We postulate that a variety of factors might contribute to proposed underrepresentation. Cutaneous infections in the community are not routinely cultured. An infected athlete may seek treatment from a private physician, walk-in clinic, or emergency room. The athlete may choose not to disclose the infection to team medical staff for personal reasons, perceived stigmatism, or concern that it might affect eligibility for play. At present, there is no working surveillance system, beyond self-reporting, to identify players with MRSA infection.
The Centers for Disease Control has issued formal recommendations regarding measures for preventing staphylococcal skin infections among sports participants (3). These include the following:
1. Cover all wounds. If a wound cannot be covered adequately, consider excluding players with potentially infectious skin lesions from practice or competitions until the lesions are healed or can be covered adequately.
2. Encourage good hygiene, including showering and washing with soap after all practices and competitions.
3. Ensure availability of adequate soap and hot water.
4. Discourage sharing of towels and personal items (e.g., clothing or equipment).
5. Establish routine cleaning schedules for shared equipment.
6. Train athletes and coaches in first aid for wounds and recognition of wounds that are potentially infected.
7. Encourage athletes to report skin lesions to coaches and encourage coaches to assess athletes regularly for skin lesions.
Given the increasingly apparent emergence of ca-MRSA in the athletic population, the standard β-lactamase-resistant synthetic penicillins may no longer suffice as adequate broad-spectrum empiric coverage for cutaneous lesions. During the outbreak in our institution, after the initial culture reports demonstrated MRSA in multiple athletes, oral trimethoprim-sulfamethoxazole was used as the first-line agent for all subsequent reported skin infections.
Any review of literature is limited by the quality of the studies that are available. Although each of the discussed outbreak descriptions contributes to a broader picture of MRSA in the collegiate football athlete, there is no standardization to the data reporting. Overall, similar information was reported but not broken down on a case-by-case basis. Because of this variation of reporting among the studies, the data are not amenable to meaningful multivariate analysis. Nonetheless, the review of literature and consolidation of data better characterizes the demographics of a disease process that can generate extensive morbidity in an otherwise healthy and robust athletic population.
Further, in two of the series, several more "suspicious" cutaneous infections, which either were not amenable to or did not necessitate drainage and culture, were mentioned. Although there are no cultures to prove MRSA as the offending organism, there is a suspicion, because of clinical presentation and temporal correlation with other culture-positive infections, that these may have been MRSA as well. The true incidence of MRSA infection among this athletic population may indeed have been even greater than what has been documented by culture and therefore reported in this study. Similarly, the high percentage of undergoing incision and drainage may seem inflated because the analyzed cases were limited to culture-proven MRSA infections; the true percentage of MRSA infections requiring drainage may actually be lower if the suspicious, but not cultured, cellulitis infections were considered as well.
As MRSA infections become more prevalent in both hospitals and the community, team physicians and athletic trainers need to remain vigilant regarding athlete and staff education, prompt diagnosis, reporting, and further investigation of associated risk factors and possible means of prevention.
The results of the present study do not constitute endorsement by the ACSM.