Healthcare-Associated Burn Wound Infections
C. Glen Mayhall
Burn patients are among the patients at highest risk for healthcare-associated infections. These patients have lost a portion of their integument that would ordinarily be a strong barrier to invasion by microorganisms. In addition, the necrotic tissue in the burn eschar combined with the presence of serum proteins provides a rich culture medium for microorganisms. Added to the loss of integument is the adverse effect of thermal injury on both local and systemic immunity (1,2). Given these effects of burn trauma, it is easily understood why burn patients are at risk for healthcare-associated burn wound infections.
There are approximately 2 million fires in the United States annually leading to 1.2 million burn injuries. About 100,000 patients with moderate to severe burns require hospitalization, and about 5,000 of these burns are fatal (3, 4, 5, 6, 7).
Data submitted from burn intensive care units (BICUs) to the National Nosocomial Infections Surveillance (NNIS) system at the Centers for Disease Control and Prevention indicate that the cumulative incidence for burn wound infections is 4.5% and the incidence rate is 6.8 cases per 1,000 patient days (R. Gaynes, personal communication, 1998). Infections are the most common cause of death in burn patients, and the most common sites of infection are the lungs and the burn wound (8). The burn wound may also initiate and perpetuate a mediator-induced septic response accompanied by multiple-organ failure in the absence of an identifiable focus of infection and with negative blood culture results (9). Thus, proliferation of microorganisms in the burn wound followed by invasion of subjacent viable tissue or the mediator-induced septic response may cause the clinical manifestations of sepsis.
Although the most important cause of death in burn patients is infection, the current overall mortality rate due to infections in the burn patient is unknown (10). However, data from the NNIS system on patients with burn wound infection who died, and for whom the relationship of infection to death was reported, indicated that 18 (12.6%) of 143 deaths were caused by burn wound infection. Burn wound infection contributed importantly to death in 104 patients (72.7%), and the burn wound infection was unrelated to death in the remaining 21 patients (14.7%) (R. Gaynes, personal communication, 1998). It has also been observed that mortality in burn patients is significantly increased by bacteremia due to gram-negative bacilli (11).
TYPES OF BURNS
Most burns are due to thermal injury. According to the National Burn Repository, for cases with a known etiology, 40% of cases are due to fire/flame injuries and 30% of cases are due to scald injuries. Nine percent of injuries are due to contact with a hot object. The remaining cases are due to electrical injury (4%) and chemical injury (3%) (12).
PATHOGENESIS OF BURN WOUND INFECTIONS
Loss of the integument combined with the immune defects that accompany thermal injury place the burn patient at high risk for burn wound infection. Microorganisms are present on the skin at the time of burning and are readily acquired from the patient’s gastrointestinal tract after the thermal injury has been sustained. Microorganisms are rapidly acquired from the environment of the burn care facility as well as from other burn patients cared for in the same unit.
In addition to the loss of the skin barrier, the rapid colonization of the burn wound from endogenous and exogenous sources, and the excellent culture medium provided by the burn wound, thermal injury has a substantial suppressive effect on the immune system. The nonspecific immune system is involved with long-term suppression of the neutrophil oxidative burst (13) and impairment of neutrophil chemotaxis (14). Abnormalities in cellular immunity are reflected by a decreased ratio of helper to suppressor lymphocytes and a decrease in natural killer cell activity (2,15). Changes in monocyte function are reflected by an early release of high mobility group box protein 1 (HMGB1) after a burn injury and release of the cytokines IL-6, IL-8, and IL-10 (16). Higher levels of HMGB1 and IL-10 on admission were predictive of a fatal outcome. Another dysfunction of monocytes that occurs after burn injury is decreased production of human leukocyte antigen-DR (mHLA-DR) (17). Severe burn injury induces a marked decrease in the production of mHLA-DR by circulating monocytes. The marked reduction of mHLA-DR expression in severely burned patients leads to septic complications and a fatal outcome.
Thermal injury leads to local accumulation of cytokines in the areas of burn injury that “spill over” into the systemic circulation (18). Thus, local accumulation of multiple cytokines in the area of injury that mediate the reparative process has a marked suppressive effect on host defenses when these cytokines enter the bloodstream and are distributed throughout the body.
Thermal injury leads to local accumulation of cytokines in the areas of burn injury that “spill over” into the systemic circulation (18). Thus, local accumulation of multiple cytokines in the area of injury that mediate the reparative process has a marked suppressive effect on host defenses when these cytokines enter the bloodstream and are distributed throughout the body.
With the loss of the integument, immunosuppression, and availability of nutrients for microbial proliferation, microorganisms contaminating the surface of the burn wound may multiply to high concentrations. Early colonization of the wound in the first 48 hours takes place with gram-positive microorganisms from within the depths of the sweat glands and hair follicles (10,15,19). Between 3 and 21 days, the wound becomes colonized with gram-negative bacilli from the patient’s own gastrointestinal tract or from other patients in the burn care facility (10,15). If microorganisms reach a concentration of at least 105 colony-forming units (CFU) per gram of tissue, they may spread from the hair follicles along the dermal subcutaneous junction (19). Perivascular colonization may result in thrombosis, vascular occlusion, and necrosis of the remaining viable elements. The resultant ischemia and bacterial autolysis may convert a partial-thickness injury to a full-thickness injury. In burn wounds with unexcised eschar, invasion of the subeschar viable subcutaneous tissue results in burn wound infection or burn wound sepsis and may be complicated by bacteremia.
CLINICAL MANIFESTATIONS OF BURN WOUND INFECTIONS
In burn wounds with unexcised eschar, clinical manifestations of burn wound infection appear when microorganisms reach high concentrations in the burn eschar and invade subjacent viable tissue. Clinical signs of infection may depend, to some extent, on the type of infecting microorganism. Hyperthermia and leukocytosis tend to be more marked in patients with gram-positive infection. Infections with gram-positive microorganisms are also more often associated with irrational behavior and mental confusion. The appearance of a wound infected by gram-positive microorganisms may be characterized by maceration with a ropy tenacious exudate and surrounding cellulitis (19).
Patients with gram-negative burn wound infection are more likely to have hypothermia and leukopenia. Although they may have altered mental status with confusion, some patients with gram-negative burn wound infection may remain lucid until near death (19,20). Patients with gramnegative infection may also have glucose intolerance with hyperglycemia, ileus and abdominal distention, respiratory distress syndrome, and oliguria (20).
The wound infected by gram-negative microorganisms is characterized by (a) focal gangrene that coalesces and spreads throughout the wound; (b) conversion of a partial-thickness wound to a full-thickness wound; (c) hemorrhagic discoloration of subeschar tissue; (d) focal, multifocal, or generalized dark brown, black, or violaceous discoloration of the burn wound; and (e) changes in unburned skin at the wound margins characterized by edema and violaceous discoloration (19,20). Bacteremia is a common complication of burn wound infection, but absence of bacteremia does not rule out burn wound infection. In fact, fatal burn wound infection may occur in the absence of bacteremia, particularly when the infection is caused by gram-negative microorganisms (19).
DIAGNOSIS OF BURN WOUND INFECTION
Clinical Diagnosis
Examination of the burn wound and clinical signs and symptoms provide important clues to the diagnosis of burn wound infection. As noted above, changes in the wound characterized by dark brown, black, or violaceous discoloration; unexpectedly rapid separation of the eschar; hemorrhagic discoloration of subeschar tissue and edema; and violaceous discoloration of unburned skin at the wound margin suggest burn wound infection (20). Clinical suspicion of infection is heightened when these local wound manifestations are accompanied by hypothermia (<36°C), hyperthermia (>38°C), hypotension (systolic blood pressure ≤90 mm Hg), oliguria (<20 mL/h), ileus with abdominal distention, glucose intolerance and hyperglycemia, or altered mental status (20,21).
Microbiologic Diagnosis of Burn Wound Infection
In a study wherein about 80% of burn wound biopsies were obtained from patients with local signs of burn wound infection, Pruitt and Foley (22) observed that 75% of patients with more than 105 CFU/g of burn wound tissue died. When this density of microorganisms in the burn eschar was combined with a grade 6 histologic diagnosis (“invasive infection with microbial penetration into viable tissue beyond the depth of original necrosis”), the mortality rate was 100%. In another study comparing quantitative burn wound cultures and histopathologic examination of the same tissue, McManus et al. (23) observed that growth of at least 105 CFU/g of tissue identified burn wound infection diagnosed by histopathologic assessment of tissue 96.1% of the time. However, 64.3% of biopsies that showed negative results histopathologically also had at least 105 CFU/g of tissue. Thus, quantitative burn wound cultures have a high sensitivity (96.1%) but a low specificity (35.7%). Stated another way, burn wounds with <105 CFU/g of tissue are highly unlikely to be infected, whereas only about one-third of burn wounds with at least 105 CFU/g of tissue will be infected.
Although the threshold of at least 105 CFU/g of burn wound tissue has a high sensitivity and low specificity for burn wound infection, a threshold of over 108 CFU/g of tissue is highly suggestive of burn wound sepsis and impending death in the untreated patient (24,25). Thus, even though the definitive diagnosis of burn wound infection may be made by histopathologic examination of a fullthickness biopsy of the burn wound (see below), it would appear that burn wound biopsies containing <105 CFU/g of tissue suggest that burn wound infection is unlikely and that biopsies with more than 108 CFU/g of tissue are highly suggestive of burn wound sepsis.
One problem with the interpretation of quantitative cultures of burn wound biopsies has been the uneven distribution of microorganisms, both qualitatively and quantitatively, throughout the burn wound. In a study wherein culture results were not correlated with clinical manifestations, appearance of the burn wound, or histopathologic examination of the tissue taken for culture, Woolfrey et al. (26) concluded, “Quantitative results derived from burn wound biopsy cultures are unreliable and may be significantly misleading when used for decision-making relative to patient care.” These authors divided each biopsy specimen and cultured the two portions separately. Between the two segments, an average of 4.8 microorganism types was recovered. At the 105 CFU/g of tissue breakpoint, the paired quantitative results agreed within the same log increment for only 38% of biopsies. Although it is clear from the data of Woolfrey et al. that there may be large qualitative and quantitative differences in burn wound microbiology between immediately adjacent areas of the wound, it is unclear how their results relate to the appearance of the burn wound (and whether biopsies were taken from areas of the wound that appeared infected on clinical examination), histopathologic examination of the tissue taken for culture, and the clinical course of the patient.
Surface swab cultures, either qualitative or quantitative, have been used in the diagnosis of burn wound infections. Steer et al. (27) compared qualitative results and quantitative bacterial counts of 141 surface swabs and 141 wound biopsies taken from 74 burn patients. They observed a significant correlation between the total bacterial counts of surface swabs and the total bacterial counts of biopsies (p < .001), but the predictive value of the counts obtained by one method to predict the counts obtained by the other method was poor. The qualitative correlation was also poor, and only 54% of the biopsy/swab pairs yielded the same microorganism on culture. There were two exceptions to the latter observation. When Staphylococcus aureus was present in the burn wound biopsy, it was present on surface culture 95% of the time, and when Pseudomonas aeruginosa was recovered from the burn wound biopsy, it was cultured from the surface swabs 92% of the time. Thus, although S. aureus and P. aeruginosa in the burn wound may be detected by surface swabs, qualitative and quantitative surface cultures are generally not useful for predicting the qualitative and quantitative microbiology of the burn wound.
When the diagnosis of burn wound infection is made by histopathologic examination of a full-thickness burn wound biopsy (see below), quantitative culture of a portion of the biopsy may identify the causative microorganism(s) and provide antimicrobial susceptibility data for the selection of appropriate antimicrobial therapy. McManus et al. (23) observed a 100% concordance between microorganisms seen on histopathologic examination and those recovered on culture. Thus, burn wound biopsies should be both cultured and examined histopathologically. Although quantitative cultures may not be necessary for identification of the causative microorganism(s), quantitation might be useful in separating microorganisms on and in the nonviable tissue from those invading viable tissue. It would be unlikely that microorganisms recovered at a concentration below 105 CFU/g of tissue would be causing burn wound infection (23).
Although some investigators have found a good correlation between the results of quantitative Gramstained preparations and quantitative cultures of biopsy tissue (28,29), others have not (30). Even with a good correlation between quantitative Gram-stain and culture results, the results of microscopic examination of Gram-stained tissue should not be used for diagnosis of burn wound infection, because definitive diagnosis of burn wound infection requires histopathologic examination and culture.
Histopathologic Diagnosis of Burn Wound Infection
In burn wounds with unexcised eschar, the diagnosis of burn wound infection may be made by histopathologic examination of a full-thickness burn wound biopsy. The biopsy should be taken from the area of the wound with the most pronounced local changes (see above). A lenticular tissue sample should be obtained using a scalpel. The biopsy should measure from 0.5 × 0.5 × 0.5 cm to 1 × 1 × 1 cm and should weigh between 100 and 500 mg. The biopsy must include underlying or adjacent unburned tissue in addition to the eschar. The specimen should be divided; one half should be cultured quantitatively, and the other half should be placed in 10% neutral buffered formalin solution for processing for histopathologic examination. The tissue should be stained with hematoxylin and eosin stain, Brown Hopps Gram stain, and periodic acid-Schiff stain (20,31).
Histopathologic examination may show microorganisms localized to the burn eschar surface or various degrees of penetration of the eschar. Burn wound infection or sepsis develops when microorganisms invade through the eschar and into viable tissue subjacent to the eschar. These histopathologic manifestations of burn wound infection were carefully described 47 years ago by Teplitz et al. (32) using rats as an experimental model. Their findings in the animal model appear to parallel those observed in humans with thermal injury.
Kim et al. (33) have described a frozen section technique for rapid evaluation of burn wound biopsies for burn wound infection. This technique reduces the time of processing from 4 hours (rapid technique) to 30 minutes. Application of the frozen section technique to burn wound biopsies had not been possible in the past because of the hardness of the eschar. However, advances in this technique, which made frozen sections of bone and cartilage possible, have permitted its application to burn wound biopsies as well. Frozen section diagnosis should always be confirmed by examination of permanent sections (rapid section technique).
DEFINITIONS OF BURN WOUND INFECTION
Although burn wound infection may be diagnosed by histopathologic examination of a full-thickness burn wound biopsy, and the causative agent may be established by culture of the biopsy or by histopathologic examination of the burn wound biopsy using special stains for microorganisms (e.g., periodic acid-Schiff stain for fungi), such studies may not be available in all burn care facilities. Further, when the
burn wound has been excised, there may be no tissue to biopsy. Thus, case definitions are needed for surveillance and outbreak investigation that make use of other, more easily obtained data such as clinical observations, blood cultures, viral cultures, and microscopic examination of lesion scrapings for viral inclusions. Table 25-1 (34) shows case definitions for burn wound infections used by the National Healthcare Safety Network (NHSN).
burn wound has been excised, there may be no tissue to biopsy. Thus, case definitions are needed for surveillance and outbreak investigation that make use of other, more easily obtained data such as clinical observations, blood cultures, viral cultures, and microscopic examination of lesion scrapings for viral inclusions. Table 25-1 (34) shows case definitions for burn wound infections used by the National Healthcare Safety Network (NHSN).
TABLE 25-1 Definitions for Burn Wound Infections | ||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ||||||||||||||||||||||||||||||||||||
One element missing from the NHSN definitions is that of the causative agent. Thus, if one were selecting a definition for burn wound infection in a suspected outbreak of burn wound infections caused by S. aureus, it would be appropriate to include culture of S. aureus from the burn wound in the case definition of infection. The source from which a culture must be taken to establish the cause of a burn wound infection is not described in the NHSN definitions. For bacterial infections, the culture should be taken from a full-thickness burn wound biopsy and not the surface of the burn wound. Another acceptable source is blood if no other possible site of infection can be identified. Although fungi may be cultured from a full-thickness burn wound biopsy, most fungal burn wound infections will probably be diagnosed by histopathologic examination of burn wound biopsies. Herpes simplex may be cultured from scrapings from the burn wound surface; viral inclusions may also be seen microscopically in burn wound scrapings.
The NHSN definitions are based on burn wounds containing unexcised eschar. Since burn wounds are treated in many centers now by early excision and coverage of the wound with autograft, cadaveric allograft, temporary biologic dressings, or dermal replacement (Integra) (35), definitions are needed for infections in surgically created wounds such as excised burns and donor sites. Further, definitions are needed for other types of infections related to the burn wound such as burn wound impetigo and burn wound cellulitis (36). For the purposes of surveillance, the NHSN lumps all wounds related to thermal injury and its treatment together as burn infections. However, when an outbreak occurs involving sites other than burn wound containing unexcised eschar, it will be necessary to use case definitions specific to the type of infection involved in the outbreak (Table 25-2).
TABLE 25-2 Proposed Definitions for Burn Wound Infections (Including Burn Wound Impetigo, Open Burn-Related Surgical Wound Infections, Cellulitis, and Infection of Unexcised Burn Wounds) | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
ETIOLOGIES OF BURN WOUND INFECTIONS
Burn wound infections may be caused by bacteria, fungi, or viruses. Although not invariably the case (37), bacteria probably cause the majority of infections in most burn care centers. Almost all burn wound infections caused by bacteria are due to aerobic microorganisms. Anaerobes cause up to 2% of all burn wound infections (38,39).
Bacteria
S. aureus continues to be one of the most important bacterial causes of burn wound infections (40,41,42). More importantly, methicillin-resistant S. aureus (MRSA) continues to be an important pathogen for burn wound infections (43,44,45). More recently, there is evidence that community-acquired MRSA is beginning to enter some burn care facilities (46) but not others (47). In one center, 25 of 206 (12.1%) of patients colonized or infected with MRSA had USA300 (46). Differences between patients colonized or infected with USA300 community-acquired MRSA and healthcare-associated MRSA were that patients with community-acquired MRSA had frequent abscesses involving their burn wounds. Nasal colonization was present in only 31.6% of patients infected with community-acquired MRSA (46). Occasionally, strains of S. aureus that produce toxic shock syndrome toxin and exfoliative toxin cause burn wound infection (48,49,50,51). Although much less common, β-hemolytic group A streptococci may occasionally cause outbreaks of burn wound infection (52). However, groups A, B, and G streptococci are the third most common cause of burn wound infections in the burn unit at the Karolinska Hospital in Stockholm (53) (see also Chapter 32).
P. aeruginosa continues to be a common cause of burn wound infections. Pseudomonas infections tend to occur more often in patients with burn wounds of >60% total body surface area (TBSA) and after 2 weeks of hospitalization (54).
Acinetobacter baumannii has become the most frequently isolated pathogen in many BICU in civilian and military populations (54, 55, 56, 57, 58,59). A. baumannii is the most common microorganism isolated from war wounds including burn wounds (57). A. baumannii becomes highly resistant to antimicrobial agents and forms biofilms that appear to increase its pathogenicity (58). It is unclear at present whether or not Acinetobacter infections in burn patients are associated with an increase in mortality (54,55,57).
Similar to healthcare-associated infections in other body sites, enterococci have become an important cause of burn wound infection. This is likely due to the widespread use of third-generation cephalosporins to which enterococci are resistant. In 1986, Jones et al. (60) reported on a series of cases of burn wound sepsis caused by enterococci. Enterococcal infection was diagnosed by recovery of at least 105 CFU/g of tissue on burn wound biopsy or by recovery of enterococci from blood cultures. They identified 38 enterococcal burn wound infections in 26 months. Twenty patients developed enterococcal bacteremia, and 10 of these patients died. Enterococci appear to be not only common but also virulent burn wound pathogens. In a more recent study, 97 isolates of vancomycin-resistant enterococci (VRE) and 652 isolates of vancomycin-susceptible enterococci (VSE) were recovered from burn patients (61). No mention was made of infections caused by VSE, but none of the patients colonized by VRE developed VRE infections. As has been noted in other patient populations in the hospital, VRE have been reported to cause an outbreak of VRE colonization and infection in a BICU (62). Four cases of bacteremia due to VRE occurred during that outbreak.
While each of the genera among the Enterobacteriaceae still contribute pathogens as a cause of burn wound infections, only Klebsiella pneumoniae joins S. aureus, A. baumannii, and P. aeruginosa as one of the four most common causes of burn wound infections (54). Those strains of K. pneumoniae that produce extended-spectrum beta-lactamase (ESBL) may result in higher mortality (63). Although the authors could not directly relate ESBL production to mortality, multivariate analysis did indicate that ESBL producing K. pneumoniae may be related to mortality in patients who are older and who are more badly burned.
Fungi
As bacterial burn wound infections have come under better control with use of topical antimicrobial agents, better isolation techniques, and, perhaps, early burn wound excision, the relative importance of fungal burn wound infections has increased. In a recently published autopsy series from the U.S. Army Institute of Surgical Research, the most common causes of fungal burn wound infection were Aspergillus species and Candida species (64). Two publications from the same institution showed that fungal burn wound infection, not burn wound colonization, is significantly associated with mortality (65,66).
Candida Species Invasive candidiasis occurs in 2% to 21% of burn patients (67). Mucosal disruption occurs in burn patients, which leads to fungal translocation. Mucosal atrophy in the gastrointestinal tract is related to the extent of the burn, and ileus complicates burns with >25% TBSA burned (67). Candida albicans is the most common species of Candida recovered from blood (65% of cases) with Candida parapsilosis causing 25% of cases and Candida tropicalis causing 10% of cases (68). The most common source for Candida bloodstream infection is the burn wound and the risk increases with increasing size of the burn wound and with delay in burn wound excision (69). The attributable mortality for candidemia in burn patients has been reported to range from 14% to 70% (67). However, whether or not Candida infections in burn patients are related to mortality remains controversial. In a recently published retrospective matched case-control study using prospectively collected data, no difference in mortality was observed between the patients with candidemia and the control group (68).
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
