Infection in Patients With Burn Injury: Sources, Pathogens, Sensitivity, and Role of Empirical Antibiotics Therapy

Injuries from burns, though preventable, continue to be a major public health challenge, especially in low- and middle-income countries.¹ These injuries commonly are caused by flame, hot liquids (scald), or contact with hot surfaces.² The outcome of these injuries has markedly improved with improvement in resuscitation, early burn wound excision and grafting, and modulation of metabolic response to trauma following burns.³ However, burn injury remains a major cause of morbidity and mortality, especially in low- and middle-income countries.¹

Despite these advances in burn care, infection has become a major challenge in the patient with burn injury. The propensity for infection has been related to the mechanism of burn, the metabolic response associated with burn, the loss of the usual protective skin covering, and the effects of inhalation injury.^4,5 Infection in patients with burn injury may be localized, or it may be systemic, with associated systemic features. Common sources of infection are the burn wound itself; instrumentation, such as intravenous and urethral catheterization; hospitalization in the intensive care unit; and as a sequela of inhalation injury.^4,5 Regardless of the source, infection in the patient with burn injury is a major public health challenge. It is the leading cause of mortality in patients with burn injury, accounting for 51% to 61% of deaths from burn injuries.^6,7

The problem of infection in burn patients is compounded by the challenge of differentiating the presence of sepsis from the already ongoing metabolic response to trauma in burns, which usually results in a hyperbolic state in patients with major burns. The sepsis criteria by the American Burn Association Consensus Conference to Define Sepsis and Infection in Burns is a useful tool for early recognition and prompt treatment.⁶ 

When infection occurs, timely management with source control, local wound care, and culture-directed antibiotic therapy is important. However, empirical antibiotics may sometimes be necessary. Having an understanding of the prevailing microorganisms in the regional burn center is useful in guiding empirical therapy while awaiting culture results. These organisms and their sensitivity pattern vary from region to region, with S aureus predominant in some geographic regions and P aeruginosa predominant in others.^8-10 With concerns of rising antibiotic resistance in burn injury, the need to develop local antibiograms becomes increasingly imperative.

The present study was conducted in a major regional burn center in western Nigeria to determine the magnitude of infection in patients with burn injury, the sources of these infections, the prevalent microorganisms and their sensitivity pattern, and the sensitivity of these microorganisms to empirical therapy.

This prospective observational study was conducted in a regional burn center in western Nigeria between October 2010 and March 2020. The study was conducted in accordance with the ethical standards enshrined in the 1964 Declaration of Helsinki and its subsequent amendments. 

All patients with burn injury admitted into the burn unit were included in the study. The burn unit admission criteria included burn injury greater than or equal to 15% TBSA, minor burn injuries in extremes of age (children younger than 10 years and adults older than 55 years), burn in special areas (perineum, feet, hands, face, and across major joints), as well as burn injuries requiring burn wound excision and grafting. Patients whose injury required outpatient care were excluded from the study. Clinical management of the patients was provided in accordance with the institution’s burn unit protocol.

The burn center is a 12-bed burn facility located in University College Hospital, the premier teaching hospital in Nigeria. This 1229-bed teaching hospital is located in Ibadan, the third largest city in Nigeria.^11,12 It serves as a referral facility for Oyo State and a number of adjoining states in southwestern Nigeria.

Each patient’s age, sex (self-reported), date of admission, and cause of injury were documented, along with the occurrence of infection, the clinical source of infection, and antibiotics administered (both empirical and sensitivity guided). Culture sources (tissue, urine, and blood), organisms cultured, and sensitivity of the organisms cultured were also documented. Wound culture was done through tissue biopsy for microscopy, culture, and sensitivity. In patients with multiple sources of infection, the clinical source of infection was defined as the first detected source of infection. Additionally, the sensitivity of the empirical antibiotics in those who received empirical antibiotics was included.

Infection was defined by the presence of 3 or more of the following parameters: temperature (<36ºC or >39ºC), tachycardia (>110 beats per minute), tachypnea (>25 breaths per minute), thrombocytopenia (<100,000/uL at least 3 days after resuscitation), hyperglycemia, feeding intolerance for more than 24 hours, and burn wound discharge.⁶ Once a clinical diagnosis of infection was made, samples were taken for microscopy, culture, and sensitivity according to the burn unit protocol. The burn center does not routinely test for anaerobes. The diagnosis of respiratory tract infection in the study was made based on the history and physical examination supplemented by chest radiography findings. Empirical antibiotics were commenced in line with the previous sensitivity patterns of the burn unit. 

Data were coded and entered into SPSS version 27 (IBM Corporation) and analyzed. A 95% CI was used for inferential statistical analysis.

A total of 364 patients were included in the study, with 213 males (58.5%) and 151 females (41.5%). The median patient age was 22 years. Flame was the most common cause of burn injury (61.5% [n = 224]) (Figure 1). The median TBSA burn was 22%.

Of the 364 patients, the majority (85.4% [n = 311]) had a clinical diagnosis of infection during the course of admission, with the burn wound (60.8% [n = 189]), respiratory tract (22.2% [n = 69]), and urinary tract (6.1% [n = 19]) constituting the clinical sources of infection. The source of infection was uncertain in 10.9% (n = 34) of the patients. Of the 189 patients with burn wound infection, 91% (n = 172) underwent wound biopsy for microbiological diagnosis and 9% (n = 17) had no tissue biopsy. Blood culture was obtained from 71 patients and urine culture from 46 patients.

Of the 172 patients who underwent wound biopsy, 70.3% (n = 121) were positive for infection and 29.7% (n = 51) showed no microbiological evidence of infection. The most prevalent causes of burn wound infection were K pneumoniae (30.2%), S aureus (24.3%), and P aeruginosa (21.0%) (Table 1). All cultured bacteria with the exception of Citrobacter freundii remained sensitive to amikacin, with K pneumoniae and S aureus being the most sensitive to amikacin; P aeruginosa was more sensitive to ceftazidime (Table 1).

Of the 71 blood cultures, 19 (26.8%) showed microbiological evidence of infection, and 52 (73.2%) were sterile cultures. S aureus was the most prevalent organism, followed by MRSA (Table 2). The majority of the organisms were sensitive to amikacin and meropenem (Table 2). Urine culture was obtained from 46 patients; results were positive for 19 (41%) and sterile for 27 (59%). The most predominant organisms were P aeruginosa (45.0%) and K pneumoniae (25.0%) (Table 3). The predominant urinary tract organisms showed sensitivity to colistin (Table 3).

The most commonly prescribed empirical antibiotics overall were amikacin (20.3% [n = 92/454]) and levofloxacin (19.8% [n = 90/454]) (Figure 2). The cultured organisms in the burn wounds showed high sensitivity and specificity to the empirical antibiotic therapy (Table 4). 

Infection in the patient with burn injury continues to be a major source of morbidity and mortality globally.¹³ A significant factor that can curtail and limit morbidity is prompt diagnosis of the infection and early commencement of the appropriate antibiotics based on culture results and sensitivity. As simple as this seems, it remains a major challenge for clinicians, especially those in low-resource countries, because of the inability of many patients to receive such care as early as is necessary, mostly because of financial reasons. The clinician is thus left with no other option than to start empirical antibiotics while waiting for the relevant microbiological tests to be done. 

Most of the patients in the present study were male and relatively young. In addition, the most common burn etiology was flame, and the median TBSA burn was greater than 20%. These epidemiological findings are in keeping with the epidemiology of burn injury in Nigeria, as documented by Ademola et al² and Abubakar et al.¹⁴

In the current study, the rate of infection was 85.4%. Following burn injury of at least partial-thickness depth, there is a loss of part of the skin with a consequent reduction in the skin protective functions against invading microbes, in addition to formation of burn eschars that become a nidus for wound infection.^15,16 The rich medium for burn wound infection provided by deep partial- or full-thickness injury can be ameliorated by early burn wound excision. Immediate grafting offers an opportunity to reestablish cutaneous integrity as well as a barrier to bacterial invasion.

The presence of inhalation injury predisposes patients with burn injury to respiratory tract infections such as pneumonia.^17,18 The 85.4% rate of infection in the present study, which includes burn wound infection, respiratory tract infection, bloodstream, and urinary tract infection, was higher than the 60% rate reported by Zhou et al,¹⁹ the 68.25% rate reported by Hamzaoui et al,²⁰ and the 27.5% rate reported by Hu et al.²¹ One of the challenges of understanding the true prevalence of infection in burns is the differences in study methodology and definition. For example, the present study incorporated all infections (burn wound, urinary tract, respiratory, bloodstream), whereas the rate reported by Zhou et al¹⁹ and Hamzaoui et al²⁰ was specifically for burn wound infection. The rate reported by Hu et al²¹ was for bloodstream infection. 

The rate of burn wound infection was 60.8% in the present study. A number of risk factors for burn wound infection have been identified, including TBSA burn, burn depth, and burn shock.¹⁹ Burn shock causes ischemia and low oxygen tension in tissues, thereby providing a rich medium for bacterial proliferation. This burn shock in the setting of burn eschar and lost skin barrier predisposes the patient to burn wound infection. With patients often presenting late or under-resuscitated in Nigeria, hypovolemia, poor first aid, and burn depth may all play a synergistic role in the development of burn wound infection. The incidence of burn wound infection in the present study is similar to that reported by Zhou et al¹⁹ and Hamzaoui et al,²⁰ however, despite the difference in geographic regions. The median TBSA burn of 22% in the present study is also similar to that reported in those studies.^19,20

The patient with burn injury is at increased risk for infection. This is due to the loss of the skin defense mechanism; the increased need for frequent instrumentation, such as intravenous and urethral catheterization; and endotracheal intubation.²² These factors have been documented as risk factors for infection in burn, especially if the aseptic barrier is breached.²³ The present study found the burn wound to be the most common source of infection, followed by the respiratory tract and urinary tract. Norbury et al⁶ also identified the respiratory tract (pneumonia), burn wound (cellulitis), and urinary tract as common sources of infection in burns. 

An understanding of the sources, the prevailing organisms, the sensitivity pattern, and the appropriateness of empirical therapy from an antibiogram is crucial in reducing the effect of infection on the patient. Having an understanding of the prevailing organisms and the sensitivity pattern is also important in formulating a local antibiogram and in evaluating an existing antibiogram. The present study shows that the prevailing microorganism and its sensitivity are dependent on the source of infection; when the burn wound is the source of infection, the likely organisms are K pneumoniae, S aureus, and P aeruginosa, in decreasing order of frequency. These organisms all had similar sensitivity to amikacin (an aminoglycoside) as the most sensitive antibiotic. They also demonstrated similar sensitivity profiles to ciprofloxacin, a quinolone. Therefore, if it is necessary to start empirical therapy for burn wound infection before the results of microbiological evaluation are available, the evidence suggests that amikacin and a quinolone should be used. 

During the study period, MRSA was not isolated from any burn wound. The leading organisms identified in the burn wounds in the present study were similar to that reported by Kehinde et al.²⁴ However, Norbury et al⁶ reported S aureus and MRSA as the leading organisms causing burn wound infection and stated that they represent a significant cause of graft loss. The results of antibiotic susceptibility testing of Pseudomonas species isolated from burn wounds in other centers shows less sensitivity to amikacin and greater sensitivity to piperacillin/tazobactam.^6,25

In burn patients with bloodstream infection, the infecting organisms in order of decreasing frequency were S aureus, MRSA, P aeruginosa, and K pneumoniae. With the exception of MRSA, these organisms were susceptible to amikacin, the cephalosporins, and penicillins. MRSA showed susceptibility to vancomycin. Therefore, amikacin and a cephalosporin should be the appropriate combination empirical antibiotic therapy for patients with burn injury and bloodstream infection. Norbury et al⁶ reported a shift from vancomycin to newer antibiotics, such as linezolid. These newer antibiotics are rarely available and are not affordable in Nigeria; therefore, rational antibiotics use in order to prevent further resistance is imperative. Additionally, the use of vancomycin should be monitored.⁶

In the present study, infections originating from the urinary tract were caused by P aeruginosa, K pneumoniae, Staphylococcus saprophyticus, and Escherichia coli, in descending order of frequency. The 2 leading causes of infection showed considerable susceptibility to colistin. These findings indicate that infection originating in the urinary tract was more likely to be caused by multiresistant organisms with a narrow range of antibiotic sensitivity. When choosing empirical antibiotics for urinary tract infection, colistin should be considered.

The present study found that the commonly prescribed empirical antibiotics for burn injury patients with infection were amikacin, levofloxacin, and the cephalosporins. It also showed that the prescribed empirical antibiotics were effective and appropriate against the cultured organisms. Therefore, a local antibiogram incorporating these agents is appropriate for empirical therapy while waiting for the microbiological susceptibility testing. The approach espoused in the present study could be easily adapted in the low-resource setting while observing the varying sensitivity pattern by geographic region of the practice. This result is contrary to the report by Hu et al,²¹ who demonstrated that 41.9% of their study population received inappropriate empirical antibiotics. These different results underscore the need to understand the prevailing microorganisms in the burn unit, the source of the infection, and the sensitivity profile of the organism.

The current study has limitations. It did not examine the relationship between the source of infection and the treatment outcome. A study examining this relationship might be useful to further elucidate the clinical implications of the role of infection source, pathogens, and microbial sensitivity in the management of burn injuries. The efficiency of the antibiotics administered was not tested. However, the patients demonstrated positive clinical response to the empirical antibiotics. The number of days postburn at which infection occurred and cultures were taken were not included in the study design. Future research incorporating this information may enhance knowledge on infection in burn and outcomes. Lastly, this study did not evaluate the role of burn depth on infection in burns. It may be beneficial to design a future study to objectively evaluate the effect of different burn depths on bacteria causing infection in burn.

The rate of infection in patients with burn injury is high. In descending order, the major sources of infection are the burn wound, the respiratory tract, and the urinary tract. The source of infection determines the most likely organism and its sensitivity profile. Prompt microbiologic studies should be done to identify the source, type of organism, and antibiotic sensitivity to commence appropriate therapy. However, where this is not feasible, a current antibiogram should be available to provide the necessary guidance for effective therapy.