Isolated Pediatric Burn Injury in Iraq and Afghanistan* Matthew A. Borgman, MD1; Renée I. Matos, MD, MPH1; Philip C. Spinella, MD2
Objectives: To characterize the epidemiology of burn injury in pediatric patients and identify factors associated with mortality based on burn severity. Design: Retrospective cohort study. Setting: U.S. military combat support hospitals and forward surgical hospitals in Iraq and Afghanistan. Patients: Iraqi and Afghan children less than 18 years old admitted with isolated burn injury. Interventions: None. Measurements and Main Results: Burn severity was classified as mild, moderate, and severe based on external Abbreviated Injury Scale score. Patient characteristics and outcomes were described according to burn severity. A multivariate logistic regression was performed on univariate associations with mortality. Of 4,743 pediatric patients, 549 (11.6%) had isolated burn injury. Overall mortality was 13%, median external Abbreviated Injury Scale was 3 (interquartile range, 2–4), and 67% were male. Variables included in the logistic regression were external Abbreviated Injury Scale score, abnormal heart rate for age, hypotension, mechanical ventilation, transfusion, Glasgow Coma Scale, international normalized ratio, base deficit, hematocrit, and platelet count. Factors independently associated with mortality were international normalized ratio (odds ratio, 2.6; 95% CI, 1.2–5.8; p = 0.021) and external Abbreviated Injury Scale (odds ratio, 2.5; 95% CI, 1.3–4.7; p = 0.004). Mortality increased with burn severity: mild 1.7%, moderate 7.2%, and severe 47% (p < 0.001). Conclusions: This is the first in-depth study of pediatric burn injuries in combat. Children with severe burns (total body surface *See also p. 197. 1 Department of Pediatrics, San Antonio Military Medical Center, Fort Sam Houston, TX. 2 Division of Critical Care, Department of Pediatrics, Washington University in St. Louis, St. Louis, MO. The opinions or assertions contained herein are the private views of the author and are not to be construed as official or as reflecting the views of the Department of the Army or the Department of Defense. Dr. Borgman conceptualized and designed the study, carried out analysis, drafted the initial article, and approved the final article as submitted. Drs. Matos and Spinella helped conceptualize the study, reviewed and revised the article, and approved the final article as submitted. Drs. Borgman and Matos disclosed government work. Dr. Spinella disclosed that he does not have any potential conflicts of interest. For information regarding this article, E-mail: [email protected] Copyright © 2015 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies DOI: 10.1097/PCC.0000000000000307
Pediatric Critical Care Medicine
area > 39% or > 29% if < 5 yr) had a high mortality and required significant resources in a setting that is not primarily resourced for long-term care of severe pediatric burn injury. Extraordinary measures are therefore used for the long-term care of these burned children within the war zones of Iraq and Afghanistan. (Pediatr Crit Care Med 2015; 16:e23–e27) Key Words: burns; combat; pediatric
B
urns account for over 300,000 deaths annually worldwide, a third of whom are under the age of 20 (1). In lowincome countries, children under 5 years old account for a disproportionate number of burns, ranging from one third to half of all cases and are typically caused by open flame or scalding injuries (2). In Iraqi civilian hospitals during the recent conflict, most burns occurred in younger females, and the most common cause was related to open flames (3). Similarly, in Afghani civilian hospitals, females have also been found to be more likely than males to die as a result of burn injuries, which were most often related to their domestic work, cooking around open flame (4). As a result of the recent conflicts in Iraq and Afghanistan, the U.S. military has had medical personnel in these countries for over a decade. By doctrine, the U.S. military treats all individuals, including local civilians, with battle injuries that threaten life, limb, or eyesight. They additionally will treat less severe and nonbattle injuries and illnesses as resources and mission requirements allow. Treatment of pediatric burns in a war environment represents a unique and challenging injury for deployed military health providers. Minimal pediatric burn training, limited resources, and the austere environment all contribute to these challenges. In addition to treating noncombat-related accidental burns such as those documented in civilian articles, children may also fall victim to burns and explosions related to ordnance and the chaotic military environment, thus increasing the prevalence of pediatric burn injuries. Previous studies have shown that younger children and burn injury are independently associated with mortality (5, 6). Our objective was to evaluate the epidemiology of wartime pediatric isolated burn injury and factors associated with mortality and severe burns.
PATIENTS AND METHODS A retrospective review of the Department of Defense Trauma Registry (DoDTR) was performed on patients admitted to www.pccmjournal.org
e23
Borgman et al
combat support hospitals and forward surgical teams in Iraq and Afghanistan from 2003 through 2011. The DoDTR is a prospective trauma registry of all patients admitted to a military treatment facility with diagnoses resulting from trauma within 72 hours. It is maintained by the Joint Theater Trauma System of the United States Army Institute of Surgical Research at Fort Sam Houston, Texas. Inclusion criteria were local Iraqi and Afghan pediatric patients less than 18 years old admitted with an isolated burn injury to combat support hospitals (role III) or forward surgical teams (role IIb). Patients were excluded if they had an external Abbreviated Injury Scale (AIS) 2005 (7) of 6 (> 89% total body surface area [TBSA] burn) as these patients are treated expectantly given the fatal prognosis in this environment. Nonisolated burn patients were also excluded, as defined by an AIS greater than 2 in any other AIS body regions. The DoDTR was queried for age, gender, injury severity score (ISS), AIS, Glasgow Coma Scale (GCS), heart rate, blood pressure, transfusion requirements, ventilator use, hematocrit, international normalized ratio (INR), base deficit, platelet count, hospital days, ICU days, and ventilator days. Patients were not excluded for missing physiologic data. Hospital-free days, ICU-free days, and ventilator-free days were calculated based on a 30-day period. Abnormal heart rate and hypotension were defined as values outside the 95% CI for age (8, 9). Burn severity was classified by the external AIS scale and categorized into three groups: mild (AIS 1–3: < 30% TBSA, or < 20% TBSA if age < 5 yr), moderate (AIS 4: 30–39% TBSA, or 20–29% if age < 5 yr), and severe (AIS 5: 39–89% TBSA, or 29–89% if age < 5 yr) (7). Patient characteristics and outcomes were evaluated based on burn severity and mortality. The chi-square test was used for categorical variables and the Mann-Whitney U test for continuous, nonparametric data. Independent associations were determined by backward stepwise regression using all variables associated with mortality with p less than 0.2. Variables were examined for colinearity based on a variance inflations factor of at least 3 or a Pearson correlation R2 greater than 0.6. Statistical analyses were performed using SPSS version 19.0 (IBM, Armonk, New York). This study was conducted under a protocol reviewed and approved by the Brooke Army Medical Center Institutional Review Board and in accordance with the approved protocol.
RESULTS Of the 4,743 pediatric patients identified in the DoDTR, 549 (11.6%) had an isolated burn injury (Fig. 1). Four patients were excluded for a TBSA greater than 89% (all fatal), and 55 patients were excluded for polytrauma (AIS > 2 in other body regions). Average age was 5.9 years (median, 4; interquartile range [IQR], 2–9). Based on external AIS scoring, the population median TBSA was approximately 15% (IQR, 15–35) and mean 25% (sd, 20). Inpatient overall mortality was 13.3% (73 of 549) and was 12% in males and 16% in females (p = 0.2). Eleven percent (62 of 549) of the cohort patients had burns related to combat, 84% (461 of 549) were noncombat, and the remaining 26 were unknown. Approximately half of the children (52%) received care in Iraq and the remainder in Afghanistan. The mean e24
www.pccmjournal.org
Figure 1. Flow diagram of included patients. TBSA = total body surface area.
hospital length of stay was 10 days (range, 0–171 d; median, 4 [IQR, 1–12]). Ten patients had a length of stay that exceeded 2 months. Table 1 compares demographic characteristics and length of stay of survivors versus nonsurvivors. Variables associated with mortality that were included in the regression were external AIS, abnormal heart rate, hypotension, ventilation, transfusion, GCS, INR, base deficit, hematocrit, and platelet count. ISS was not considered due to colinearity with external AIS. After logistic regression, two variables were independently associated with mortality: external AIS (odds ratio [OR], 2.5; 95% CI, 1.3–4.7; p < 0.05) and INR (OR, 2.6; 95% CI, 1.2–5.8; p < 0.05) (Table 2). Table 3 compares pediatric patients according to burn severity. When compared with mild and moderate burn patients, severely burned patients presented with more hypotension, had a higher base deficit, and had a higher INR (p < 0.05). Severely burned patients also had significantly worse outcomes, as evidenced by a higher mortality (47.3%) and much fewer hospital-, ICU-, and ventilator-free days (p < 0.05). Moderate burns had the longest length of stay with a median of 10 days (3–21 d).
DISCUSSION This is the largest and most comprehensive study to date that evaluates pediatric burn injuries independently in a combat environment. Previous cohort studies examining military hospitalized pediatric patients in Iraq and Afghanistan have found that patients were less likely to survive to hospital discharge if they were younger, female, or sustained burn injury (5, 10). However, these studies did not independently investigate burn injuries. In a civilian study investigating Iraqi burn victims, 38% of burn deaths occurred in patients less than 21 years, of which 81% were females (3). These authors found that 90% of the burns were caused by flame (predominantly related to kerosene heaters), with some scald injuries occurring in the very young. In a similar study of Afghan civilian burn injuries, 59% were females, and 21% were children less than 15 years old (4). Unlike the Iraqi study, only 46% were due to flame and 36% were due to gas explosion. The female predominance of burn injury noted in these studies seems to be related to kerosene and petrol use in domestic appliances. Additionally, sarees February 2015 • Volume 16 • Number 2
Online Clinical Investigations
Table 1.
Comparison of Survivors and Nonsurvivors in Pediatric Isolated Burn Patients n
Variable
Overall
Survivors (n = 476)
Nonsurvivors (n = 73)
p
Age
548
4 [2–9]
4 [2–9]
Male
549
367 (67)
323 (68)
44 (60)
0.23
Injury Severity Score
549
9 [4–16]
9 [4–16]
25 [25–25]
< 0.0001
Burn Abbreviated Injury Scale
549
3 [2–4]
3 [1–4]
5 [5–5]
< 0.0001
Glasgow Coma Scale
429
15 [15–15]
15 [15–15]
11 [3–15]
< 0.0001
Abnormal heart rate
508
267 (53)
223 (51)
44 (65)
0.037
Hypotension
445
21 (5)
12 (3)
9 (18)
< 0.001
Transfused
549
48 (9)
38 (8)
10 (14)
0.12
Ventilated
549
105 (19)
83 (17)
22 (30)
0.016
Hematocrit (mg/dL)
342
38 [34–43]
38 [34–42]
40 [32–47]
< 0.001
International normalized ratio
223
1.1 [0.9–1.4]
1.1 [0.9–1.3]
1.4 [1.2–1.9]
< 0.001
Base deficit
267
10 [6–16]
< 0.001
Platelet count (×100,000)
323
399 [291–518]
402 [300–518]
Hospital days
549
4 [1–12]
4 [2–15]
2 [1–8]
0.004
Ventilator days
549
0 [0–1]
0 [0–1]
2 [1–5]
0.018
ICU days
549
1 [0–5]
1 [0–4.5]
2 [1–6]
< 0.001
4 [1–8]
4 [1–7]
5 [2–9]
0.37
322 [237–510]
0.13
All reported as either median [interquartile range] or “n” (% of group).
Regression Results of Factors Associated With Nonsurvival Table 2.
p
β
OR (95% CI)
External Abbreviated Injury Scale
0.92
2.5 (1.3–4.7)
0.004
International normalized ratio
0.95
2.6 (1.2–5.8)
0.021
Abnormal heart rate
1.1
3.1 (0.78–13.0)
0.115
Variable
Constant
–7.9
< 0.001
OR = odds ratio.
or hijabs, customary scarves of females, have been known to catch fire when dangling near open flame (1, 2, 4). Gas or fuel explosions are often the result of refueling kerosene lantern or stoves, a job typically carried out by females (2, 4). The distinction between these civilian studies and the patients who presented to our military hospitals is that our cohort analysis did not reveal any significant gender differences. It is unclear whether the combat environment may have played a role in this difference as some younger males, compared with females, could be more often involved as combatants or in close proximity to hostile events. A large, single burn center study in the United States of 952 severely (> 30% TBSA) burned pediatric patients reported an overall mortality of 13% (11). The mortality rate in the U.S. study is notably lower than the comparable moderate and Pediatric Critical Care Medicine
severe burn groups (based on TBSA) in our study with a combined mortality rate of 33.5% (67 of 200). There are several important factors that may have contributed to this difference. The U.S. children were slightly older (7.3 yr vs 5.9 yr), were from a developed country and therefore may have had better baseline health and nourishment, had no combat-related burns, and were all treated at a tertiary, specialized pediatric burn center as opposed to a field, combat support hospital. Although we are unable to perform a direct comparison of these distinct populations, there is also the potential that lack of experience or resources in an austere combat environment contributed to this striking mortality difference. We were unable to account for differences in provider experience and training as we had no documentation on whether the surgeon was a general surgeon or whether they had any experience with burn injuries. In recent years, burn care is no longer a requirement for general surgery residents, and many military physicians are early in their careers. Although some physicians might receive education about burn care prior to deployment, this is not universal, and it is possible that more robust burn training predeployment could translate to better outcomes. Our study also illustrates the significant resources, primarily in hospital and ICU bed-days, necessary to take care of these burned children. Because of the high mortality in the severely burned group, the moderately burned group used the most resources based on hospital length of stay. Schmidt et al (12) described the care of eight Iraqi children who were cared for in military combat support hospitals that were subsequently transported to Shriners Institute for Burned www.pccmjournal.org
e25
Borgman et al
Table 3.
Comparison of Pediatric Patients by Burn Severity n
Mild
Moderate
Severe
n (%)
549
349 (63.6)
69 (12.6)
131 (23.8)
Age (yr)
548
4 [2–9]
4 [2–9]
5 [2–10]
Male, %
548
68
65
Glasgow Coma Scale
429
15 [15–15]
15 [15–15]
Abnormal heart rate, %
508
46
Hypotension, %
283
Transfused, %
Variable
66 15 [6–15]a,b
60a
67a
2.8
1.8
11a,b
549
3
7.5
10a
Ventilated, %
549
16
16
29a,b
Base deficit
267
3 [0–6]
5 [3–9]a
8 [5–12]a,b
International normalized ratio
223
1.0 [0.9–1.2]
1.1 [1.0–1.4]
1.4 [1.2–1.7]a,b
Hematocrit (mg/dL)
342
38 [34–41]
40 [32–43]
41 [33–48]a
Platelet count (×100,000)
323
408 [311–513]
Mortality, %
549
1.7
Hospital days
549
3 [1–9]
10 [3–21]a
5 [1–19]a
ICU days
549
1 [0–3]
3 [1–9]a
4 [1–13]a
Ventilator days
549
0 [0–0]
0 [0–2]a
1 [1–8]a,b
Hospital-free days
549
27 [21–29]
19 [4–27]a
0 [0–17]a,b
ICU-free days
549
29 [27–30]
27 [18–29]a
0 [0–22]a,b
Ventilator-free days
549
30 [30–30]
30 [28–30]a
6 [0–29]a,b
357 [261–460]
389 [248–553]a
7.2a
47.3a,b
p < 0.05 compared to mild burn. p < 0.05 compared to moderate burn. Values reported as either median [interquartile range] or percent. a
b
Children in Boston, MA. They describe the immense multidisciplinary resources required to treat severely burned children, some requiring up to 23 visits to the operating room. These resources were obviously only available to a small minority. In our study, most children received their definitive care from our deployed military facilities. The fact that there were 10 children who spent at least 2 months admitted to military treatment facilities for burn-related injuries is a testament to the extensive resources dedicated to caring for these children. There have been several reviews of burned soldiers and adult civilians in the recent Iraq and Afghanistan conflicts, which have resulted in guidelines for their management, including resuscitation, wound care, and antibiotic use (13–16). In an epidemiologic review, British military medical forces report that burns account for 6% of injuries, 50% of which are combat related (17). Seventy-five percent of the combat burns are due to blast injury, and 86% of noncombat burns are due to flame. Another British military analysis indicated that 60% of burns in their soldiers were accidental in nature (18). Kauvar et al (19) reported that 63% of military burned casualties were due to combat. In a separate review including civilians, they report that 52% are due to hostile action (20). Our analysis indicates that the majority (84%) of care the military provided to burned children was for noncombat-related burns, a proportion higher than that e26
www.pccmjournal.org
was noted in the adult studies. This is a further testament to the important humanitarian care provided by military medicine, particularly when the prevalence of civilian burns may increase due to the chaotic war environment. We have also shown that the degree of burn, as measured by external AIS, and INR were independently associated with mortality. The study by Kauvar et al (19) compared civilian and military burn mortality and found that TBSA, age greater than 40 years, inhalational injury, and ventilator days were independently associated with mortality, although this study lacked physiologic and laboratory variables (19). Although our analysis did not include inhalational injury, it did include use of invasive ventilation. However, it is not clear whether patients were intubated due to respiratory failure from acute respiratory distress syndrome, pneumonia, inhalational injury, or simply for procedures. In addition, the use of physiologic variables in our study is a more comprehensive way to adjust for injury severity and evaluate predictors of mortality. Although admission INR is associated with mortality, we did not account for the availability and the changes in practice patterns in the use of fresh frozen plasma, so it is unclear if this played a role in affecting the data. Limitations in our study include those inherent to retrospective reviews, particularly with missing or incorrect data. However, the DoDTR staff who input data undergo extensive February 2015 • Volume 16 • Number 2
Online Clinical Investigations
training, interrater reliability audits, and other quality assurance measures to ensure high data quality. We also chose to examine the severity of burns based on external AIS, as opposed to percent of TBSA, as this was how the data were available to us in the registry. However, by using the AIS external score as a surrogate for TBSA, we were still able to categorize burn severity as mild, moderate, or severe. One study has compared external AIS with TBSA for burns, finding a near-perfect coding agreement between AIS coding and % TBSA calculated by expert burn clinician (weighted κ, 0.83), which is sufficient for registry and research purposes (21). Using external AIS within the multivariate logistic regression has not been done before, and this would need to be validated in other studies before being adopted. Unfortunately, we were unable to characterize the cause of burn because of lack of documentation with most etiologies listed as “unknown” or simply “burn.” It has also been reported that there is a significant number of suicide attempts by teenage females in both countries (3, 4) although we do not have this categorized in the database. We also do not have any data on prehospital transport times or time of injury prior to admission. Although our study had predominantly noncombat-related burn injuries, it would have been interesting to compare the civilian studies with our military study to evaluate the causes of the burns (i.e., scald vs flame vs blast). In the future, improved documentation of this field would improve our understanding of the epidemiology of pediatric burns in an austere environment. Unfortunately, we have no data on patients after discharge, and burn patients frequently require multiple follow-up surgeries, reconstruction, and rehabilitation. Although this was provided on a limited basis for some children prior to ultimate discharge, follow-up medical missions would be necessary to provide this care and identify some long-term outcomes. In the future, military medical leaders could use these data to plan for resource utilization and consider training, potentially at specialized pediatric burn centers, to prepare medical staff to best care for this spectrum of injuries. The higher mortality in these patients, relative to reported mortality in civilian pediatric burn centers, possibly reflects a significant potential to save lives with additional training and resources. It is unknown whether additional training alone would improve outcomes, and the mechanisms of burns (both from combat and noncombat injuries) likely represent a worse injury than those occurring in the developed world in healthier children. However, given the challenge of caring for these complex cases and the numbers at which they occur, it seems prudent to prepare our physicians as best as possible with training opportunities already available.
CONCLUSIONS In this first study of pediatric burn care by U.S. military treatment facilities in Iraq and Afghanistan, we examined factors associated with mortality and we found evidence that significant resources are used in caring for children, sometimes for months and most commonly for noncombat-related injuries. Children with severe burns (TBSA > 39% or > 29% if < 5 yr) had a high mortality and required significant resources in a setting that is not primarily resourced for long-term care of Pediatric Critical Care Medicine
severe pediatric burn injury. Extraordinary measures are therefore used for the long-term care of these burned children within the war zones of Iraq and Afghanistan.
REFERENCES
1. Peden M, Oyegbite K, Ozanne-Smith J, et al: World Report on Child Injury Prevention: Burns. Geneva, Switzerland, WHO, 2008 2. Forjuoh SN: Burns in low- and middle-income countries: A review of available literature on descriptive epidemiology, risk factors, treatment, and prevention. Burns 2006; 32:529–537 3. Qader AR: Burn mortality in Iraq. Burns 2012; 38:772–775 4. Padovese V, De Martino R, Eshan MA, et al: Epidemiology and outcome of burns in Esteqlal Hospital of Kabul, Afghanistan. Burns 2010; 36:1101–1106 5. Borgman M, Matos RI, Blackbourne LH, et al: Ten years of military pediatric care in Afghanistan and Iraq. J Trauma Acute Care Surg 2012; 73:S509–S513 6. Matos RI, Holcomb JB, Callahan C, et al: Increased mortality rates of young children with traumatic injuries at a US army combat support hospital in Baghdad, Iraq, 2004. Pediatrics 2008; 122:e959–e966 7. Gennarelli TA, Wodzin E: AIS 2005: A contemporary injury scale. Injury 2006; 37:1083–1091 8. ECC Committee, Subcommittees and Task Forces of the American Heart Association: 2005 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Circulation 2005; 112(24 Suppl):IV1–IV203 9. Fleming S, Thompson M, Stevens R, et al: Normal ranges of heart rate and respiratory rate in children from birth to 18 years of age: A systematic review of observational studies. Lancet 2011; 377:1011–1018 10. Creamer KM, Edwards MJ, Shields CH, et al: Pediatric wartime admissions to US military combat support hospitals in Afghanistan and Iraq: Learning from the first 2,000 admissions. J Trauma 2009; 67:762–768 11. Kraft R, Herndon DN, Al-Mousawi AM, et al: Burn size and survival probability in paediatric patients in modern burn care: A prospective observational cohort study. Lancet 2012; 379:1013–1021 12. Schmidt PM, Sheridan RL, Moore CL, et al: From Baghdad to Boston: International transfer of burned children in time of war. J Burn Care Res 2014; 35:369–373 13. Cancio LC, Lundy JB, Sheridan RL: Evolving changes in the management of burns and environmental injuries. Surg Clin North Am 2012; 92:959–986, ix 14. Chung KK, Salinas J, Renz EM, et al: Simple derivation of the initial fluid rate for the resuscitation of severely burned adult combat casualties: In silico validation of the rule of 10. J Trauma 2010; 69(Suppl 1):S49–S54 15. Hospenthal DR, Murray CK, Andersen RC, et al; Infectious Diseases Society of America; Surgical Infection Society: Guidelines for the prevention of infections associated with combat-related injuries: 2011 update: Endorsed by the Infectious Diseases Society of America and the Surgical Infection Society. J Trauma 2011; 71(2, Suppl 2):S210–S234 16. USAISR: Burn care. Joint Theater Trauma System Clinical Practice Guideline 2012. Available at: http://www.usaisr.amedd.army.mil/assets/ cpgs/burn_care_13_nov_13.pdf. Accessed November 22, 2013 17. Jeevaratnam JA, Pandya AN: One year of burns at a role 3 Medical Treatment Facility in Afghanistan. J R Army Med Corps 2014; 160:22–26 18. Foster MA, Moledina J, Jeffery SL: Epidemiology of U.K. military burns. J Burn Care Res 2011; 32:415–420 19. Kauvar DS, Cancio LC, Wolf SE, et al: Comparison of combat and non-combat burns from ongoing U.S. military operations. J Surg Res 2006; 132:195–200 20. Kauvar DS, Wolf SE, Wade CE, et al: Burns sustained in combat explosions in Operations Iraqi and Enduring Freedom (OIF/OEF explosion burns). Burns 2006; 32:853–857 21. Watterson D, Cleland H, Picton N, et al: Level of agreement between coding sources of percentage total body surface area burnt (%TBSA). HIM J 2011; 40:21–24 www.pccmjournal.org
e27
Objectives: To characterize the epidemiology of burn injury in pediatric patients and identify factors associated with mortality based on burn severity. Design: Retrospective cohort study. Setting: U.S. military combat support hospitals and forward surgical hospitals in Iraq and Afghanistan. Patients: Iraqi and Afghan children less than 18 years old admitted with isolated burn injury. Interventions: None. Measurements and Main Results: Burn severity was classified as mild, moderate, and severe based on external Abbreviated Injury Scale score. Patient characteristics and outcomes were described according to burn severity. A multivariate logistic regression was performed on univariate associations with mortality. Of 4,743 pediatric patients, 549 (11.6%) had isolated burn injury. Overall mortality was 13%, median external Abbreviated Injury Scale was 3 (interquartile range, 2–4), and 67% were male. Variables included in the logistic regression were external Abbreviated Injury Scale score, abnormal heart rate for age, hypotension, mechanical ventilation, transfusion, Glasgow Coma Scale, international normalized ratio, base deficit, hematocrit, and platelet count. Factors independently associated with mortality were international normalized ratio (odds ratio, 2.6; 95% CI, 1.2–5.8; p = 0.021) and external Abbreviated Injury Scale (odds ratio, 2.5; 95% CI, 1.3–4.7; p = 0.004). Mortality increased with burn severity: mild 1.7%, moderate 7.2%, and severe 47% (p < 0.001). Conclusions: This is the first in-depth study of pediatric burn injuries in combat. Children with severe burns (total body surface *See also p. 197. 1 Department of Pediatrics, San Antonio Military Medical Center, Fort Sam Houston, TX. 2 Division of Critical Care, Department of Pediatrics, Washington University in St. Louis, St. Louis, MO. The opinions or assertions contained herein are the private views of the author and are not to be construed as official or as reflecting the views of the Department of the Army or the Department of Defense. Dr. Borgman conceptualized and designed the study, carried out analysis, drafted the initial article, and approved the final article as submitted. Drs. Matos and Spinella helped conceptualize the study, reviewed and revised the article, and approved the final article as submitted. Drs. Borgman and Matos disclosed government work. Dr. Spinella disclosed that he does not have any potential conflicts of interest. For information regarding this article, E-mail: [email protected] Copyright © 2015 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies DOI: 10.1097/PCC.0000000000000307
Pediatric Critical Care Medicine
area > 39% or > 29% if < 5 yr) had a high mortality and required significant resources in a setting that is not primarily resourced for long-term care of severe pediatric burn injury. Extraordinary measures are therefore used for the long-term care of these burned children within the war zones of Iraq and Afghanistan. (Pediatr Crit Care Med 2015; 16:e23–e27) Key Words: burns; combat; pediatric
B
urns account for over 300,000 deaths annually worldwide, a third of whom are under the age of 20 (1). In lowincome countries, children under 5 years old account for a disproportionate number of burns, ranging from one third to half of all cases and are typically caused by open flame or scalding injuries (2). In Iraqi civilian hospitals during the recent conflict, most burns occurred in younger females, and the most common cause was related to open flames (3). Similarly, in Afghani civilian hospitals, females have also been found to be more likely than males to die as a result of burn injuries, which were most often related to their domestic work, cooking around open flame (4). As a result of the recent conflicts in Iraq and Afghanistan, the U.S. military has had medical personnel in these countries for over a decade. By doctrine, the U.S. military treats all individuals, including local civilians, with battle injuries that threaten life, limb, or eyesight. They additionally will treat less severe and nonbattle injuries and illnesses as resources and mission requirements allow. Treatment of pediatric burns in a war environment represents a unique and challenging injury for deployed military health providers. Minimal pediatric burn training, limited resources, and the austere environment all contribute to these challenges. In addition to treating noncombat-related accidental burns such as those documented in civilian articles, children may also fall victim to burns and explosions related to ordnance and the chaotic military environment, thus increasing the prevalence of pediatric burn injuries. Previous studies have shown that younger children and burn injury are independently associated with mortality (5, 6). Our objective was to evaluate the epidemiology of wartime pediatric isolated burn injury and factors associated with mortality and severe burns.
PATIENTS AND METHODS A retrospective review of the Department of Defense Trauma Registry (DoDTR) was performed on patients admitted to www.pccmjournal.org
e23
Borgman et al
combat support hospitals and forward surgical teams in Iraq and Afghanistan from 2003 through 2011. The DoDTR is a prospective trauma registry of all patients admitted to a military treatment facility with diagnoses resulting from trauma within 72 hours. It is maintained by the Joint Theater Trauma System of the United States Army Institute of Surgical Research at Fort Sam Houston, Texas. Inclusion criteria were local Iraqi and Afghan pediatric patients less than 18 years old admitted with an isolated burn injury to combat support hospitals (role III) or forward surgical teams (role IIb). Patients were excluded if they had an external Abbreviated Injury Scale (AIS) 2005 (7) of 6 (> 89% total body surface area [TBSA] burn) as these patients are treated expectantly given the fatal prognosis in this environment. Nonisolated burn patients were also excluded, as defined by an AIS greater than 2 in any other AIS body regions. The DoDTR was queried for age, gender, injury severity score (ISS), AIS, Glasgow Coma Scale (GCS), heart rate, blood pressure, transfusion requirements, ventilator use, hematocrit, international normalized ratio (INR), base deficit, platelet count, hospital days, ICU days, and ventilator days. Patients were not excluded for missing physiologic data. Hospital-free days, ICU-free days, and ventilator-free days were calculated based on a 30-day period. Abnormal heart rate and hypotension were defined as values outside the 95% CI for age (8, 9). Burn severity was classified by the external AIS scale and categorized into three groups: mild (AIS 1–3: < 30% TBSA, or < 20% TBSA if age < 5 yr), moderate (AIS 4: 30–39% TBSA, or 20–29% if age < 5 yr), and severe (AIS 5: 39–89% TBSA, or 29–89% if age < 5 yr) (7). Patient characteristics and outcomes were evaluated based on burn severity and mortality. The chi-square test was used for categorical variables and the Mann-Whitney U test for continuous, nonparametric data. Independent associations were determined by backward stepwise regression using all variables associated with mortality with p less than 0.2. Variables were examined for colinearity based on a variance inflations factor of at least 3 or a Pearson correlation R2 greater than 0.6. Statistical analyses were performed using SPSS version 19.0 (IBM, Armonk, New York). This study was conducted under a protocol reviewed and approved by the Brooke Army Medical Center Institutional Review Board and in accordance with the approved protocol.
RESULTS Of the 4,743 pediatric patients identified in the DoDTR, 549 (11.6%) had an isolated burn injury (Fig. 1). Four patients were excluded for a TBSA greater than 89% (all fatal), and 55 patients were excluded for polytrauma (AIS > 2 in other body regions). Average age was 5.9 years (median, 4; interquartile range [IQR], 2–9). Based on external AIS scoring, the population median TBSA was approximately 15% (IQR, 15–35) and mean 25% (sd, 20). Inpatient overall mortality was 13.3% (73 of 549) and was 12% in males and 16% in females (p = 0.2). Eleven percent (62 of 549) of the cohort patients had burns related to combat, 84% (461 of 549) were noncombat, and the remaining 26 were unknown. Approximately half of the children (52%) received care in Iraq and the remainder in Afghanistan. The mean e24
www.pccmjournal.org
Figure 1. Flow diagram of included patients. TBSA = total body surface area.
hospital length of stay was 10 days (range, 0–171 d; median, 4 [IQR, 1–12]). Ten patients had a length of stay that exceeded 2 months. Table 1 compares demographic characteristics and length of stay of survivors versus nonsurvivors. Variables associated with mortality that were included in the regression were external AIS, abnormal heart rate, hypotension, ventilation, transfusion, GCS, INR, base deficit, hematocrit, and platelet count. ISS was not considered due to colinearity with external AIS. After logistic regression, two variables were independently associated with mortality: external AIS (odds ratio [OR], 2.5; 95% CI, 1.3–4.7; p < 0.05) and INR (OR, 2.6; 95% CI, 1.2–5.8; p < 0.05) (Table 2). Table 3 compares pediatric patients according to burn severity. When compared with mild and moderate burn patients, severely burned patients presented with more hypotension, had a higher base deficit, and had a higher INR (p < 0.05). Severely burned patients also had significantly worse outcomes, as evidenced by a higher mortality (47.3%) and much fewer hospital-, ICU-, and ventilator-free days (p < 0.05). Moderate burns had the longest length of stay with a median of 10 days (3–21 d).
DISCUSSION This is the largest and most comprehensive study to date that evaluates pediatric burn injuries independently in a combat environment. Previous cohort studies examining military hospitalized pediatric patients in Iraq and Afghanistan have found that patients were less likely to survive to hospital discharge if they were younger, female, or sustained burn injury (5, 10). However, these studies did not independently investigate burn injuries. In a civilian study investigating Iraqi burn victims, 38% of burn deaths occurred in patients less than 21 years, of which 81% were females (3). These authors found that 90% of the burns were caused by flame (predominantly related to kerosene heaters), with some scald injuries occurring in the very young. In a similar study of Afghan civilian burn injuries, 59% were females, and 21% were children less than 15 years old (4). Unlike the Iraqi study, only 46% were due to flame and 36% were due to gas explosion. The female predominance of burn injury noted in these studies seems to be related to kerosene and petrol use in domestic appliances. Additionally, sarees February 2015 • Volume 16 • Number 2
Online Clinical Investigations
Table 1.
Comparison of Survivors and Nonsurvivors in Pediatric Isolated Burn Patients n
Variable
Overall
Survivors (n = 476)
Nonsurvivors (n = 73)
p
Age
548
4 [2–9]
4 [2–9]
Male
549
367 (67)
323 (68)
44 (60)
0.23
Injury Severity Score
549
9 [4–16]
9 [4–16]
25 [25–25]
< 0.0001
Burn Abbreviated Injury Scale
549
3 [2–4]
3 [1–4]
5 [5–5]
< 0.0001
Glasgow Coma Scale
429
15 [15–15]
15 [15–15]
11 [3–15]
< 0.0001
Abnormal heart rate
508
267 (53)
223 (51)
44 (65)
0.037
Hypotension
445
21 (5)
12 (3)
9 (18)
< 0.001
Transfused
549
48 (9)
38 (8)
10 (14)
0.12
Ventilated
549
105 (19)
83 (17)
22 (30)
0.016
Hematocrit (mg/dL)
342
38 [34–43]
38 [34–42]
40 [32–47]
< 0.001
International normalized ratio
223
1.1 [0.9–1.4]
1.1 [0.9–1.3]
1.4 [1.2–1.9]
< 0.001
Base deficit
267
10 [6–16]
< 0.001
Platelet count (×100,000)
323
399 [291–518]
402 [300–518]
Hospital days
549
4 [1–12]
4 [2–15]
2 [1–8]
0.004
Ventilator days
549
0 [0–1]
0 [0–1]
2 [1–5]
0.018
ICU days
549
1 [0–5]
1 [0–4.5]
2 [1–6]
< 0.001
4 [1–8]
4 [1–7]
5 [2–9]
0.37
322 [237–510]
0.13
All reported as either median [interquartile range] or “n” (% of group).
Regression Results of Factors Associated With Nonsurvival Table 2.
p
β
OR (95% CI)
External Abbreviated Injury Scale
0.92
2.5 (1.3–4.7)
0.004
International normalized ratio
0.95
2.6 (1.2–5.8)
0.021
Abnormal heart rate
1.1
3.1 (0.78–13.0)
0.115
Variable
Constant
–7.9
< 0.001
OR = odds ratio.
or hijabs, customary scarves of females, have been known to catch fire when dangling near open flame (1, 2, 4). Gas or fuel explosions are often the result of refueling kerosene lantern or stoves, a job typically carried out by females (2, 4). The distinction between these civilian studies and the patients who presented to our military hospitals is that our cohort analysis did not reveal any significant gender differences. It is unclear whether the combat environment may have played a role in this difference as some younger males, compared with females, could be more often involved as combatants or in close proximity to hostile events. A large, single burn center study in the United States of 952 severely (> 30% TBSA) burned pediatric patients reported an overall mortality of 13% (11). The mortality rate in the U.S. study is notably lower than the comparable moderate and Pediatric Critical Care Medicine
severe burn groups (based on TBSA) in our study with a combined mortality rate of 33.5% (67 of 200). There are several important factors that may have contributed to this difference. The U.S. children were slightly older (7.3 yr vs 5.9 yr), were from a developed country and therefore may have had better baseline health and nourishment, had no combat-related burns, and were all treated at a tertiary, specialized pediatric burn center as opposed to a field, combat support hospital. Although we are unable to perform a direct comparison of these distinct populations, there is also the potential that lack of experience or resources in an austere combat environment contributed to this striking mortality difference. We were unable to account for differences in provider experience and training as we had no documentation on whether the surgeon was a general surgeon or whether they had any experience with burn injuries. In recent years, burn care is no longer a requirement for general surgery residents, and many military physicians are early in their careers. Although some physicians might receive education about burn care prior to deployment, this is not universal, and it is possible that more robust burn training predeployment could translate to better outcomes. Our study also illustrates the significant resources, primarily in hospital and ICU bed-days, necessary to take care of these burned children. Because of the high mortality in the severely burned group, the moderately burned group used the most resources based on hospital length of stay. Schmidt et al (12) described the care of eight Iraqi children who were cared for in military combat support hospitals that were subsequently transported to Shriners Institute for Burned www.pccmjournal.org
e25
Borgman et al
Table 3.
Comparison of Pediatric Patients by Burn Severity n
Mild
Moderate
Severe
n (%)
549
349 (63.6)
69 (12.6)
131 (23.8)
Age (yr)
548
4 [2–9]
4 [2–9]
5 [2–10]
Male, %
548
68
65
Glasgow Coma Scale
429
15 [15–15]
15 [15–15]
Abnormal heart rate, %
508
46
Hypotension, %
283
Transfused, %
Variable
66 15 [6–15]a,b
60a
67a
2.8
1.8
11a,b
549
3
7.5
10a
Ventilated, %
549
16
16
29a,b
Base deficit
267
3 [0–6]
5 [3–9]a
8 [5–12]a,b
International normalized ratio
223
1.0 [0.9–1.2]
1.1 [1.0–1.4]
1.4 [1.2–1.7]a,b
Hematocrit (mg/dL)
342
38 [34–41]
40 [32–43]
41 [33–48]a
Platelet count (×100,000)
323
408 [311–513]
Mortality, %
549
1.7
Hospital days
549
3 [1–9]
10 [3–21]a
5 [1–19]a
ICU days
549
1 [0–3]
3 [1–9]a
4 [1–13]a
Ventilator days
549
0 [0–0]
0 [0–2]a
1 [1–8]a,b
Hospital-free days
549
27 [21–29]
19 [4–27]a
0 [0–17]a,b
ICU-free days
549
29 [27–30]
27 [18–29]a
0 [0–22]a,b
Ventilator-free days
549
30 [30–30]
30 [28–30]a
6 [0–29]a,b
357 [261–460]
389 [248–553]a
7.2a
47.3a,b
p < 0.05 compared to mild burn. p < 0.05 compared to moderate burn. Values reported as either median [interquartile range] or percent. a
b
Children in Boston, MA. They describe the immense multidisciplinary resources required to treat severely burned children, some requiring up to 23 visits to the operating room. These resources were obviously only available to a small minority. In our study, most children received their definitive care from our deployed military facilities. The fact that there were 10 children who spent at least 2 months admitted to military treatment facilities for burn-related injuries is a testament to the extensive resources dedicated to caring for these children. There have been several reviews of burned soldiers and adult civilians in the recent Iraq and Afghanistan conflicts, which have resulted in guidelines for their management, including resuscitation, wound care, and antibiotic use (13–16). In an epidemiologic review, British military medical forces report that burns account for 6% of injuries, 50% of which are combat related (17). Seventy-five percent of the combat burns are due to blast injury, and 86% of noncombat burns are due to flame. Another British military analysis indicated that 60% of burns in their soldiers were accidental in nature (18). Kauvar et al (19) reported that 63% of military burned casualties were due to combat. In a separate review including civilians, they report that 52% are due to hostile action (20). Our analysis indicates that the majority (84%) of care the military provided to burned children was for noncombat-related burns, a proportion higher than that e26
www.pccmjournal.org
was noted in the adult studies. This is a further testament to the important humanitarian care provided by military medicine, particularly when the prevalence of civilian burns may increase due to the chaotic war environment. We have also shown that the degree of burn, as measured by external AIS, and INR were independently associated with mortality. The study by Kauvar et al (19) compared civilian and military burn mortality and found that TBSA, age greater than 40 years, inhalational injury, and ventilator days were independently associated with mortality, although this study lacked physiologic and laboratory variables (19). Although our analysis did not include inhalational injury, it did include use of invasive ventilation. However, it is not clear whether patients were intubated due to respiratory failure from acute respiratory distress syndrome, pneumonia, inhalational injury, or simply for procedures. In addition, the use of physiologic variables in our study is a more comprehensive way to adjust for injury severity and evaluate predictors of mortality. Although admission INR is associated with mortality, we did not account for the availability and the changes in practice patterns in the use of fresh frozen plasma, so it is unclear if this played a role in affecting the data. Limitations in our study include those inherent to retrospective reviews, particularly with missing or incorrect data. However, the DoDTR staff who input data undergo extensive February 2015 • Volume 16 • Number 2
Online Clinical Investigations
training, interrater reliability audits, and other quality assurance measures to ensure high data quality. We also chose to examine the severity of burns based on external AIS, as opposed to percent of TBSA, as this was how the data were available to us in the registry. However, by using the AIS external score as a surrogate for TBSA, we were still able to categorize burn severity as mild, moderate, or severe. One study has compared external AIS with TBSA for burns, finding a near-perfect coding agreement between AIS coding and % TBSA calculated by expert burn clinician (weighted κ, 0.83), which is sufficient for registry and research purposes (21). Using external AIS within the multivariate logistic regression has not been done before, and this would need to be validated in other studies before being adopted. Unfortunately, we were unable to characterize the cause of burn because of lack of documentation with most etiologies listed as “unknown” or simply “burn.” It has also been reported that there is a significant number of suicide attempts by teenage females in both countries (3, 4) although we do not have this categorized in the database. We also do not have any data on prehospital transport times or time of injury prior to admission. Although our study had predominantly noncombat-related burn injuries, it would have been interesting to compare the civilian studies with our military study to evaluate the causes of the burns (i.e., scald vs flame vs blast). In the future, improved documentation of this field would improve our understanding of the epidemiology of pediatric burns in an austere environment. Unfortunately, we have no data on patients after discharge, and burn patients frequently require multiple follow-up surgeries, reconstruction, and rehabilitation. Although this was provided on a limited basis for some children prior to ultimate discharge, follow-up medical missions would be necessary to provide this care and identify some long-term outcomes. In the future, military medical leaders could use these data to plan for resource utilization and consider training, potentially at specialized pediatric burn centers, to prepare medical staff to best care for this spectrum of injuries. The higher mortality in these patients, relative to reported mortality in civilian pediatric burn centers, possibly reflects a significant potential to save lives with additional training and resources. It is unknown whether additional training alone would improve outcomes, and the mechanisms of burns (both from combat and noncombat injuries) likely represent a worse injury than those occurring in the developed world in healthier children. However, given the challenge of caring for these complex cases and the numbers at which they occur, it seems prudent to prepare our physicians as best as possible with training opportunities already available.
CONCLUSIONS In this first study of pediatric burn care by U.S. military treatment facilities in Iraq and Afghanistan, we examined factors associated with mortality and we found evidence that significant resources are used in caring for children, sometimes for months and most commonly for noncombat-related injuries. Children with severe burns (TBSA > 39% or > 29% if < 5 yr) had a high mortality and required significant resources in a setting that is not primarily resourced for long-term care of Pediatric Critical Care Medicine
severe pediatric burn injury. Extraordinary measures are therefore used for the long-term care of these burned children within the war zones of Iraq and Afghanistan.
REFERENCES
1. Peden M, Oyegbite K, Ozanne-Smith J, et al: World Report on Child Injury Prevention: Burns. Geneva, Switzerland, WHO, 2008 2. Forjuoh SN: Burns in low- and middle-income countries: A review of available literature on descriptive epidemiology, risk factors, treatment, and prevention. Burns 2006; 32:529–537 3. Qader AR: Burn mortality in Iraq. Burns 2012; 38:772–775 4. Padovese V, De Martino R, Eshan MA, et al: Epidemiology and outcome of burns in Esteqlal Hospital of Kabul, Afghanistan. Burns 2010; 36:1101–1106 5. Borgman M, Matos RI, Blackbourne LH, et al: Ten years of military pediatric care in Afghanistan and Iraq. J Trauma Acute Care Surg 2012; 73:S509–S513 6. Matos RI, Holcomb JB, Callahan C, et al: Increased mortality rates of young children with traumatic injuries at a US army combat support hospital in Baghdad, Iraq, 2004. Pediatrics 2008; 122:e959–e966 7. Gennarelli TA, Wodzin E: AIS 2005: A contemporary injury scale. Injury 2006; 37:1083–1091 8. ECC Committee, Subcommittees and Task Forces of the American Heart Association: 2005 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Circulation 2005; 112(24 Suppl):IV1–IV203 9. Fleming S, Thompson M, Stevens R, et al: Normal ranges of heart rate and respiratory rate in children from birth to 18 years of age: A systematic review of observational studies. Lancet 2011; 377:1011–1018 10. Creamer KM, Edwards MJ, Shields CH, et al: Pediatric wartime admissions to US military combat support hospitals in Afghanistan and Iraq: Learning from the first 2,000 admissions. J Trauma 2009; 67:762–768 11. Kraft R, Herndon DN, Al-Mousawi AM, et al: Burn size and survival probability in paediatric patients in modern burn care: A prospective observational cohort study. Lancet 2012; 379:1013–1021 12. Schmidt PM, Sheridan RL, Moore CL, et al: From Baghdad to Boston: International transfer of burned children in time of war. J Burn Care Res 2014; 35:369–373 13. Cancio LC, Lundy JB, Sheridan RL: Evolving changes in the management of burns and environmental injuries. Surg Clin North Am 2012; 92:959–986, ix 14. Chung KK, Salinas J, Renz EM, et al: Simple derivation of the initial fluid rate for the resuscitation of severely burned adult combat casualties: In silico validation of the rule of 10. J Trauma 2010; 69(Suppl 1):S49–S54 15. Hospenthal DR, Murray CK, Andersen RC, et al; Infectious Diseases Society of America; Surgical Infection Society: Guidelines for the prevention of infections associated with combat-related injuries: 2011 update: Endorsed by the Infectious Diseases Society of America and the Surgical Infection Society. J Trauma 2011; 71(2, Suppl 2):S210–S234 16. USAISR: Burn care. Joint Theater Trauma System Clinical Practice Guideline 2012. Available at: http://www.usaisr.amedd.army.mil/assets/ cpgs/burn_care_13_nov_13.pdf. Accessed November 22, 2013 17. Jeevaratnam JA, Pandya AN: One year of burns at a role 3 Medical Treatment Facility in Afghanistan. J R Army Med Corps 2014; 160:22–26 18. Foster MA, Moledina J, Jeffery SL: Epidemiology of U.K. military burns. J Burn Care Res 2011; 32:415–420 19. Kauvar DS, Cancio LC, Wolf SE, et al: Comparison of combat and non-combat burns from ongoing U.S. military operations. J Surg Res 2006; 132:195–200 20. Kauvar DS, Wolf SE, Wade CE, et al: Burns sustained in combat explosions in Operations Iraqi and Enduring Freedom (OIF/OEF explosion burns). Burns 2006; 32:853–857 21. Watterson D, Cleland H, Picton N, et al: Level of agreement between coding sources of percentage total body surface area burnt (%TBSA). HIM J 2011; 40:21–24 www.pccmjournal.org
e27
