Etiology and outcome of pediatric burns

Journal

of Pediatric

Surgery MARCH

VOL 31, NO 3

1996

Etiology and Outcome of Pediatric Burns By Stephen

E. Morrow,

David L. Smith,

Bruce A. Cairns, Patrick D. Howell, Chapel Hill, North Carolina

0 A 6-year retrospective review of burn victims hospitalized at a major burn center was conducted to determine the etiology and outcome of pediatric burns. Four hundred forty-nine patients under age 16 years were identified and stratified by age, sex, burn size, presence or absence of inhalation injury, cause of burn, and county of residence. The mean patient age was 4.3 * 0.2 years, and the male:female ratio was 1.9: 1. There were 21 deaths overall (4.7%). the majority of which (18) were among children under 4 years of age. With respect to large burns, defined as 230% total body surface area (TBSA), the mortality rate for children under age 4 was significantly higher than that for older children (46.9% v 12.5%; P < .Ol), despite the nearly identical mean burn size of the two groups. Except for burn incidence, there were no significant differences between males and females. The mean burn size was 15.1% f 0.7%, and was significantly larger for nonsurvivors than survivors (55.3% YZ5.7 v 13.1% f 0.5%; P < .Ol). Inhalation injuries were strongly associated with large burns and were present in all 15 flame-burn fatalities. Scalds were the most common type of burn among children under 4 years of age; flame burns predominated in older children. There were 6 deaths related to scalds, all of which occurred in children under 4. Burn type, size, and mortality rate did not differ between children from urban and rural counties. Large burn size was the strongest predictor of mortality, followed by (in order) age less than 4 and the presence of inhalation injury. Infants and young children have the highest risk of death from burn injury. Burns smaller than 30% TBSA without an inhalation injury (such as small scald injuries) occasionally are lethal in infants and small children, despite modern therapy. Copyright o 1996 by W.B. Saunders Company INDEX

WORDS:

Burn,

etiology.

A

LTHOUGH improved survival rates for pediatric burn victims have been reported in recent years,lB2burns continue to pose a major threat to the heahh of children in the United States, For the past 20 years, burns have ranked second among the causes of accidental death in children under age 4, and third for older children3s4 Although the direct impact of patient age and inhalation injury on mortality in adult burn victims has been well established, the role of these factors in pediatric burn victims is less clear. Additionally, there are few reports concerning the

./ournaloffediarric

Surgery,

Vol31,

No 3 (March),

1996: pp 329-333

Don K. Nakayama,

and H.D. Peterson

influence of burn etiology on outcome for pediatric burn victims. The purpose of this study was to review the recent experience at a major burn center to determine the effect of burn etiology on outcome and to identify predictors of mortality for pediatric burn victims. MATERIALS

AND METHODS

A retrospective review of all pediatric admissions to the Jaycee Burn Center from January 1988 through January 1994 was conducted using a computerized database. A total of 467 children (under age 16 years) were identified, 449 (96.1%) of whom had statistical data sufficient to complete the study. According to 1990 census data, the population of children in North Carolina aged 16 and under was 1,425,407 (21.5% of the state’s total population). This figure represents an estimate of the pediatric population at risk for burn injury during the study period, when the Jaycee Burn Center hospitalized approximately 80 pediatric burn victims per year. An unknown (but certainly much smaller) number of pediatric burn victims may have been hospitalized at other facilities in North Carolina. The patients were stratified by age, sex, burn size, presence or absence of inhalation injury, cause of burn, outcome, and county of residence. To facilitate comparison with other studies, the children were divided into two age groups: those who had not surpassed their fourth birthday (“younger” children) and those who had surpassed it (“older” children). Burn size was determined clinically and refers to the total body surface area (TBSA) sustaining full- or partial-thickness injury. Burn sizes were arbitrarily divided into “small” (< 30% TBSA) and “large” ( > 30% TBSA). Fluid requirements were estimated using the Parkland formula. Urine output was used as the principle resuscitation guideline. Inhalation injury was determined by one of two attending surgeons, based on the historical details of each case and the clinical findings. Burn cause was determined by history; standard categories were used and included flame burns, scalds (spills v immersions), electrical, chemical, contact, and burns of unknown causes. Outcome was recorded as patient survival or death. From the Division of Pediatric Surgery, Department of Sutgev, and the Jaycee Bum Center, The University of North Carolina at Chapel Hill, Chapel Hill, NC. Address reprint requests to Don K. Nakayama, MD, Chief Division of Pediatric Surgery, The University of North Carolina at Chapel Hill, Chapel Hill, NC 27599-7210. Copyright o 1996 by W B. Saunders Company 0022-3468/96/3103-0001$03.00/0 329

330

MORROW

North Carolina has 100 counties, 78 of which are currently defined as “rural” by the Office of State Planning, based on 1990 census data. The other 22 are “urban.” North Carolina uses the term “urban” to describe an area with 50,000 or more inhabitants and a population density of at least 1,000 persons per square mile. Patients originate in the entire state of North Carolina and generally are transferred to the Jaycee Burn Center according to Advanced Trauma Life Support (ATLS) guidelines. The Jaycee Burn Center does not have specific acceptance criteria; all of North Carolina’s burn patients are accepted, regardless of burn size. The vast majority of initial resuscitations were conducted at the Jaycee Burn Center because it is centrally located in North Carolina and burn victims typically arrive from outlying facilities within a few hours of injury. Data were stored in a computer registry that was maintained by designated burn center registrars and were cross-checked against hospital records for accuracy. Statistical analysis (logistic regression analysis, t test, x2 square, or comparison of proportions where appropriate) was performed using Statistical Analysis Software (Gary, NC). All data are expressed as mean f standard error of the mean unless otherwise noted. A P value of less than .05 was considered statistically significant. RESULTS

Patient Age and Gender

The mean age was 4.3 years, with a standard deviation of 4.3 years and a range of 6 weeks to 15 years 11 months. There were 296 males and 153 females (male:female ratio, 1.9:1). Children under 2 years of age represented the largest single group of patients (Table 1). The patients’ ages burn sizes, burn types, and the presence or absence of inhalation injury are noted in Table 1. Although there were 21 deaths (4.7%) among the entire group, the majority (18) occurred in children under 4 years of age. The mortality rate among children aged 0 to 4 years was significantly higher than that of older children (6.1% v 1.9%; P < .0.5). This higher mortality rate among children aged 0 to 4 persisted for large, but not small, burns (Table 2). There were no significant gender differences with respect to mortality, burn size, burn type, presence of inhalation injury, or county of origin. Bum Size

The mean TBSA burn size was 15.1% + 0.7% (range, 0.5% to 98%). The burn size distribution was skewed toward sizes smaller than 30% TBSA, as reflected by the standard deviation of 15.1%. The average size of lethal burns was significantly larger than that of nonfatal burns (55.3% +- 5.7% v 13.1% ? 0.5%; P < .OOl). Only two of the 15 fatal flame burns and one of the six fatal scalds were less than 30% in size and each of these lethal small burns occurred in children under age 4. Burn size did not vary significantly between age groups. Among the fatalities in 0 to 4 year olds, the average burn was

Table

1. Overall

ET AL

Mortality Rates by Age, Burn Size, Burn Type, Presence of Inhalation Injury No. of Patients

and

Mortality

Fatalities

Survivors

Age (vi

Rate (%I

O-2

197

187

10

5.1

2-4 All 14

97 294

a9 276

a ia

a.2 6.1 2.9

4-a a-12 12-16 All >4

68 34

66 33

2 1

53 155

53 152

0 3

All children

449

428

21

Burn size <30% 2 30%

393 56

390 38

18

Burn type Flame

163

148

155 62

152 59

40 16 6 7

40 16 6 7

Spill scald Immersion Contact Chemical Electrical Unknown Inhalation injury Present

NOTE. The mortality

1.9 4.7

3

0.8 32.1 9.2 1.9 4.8 0 0 0 0

41

56 393

2.9 0

15 6

387

rate is significantly

higher

26.8 1.5

for children

14 years

of age, burns of 230% TBSA, and presence of inhalation injury. Flame burns collectively are associated with a significantly higher mortality rate than that for scalds, but their greater. When burn size is controlled, significantly between burn types.

mean TBSA the mortality

also is significantly rate does not differ

significantly smaller than that of older nonsurvivors (50.7% f 5.6% v 83.0% +- 14.0%; P = .04). Inhalation Injury

Inhalation injury was present in 12.2% of the total population and was significantly more common among nonsurvivors than survivors (66.7% v 9.6%; P < .OOl). All 15 fatal flame burns had associated inhalation injury. Collectively, the mortality rate for all children with inhalation injury was 26.8%; the rate for those without inhalation injury was only 1.5% (P < .OOl). However, inhalation injury was strongly associated Table AW

14yr

>4yr

2. Burn Mortality

Data

by Age Group

and Burn Size

No. of Patients

Survivors

Fatalities

<30% 230% All burns <30%

262 32 294 131

259 17 276 131

3 15 18 0

230% All burns

24 155

21 152

3 3

TBSA

*P < .Ol, mortality rate for patients age > 4 (x2 analysis). tP < .05, mortality rate for all patients (x2 analysis).

with

large

with burns:

burns:

Mortality Rate (%I

0.4 46.9* 6.lt 0 12.5* 1.9t age

14

age ~4 v age >4

v

OUTCOME

OF PEDIATRIC

331

BURNS

10.7% i 0.4%; P = .048; and 51.7 + 3.2 v 39.2 + 2.6%; P = .02). However, the difference was not significant when the burns were further subdivided between younger and older children. Among patients with large burns, the mortality rate for flame injuries did not differ significantly from that for scalds (34% v 33%).

with large burns. The average size of flame burns complicated by inhalation injury was three times greater than that of flame burns without inhalation injury (35.6% + 3.5% v 12.1% -+ 0.5%; P < .OOl). Young children with flame burns had a higher incidence of inhalation injury than did older children with flame burns (50% v 22.7%; P < .OOl), even though their average burn sizes were similar.

County of Residence

Etiology of Bum

There was an even distribution of patients among the urban and rural counties (209 urban, 204 rural). The other 36 patients were from outside the state of North Carolina or had an unknown address. Among young children, the mortality rate for urban patients did not differ significantly from that of rural patients (6.3% v 5.2%). Th ere was no significant difference between urban and rural children with respect to mean patient age (4.2 ‘-t-0.3 years v 4.4 * 0.3 years), burn size (15.0% + 1.0% v 14.1% 2 l.O%), presence of inhalation injury (11.0% v 10.8%), or distribution of burn etiology.

Scalds were the most common burns (58%) among the 294 younger children, whereas two thirds of older children had flame injuries (Fig 1). Relatively few injuries were related to other mechanisms; generally these burns were small (7.8% -t 1.1% TBSA), and they caused no fatalities. Among scald injuries, spills outnumbered immersions by more than 2 to 1 and were similar in incidence (12.1% + 0.6% v 14.0% f 1.6%). There were six scald fatalities in this series. All these deaths involved young children; there were three were immersions and three spills. Among the six scald fatalities, the average burn size for immersions was 47.3% t11.5%; that for spills was 34.0% & 8.4% (P = NS). The mean fatal scald size was 40.7 + 7.1%, which was not significantly smaller than the mean fatal flame injury (61.1% +- 7.0%). Flame burns caused the remaining fatalities in this series; all but three of these deaths involved younger children. Flame burns were significantly larger than scalds, both collectively (21.3% f 1.6% v 12.7% +- 0.6% TBSA, P < .Ol) and within the small and large burn subgroups (12.1% + 0.6% v

Predictors of Outcome

Logistic regression analysis was used to determine whether any of the variables associated with burn patients were predictors of fatal outcome. Large burn size was the strongest predictor of mortality, followed by patient age of less than 4 years (Fig 2). Inhalation injury, strongly linked to large burns, was found to be an independent predictor of death, although this fact became apparent only when the skewed burn size distribution was adjusted logarithmically. Patient gen-

100 90 80

( mUk

q

El

60

q

Ch

50

Elct

70

I

Elm SP

n

4-8

8-12

FI

12-16

Age group Fig 1. Distribution Uk, unknown cause.

of burn etiology by age group. Scalds (Sp and Im) predominate

FI, flame; Sp, spill scald; Im, immersion in younger age groups, whereas flame

scald; Ct, contact burn; Ch, chemical; El, alactrical; burns are more common among older children.

332

MORROW

1 0.9 0.8 y$ 0.7 ,> 0.6 $ m

3 0.3 0.2 0.1 0

TBSA

100

‘0

Fig 2. Logistic regression analysis model for estimating mortality rates among pediatric burn patients. Probability of survival is depicted as a function of age and burn size (TBSA). The survival rate begins to decrease to below 100% for small burn sizes in very young children (22 years of age), but it remains near 100% for older children without large burns.

der, cause of burn, and county of residence were not predictive of mortality. DISCUSSION

As expected, we found that increased burn size leads to increased risk of mortality among children, a finding confirmed by other studies.1,5,6 The average lethal burn size in the present series was 55%, and only three of the 21 deaths were caused by burns of less than 30% TBSA. While laudable improvements in survival rates for burned children have been reported in recent years, 1,2,7the three nonsurvivors with small burns in this series are an important reminder that children can die from small burns. Logistic regression analysis also found that children aged 0 to 4 years had a higher risk of death, independent of burn size. Although it is well known that young children represent the largest single age group admitted to most burn centers8 the issue of whether young age predisposes a burn victim to increased mortality has been debated in the literature. There are several reports that infants and toddlers have a higher mortality rate than older children when burn size is controlled2,6,g; other investigators have observed no difference.rJOJ1 Reasons cited for the higher pediatric burn mortality rate include an immature immune system and increased fluid requirements, which place children at a higher risk for sepsis and hypovolemic shock after burn injury.gJ1J2 On the other hand, the relative absence of coexistent disease in the majority of pediatric burn victims, when compared with older burn patients, could negate these physiological handicaps. The higher

ET AL

mortality rate for the young children in our series might be explained in part by the fact that they had an higher incidence of inhalation injury, even when burn size was controlled. This finding is not surprising given the relative immobility and lack of situation awareness that might make young children more prone to inhalation injury. However, all six scald fatalities, which by nature are not associated with inhalation injury, occurred in children under age 4. Moreover, no deaths from small burns occurred among children over age 4, but three such deaths occurred in younger children. These observations, coupled with the significantly higher mortality rate for children under age 4 despite their significantly smaller mean fatal burn size, provide reasonable evidence that the risk of death from thermal injury among young children is higher than that for older children. Logistic regression analysis did validate inhalation injury as an independent predictor of mortality in our series, although it was the weakest predictor. In children, the diagnosis of inhalation injury is usually made clinically because bronchoscopy is technically difficult in the pediatric burn patient and has not been shown to predict the need for ventilatory support.‘3 First recognized as an important entity in burn injury after the Cocoanut Grove fire in 1942, inhalation injury has since been cited by some investigators as the single most important predictor of mortality in burn victims.14-l7 Several observations support our finding of a lesser importance for inhalation injury in children. First, we found, as have others,14J7 that inhalation injuries are strongly associated with large burns, which alone are an independent mortality predictor. All fatal flame burns in this series were large (61.1% + 7.0% TBSA) and had associated inhalation injury. If inhalation injury were a strong independent predictor of death, one would expect an appreciable mortality rate among children with small burns and coexistent inhalation injuries. In the present series, only two of 31 children with this combination died. Second, one would expect that the average fatal scald burn would be larger than the average fatal flame burn, if inhalation injuries were included with the flame burns. The converse was true in our series. Finally, our finding that inhalation injury was the weakest predictor of mortality is consistent with previous work showing that the greatest impact of inhalation injury is among patients with burns of 40% to 60% TBSA.15J7J8 Only three of the 15 deaths from flame burns in this series involved burn sizes within this range. We determined that burn cause was not an independent predictor of mortality, that is, a flame injury was neither worse nor better per se than a scald injury,

OUTCOME

OF PEDIATRIC

333

BURNS

when burn size was controlled. Other studies concur with our findings that young children suffer mainly from scald injuries and that older children are more prone to flame burns. 19-25We were surprised by the finding that the average burn size for the six fatal scalds, all of which occurred in children under age 4, was only 40.7% + 7.1%. The burn sizes for these cases were l&5%, 32%, 36%, 40%, 47.5%, and 70% TBSA. A 7-month-old patient had a hot grease scald to the head and neck and required intubation. Respiratory distress syndrome with barotrauma developed quickly, and he died of multisystem organ failure. The fact that a 7-month-old died from an 18% scald injury is sobering testimony to the potential gravity of relatively small burns in these young patients. Although there were numerous cases of suspected child abuse, particularly among young children with scald injuries, it was often difficult to distinguish, with certainty, abuse from negligence. Not surprisingly, all

21 deaths in this series resulted from septic complications (primarily pneumonia and respiratory distress syndrome) that led to multisystem organ failure. In 1990 it was reported that boys who had burns of greater than 30% TBSA without inhalation injury had a higher mortality rate than girls with the same degree of injury. 26 We were unable to confirm this finding, nor did we detect any gender differences for any other parameters analyzed herein, except for the higher incidence of burns among males. Our patients were evenly distributed among urban and rural counties, and there were no difference between these two groups with respect to age, burn etiology, burn size, presence of inhalation injury, or mortality rate. These findings suggest that all children are at risk for burn injury, and that a regional burn center, with an efficient patient transfer apparatus, can deliver burn care to urban and rural children, with equivalent outcome.

REFERENCES 1. Tompkins RG, Remensnyder JP, Burke JF, et al: Significant reductions in mortality for children with burn injuries through the use of prompt eschar excision. Ann Surg 208:577-585,1988 2. Erickson EJ, Merrell SW, Saffle JR, et al: Differences in mortality from thermal injury between pediatric and adult patients. J Pediatr Surg 26:821-825,199l 3. Berman W, Reichfelderfer T, Mofenson HC: Childhood burn injuries and deaths. Pediatrics 51:1069-1071, 1973 4. National Safety Council: Accident Facts. 1993 edition, Itasca, P 22 5. Bruck HM, Asch MJ, Pruitt BA: Burns in children: A lo-year experience with 412 patients. J Trauma 10:658-662,197O 6. Durtschi MB, Kohler TR, Finley A, et al: Burn injury in infants and young children. Surg Gynecol Obstet 130:651-656,198O 7. Caravajal HF, Parks DH: Survival statistics in burned children. J Burn Care Rehabil3:81-84,1982 8. Herndon DN, Gore D, Cole M, et al: Determinants of mortality in pediatric patients with greater than 70% full-thickness total body surface area thermal injury treated by early total excision and grafting. J Trauma 27:208-212,1987 9. McLaughlin E, McGuire A: The causes, cost, and prevention of childhood burn injuries. Am J Dis Child 144:677-683,199O 10. Bull JP, Fisher AJ: A study of mortality in a burns unit: A revised estimate. Ann Surg 139:269-273,1954 11. Merrell SW, Saffle JR, Sullivan JJ, et al: Fluid resuscitation in thermally injured children. Am J Surg 152664-669, 1986 12. O’Neill JA: Fluid resuscitation in the burned child-A reappraisal. J Pediatr Surg 17:604-607, 1982 13. Bingham HG, Gallagher TJ, Powell MD: Early bronchoscopy as a predictor of ventilatory support for burned patients. J Trauma 27:1286-1288,1987

14. Thompson PB, Herndon DN, Traber DL, et al: Effect on mortality of inhalation injury. J Trauma 26:163-165, 1986 15. Shirani KZ, Pruitt BA, Mason AD: The influence of inhalation injury and pneumonia on burn mortality. Ann Surg 205:82-87, 1987 16. Herndon DL, Langner F, Thompson P, et al: Pulmonary injury in burned patients. Surg Clin North Am 67:31-43,1987 17. Tredget EE, Shankowsky HA, Taerum TV, et al: The role of inhalation injury in burn trauma. Ann Surg 212:720-727,199O 18. Agee RN, Long JM, Hunt JL, et al: Use of xenon in early diagnosis of inhalation injury. J Trauma 16:218-224, 1976 19. Stuart SP, Pegg JJ, Gregory PG, et al: Burns in childhood: An epidemiological survey. Aust N Z J Surg 48:365-373,1978 20. Feller I, James MH, Jones CA: Burn epidemiology: Focus on youngsters and the aged. J Burn Care Rehabil3:285-288,1982 21. Glasheen WP, Attinger EO, Anne A, et al: Identification of the high-risk population for serious burn injuries. Burns 9:193-200, 1983 22. Raine AM, Azmy A: A review of thermal injuries in young children. J Pediatr Surg 18:21-26, 1983 23. Green AR, Fairclough J, Sykes PJ: Epidemiology of burns in childhood. Burns 10:369-371,1984 24. Carrigan L, Heimbach DM, Marvin JA: Risk management in children with burn injuries. J Burn Care Rehabil9:75-78, 1988 25. Ryan CA, Shankowsky HA, Tredget EE: Profile of the paediatric burn patient in a Canadian burn centre. Burns 18:267272,1992

26. Barrow RE, Herndon DN: Incidence of mortality in boys and girls after severe thermal burns. Surg Gynecol Obstet 170:295298,199O