- Email: [email protected]
Epidemiology of burn patients presenting to a tertiary hospital emergency department in Lebanon
JBUR 5312 No. of Pages 8
burns xxx (2017) xxx –xxx
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/burns
Epidemiology of burn patients presenting to a tertiary hospital emergency department in Lebanon Abou Dagher Gilbert a, Eva Rajha a, Christopher El Khuri a, Ralph Bou Chebl a, Aurelie Mailhac b , Maha Makki b, Mazen El Sayed a, * a b
Department of Emergency Medicine, American University of Beirut Medical Center, Beirut, Lebanon Clinical Research Institute, American University of Beirut Medical Center, Beirut, Lebanon
article info
abstract
Article history:
Background: The epidemiology of burns is well described in the United States, Europe and
Accepted 24 June 2017
Asia. However, few studies address this topic in the Levant region. This study aims to
Available online xxx
describe characteristics of burn victims treated at the emergency department of a tertiary care center in Beirut, Lebanon and to report on factors that affect ED disposition.
Keywords: Emergency department Burn ED disposition Levant
Methods: A retrospective cohort study was conducted in the ED of a tertiary care hospital, in Beirut, Lebanon, between 2009 and 2013. Patients were recruited if their final ED diagnosis was burn. A descriptive analysis of patients’ characteristics including burns was done, followed by a bivariate analysis to compare two groups (admitted vs discharged). Statistical analysis included the use of Student t-test and Pearson chi-square where appropriate. A multivariate analysis was then conducted to determine the predictors of hospital admission. Results: A total of 366 patients had their charts queried. Age category distributions of the patients were 73 (19.9%) <5 years, 39 (10.7%) 5–14 years, 236 (64.5%) 15–65 years and 18 (4.9%) >65 years of age. Around half of the patients (47.3%) were males, with scalding being the most common mechanism of burn (53.9%), followed by contact with hot object (16.8%) and flame (11.9%). In terms of disposition from the ED, 58 (15.8%) patients were admitted to the hospital with 42 (72.4%) going to GPU, 12 (20.6%) to ICU and 4(6.9%) transferred to either another hospital or to an acute burn facility. Admitted patients tended to be at the extremes of ages (<5 years or 65 years), male dominant, more likely to be brought in by family, with higher proportions of more severe mechanisms of injury (flame, electric, chemical). Admitted patients also sustained higher degrees of burns with more TBSA injured. Predictors of hospital admission included the aforementioned factors with the strongest predictors of admission being full-thickness degree burn (OR 18.56 (4.67–73.72 CI95% p 0.001)) and mechanism of injury such as electrical (OR 23.01 (3.23–163.89 CI95% p=0.002)) and chemical (OR 17.43 (2.33–130.14 CI95% p=0.003)).
Abbreviations: ABA, American Burn Association; ED, emergency department; EHR, electronic health record; EMS, emergency medical services; GPU, general practitioner unit; ICD, International Statistical Classification of Diseases and Related Health Problems; ICU, intensive care unit; TBSA, total body surface area; WHO, World Health Organization. * Corresponding author. E-mail address: [email protected] (M. El Sayed). http://dx.doi.org/10.1016/j.burns.2017.06.015 0305-4179/© 2017 Elsevier Ltd and ISBI. All rights reserved.
Please cite this article in press as: A.D. Gilbert, et al., Epidemiology of burn patients presenting to a tertiary hospital emergency department in Lebanon, Burns (2017), http://dx.doi.org/10.1016/j.burns.2017.06.015
JBUR 5312 No. of Pages 8
2
burns xxx (2017) xxx –xxx
Conclusion: Characteristics of burn patients treated in an urban ED in Lebanon mirror those of burn victims in other international studies. Future larger epidemiologic studies are needed to better quantify the impact of burns in Lebanon. © 2017 Elsevier Ltd and ISBI. All rights reserved.
1.
Introduction
The World Health Organization (WHO) global burden of disease data, updated in 2008, reports that approximately 11 million people per year sustain burns requiring medical attention. The report places burn as the fourth most common substantial injury after road traffic accidents, falls and interpersonal violence [1]. In the US, approximately 450,000 individuals require medical treatment for burns per year, with most of these patients treated in the emergency department (ED), and discharged with outpatient follow-up [2,3]. On a global scale, 90% of all burns occur in low and middle income countries [1]. There is a growing international literature regarding the epidemiology of burns in the US, Europe and Asia [2,4–6]. Pakistan, Turkey and Iran have contributed to the database of burn patients in the Levant region [7–9] with studies reporting on burn patients that present to emergency departments and not just to dedicated burn centers. Studies pertaining to burns are lacking in Lebanon. This study was conducted in a tertiary care center in Lebanon to contribute to the international and local literature, by describing characteristics of burn victims treated at the emergency department and by reporting on factors that affect ED disposition.
calculation was performed. However, given that multiple studies for burn patients presenting to emergency departments recruited between 300 and 400 patients, we aimed to recruit a similar number of patients [6–9].
2.4.
Trained research fellows extracted available clinical information from scanned charts and electronic laboratory reports: Patients’ demographic information, past medical history and burn characteristics (etiology, injury site, circumstance, suspected motive, time from injury, degree, percentage surface area affected and degree of burn) were extracted. Disposition from the ED, length of stay in the ED, general practitioner unit (GPU), and intensive care unit (ICU) were noted. If the patient was transferred to another hospital or the country’s specialized burn center, the patient was considered “admitted” though physically discharged from our hospital. This is due to the similar clinical profile shared by those admitted to our hospital and those requiring hospital treatment elsewhere. The reason In-hospital mortality was collected as well. Finally, the American Burn Association (ABA) severity score was calculated for all burns as per the definition set by the ABA [10].
2.5.
2.
Methods
2.1.
Study design and setting
This is an IRB approved (ER.GA.06), single center, retrospective cohort study conducted in the ED of the American University of Beirut Medical Center (AUBMC). AUBMC is the largest tertiary care center in Lebanon with over 56,000 ED visits per year. All patients presenting from July 2009 to December 2013 had their medical records queried via the hospital’s Electronic Health Record (EHR).
2.2.
Patient selection
Patients’ ED encounters were filtered by an experienced data user using the hospital’s EHR via structured keyword searches and ICD-9 coding (International Statistical Classification of Diseases and Related Health Problems). Patients were included in the study if their charts included the final ED diagnosis of “Burn” regardless of severity. No age restriction was implemented; and no exclusion criteria were used.
2.3.
Sample size calculation
This study serves as a foundation for future epidemiologic studies of burn patients in Lebanon. No formal sample size
Variables
Statistical analysis
Statistical analyses were performed using SPSS version 24.0 (Armonk, NY: IBM Corp). The distributions of the continuous and categorical variables were presented as meanstandard deviation and frequency/percentages respectively. Selected characteristics were then stratified by admission status (Discharged or Admitted) and variable differences between the two groups were calculated by Pearson’s chi square/ Fischer test and Student’s t-test where appropriate. Tests were interpreted at a significance level alpha=0.05. A multivariate logistic regression analysis was performed to ascertain the predictive factors of hospital admission in the population. A backward selection procedure, with significance level for variable removal from the model set at 0.1, was conducted. The independent variables chosen for modelling were those found to be significant at the bivariate analysis level in addition to those considered clinically meaningful. The variables included in the model were: Age, gender, method of transportation, degree of burn, location of burn, mechanism of injury, Total Body Surface Area (TBSA), and ABA score. The results were described as odds ratios (OR) and their corresponding 95% confidence intervals (CI). A sub-analysis was performed to describe the association between gender and injury site. All patients were stratified by gender and a Pearson’s chi square/Fischer test was performed where appropriate (Supplement 1). A second sub-analysis was done to evaluate the association between the degree of burns
Please cite this article in press as: A.D. Gilbert, et al., Epidemiology of burn patients presenting to a tertiary hospital emergency department in Lebanon, Burns (2017), http://dx.doi.org/10.1016/j.burns.2017.06.015
JBUR 5312 No. of Pages 8
3
burns xxx (2017) xxx –xxx
and hospital length of stay (LOS). All admitted patients were stratified by degree of burn and their respective hospital LOS averages were compared by one-way ANOVA (Supplement 2).
3.
Results
3.1.
Total patient demographics and disposition
A total of 366 patients were included in this study. The age distribution of the population was 73 (19.9%) <5 years, 39 (10.7%) 5–14 years, 236 (64.5%) 15–65 years and 18 (4.9%) >65 years of age, with an age range of 1month–89years. Around half (47.3%) of patients were male, with most patients presenting to the ED with family (57.2%) and coming from home (70.9%). Additionally, the mean time from injury to ED presentation was 33.9473.81min with most patients either presenting within 1h (44.7%) or after 24h (39.1%). In terms of disposition from the ED, 58 (15.8%) patients were admitted to the hospital, including 42 (72.4%) to the GPU, 12 (20.6%) to ICU and 4 (6.9%) to either another hospital or an acute burn facility. The mean length of stay (LOS) in the ICU and hospital were found to be 7.186.76 and 9.5411.85days respectively. Of those discharged from the ED, most were sent home, with only a few patients leaving against medical advice (AMA). From the 366 patients included in the study, one in-hospital ICU mortality was noted. The data is summarized in Table 1.
3.2.
Total population burn characteristics
The most common mechanism of burn was scalding (53.9%) followed by contact with hot object (16.8%) and flame (11.9%). The degree of burn ranged from superficial (82%) to deep partial thickness (10.2%) to full-thickness (7.7%). The total body surface area affected was mostly <5% and accounted for 84.4% of all injuries in this study. In terms of location of the burn, the upper extremity was the most affected (57.9%) followed by the lower extremity (30.6%) and the head (18.6%). The ABA severity score for all burns was calculated and found to have the following distribution: Minor (77.6%), moderate (10.2%) and major (5.7%) (Table 2).
3.3.
Admitted vs. discharged patients: demographics
When compared to discharged patients, admitted patients tended to belong more to the youngest (<5 years group 41.4% vs 15.9%) and eldest (65 years group 10.3% vs 3.9%) age distributions. Additionally, admitted patients were predominantly males and a higher proportion of them were brought in by family compared to discharged patients (77.6% vs 53.2%). Injury site and time from injury to presentation did not differ between the two groups (Table 3).
Table 1 – Total population demographics (N=366). Male no. (%) Female no. (%)
173 (47.3) 193 (52.7)
Age no. (%) <5 years 5–14 years 15–65 years 65 years
73 (19.9) 39 (10.7) 236 (64.5) 18 (4.9)
Method of transportation no. (%) Self Family EMS
137 (38.8) 202 (57.2) 14 (4.0)
Injury site no. (%) Home Street Recreation and sport Other
180 (70.9) 35 (13.8) 15 (5.9) 24 (9.4)
Time from injury no. (%) 1 h 2 h–1 day >1 day
127 (44.7) 46 (16.2) 111 (39.1)
Disposition no. (%)a Discharged (N =308)b Home Left AMA Admitted (N =58)b CCU GPU Transferred to another hospital Transferred to an acute burn facility LOS in days N=54 (Mean SD)c Hospital ICU
12 (20.6) 42 (72.4) 3 (5.1) 1 (1.8)
9.54 11.85 7.18 6.76
a
One ICU mortality was noted. Percentages for each group breakdown is based on each respective N and not the total 366 patients. c 4 patients that left our hospital were not followed up in terms of hospital LOS. b
predominantly superficial burns, 29.8% having deep partialthickness burns and 36.8% sustaining full-thickness burns. In terms of location, admitted patients had more head and lower extremity burns than their discharged counterparts. More patients in the admission group had >10% BSA burnt compared to discharged patients (22.4% vs 2.9% respectively, p0.05), and most of the discharged patients suffered burns that were 5% of BSA (90.4%). Finally, ABA scores were more frequently “major” in the admitted group (17.2% vs 3.6%) and “minor” in the discharged group (82.8% vs 50%). (Table 3).
3.5. 3.4.
304 (98.7) 4 (1.3)
Predictors of hospital admission
Admitted vs. discharged patients: burn characteristics
Admitted patients had higher proportions of more severe mechanisms of injury such as flame (20.7% vs 10.1%), electrical (12.1% vs 2.4%) and chemical (6.9% vs 3.1%) burns. They also had a more uniform distribution of burn degrees as compared to those discharged, with 33.3% of admitted patients having
Table 4 presents the variables that were found to be statistically significant predictors of hospital admission in our burn population. With respect to demographic predictors: Age was found to be statistically significant, as patients at the extremes of ages (<5 years and 65 years) had increased odds of being admitted when compared to the 15–65 years age
Please cite this article in press as: A.D. Gilbert, et al., Epidemiology of burn patients presenting to a tertiary hospital emergency department in Lebanon, Burns (2017), http://dx.doi.org/10.1016/j.burns.2017.06.015
JBUR 5312 No. of Pages 8
4
burns xxx (2017) xxx –xxx
Table 2 – Total population burn characteristics (N=366). Mechanism of injury no. (%) Flame Scald Contact with hot object Electrical Chemical Other
41 (11.9) 186 (53.9) 58 (16.8) 14 (4.1) 13 (3.8) 33 (9.6)
Degree of burn no. (%)a Superficialb Deep partial-thickness Full-thickness
297 (82.0) 37 (10.2) 28 (7.7)
Total body surface area no. (%) 5% >5–10% >10%
304 (84.4) 37 (10.3) 19 (5.3)
Location of burn no. (%) Head Upper extremity Lower extremity Chest, back and abdomen Perineum
68 (18.6) 212 (57.9) 112 (30.6) 65 (17.8) 14 (3.8)
ABA severity score no. (%) Minor Moderate Major
127 (44.7) 46 (16.2) 111 (39.1)
a Degree of burn was described for the most clinically relevant burn during each encounter. b Superficial burns include first degree and superficial partialthickness burns.
category. Meanwhile, in terms of burn characteristics: there was a trend for higher odds of admission if the location of the burn was the head, chest/back/abdomen or lower limbs. Patients with increasing total body surface area affected such as 5–10% BSA and >10% had 3.67 (95% CI: 1.08–12.47) and 5.45 (95% CI: 1.01–29.52) increased odds of admission compared to patients with <5% BSA respectively. When the mechanism of burn was considered, more violent injuries were strong predictors of admission: Flame (OR=5.83, 95% CI (1.54–22.12)), electric (OR=19.15, 95% CI (3.02–121.41)) and chemical (OR=18.83, 95% CI (2.78–127.57)) injuries were all statistically significant and positively correlated with admission. Finally, with increasing degrees of burn there was higher odds of being admitted. Deep partial thickness burns held a 10.57 (95% CI: 3.42–32.70) higher odds of admission taking superficial injuries as controls, and full-thickness burns were twice that at 26.48 (95% CI: 4.67–73.72).
4.
Discussion
Burn is noted by the WHO to be among the most devastating of human injuries and is considered a leading cause of preventable injuries, especially in low and middle-income countries [11]. Epidemiological studies are required for future management planning and prevention strategies on national and international levels. In Lebanon, there is a lack of organized burn care and data is limited by the paucity of studies. These
studies include a retrospective description of 165 cases of burn patients at the American University of Beirut Medical Center between 1982 and 1987 [12] and another study between 1992 and 2012 looking at 1,524 patients admitted to our national burn center [13]. While other papers have looked at burns admitted to the hospital, this is the first study reviewing all burn patients presenting acutely to a tertiary care academic emergency department during a four-year period. Our study describes characteristics of burn victims treated at an urban ED in a developing country. In terms of demographics, the age groups that most frequently presented for burn were <5 years of age (19.9%) and those between 15–65 years (64.5%), with our observations correlating with international trends [14]. Children are known to be a vulnerable population, and are considered to be at a higher risk for burns due to their curiosity and lack of understanding of danger [6]. Likewise, the increase of burns in the adult population can be attributed to the lack of safety measures and awareness in homes and the workplace [15]. With respect to gender, our distribution was female predominant at 57.2% which is similar to the distribution reported in several international studies [15,16]. In contrast, studies from Oman, Brazil and Europe showed a 2:1 male to female ratio [17–19]. This can potentially be explained by women having lower threshold to seek medical care in our Lebanese population. Of note, the literature holds conflicting data regarding gender distribution and the reasons behind male or female predominance [17]. Most patients in our study sustained their injuries at home (70.9%). This is in keeping with a systematic review of unintentional burn cases in South Asia which reported that 80% of burns occurred at home [20]. This has been also echoed by a recent review on the epidemiology on burns in the UK [4]. Additionally, we found that a higher percentage of females than males (76.7% vs 64.5% p=0.01) reported home injuries in a sub-analysis (Supplement 1). This finding correlates with multiple studies noting that most female victims were housewives and burns were sustained in the kitchen [15,18]. In our setting, women contribute to only 27.8% of the workforce per the National Commission for Lebanese Women and spend more time in the household exposing themselves to more accidental scald and flame burns. Most of our population presented to the ED via private transport, mainly with family (57.2%). This correlates with data from Brazil where 54% of cases arrived to the hospital via personal vehicles [18] and from Pakistan where most patients arrived to the ED via non-ambulance transport [17]. The reason for this may be underutilization of the local Emergency Medical Transport (EMS) system and its under preparedness with patients relying on private mode of transport to reach healthcare providers [21]. While considering burn characteristics, the predominant mechanisms of burn were scalding, contact with hot object and flame and were mostly of a superficial degree (82%) with a TBSA affected of <5% (84.4%). These findings correlate with data from Iran, Taiwan and the East Mediterranean region [6,8,11]. However, studies from India, Brazil as well as the international burn database from England and Wales cite contact and flame, followed by scald injuries, as the most common burn etiologies [4,15,18]. This variability can be attributed to varying cultural backgrounds. For example, in the
Please cite this article in press as: A.D. Gilbert, et al., Epidemiology of burn patients presenting to a tertiary hospital emergency department in Lebanon, Burns (2017), http://dx.doi.org/10.1016/j.burns.2017.06.015
JBUR 5312 No. of Pages 8
5
burns xxx (2017) xxx –xxx
Table 3 – Comparison between discharged and admitted patients. Discharged N=308
Admitted N=58
P-value
Male no. (%) Female no. (%)
135 (43.8) 173 (56.2)
38 (65.5) 20 (34.5)
0.002*
Age no. (%) <5 years 5–14 years 15–65 years 65 years
49 (15.9) 33 (10.7) 214 (69.5) 12 (3.9)
24 (41.4) 6 (10.3) 22 (37.9) 6 (10.3)
<0.0001*
Method of transportation no. (%) Self Family EMS
134 (45.4) 157 (53.2) 4 (1.4)
3 (5.2) 45 (77.6) 10 (17.2)
<0.0001*
Injury site no. (%) Home Street Recreation and sport Other
140 (69.7) 30 (14.9) 13 (6.5) 18 (9.0)
40 (75.5) 5 (9.4) 2 (3.8) 6 (11.3)
0.6
Time from injury no. (%) 1 h 2 h–1 day >1 day
107 (45.7) 38 (16.2) 89 (38.0)
20 (40.0) 8 (16.0) 22 (44.0)
0.71
Mechanism of injury no. (%) Flame Scald Contact with hot object Electrical Chemical Other
29 (10.1) 161 (56.1) 55 (19.2) 7 (2.4) 9 (3.1) 26 (9.1)
12 (20.7) 25 (43.1) 3 (5.2) 7 (12.1) 4 (6.9) 7 (12.1)
<0.0001*
Degree of burn no. (%)a Superficialb Deep partial-thickness Full-thickness
278 (91.1) 20 (6.6) 7 (2.3)
19 (33.3) 17 (29.8) 21 (36.8)
<0.0001*
Total body surface area no. (%) 5% >5–10% >10%
273 (90.4) 23 (7.6) 6 (2.0)
31 (53.4) 14 (24.1) 13 (22.4)
<0.0001*
Location of burn no. (%) Head Upper extremity Lower extremity Chest, back and abdomen Perineum
50 (16.2) 177 (57.5) 84 (27.3) 45 (14.6) 9 (2.9)
18 (31.0) 35 (60.3) 28 (48.3) 20 (34.5) 5 (8.6)
0.008* 0.68 0.001* <0.0001* 0.05*
ABA severity score no. (%) Minor Moderate Major
255 (82.8) 42 (13.6) 11 (3.6)
29 (50.0) 19 (32.8) 10 (17.2)
<0.0001*
* a b
p 0.05 considered significant. Degree of burn was described for the most clinically relevant burn during each encounter. Superficial burns include first degree and superficial partial-thickness burns.
East Mediterranean region (North Africa, Middle-East, Gulf region, Afghanistan and Pakistan) and during the winter seasons, there is a common practice of heating water on stoves in households and preparing hot drinks as a way of keeping
warm. The cold seasons also results in households spending more time indoors thus increasing the chance of burn [11]. In terms of disposition from the ED, statistics from the East Mediterranean region report that 112–518 per 100,000 burn
Please cite this article in press as: A.D. Gilbert, et al., Epidemiology of burn patients presenting to a tertiary hospital emergency department in Lebanon, Burns (2017), http://dx.doi.org/10.1016/j.burns.2017.06.015
JBUR 5312 No. of Pages 8
6
burns xxx (2017) xxx –xxx
Table 4 – Multivariate regression: predictors of admission. a
OR (95% CI)
p-value
14.38 (3.88–53.36) 9.46 (1.57–57.07)
<0.0001 0.01
Location of burn Lower limbs
6.22 (2.14–18.07)
0.001
TBSAc >5–10% >10%
3.67 (1.08–12.47) 5.45 (1.01–29.52)
0.04 0.05
Mechanism of injuryd Flame Electrical Chemical
6.19 (1.60–23.85) 23.01 (3.23–163.89) 17.43 (2.33–130.14)
0.008 0.002 0.005
Degree of burne Deep partial-thickness Deep full-thickness
10.57 (3.42–32.70) 18.56 (4.67–73.72)
<0.0001 <0.0001
Variables b
Age <5 years 65 years
a Variables included in the model were: age, gender, method of transportation, degree of burn, location of burn, mechanism of injury, TBSA, ABA score. Only variables found to be statistically significant (p 0.05) are shown. b Reference is 15–65 years of age. c Reference is 5% TBSA. d Reference is scald injury. e Reference is superficial degree.
cases are admitted to the hospital [11]. In our study, 58 out of 366 (14.8%) were admitted to the hospital with 20.6% of admitted cases managed in a critical care unit. This discrepancy in admission rates can be explained by two factors: our tertiary care center’s conservative admission criteria and its status as a referral center for skin graft procedures. To note, our study did not include referrals for skin grafting from other hospitals or health care facilities (estimated at about 38 cases per year). Instead, because our center has a specialized wound management service with a wide range of capabilities including skin grafting, patients who presented to the ED with acute burns requiring wound debridement and grafting were admitted. This may have increased our study’s admission rates. In fact, out of the fifty-eight admitted patients, ten (17.2%) fell into this category. With respect to the length of stay, our results showed an average of 9.5411.85 hospital days and 7.186.76 ICU days. In the literature, hospital LOS was shown to increase with increasing degrees of burns [22]. A sub-analysis of our admitted patients echoed this trend with increasing days spent in hospital moving from superficial to deep partial to full-thickness burns (4.005.32days vs 8.06 6.80days vs 13.8115.95days respectively p=0.003) (Supplement 2). Additionally, out of the 366 patients included in the study, only one mortality was noted and was attributed to traumatic injuries associated with the burn. This number is not equivalent to the regional average. In fact, as per WHO data, in the Mediterranean region, burns are responsible for 5.6 deaths per 100,000 cases. Mortality rates found in the literature are significantly higher in some studies [14,15,19]. This could be related to multiple factors, including differences in hospital management, study setting, TBSA and burn depths in the other
study populations. Lebanon has a quasi-organized system for burn management with only one designated burn center which has a higher capacity for larger and deeper burns, and a dedicated burn critical care unit with specialized staff. In Lebanon, there are no pre-hospital criteria for admission of a patient to the burn center and, unfortunately, no official assessment of needs for burn centers in terms of capacity and number has been conducted to date. Thus, not all cases are referred to this center, and end up admitted to local hospitals due to lack of capacity in the burn center or a better organization and/or appeal in the other hospitals. In our population, 84.4% of the burns described had a surface area of <5%, which are considered minor and are associated with low mortality. Across the world, especially in high-income countries, there is a decline in ED visits and burn mortality with countries such as Australia describing a 0.7–1.5 per 100,000 deaths. In the East Mediterranean region, a study from Pakistan echoed our results with a majority of patients being discharged from the ED and a low total mortality rate of 6.5% [9]. On the other hand, in India, one study noted a 45% mean TBSA burnt with a mortality rate was 83.3% among cases, with TBSA involvement between 61 and 80%, while a 100% mortality was observed in the cases with >80% TBSA involvement [15]. However it must be noted that the latter article focused on suicide and homicide patients with related burns which could have inflated the mortality beyond what is expected in the case of an unintentional burn. The literature also focuses on high-risk, severely injured, hospitalized burn patients. A 2010 European systematic review estimated the mortality rate of this sub-group of patients between 1.4% and 18% [23]. In our study, these critically ill patients were poorly represented. We attribute this to the relatively affluent community that our tertiary care center cares for and to direct referrals to the burn center that bypass our hospital or to non-transport of burn related dead patients from scene. However, considering our data, a total of 10 out of 58 admitted (17.2%) patients suffered from “major burns” as defined by ABA. Of these patients, 1 passed away, yielding a mortality rate of 10%, which is consistent with the aforementioned systematic review. Thus, when considering the total population, our mortality rates seem low (due to the multiple reasons discussed above) but are in line with international literature when high-risk admitted patients are considered. Finally, our study identified predictors of hospital admission in all burn patients presenting to the ED. When compared to discharged patients, admitted patients tended to belong to the more extreme age categories (as compared to ages 15–65 years), have a higher TBSA of 5–10% and >10% (as compared to <5%), be subject to more violent mechanism of action (electrical, chemical, flame) and have higher degrees of burn. In fact, the most predictive variable of hospital admission was the degree of burn with 10.57 higher odds of admission for deep partial-thickness compared to superficial thickness burns and almost double the odds of admission for full-thickness burn patients (OR 18.56). This was expected, since the evidence behind this finding is well demonstrated in the literature, with high degree and large surface area burns requiring shock resuscitation, aggressive nutritional support, and specialized care; care provided only in hospital settings and dedicated burn care centers [24]. The greater likelihood of admission for burns for extremes of ages can be attributed to
Please cite this article in press as: A.D. Gilbert, et al., Epidemiology of burn patients presenting to a tertiary hospital emergency department in Lebanon, Burns (2017), http://dx.doi.org/10.1016/j.burns.2017.06.015
JBUR 5312 No. of Pages 8
burns xxx (2017) xxx –xxx
their vulnerability, delayed inflammatory responses and comorbidities [4,25] and portrays the cautious attitude of healthcare providers when managing these populations. Moreover, regional data supports our findings regarding mechanism of injury, and shows that when flame is the cause of burn, patients are more likely to be managed in-hospital compared to patients with scalds burns since they were associated with larger and deeper burns and occur frequently with inhalation injuries [7,11].
4.1.
studies. Future larger epidemiologic studies are required to better quantify the impact of burns in Lebanon. Such studies would allow more refined approaches to the development of preventive programs and burn management algorithms based on clinically relevant variables as well as the establishment of a national burn registry. We hope our paper adds to the growing literature of burn patients, specifically those that present to the emergency department and sheds light on the relevant predictors of hospital admission in this setting.
Limitations
This was a retrospective cohort study and the authors are aware of the inherent limitations with possible misclassification and ascertainment bias. The patient charts queried were scanned hand-written charts and thus subject to possible errors in data collection on the part of the researcher. Additionally, ICD-9 coding/discharge diagnoses errors may occur with the electronic health record chart review process. To minimize all these biases, regular meetings with the primary investigator of the study were conducted to standardize the interpretation and collection of the information. Some variables had missing data points such as “Injury site” and “Time of Injury”. In our setting, prehospital clinical data is collected in the ED upon patient arrival and not by emergency medical services (EMS) agencies. This, in addition to the retrospective nature of the study, may have resulted in some missing data. Nevertheless, the authors would like to highlight that this is the first study in Lebanon looking at the epidemiology of ED burn patients and hope it would stimulate future prospective research on the topic, with better structured prehospital data collection in the field, as to avoid the potential missing of data points. The 4 patients that were transferred from our ED to other hospitals were not followed up in terms of in-hospital mortality and hospital LOS due to logistic difficulty as the hospital data infrastructure in Lebanon is still in development. To note, these patients were transferred mainly due to financial reasons and preference of being treated by their primary care physician in another hospital and not due to increased resource requirement or increased critical illness. At our institution, ED care and admission for unstable patients are ensured regardless of financial status. Coordinated transport to appropriate facilities is arranged for patients who are unable to afford admission and who are deemed stable for transport. Additionally, all other data points were collected for these patients as they were first resuscitated in our emergency department. The population sample was also taken from a single tertiary care center which may affect the generalizability of the results, especially that major burns were under-represented. Our center is however the largest tertiary care center in Lebanon, acting as a referral center for patients from all over Lebanon and the region.
5.
7
Conclusion
Characteristics of burn patients treated in an urban ED in Lebanon mirror those of burn victims in other international
Conflict of interest Authors have no conflict of interest to disclose.
Funding This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors
Authors’ contribution GAD and MES have made substantial contributions to conception and design of the study. CEK, ER, AM and MM for acquisition of data, analysis and interpretation of data. CEK, ER and RBC have been involved in drafting the manuscript. CEK, GAD, RBC and MES for revising manuscript critically for important intellectual content. All authors contributed substantially to its revision. GAD and MES take responsibility for the paper as a whole
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.burns.2017. 06.015. REFERENCES
[1] Global Burden of Disease: 2004 Update. World Health Organization Website http://www.who.int/healthinfo/ global_burden_disease/2004_report_update/en Published 2004. Updated 2008 [Accessed 5 February 2014]. [2] Burn Incidence and Treatment in the U.S.: 2013 Fact Sheet. American Burn Association Website. https://www.ameriburn. org/resources_factsheet.php Published 2013 [Accessed 30 December 2013]. [3] Smith S, Duncan M, Mobley J, Kagan R. Emergency room management of minor burn injuries: a quality management evaluation. J Burn Care Rehabil 1997;18(1 Pt. 1):76–80. [4] Stylianou N, Buchan I, Dunn KW. A review of the international Burn Injury Database (iBID) for England and Wales: descriptive analysis of burn injuries 2003–2011. BMJ Open 2015;5(2): e006184. [5] Wasiak J, Spinks A, Ashby K, Clapperton A, Cleland H, Gabbe B. The epidemiology of burn injuries in an Australian setting, 2000–2006. Burns 2009;35(8):1124–32.
Please cite this article in press as: A.D. Gilbert, et al., Epidemiology of burn patients presenting to a tertiary hospital emergency department in Lebanon, Burns (2017), http://dx.doi.org/10.1016/j.burns.2017.06.015
JBUR 5312 No. of Pages 8
8
burns xxx (2017) xxx –xxx
[6] Chien WC, Pai L, Lin CC, Chen HC. Epidemiology of hospitalized burns patients in Taiwan. Burns 2003;29(6):582–8. [7] Avsarogullari L, Sozuer E, Ikizceli I, Kekec Z, Yurumez Y, Ozkan S. Adult burn injuries in an Emergency Department in Central Anatolia, Turkey: a 5-year analysis. Burns 2003;29(6):571–7. [8] Ansari-Lari M, Askarian M. Epidemiology of burns presenting to an emergency department in Shiraz, South Iran. Burns 2003;29(6):579–81. [9] Siddiqui E, Zia N, Feroze A, Awan S, Ali A, Razzak J, et al. Burn injury characteristics: findings from Pakistan National Emergency Department Surveillance Study. BMC Emerg Med 2015;15(Suppl. 2):S5. [10] Singer AJ, Brebbia J, Soroff HH. Management of local burn wounds in the ED. Am J Emerg Med 2007;25(6):666–71. [11] Othman N, Kendrick D. Epidemiology of burn injuries in the East Mediterranean Region: a systematic review. BMC Public Health 2010;10:83. [12] Jiz F, Kadourra I, Saba M. Statistical restrospective analysis of burn patients admitted to AUBMC between 1982–87 (towards a changing future in burn management). Ann MBC 1989;2:1. [13] Ghanime G, Rizkallah N, Said JM. Epidemiology of major burns at the Lebanese Burn Center in Geitawi, Lebanon. Ann Burns Fire Disasters 2013;26(2):59–62. [14] Peck MD. Epidemiology of burns throughout the world. Part I: distribution and risk factors. Burns 2011;37(7):1087–100. [15] Kumar N, Kanchan T, Unnikrishnan B, Rekha T, Mithra P, Venugopal A, et al. Clinico-epidemiological profile of burn patients admitted in a tertiary care hospital in coastal South India. J Burn Care Res 2012;33(5):660–7. [16] Chen SH, Chen YC, Chen TJ, Ma H. Epidemiology of burns in Taiwan: a nationwide report including inpatients and outpatients. Burns 2014;40(7):1397–405.
[17] Al-Shaqsi S, Al-Kashmiri A, Al-Bulushi T. Epidemiology of burns undergoing hospitalization to the National Burns Unit in the Sultanate of Oman: a 25-year review. Burns 2013;39 (8):1606–11. [18] Gawryszewski P, Bernal RT, Silva NN, Morais Neto OL, Silva MM, Mascarenhas MD, et al. Atendimentos decorrentes de queimaduras em serviços públicos de emergência no Brasil. Cad Saúde Pública 2012;28(4):629–40. [19] Brusselaers N, Monstrey S, Vogelaers D, Hoste E, Blot S. Severe burn injury in Europe: a systematic review of the incidence, etiology, morbidity, and mortality. Crit Care 2010;14 (5):R188. [20] Golshan A, Patel C, Hyder AA. A systematic review of the epidemiology of unintentional burn injuries in South Asia. J Public Health (Oxf) 2013;35(3):384–96. [21] El Sayed M, Tamim H, Chehadeh AA, Kazzi AA. Emergency medical services utilization in EMS priority conditions in Beirut, Lebanon. Prehosp Disaster Med 2016;31(6):621–7. [22] Sierra Zuniga MF, Castro Delgado OE, Merchan-Galvis AM, Caicedo JC, Calvache JA, Delgado-Noguera M. Factors associated with length of hospital stay in minor and moderate burns at Popayan, Colombia. Analysis of a cohort study. Burns 2016;42(1):190–5. [23] Brusselaers N, Monstrey S, Vogelaers D, Hoste E, Blot S. Severe burn injury in Europe: a systematic review of the incidence, etiology, morbidity, and mortality. Crit Care 2010;14(5):R188. [24] ISBI Practice Guidelines Committee, Steering Subcommittee, Advisory Subcommittee. ISBI practice guidelines for burn care. Burns 2016;42(5):953–1021. [25] Jeschke MG, Pinto R, Costford SR, Amini-Nik S. Threshold age and burn size associated with poor outcomes in the elderly after burn injury. Burns 2016;42(2):276–81.
Please cite this article in press as: A.D. Gilbert, et al., Epidemiology of burn patients presenting to a tertiary hospital emergency department in Lebanon, Burns (2017), http://dx.doi.org/10.1016/j.burns.2017.06.015
burns xxx (2017) xxx –xxx
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/burns
Epidemiology of burn patients presenting to a tertiary hospital emergency department in Lebanon Abou Dagher Gilbert a, Eva Rajha a, Christopher El Khuri a, Ralph Bou Chebl a, Aurelie Mailhac b , Maha Makki b, Mazen El Sayed a, * a b
Department of Emergency Medicine, American University of Beirut Medical Center, Beirut, Lebanon Clinical Research Institute, American University of Beirut Medical Center, Beirut, Lebanon
article info
abstract
Article history:
Background: The epidemiology of burns is well described in the United States, Europe and
Accepted 24 June 2017
Asia. However, few studies address this topic in the Levant region. This study aims to
Available online xxx
describe characteristics of burn victims treated at the emergency department of a tertiary care center in Beirut, Lebanon and to report on factors that affect ED disposition.
Keywords: Emergency department Burn ED disposition Levant
Methods: A retrospective cohort study was conducted in the ED of a tertiary care hospital, in Beirut, Lebanon, between 2009 and 2013. Patients were recruited if their final ED diagnosis was burn. A descriptive analysis of patients’ characteristics including burns was done, followed by a bivariate analysis to compare two groups (admitted vs discharged). Statistical analysis included the use of Student t-test and Pearson chi-square where appropriate. A multivariate analysis was then conducted to determine the predictors of hospital admission. Results: A total of 366 patients had their charts queried. Age category distributions of the patients were 73 (19.9%) <5 years, 39 (10.7%) 5–14 years, 236 (64.5%) 15–65 years and 18 (4.9%) >65 years of age. Around half of the patients (47.3%) were males, with scalding being the most common mechanism of burn (53.9%), followed by contact with hot object (16.8%) and flame (11.9%). In terms of disposition from the ED, 58 (15.8%) patients were admitted to the hospital with 42 (72.4%) going to GPU, 12 (20.6%) to ICU and 4(6.9%) transferred to either another hospital or to an acute burn facility. Admitted patients tended to be at the extremes of ages (<5 years or 65 years), male dominant, more likely to be brought in by family, with higher proportions of more severe mechanisms of injury (flame, electric, chemical). Admitted patients also sustained higher degrees of burns with more TBSA injured. Predictors of hospital admission included the aforementioned factors with the strongest predictors of admission being full-thickness degree burn (OR 18.56 (4.67–73.72 CI95% p 0.001)) and mechanism of injury such as electrical (OR 23.01 (3.23–163.89 CI95% p=0.002)) and chemical (OR 17.43 (2.33–130.14 CI95% p=0.003)).
Abbreviations: ABA, American Burn Association; ED, emergency department; EHR, electronic health record; EMS, emergency medical services; GPU, general practitioner unit; ICD, International Statistical Classification of Diseases and Related Health Problems; ICU, intensive care unit; TBSA, total body surface area; WHO, World Health Organization. * Corresponding author. E-mail address: [email protected] (M. El Sayed). http://dx.doi.org/10.1016/j.burns.2017.06.015 0305-4179/© 2017 Elsevier Ltd and ISBI. All rights reserved.
Please cite this article in press as: A.D. Gilbert, et al., Epidemiology of burn patients presenting to a tertiary hospital emergency department in Lebanon, Burns (2017), http://dx.doi.org/10.1016/j.burns.2017.06.015
JBUR 5312 No. of Pages 8
2
burns xxx (2017) xxx –xxx
Conclusion: Characteristics of burn patients treated in an urban ED in Lebanon mirror those of burn victims in other international studies. Future larger epidemiologic studies are needed to better quantify the impact of burns in Lebanon. © 2017 Elsevier Ltd and ISBI. All rights reserved.
1.
Introduction
The World Health Organization (WHO) global burden of disease data, updated in 2008, reports that approximately 11 million people per year sustain burns requiring medical attention. The report places burn as the fourth most common substantial injury after road traffic accidents, falls and interpersonal violence [1]. In the US, approximately 450,000 individuals require medical treatment for burns per year, with most of these patients treated in the emergency department (ED), and discharged with outpatient follow-up [2,3]. On a global scale, 90% of all burns occur in low and middle income countries [1]. There is a growing international literature regarding the epidemiology of burns in the US, Europe and Asia [2,4–6]. Pakistan, Turkey and Iran have contributed to the database of burn patients in the Levant region [7–9] with studies reporting on burn patients that present to emergency departments and not just to dedicated burn centers. Studies pertaining to burns are lacking in Lebanon. This study was conducted in a tertiary care center in Lebanon to contribute to the international and local literature, by describing characteristics of burn victims treated at the emergency department and by reporting on factors that affect ED disposition.
calculation was performed. However, given that multiple studies for burn patients presenting to emergency departments recruited between 300 and 400 patients, we aimed to recruit a similar number of patients [6–9].
2.4.
Trained research fellows extracted available clinical information from scanned charts and electronic laboratory reports: Patients’ demographic information, past medical history and burn characteristics (etiology, injury site, circumstance, suspected motive, time from injury, degree, percentage surface area affected and degree of burn) were extracted. Disposition from the ED, length of stay in the ED, general practitioner unit (GPU), and intensive care unit (ICU) were noted. If the patient was transferred to another hospital or the country’s specialized burn center, the patient was considered “admitted” though physically discharged from our hospital. This is due to the similar clinical profile shared by those admitted to our hospital and those requiring hospital treatment elsewhere. The reason In-hospital mortality was collected as well. Finally, the American Burn Association (ABA) severity score was calculated for all burns as per the definition set by the ABA [10].
2.5.
2.
Methods
2.1.
Study design and setting
This is an IRB approved (ER.GA.06), single center, retrospective cohort study conducted in the ED of the American University of Beirut Medical Center (AUBMC). AUBMC is the largest tertiary care center in Lebanon with over 56,000 ED visits per year. All patients presenting from July 2009 to December 2013 had their medical records queried via the hospital’s Electronic Health Record (EHR).
2.2.
Patient selection
Patients’ ED encounters were filtered by an experienced data user using the hospital’s EHR via structured keyword searches and ICD-9 coding (International Statistical Classification of Diseases and Related Health Problems). Patients were included in the study if their charts included the final ED diagnosis of “Burn” regardless of severity. No age restriction was implemented; and no exclusion criteria were used.
2.3.
Sample size calculation
This study serves as a foundation for future epidemiologic studies of burn patients in Lebanon. No formal sample size
Variables
Statistical analysis
Statistical analyses were performed using SPSS version 24.0 (Armonk, NY: IBM Corp). The distributions of the continuous and categorical variables were presented as meanstandard deviation and frequency/percentages respectively. Selected characteristics were then stratified by admission status (Discharged or Admitted) and variable differences between the two groups were calculated by Pearson’s chi square/ Fischer test and Student’s t-test where appropriate. Tests were interpreted at a significance level alpha=0.05. A multivariate logistic regression analysis was performed to ascertain the predictive factors of hospital admission in the population. A backward selection procedure, with significance level for variable removal from the model set at 0.1, was conducted. The independent variables chosen for modelling were those found to be significant at the bivariate analysis level in addition to those considered clinically meaningful. The variables included in the model were: Age, gender, method of transportation, degree of burn, location of burn, mechanism of injury, Total Body Surface Area (TBSA), and ABA score. The results were described as odds ratios (OR) and their corresponding 95% confidence intervals (CI). A sub-analysis was performed to describe the association between gender and injury site. All patients were stratified by gender and a Pearson’s chi square/Fischer test was performed where appropriate (Supplement 1). A second sub-analysis was done to evaluate the association between the degree of burns
Please cite this article in press as: A.D. Gilbert, et al., Epidemiology of burn patients presenting to a tertiary hospital emergency department in Lebanon, Burns (2017), http://dx.doi.org/10.1016/j.burns.2017.06.015
JBUR 5312 No. of Pages 8
3
burns xxx (2017) xxx –xxx
and hospital length of stay (LOS). All admitted patients were stratified by degree of burn and their respective hospital LOS averages were compared by one-way ANOVA (Supplement 2).
3.
Results
3.1.
Total patient demographics and disposition
A total of 366 patients were included in this study. The age distribution of the population was 73 (19.9%) <5 years, 39 (10.7%) 5–14 years, 236 (64.5%) 15–65 years and 18 (4.9%) >65 years of age, with an age range of 1month–89years. Around half (47.3%) of patients were male, with most patients presenting to the ED with family (57.2%) and coming from home (70.9%). Additionally, the mean time from injury to ED presentation was 33.9473.81min with most patients either presenting within 1h (44.7%) or after 24h (39.1%). In terms of disposition from the ED, 58 (15.8%) patients were admitted to the hospital, including 42 (72.4%) to the GPU, 12 (20.6%) to ICU and 4 (6.9%) to either another hospital or an acute burn facility. The mean length of stay (LOS) in the ICU and hospital were found to be 7.186.76 and 9.5411.85days respectively. Of those discharged from the ED, most were sent home, with only a few patients leaving against medical advice (AMA). From the 366 patients included in the study, one in-hospital ICU mortality was noted. The data is summarized in Table 1.
3.2.
Total population burn characteristics
The most common mechanism of burn was scalding (53.9%) followed by contact with hot object (16.8%) and flame (11.9%). The degree of burn ranged from superficial (82%) to deep partial thickness (10.2%) to full-thickness (7.7%). The total body surface area affected was mostly <5% and accounted for 84.4% of all injuries in this study. In terms of location of the burn, the upper extremity was the most affected (57.9%) followed by the lower extremity (30.6%) and the head (18.6%). The ABA severity score for all burns was calculated and found to have the following distribution: Minor (77.6%), moderate (10.2%) and major (5.7%) (Table 2).
3.3.
Admitted vs. discharged patients: demographics
When compared to discharged patients, admitted patients tended to belong more to the youngest (<5 years group 41.4% vs 15.9%) and eldest (65 years group 10.3% vs 3.9%) age distributions. Additionally, admitted patients were predominantly males and a higher proportion of them were brought in by family compared to discharged patients (77.6% vs 53.2%). Injury site and time from injury to presentation did not differ between the two groups (Table 3).
Table 1 – Total population demographics (N=366). Male no. (%) Female no. (%)
173 (47.3) 193 (52.7)
Age no. (%) <5 years 5–14 years 15–65 years 65 years
73 (19.9) 39 (10.7) 236 (64.5) 18 (4.9)
Method of transportation no. (%) Self Family EMS
137 (38.8) 202 (57.2) 14 (4.0)
Injury site no. (%) Home Street Recreation and sport Other
180 (70.9) 35 (13.8) 15 (5.9) 24 (9.4)
Time from injury no. (%) 1 h 2 h–1 day >1 day
127 (44.7) 46 (16.2) 111 (39.1)
Disposition no. (%)a Discharged (N =308)b Home Left AMA Admitted (N =58)b CCU GPU Transferred to another hospital Transferred to an acute burn facility LOS in days N=54 (Mean SD)c Hospital ICU
12 (20.6) 42 (72.4) 3 (5.1) 1 (1.8)
9.54 11.85 7.18 6.76
a
One ICU mortality was noted. Percentages for each group breakdown is based on each respective N and not the total 366 patients. c 4 patients that left our hospital were not followed up in terms of hospital LOS. b
predominantly superficial burns, 29.8% having deep partialthickness burns and 36.8% sustaining full-thickness burns. In terms of location, admitted patients had more head and lower extremity burns than their discharged counterparts. More patients in the admission group had >10% BSA burnt compared to discharged patients (22.4% vs 2.9% respectively, p0.05), and most of the discharged patients suffered burns that were 5% of BSA (90.4%). Finally, ABA scores were more frequently “major” in the admitted group (17.2% vs 3.6%) and “minor” in the discharged group (82.8% vs 50%). (Table 3).
3.5. 3.4.
304 (98.7) 4 (1.3)
Predictors of hospital admission
Admitted vs. discharged patients: burn characteristics
Admitted patients had higher proportions of more severe mechanisms of injury such as flame (20.7% vs 10.1%), electrical (12.1% vs 2.4%) and chemical (6.9% vs 3.1%) burns. They also had a more uniform distribution of burn degrees as compared to those discharged, with 33.3% of admitted patients having
Table 4 presents the variables that were found to be statistically significant predictors of hospital admission in our burn population. With respect to demographic predictors: Age was found to be statistically significant, as patients at the extremes of ages (<5 years and 65 years) had increased odds of being admitted when compared to the 15–65 years age
Please cite this article in press as: A.D. Gilbert, et al., Epidemiology of burn patients presenting to a tertiary hospital emergency department in Lebanon, Burns (2017), http://dx.doi.org/10.1016/j.burns.2017.06.015
JBUR 5312 No. of Pages 8
4
burns xxx (2017) xxx –xxx
Table 2 – Total population burn characteristics (N=366). Mechanism of injury no. (%) Flame Scald Contact with hot object Electrical Chemical Other
41 (11.9) 186 (53.9) 58 (16.8) 14 (4.1) 13 (3.8) 33 (9.6)
Degree of burn no. (%)a Superficialb Deep partial-thickness Full-thickness
297 (82.0) 37 (10.2) 28 (7.7)
Total body surface area no. (%) 5% >5–10% >10%
304 (84.4) 37 (10.3) 19 (5.3)
Location of burn no. (%) Head Upper extremity Lower extremity Chest, back and abdomen Perineum
68 (18.6) 212 (57.9) 112 (30.6) 65 (17.8) 14 (3.8)
ABA severity score no. (%) Minor Moderate Major
127 (44.7) 46 (16.2) 111 (39.1)
a Degree of burn was described for the most clinically relevant burn during each encounter. b Superficial burns include first degree and superficial partialthickness burns.
category. Meanwhile, in terms of burn characteristics: there was a trend for higher odds of admission if the location of the burn was the head, chest/back/abdomen or lower limbs. Patients with increasing total body surface area affected such as 5–10% BSA and >10% had 3.67 (95% CI: 1.08–12.47) and 5.45 (95% CI: 1.01–29.52) increased odds of admission compared to patients with <5% BSA respectively. When the mechanism of burn was considered, more violent injuries were strong predictors of admission: Flame (OR=5.83, 95% CI (1.54–22.12)), electric (OR=19.15, 95% CI (3.02–121.41)) and chemical (OR=18.83, 95% CI (2.78–127.57)) injuries were all statistically significant and positively correlated with admission. Finally, with increasing degrees of burn there was higher odds of being admitted. Deep partial thickness burns held a 10.57 (95% CI: 3.42–32.70) higher odds of admission taking superficial injuries as controls, and full-thickness burns were twice that at 26.48 (95% CI: 4.67–73.72).
4.
Discussion
Burn is noted by the WHO to be among the most devastating of human injuries and is considered a leading cause of preventable injuries, especially in low and middle-income countries [11]. Epidemiological studies are required for future management planning and prevention strategies on national and international levels. In Lebanon, there is a lack of organized burn care and data is limited by the paucity of studies. These
studies include a retrospective description of 165 cases of burn patients at the American University of Beirut Medical Center between 1982 and 1987 [12] and another study between 1992 and 2012 looking at 1,524 patients admitted to our national burn center [13]. While other papers have looked at burns admitted to the hospital, this is the first study reviewing all burn patients presenting acutely to a tertiary care academic emergency department during a four-year period. Our study describes characteristics of burn victims treated at an urban ED in a developing country. In terms of demographics, the age groups that most frequently presented for burn were <5 years of age (19.9%) and those between 15–65 years (64.5%), with our observations correlating with international trends [14]. Children are known to be a vulnerable population, and are considered to be at a higher risk for burns due to their curiosity and lack of understanding of danger [6]. Likewise, the increase of burns in the adult population can be attributed to the lack of safety measures and awareness in homes and the workplace [15]. With respect to gender, our distribution was female predominant at 57.2% which is similar to the distribution reported in several international studies [15,16]. In contrast, studies from Oman, Brazil and Europe showed a 2:1 male to female ratio [17–19]. This can potentially be explained by women having lower threshold to seek medical care in our Lebanese population. Of note, the literature holds conflicting data regarding gender distribution and the reasons behind male or female predominance [17]. Most patients in our study sustained their injuries at home (70.9%). This is in keeping with a systematic review of unintentional burn cases in South Asia which reported that 80% of burns occurred at home [20]. This has been also echoed by a recent review on the epidemiology on burns in the UK [4]. Additionally, we found that a higher percentage of females than males (76.7% vs 64.5% p=0.01) reported home injuries in a sub-analysis (Supplement 1). This finding correlates with multiple studies noting that most female victims were housewives and burns were sustained in the kitchen [15,18]. In our setting, women contribute to only 27.8% of the workforce per the National Commission for Lebanese Women and spend more time in the household exposing themselves to more accidental scald and flame burns. Most of our population presented to the ED via private transport, mainly with family (57.2%). This correlates with data from Brazil where 54% of cases arrived to the hospital via personal vehicles [18] and from Pakistan where most patients arrived to the ED via non-ambulance transport [17]. The reason for this may be underutilization of the local Emergency Medical Transport (EMS) system and its under preparedness with patients relying on private mode of transport to reach healthcare providers [21]. While considering burn characteristics, the predominant mechanisms of burn were scalding, contact with hot object and flame and were mostly of a superficial degree (82%) with a TBSA affected of <5% (84.4%). These findings correlate with data from Iran, Taiwan and the East Mediterranean region [6,8,11]. However, studies from India, Brazil as well as the international burn database from England and Wales cite contact and flame, followed by scald injuries, as the most common burn etiologies [4,15,18]. This variability can be attributed to varying cultural backgrounds. For example, in the
Please cite this article in press as: A.D. Gilbert, et al., Epidemiology of burn patients presenting to a tertiary hospital emergency department in Lebanon, Burns (2017), http://dx.doi.org/10.1016/j.burns.2017.06.015
JBUR 5312 No. of Pages 8
5
burns xxx (2017) xxx –xxx
Table 3 – Comparison between discharged and admitted patients. Discharged N=308
Admitted N=58
P-value
Male no. (%) Female no. (%)
135 (43.8) 173 (56.2)
38 (65.5) 20 (34.5)
0.002*
Age no. (%) <5 years 5–14 years 15–65 years 65 years
49 (15.9) 33 (10.7) 214 (69.5) 12 (3.9)
24 (41.4) 6 (10.3) 22 (37.9) 6 (10.3)
<0.0001*
Method of transportation no. (%) Self Family EMS
134 (45.4) 157 (53.2) 4 (1.4)
3 (5.2) 45 (77.6) 10 (17.2)
<0.0001*
Injury site no. (%) Home Street Recreation and sport Other
140 (69.7) 30 (14.9) 13 (6.5) 18 (9.0)
40 (75.5) 5 (9.4) 2 (3.8) 6 (11.3)
0.6
Time from injury no. (%) 1 h 2 h–1 day >1 day
107 (45.7) 38 (16.2) 89 (38.0)
20 (40.0) 8 (16.0) 22 (44.0)
0.71
Mechanism of injury no. (%) Flame Scald Contact with hot object Electrical Chemical Other
29 (10.1) 161 (56.1) 55 (19.2) 7 (2.4) 9 (3.1) 26 (9.1)
12 (20.7) 25 (43.1) 3 (5.2) 7 (12.1) 4 (6.9) 7 (12.1)
<0.0001*
Degree of burn no. (%)a Superficialb Deep partial-thickness Full-thickness
278 (91.1) 20 (6.6) 7 (2.3)
19 (33.3) 17 (29.8) 21 (36.8)
<0.0001*
Total body surface area no. (%) 5% >5–10% >10%
273 (90.4) 23 (7.6) 6 (2.0)
31 (53.4) 14 (24.1) 13 (22.4)
<0.0001*
Location of burn no. (%) Head Upper extremity Lower extremity Chest, back and abdomen Perineum
50 (16.2) 177 (57.5) 84 (27.3) 45 (14.6) 9 (2.9)
18 (31.0) 35 (60.3) 28 (48.3) 20 (34.5) 5 (8.6)
0.008* 0.68 0.001* <0.0001* 0.05*
ABA severity score no. (%) Minor Moderate Major
255 (82.8) 42 (13.6) 11 (3.6)
29 (50.0) 19 (32.8) 10 (17.2)
<0.0001*
* a b
p 0.05 considered significant. Degree of burn was described for the most clinically relevant burn during each encounter. Superficial burns include first degree and superficial partial-thickness burns.
East Mediterranean region (North Africa, Middle-East, Gulf region, Afghanistan and Pakistan) and during the winter seasons, there is a common practice of heating water on stoves in households and preparing hot drinks as a way of keeping
warm. The cold seasons also results in households spending more time indoors thus increasing the chance of burn [11]. In terms of disposition from the ED, statistics from the East Mediterranean region report that 112–518 per 100,000 burn
Please cite this article in press as: A.D. Gilbert, et al., Epidemiology of burn patients presenting to a tertiary hospital emergency department in Lebanon, Burns (2017), http://dx.doi.org/10.1016/j.burns.2017.06.015
JBUR 5312 No. of Pages 8
6
burns xxx (2017) xxx –xxx
Table 4 – Multivariate regression: predictors of admission. a
OR (95% CI)
p-value
14.38 (3.88–53.36) 9.46 (1.57–57.07)
<0.0001 0.01
Location of burn Lower limbs
6.22 (2.14–18.07)
0.001
TBSAc >5–10% >10%
3.67 (1.08–12.47) 5.45 (1.01–29.52)
0.04 0.05
Mechanism of injuryd Flame Electrical Chemical
6.19 (1.60–23.85) 23.01 (3.23–163.89) 17.43 (2.33–130.14)
0.008 0.002 0.005
Degree of burne Deep partial-thickness Deep full-thickness
10.57 (3.42–32.70) 18.56 (4.67–73.72)
<0.0001 <0.0001
Variables b
Age <5 years 65 years
a Variables included in the model were: age, gender, method of transportation, degree of burn, location of burn, mechanism of injury, TBSA, ABA score. Only variables found to be statistically significant (p 0.05) are shown. b Reference is 15–65 years of age. c Reference is 5% TBSA. d Reference is scald injury. e Reference is superficial degree.
cases are admitted to the hospital [11]. In our study, 58 out of 366 (14.8%) were admitted to the hospital with 20.6% of admitted cases managed in a critical care unit. This discrepancy in admission rates can be explained by two factors: our tertiary care center’s conservative admission criteria and its status as a referral center for skin graft procedures. To note, our study did not include referrals for skin grafting from other hospitals or health care facilities (estimated at about 38 cases per year). Instead, because our center has a specialized wound management service with a wide range of capabilities including skin grafting, patients who presented to the ED with acute burns requiring wound debridement and grafting were admitted. This may have increased our study’s admission rates. In fact, out of the fifty-eight admitted patients, ten (17.2%) fell into this category. With respect to the length of stay, our results showed an average of 9.5411.85 hospital days and 7.186.76 ICU days. In the literature, hospital LOS was shown to increase with increasing degrees of burns [22]. A sub-analysis of our admitted patients echoed this trend with increasing days spent in hospital moving from superficial to deep partial to full-thickness burns (4.005.32days vs 8.06 6.80days vs 13.8115.95days respectively p=0.003) (Supplement 2). Additionally, out of the 366 patients included in the study, only one mortality was noted and was attributed to traumatic injuries associated with the burn. This number is not equivalent to the regional average. In fact, as per WHO data, in the Mediterranean region, burns are responsible for 5.6 deaths per 100,000 cases. Mortality rates found in the literature are significantly higher in some studies [14,15,19]. This could be related to multiple factors, including differences in hospital management, study setting, TBSA and burn depths in the other
study populations. Lebanon has a quasi-organized system for burn management with only one designated burn center which has a higher capacity for larger and deeper burns, and a dedicated burn critical care unit with specialized staff. In Lebanon, there are no pre-hospital criteria for admission of a patient to the burn center and, unfortunately, no official assessment of needs for burn centers in terms of capacity and number has been conducted to date. Thus, not all cases are referred to this center, and end up admitted to local hospitals due to lack of capacity in the burn center or a better organization and/or appeal in the other hospitals. In our population, 84.4% of the burns described had a surface area of <5%, which are considered minor and are associated with low mortality. Across the world, especially in high-income countries, there is a decline in ED visits and burn mortality with countries such as Australia describing a 0.7–1.5 per 100,000 deaths. In the East Mediterranean region, a study from Pakistan echoed our results with a majority of patients being discharged from the ED and a low total mortality rate of 6.5% [9]. On the other hand, in India, one study noted a 45% mean TBSA burnt with a mortality rate was 83.3% among cases, with TBSA involvement between 61 and 80%, while a 100% mortality was observed in the cases with >80% TBSA involvement [15]. However it must be noted that the latter article focused on suicide and homicide patients with related burns which could have inflated the mortality beyond what is expected in the case of an unintentional burn. The literature also focuses on high-risk, severely injured, hospitalized burn patients. A 2010 European systematic review estimated the mortality rate of this sub-group of patients between 1.4% and 18% [23]. In our study, these critically ill patients were poorly represented. We attribute this to the relatively affluent community that our tertiary care center cares for and to direct referrals to the burn center that bypass our hospital or to non-transport of burn related dead patients from scene. However, considering our data, a total of 10 out of 58 admitted (17.2%) patients suffered from “major burns” as defined by ABA. Of these patients, 1 passed away, yielding a mortality rate of 10%, which is consistent with the aforementioned systematic review. Thus, when considering the total population, our mortality rates seem low (due to the multiple reasons discussed above) but are in line with international literature when high-risk admitted patients are considered. Finally, our study identified predictors of hospital admission in all burn patients presenting to the ED. When compared to discharged patients, admitted patients tended to belong to the more extreme age categories (as compared to ages 15–65 years), have a higher TBSA of 5–10% and >10% (as compared to <5%), be subject to more violent mechanism of action (electrical, chemical, flame) and have higher degrees of burn. In fact, the most predictive variable of hospital admission was the degree of burn with 10.57 higher odds of admission for deep partial-thickness compared to superficial thickness burns and almost double the odds of admission for full-thickness burn patients (OR 18.56). This was expected, since the evidence behind this finding is well demonstrated in the literature, with high degree and large surface area burns requiring shock resuscitation, aggressive nutritional support, and specialized care; care provided only in hospital settings and dedicated burn care centers [24]. The greater likelihood of admission for burns for extremes of ages can be attributed to
Please cite this article in press as: A.D. Gilbert, et al., Epidemiology of burn patients presenting to a tertiary hospital emergency department in Lebanon, Burns (2017), http://dx.doi.org/10.1016/j.burns.2017.06.015
JBUR 5312 No. of Pages 8
burns xxx (2017) xxx –xxx
their vulnerability, delayed inflammatory responses and comorbidities [4,25] and portrays the cautious attitude of healthcare providers when managing these populations. Moreover, regional data supports our findings regarding mechanism of injury, and shows that when flame is the cause of burn, patients are more likely to be managed in-hospital compared to patients with scalds burns since they were associated with larger and deeper burns and occur frequently with inhalation injuries [7,11].
4.1.
studies. Future larger epidemiologic studies are required to better quantify the impact of burns in Lebanon. Such studies would allow more refined approaches to the development of preventive programs and burn management algorithms based on clinically relevant variables as well as the establishment of a national burn registry. We hope our paper adds to the growing literature of burn patients, specifically those that present to the emergency department and sheds light on the relevant predictors of hospital admission in this setting.
Limitations
This was a retrospective cohort study and the authors are aware of the inherent limitations with possible misclassification and ascertainment bias. The patient charts queried were scanned hand-written charts and thus subject to possible errors in data collection on the part of the researcher. Additionally, ICD-9 coding/discharge diagnoses errors may occur with the electronic health record chart review process. To minimize all these biases, regular meetings with the primary investigator of the study were conducted to standardize the interpretation and collection of the information. Some variables had missing data points such as “Injury site” and “Time of Injury”. In our setting, prehospital clinical data is collected in the ED upon patient arrival and not by emergency medical services (EMS) agencies. This, in addition to the retrospective nature of the study, may have resulted in some missing data. Nevertheless, the authors would like to highlight that this is the first study in Lebanon looking at the epidemiology of ED burn patients and hope it would stimulate future prospective research on the topic, with better structured prehospital data collection in the field, as to avoid the potential missing of data points. The 4 patients that were transferred from our ED to other hospitals were not followed up in terms of in-hospital mortality and hospital LOS due to logistic difficulty as the hospital data infrastructure in Lebanon is still in development. To note, these patients were transferred mainly due to financial reasons and preference of being treated by their primary care physician in another hospital and not due to increased resource requirement or increased critical illness. At our institution, ED care and admission for unstable patients are ensured regardless of financial status. Coordinated transport to appropriate facilities is arranged for patients who are unable to afford admission and who are deemed stable for transport. Additionally, all other data points were collected for these patients as they were first resuscitated in our emergency department. The population sample was also taken from a single tertiary care center which may affect the generalizability of the results, especially that major burns were under-represented. Our center is however the largest tertiary care center in Lebanon, acting as a referral center for patients from all over Lebanon and the region.
5.
7
Conclusion
Characteristics of burn patients treated in an urban ED in Lebanon mirror those of burn victims in other international
Conflict of interest Authors have no conflict of interest to disclose.
Funding This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors
Authors’ contribution GAD and MES have made substantial contributions to conception and design of the study. CEK, ER, AM and MM for acquisition of data, analysis and interpretation of data. CEK, ER and RBC have been involved in drafting the manuscript. CEK, GAD, RBC and MES for revising manuscript critically for important intellectual content. All authors contributed substantially to its revision. GAD and MES take responsibility for the paper as a whole
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.burns.2017. 06.015. REFERENCES
[1] Global Burden of Disease: 2004 Update. World Health Organization Website http://www.who.int/healthinfo/ global_burden_disease/2004_report_update/en Published 2004. Updated 2008 [Accessed 5 February 2014]. [2] Burn Incidence and Treatment in the U.S.: 2013 Fact Sheet. American Burn Association Website. https://www.ameriburn. org/resources_factsheet.php Published 2013 [Accessed 30 December 2013]. [3] Smith S, Duncan M, Mobley J, Kagan R. Emergency room management of minor burn injuries: a quality management evaluation. J Burn Care Rehabil 1997;18(1 Pt. 1):76–80. [4] Stylianou N, Buchan I, Dunn KW. A review of the international Burn Injury Database (iBID) for England and Wales: descriptive analysis of burn injuries 2003–2011. BMJ Open 2015;5(2): e006184. [5] Wasiak J, Spinks A, Ashby K, Clapperton A, Cleland H, Gabbe B. The epidemiology of burn injuries in an Australian setting, 2000–2006. Burns 2009;35(8):1124–32.
Please cite this article in press as: A.D. Gilbert, et al., Epidemiology of burn patients presenting to a tertiary hospital emergency department in Lebanon, Burns (2017), http://dx.doi.org/10.1016/j.burns.2017.06.015
JBUR 5312 No. of Pages 8
8
burns xxx (2017) xxx –xxx
[6] Chien WC, Pai L, Lin CC, Chen HC. Epidemiology of hospitalized burns patients in Taiwan. Burns 2003;29(6):582–8. [7] Avsarogullari L, Sozuer E, Ikizceli I, Kekec Z, Yurumez Y, Ozkan S. Adult burn injuries in an Emergency Department in Central Anatolia, Turkey: a 5-year analysis. Burns 2003;29(6):571–7. [8] Ansari-Lari M, Askarian M. Epidemiology of burns presenting to an emergency department in Shiraz, South Iran. Burns 2003;29(6):579–81. [9] Siddiqui E, Zia N, Feroze A, Awan S, Ali A, Razzak J, et al. Burn injury characteristics: findings from Pakistan National Emergency Department Surveillance Study. BMC Emerg Med 2015;15(Suppl. 2):S5. [10] Singer AJ, Brebbia J, Soroff HH. Management of local burn wounds in the ED. Am J Emerg Med 2007;25(6):666–71. [11] Othman N, Kendrick D. Epidemiology of burn injuries in the East Mediterranean Region: a systematic review. BMC Public Health 2010;10:83. [12] Jiz F, Kadourra I, Saba M. Statistical restrospective analysis of burn patients admitted to AUBMC between 1982–87 (towards a changing future in burn management). Ann MBC 1989;2:1. [13] Ghanime G, Rizkallah N, Said JM. Epidemiology of major burns at the Lebanese Burn Center in Geitawi, Lebanon. Ann Burns Fire Disasters 2013;26(2):59–62. [14] Peck MD. Epidemiology of burns throughout the world. Part I: distribution and risk factors. Burns 2011;37(7):1087–100. [15] Kumar N, Kanchan T, Unnikrishnan B, Rekha T, Mithra P, Venugopal A, et al. Clinico-epidemiological profile of burn patients admitted in a tertiary care hospital in coastal South India. J Burn Care Res 2012;33(5):660–7. [16] Chen SH, Chen YC, Chen TJ, Ma H. Epidemiology of burns in Taiwan: a nationwide report including inpatients and outpatients. Burns 2014;40(7):1397–405.
[17] Al-Shaqsi S, Al-Kashmiri A, Al-Bulushi T. Epidemiology of burns undergoing hospitalization to the National Burns Unit in the Sultanate of Oman: a 25-year review. Burns 2013;39 (8):1606–11. [18] Gawryszewski P, Bernal RT, Silva NN, Morais Neto OL, Silva MM, Mascarenhas MD, et al. Atendimentos decorrentes de queimaduras em serviços públicos de emergência no Brasil. Cad Saúde Pública 2012;28(4):629–40. [19] Brusselaers N, Monstrey S, Vogelaers D, Hoste E, Blot S. Severe burn injury in Europe: a systematic review of the incidence, etiology, morbidity, and mortality. Crit Care 2010;14 (5):R188. [20] Golshan A, Patel C, Hyder AA. A systematic review of the epidemiology of unintentional burn injuries in South Asia. J Public Health (Oxf) 2013;35(3):384–96. [21] El Sayed M, Tamim H, Chehadeh AA, Kazzi AA. Emergency medical services utilization in EMS priority conditions in Beirut, Lebanon. Prehosp Disaster Med 2016;31(6):621–7. [22] Sierra Zuniga MF, Castro Delgado OE, Merchan-Galvis AM, Caicedo JC, Calvache JA, Delgado-Noguera M. Factors associated with length of hospital stay in minor and moderate burns at Popayan, Colombia. Analysis of a cohort study. Burns 2016;42(1):190–5. [23] Brusselaers N, Monstrey S, Vogelaers D, Hoste E, Blot S. Severe burn injury in Europe: a systematic review of the incidence, etiology, morbidity, and mortality. Crit Care 2010;14(5):R188. [24] ISBI Practice Guidelines Committee, Steering Subcommittee, Advisory Subcommittee. ISBI practice guidelines for burn care. Burns 2016;42(5):953–1021. [25] Jeschke MG, Pinto R, Costford SR, Amini-Nik S. Threshold age and burn size associated with poor outcomes in the elderly after burn injury. Burns 2016;42(2):276–81.
Please cite this article in press as: A.D. Gilbert, et al., Epidemiology of burn patients presenting to a tertiary hospital emergency department in Lebanon, Burns (2017), http://dx.doi.org/10.1016/j.burns.2017.06.015