RISK FACTORS FOR HYPOTHERMIA IN EMS-TREATED BURN PATIENTS Matthew D. Weaver, MPH, EMT-P, Jon C. Rittenberger, MD, MS, P. Daniel Patterson, PhD, EMT-P, Serina J. McEntire, PhD, Alain C. Corcos, MD, Jenny A. Ziembicki, MD, David Hostler, PhD, EMT-P
INTRODUCTION
Prehosp Emerg Care Downloaded from informahealthcare.com by Nyu Medical Center on 05/15/15 For personal use only.
ABSTRACT Objective. Hypothermia has been associated with increased mortality in burn patients. We sought to characterize the body temperature of burn patients transported directly to a burn center by emergency medical services (EMS) personnel and identify the factors independently associated with hypothermia. Methods. We utilized prospective data collected by a statewide trauma registry to carry out a nested casecontrol study of burn patients transported by EMS directly to an accredited burn center between 2000 and 2011. Temperature at hospital admission ≤36.5◦ C was defined as hypothermia. We utilized registry data abstracted from prehospital care reports and hospital records in building a multivariable regression model to identify the factors associated with hypothermia. Results. Forty-two percent of the sample was hypothermic. Burns of 20–39% total body surface area (TBSA) (OR 1.44; 1.17–1.79) and ≥40% TBSA (OR 2.39; 1.57–3.64) were associated with hypothermia. Hypothermia was also associated with age > 60 (OR 1.50; 1.30–1.74), polytrauma (OR 1.58; 1.19–2.09), prehospital Glasgow Coma Scale 10% surface area, third
Received June 18 2013 from the Department of Emergency Medicine, University of Pittsburgh, Pittsburgh, Pennsylvania (MDW, JCR, PDP, SJM), UPMC Mercy Hospital Trauma and Burn Services (ACC, JAZ), and Exercise and Nutrition Sciences, SUNY University at Buffalo, Buffalo, New York (DH). Revision received September 20, 2013; accepted for publication October 4, 2013. These data were provided by the Pennsylvania Trauma Systems Foundation, Mechanicsburg, PA. The foundation specifically disclaims responsibility for any analyses, interpretations, or conclusions. The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper. Address correspondence to David Hostler, PhD, Exercise and Nutrition Sciences, SUNY University at Buffalo, 212 Kimball Tower, Buffalo, NY 14214, USA. e-mail: [email protected] doi: 10.3109/10903127.2013.864354
335
336 degree burn, chemical burn, electrical injury, burn of face, hands, feet, perineum, or over major joints, inhalation injury, burn accompanied by significant associated injury or preexisting disease, or suspected child abuse. The registry includes patient demographics, prehospital presenting condition and treatment, process of care measures, diagnoses, major medical procedures, and discharge disposition. Personnel at specialty trauma centers are trained to perform abstraction of the data points from medical records.
Prehosp Emerg Care Downloaded from informahealthcare.com by Nyu Medical Center on 05/15/15 For personal use only.
Inclusion Criteria We limited the dataset to include only subjects transported directly to a PTOS-verified burn center by public safety personnel. Subjects not treated at a burn center, referrals to a burn center from another hospital, and those transported by private vehicle were excluded. Documentation of temperature and burn size was required for inclusion. Records where death was declared in the emergency department or when the patient’s temperature was outside a physiologic range (26.0–42.0◦ C) were excluded, along with records in which vital signs on admission were missing.
Description and Classification of Available Data We explored variables related to patient characteristics, situation and severity of injury, prehospital treatment, and temperature on hospital admission. Age was transformed from a continuous to categorical variable to better illustrate groups of patients at different points along the age spectrum. Temperature was reported in the data set as a continuous variable but was categorized into hypothermic and normothermic. We used the patient’s first temperature as recorded upon admission to the emergency department. We defined a subject as hypothermic if the subject had a temperature on hospital admission ≤36.5◦ C. This has been associated with increased mortality in prior work.16 Total burn surface area (TBSA) was categorized into clinically meaningful categories of 2 in at least two different body regions.17
Statistical Analysis The temperature on admission is presented using descriptive statistics. We examined differences between hypothermic and normothermic subjects using unadjusted logistic regression. We then constructed a multivariate model to identify factors independently associated with hypothermia after adjustment for confounding. The model was built using a backward selection process, using a threshold of 0.20 for entry into the model and 0.15 for exit. Interactions were considered where appropriate. The likelihood of type 1 error was set at 0.05 for all statistical tests, with the standard errors adjusted for clustering of observations at the hospital level. Variance inflation factors and correlation matrices were used to check for multicollinearity. Model fit was assessed using the Hosmer-Lemeshow global test of goodness of fit.
RESULTS The final sample was composed of 2,770 visits to five accredited burn centers (Figure 1). The median age was 42 years (IQR 24–57). Subjects were primarily male (72.3%) and Caucasian (67.6%) (Table 1). The majority were burned on less than 20% of their body area (81.0%), though 7% had at least 40% TBSA. Thermal burns were the most commonly reported type of injury (67.8%). Less than 3% of the sample suffered polytrauma. Ground ambulance was the most common mode of transportation (64.2%).
Temperature The median temperature at admission for all subjects was 36.7◦ C (IQR 36.2–37.0). Nearly half (42%) of the patients in the sample were recorded as ≤36.5◦ C, with 3.5% of the sample below 35◦ C (Figure 2). The proportion subjects who were hypothermic increased with burn size. Thirty-nine percent of patients with 60 Male sex Race White Black Other Unknown Weight category (kg)
INTRODUCTION
Prehosp Emerg Care Downloaded from informahealthcare.com by Nyu Medical Center on 05/15/15 For personal use only.
ABSTRACT Objective. Hypothermia has been associated with increased mortality in burn patients. We sought to characterize the body temperature of burn patients transported directly to a burn center by emergency medical services (EMS) personnel and identify the factors independently associated with hypothermia. Methods. We utilized prospective data collected by a statewide trauma registry to carry out a nested casecontrol study of burn patients transported by EMS directly to an accredited burn center between 2000 and 2011. Temperature at hospital admission ≤36.5◦ C was defined as hypothermia. We utilized registry data abstracted from prehospital care reports and hospital records in building a multivariable regression model to identify the factors associated with hypothermia. Results. Forty-two percent of the sample was hypothermic. Burns of 20–39% total body surface area (TBSA) (OR 1.44; 1.17–1.79) and ≥40% TBSA (OR 2.39; 1.57–3.64) were associated with hypothermia. Hypothermia was also associated with age > 60 (OR 1.50; 1.30–1.74), polytrauma (OR 1.58; 1.19–2.09), prehospital Glasgow Coma Scale 10% surface area, third
Received June 18 2013 from the Department of Emergency Medicine, University of Pittsburgh, Pittsburgh, Pennsylvania (MDW, JCR, PDP, SJM), UPMC Mercy Hospital Trauma and Burn Services (ACC, JAZ), and Exercise and Nutrition Sciences, SUNY University at Buffalo, Buffalo, New York (DH). Revision received September 20, 2013; accepted for publication October 4, 2013. These data were provided by the Pennsylvania Trauma Systems Foundation, Mechanicsburg, PA. The foundation specifically disclaims responsibility for any analyses, interpretations, or conclusions. The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper. Address correspondence to David Hostler, PhD, Exercise and Nutrition Sciences, SUNY University at Buffalo, 212 Kimball Tower, Buffalo, NY 14214, USA. e-mail: [email protected] doi: 10.3109/10903127.2013.864354
335
336 degree burn, chemical burn, electrical injury, burn of face, hands, feet, perineum, or over major joints, inhalation injury, burn accompanied by significant associated injury or preexisting disease, or suspected child abuse. The registry includes patient demographics, prehospital presenting condition and treatment, process of care measures, diagnoses, major medical procedures, and discharge disposition. Personnel at specialty trauma centers are trained to perform abstraction of the data points from medical records.
Prehosp Emerg Care Downloaded from informahealthcare.com by Nyu Medical Center on 05/15/15 For personal use only.
Inclusion Criteria We limited the dataset to include only subjects transported directly to a PTOS-verified burn center by public safety personnel. Subjects not treated at a burn center, referrals to a burn center from another hospital, and those transported by private vehicle were excluded. Documentation of temperature and burn size was required for inclusion. Records where death was declared in the emergency department or when the patient’s temperature was outside a physiologic range (26.0–42.0◦ C) were excluded, along with records in which vital signs on admission were missing.
Description and Classification of Available Data We explored variables related to patient characteristics, situation and severity of injury, prehospital treatment, and temperature on hospital admission. Age was transformed from a continuous to categorical variable to better illustrate groups of patients at different points along the age spectrum. Temperature was reported in the data set as a continuous variable but was categorized into hypothermic and normothermic. We used the patient’s first temperature as recorded upon admission to the emergency department. We defined a subject as hypothermic if the subject had a temperature on hospital admission ≤36.5◦ C. This has been associated with increased mortality in prior work.16 Total burn surface area (TBSA) was categorized into clinically meaningful categories of 2 in at least two different body regions.17
Statistical Analysis The temperature on admission is presented using descriptive statistics. We examined differences between hypothermic and normothermic subjects using unadjusted logistic regression. We then constructed a multivariate model to identify factors independently associated with hypothermia after adjustment for confounding. The model was built using a backward selection process, using a threshold of 0.20 for entry into the model and 0.15 for exit. Interactions were considered where appropriate. The likelihood of type 1 error was set at 0.05 for all statistical tests, with the standard errors adjusted for clustering of observations at the hospital level. Variance inflation factors and correlation matrices were used to check for multicollinearity. Model fit was assessed using the Hosmer-Lemeshow global test of goodness of fit.
RESULTS The final sample was composed of 2,770 visits to five accredited burn centers (Figure 1). The median age was 42 years (IQR 24–57). Subjects were primarily male (72.3%) and Caucasian (67.6%) (Table 1). The majority were burned on less than 20% of their body area (81.0%), though 7% had at least 40% TBSA. Thermal burns were the most commonly reported type of injury (67.8%). Less than 3% of the sample suffered polytrauma. Ground ambulance was the most common mode of transportation (64.2%).
Temperature The median temperature at admission for all subjects was 36.7◦ C (IQR 36.2–37.0). Nearly half (42%) of the patients in the sample were recorded as ≤36.5◦ C, with 3.5% of the sample below 35◦ C (Figure 2). The proportion subjects who were hypothermic increased with burn size. Thirty-nine percent of patients with 60 Male sex Race White Black Other Unknown Weight category (kg)
