ORIGINAL ARTICLE
Morbid obesity predisposes trauma patients to worse outcomes: A National Trauma Data Bank analysis Michael Ditillo, DO, Viraj Pandit, MD, Peter Rhee, MD, Hassan Aziz, MD, Steven Hadeed, MPH, Bishwajit Bhattacharya, MD, Randall S. Friese, MD, Kimberly Davis, MD, and Bellal Joseph, MD, Tucson, Arizona
One third of US adults are obese. The impact of obesity on outcomes after blunt traumatic injury has been studied with discrepant results. The aim of our study was to evaluate outcomes in morbidly obese patients after blunt trauma. We hypothesized that morbidly obese patients have adverse outcomes as compared with nonobese patients after blunt traumatic injury. METHODS: We performed a retrospective analysis of all blunt trauma patients (Q18 years) using the National Trauma Data Bank for years 2007 to 2010. Patients with recorded comorbidity of morbid obesity (body mass index Q 40) were identified. Patients transferred, dead on arrival, and with isolated traumatic brain injury were excluded. Propensity score matching was used to match morbidly obese patients to nonYmorbidly obese patients (body mass index G 40) in a 1:1 ratio based on age, sex, Injury Severity Score (ISS), Glasgow Coma Scale (GCS), and systolic blood pressure on presentation. The primary outcome was mortality, and the secondary outcome was hospital complications. RESULTS: A total of 32,780 patients (morbidly obese, 16,390; nonobese, 16,390) were included in the study. Morbidly obese patients were more likely to have in-hospital complications (odds ratio [OR], 1.8, 95% confidence interval [CI], 1.6Y1.9), longer hospital stay (OR, 1.2; 95% CI, 1.1Y1.3), and longer intensive care unit stay (OR, 1.15; 95% CI, 1.09Y1.2). The overall mortality rate was 2.8% (n = 851). Mortality was higher in morbidly obese patients compared with the nonobese patients (3.0 vs. 2.2; OR, 1.4; 95% CI, 1.1Y1.5). CONCLUSION: In a cohort of matched patients, morbid obesity is a risk factor for the development of in-hospital complications and mortality after blunt traumatic injury. The results of our study call for attention through focused injury prevention efforts. Future studies are needed to help define the consequences of obesity that influence outcomes. (J Trauma Acute Care Surg. 2014;76: 176Y179. Copyright * 2014 by Lippincott Williams & Wilkins) LEVEL OF EVIDENCE: Prognostic study, level III. KEY WORDS: Obesity; trauma; adverse outcomes in obese patients; mortality in obese patients; obesity and trauma. BACKGROUND:
T
he percentage of obese individuals in the United States has been increasing at an alarming rate. According to a 2012 census by the Center of Disease Control and Prevention, 28.9% of the adult US population was categorized as obese (body mass index [BMI] Q 30).1 Obesity is a complex disease state, which is commonly associated with cardiovascular, metabolic, and endocrine abnormalities.2 The adverse impact of obesity has been extensively studied in the general population; however, with the increasing burden of trauma, the definite impact of morbid obesity on trauma patients still remains unclear. Obesity has been largely considered a public health issue, and the primary focus has been on its prevention and treatment
Submitted: August 2, 2013, Revised: September 21, 2013, Accepted: September 23, 2013. From the Division of Trauma, Critical Care, Burns and Emergency Surgery (V.P., P.R., H.A., S.H., R.S.F., B.J.), Department of Surgery, University of Arizona, Tucson, Arizona; and Division of Trauma, Critical Care, Burns and Emergency Surgery (M.D., B.B., K.D.), Department of Surgery, Yale-New Haven Medical Center, New Haven, Connecticut. This study was presented as a poster at the 72nd annual meeting of the American Association for the Surgery of Trauma, September 18Y21, 2013, in San Francisco, California. Address for reprints: Bellal Joseph, MD, Division of Trauma, Critical Care, Burns and Emergency Surgery, Department of Surgery, University of Arizona, 1501 N. Campbell Ave, Room.5411, PO Box 245063, Tucson, AZ 85724; email: [email protected]. DOI: 10.1097/TA.0b013e3182ab0d7c
176
of associated medical conditions.3 Although obesity is a critical topic, few studies have been published assessing the impact of obesity on outcomes in critically ill and injured surgical patients.4,5 Studies have demonstrated higher rates of in-hospital complications and increased mortality among obese patients after blunt traumatic injury.4Y8 However, all previous studies are single-institution studies with small sample sizes. Furthermore, the lack of reliability in identifying obese patients and lack of a standard BMI cutoff to define obese patients make the generalization of the results from the previous studies difficult. The aim of our study was to evaluate outcomes in morbidly obese trauma patients after blunt trauma using the National Trauma Data Bank (NTDB). We hypothesized that morbidly obese patients have worse outcomes as compared with nonobese patients after blunt traumatic injury.
PATIENTS AND METHODS This is a 4-year (2007Y2010) retrospective analysis of the NTDB (version 7.2). The NTDB is the largest collection of trauma index cases, which is maintained by the American College of Surgeons (Chicago, IL). The NTDB contains data on more than 1.8 million patients, which are contributed by more than 900 trauma centers across the United States. J Trauma Acute Care Surg Volume 76, Number 1
Copyright © 2013 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 76, Number 1
Ditillo et al.
This study includes all patients with blunt traumatic injury and those 18 years and older. Patients transferred from another facility, patients dead on presentation, patients with burn injuries or isolated traumatic brain injury (head Abbreviated Injury Scale [AIS] score 9 3 and other body regions AIS score G 3), and patients with missing data of comorbidities were excluded from our study. We defined blunt mechanism of injury using E-codes. The following data points were abstracted from the NTDB database: demographics (age, sex, race, and ethnicity), vitals on presentation (heart rate [HR], systolic blood pressure [SBP], respiratory rate, and temperature), mechanism of injury, Glasgow Coma Scale (GCS) score on presentation, comorbidities, in-hospital complications, hospital and intensive care unit length of stay, and in-hospital mortality. Patient’s injury characteristics were abstracted using the Injury Severity Score (ISS) and the AIS score. Morbidly obese patients were identified from the NTDB based on the recorded comorbid condition of obesity. The NTDB defines morbid obesity as a BMI of 40 or greater. Propensity scoring matched patients with recorded comorbidity of morbid obesity to nonobese patients in a 1:1 ratio based on the following factors: age, sex, race, SBP, HR, and GCS score on presentation, mechanism of injury, and ISS. Propensity matching is an analog to the process of randomization of a clinical trial that is commonly used in observational studies. The propensity score denotes the conditional probability of an individual to receive a certain treatment. A propensity score is generated for each patient based on all the confounding factors using a logistic regression model. In our study, morbidly obese patients were matched to nonobese patients based on their propensity scores within 0.00001 of the estimated score. We then quantified the accuracy of the model based on the area under the receiver operating characteristic curve. The primary outcome measure was in-hospital mortality. The secondary outcome measure was in-hospital complications. In-hospital complications are defined as infectious complications (sepsis, pneumonia, urinary tract infections), hematologic complications (deep venous thrombosis and disseminated intravascular coagulation), adult respiratory distress syndrome (ARDS), and decubitus ulcer. Data are reported as mean (SD) for continuous variables, median (range) for ordinal variables and as proportions for categorical variables. To explore for the differences between obese and nonobese patients, we performed Mann-Whitney U-test and Student’s t test for continuous variables and W2 test for categorical variables. Univariate analysis was performed to compare the outcomes between obese and nonobese patients. p e 0.05 was considered statistically significant. All statistical analyses were performed using SPSS (version 20, SPSS, Inc., Chicago, IL).
RESULTS A total of 1,699,373 patients with blunt injuries were identified, of whom 32,780 (morbidly obese, 16,390; nonobese, 16,390) were included in the study. The mean (SD) age was 53.35 (19.3) years, 73 % were male, mean (SD) SBP was 130 (27.3) mm Hg, and the median ISS was 9 (range, 4Y13) for all patients. There were no differences between the two groups
in demographics, vitals on presentation, and injury severity parameters. Table 1 highlights the demographics of the study populations. Morbidly obese patients were more likely to have inhospital complications (odds ratio [OR], 1.8; 95% confidence interval [CI], 1.6Y1.9). ARDS ( p = 0.001), urinary tract infections ( p = 0.01), and deep venous thromboembolism ( p = 0.001) were the common complications in the morbidly obese population. Table 2 has also been edited to add the proportion of patients in each group (obese and nonobese) who developed urinary tract infection. After stratifying patients based on age, morbidly obese patients were more likely to have in-hospital complications. Figure 1 highlights the complications in the patients stratified by age. TABLE 1. Patient Characteristic by Group Characteristic Demographics Age, mean (SD), y e25, % 26Y50, % 51Y75, % Q75, % Male, % Race, % White Black Hispanics Intoxication Vital parameters GCS score, median (range) 13Y15 9Y12 e8 ED SBP, mean (SD) Hypotensive (SBP e 90), % ED HR, mean (SD) Tachycardia (HR 9 90), % ED temperature, mean (SD) Injury parameters Mechanism of injury Fall, % MVC, % Head AIS, median (range) Head AIS score Q 3, % ISS, median (range) e8, % 9Y15, % 16Y24, % Q25, %
Morbidly-Obese (n = 16,390) 53.4 (17.3) 13.1 32.5 34 20.4 71
Nonobese (n = 16,390) 53.3 (21) 7.4 34 46.9 11.7 75
48.2 27.4 37.2 20.6 15 (13Y15)
p
45.7 23.6 39.1 18.7
0.8 V V V V 0.7 0.5 V V 0.4 0.8
15 (13Y15)
0.7
94.4 2.5 3.1 141 (26.3) 2.5
94.6 2.1 3.4 140 (26) 3.7
V V V 0.8 V
91.3 (19) 50.5
91.7 (19) 50.1
0.7 V
36.2 (0.7)
36.4 (0.6)
0.6
36.4 53.5 2 (3Y4)
32.7 51.8 2 (3Y4)
0.6 V V 0.7
22.9 9 (4Y13) 42.7 37.7 12.1 8.1
21
V
9 (4Y13) 42.1 38 12.1 7.8
0.5 V V V V
ED, emergency department.
* 2014 Lippincott Williams & Wilkins
Copyright © 2013 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
177
J Trauma Acute Care Surg Volume 76, Number 1
Ditillo et al.
TABLE 2. In-Hospital Complications
TABLE 3. Outcomes
Morbidly Obese Nonbese OR (n = 16,390) (n = 16,390) (95%CI) In-hospital complications, % (1.6Y1.9) ARDS, % (n) (1.5Y3.6) Decubitus Ulcer, % (n) (1.6Y2.9) Infectious complications Pneumonia, % (n) (1.2Y1.7) Surgical site infection, % (n) (1.6Y3.6) Sepsis, % (n) (1.3Y3.45)
9.3 (1,525) 0.01 2.1 (344) 0.001 0.8 (131) 0.01
5.8 (953)
1.8
1 (164)
2.4
0.4 (65)
Morbidly Obese Nonbese (n = 16,390) (n = 16,390)
p
2.1
Hospital LOS, mean (SD) ICU LOS, mean (SD) Ventilation days, mean (SD) Mortality, % (n)
OR (95%CI)
p
8.3 (10.8)
6.1 (8.3)
1.2 (1.1Y1.3)
0.03
7.5 (9.5)
5.8 (7.4)
1.15 (1.09Y1.2) 0.04
6.8 (5.4)
4.1 (6.3)
1.1 (0.9Y1.6)
0.1
3 (491)
2.2 (360)
1.4 (1.1Y1.5)
0.01
ICU, intensive care unit; LOS, length of stay.
1.7 (279) 0.02 0.4 (65) 0.01 0.3 (49) 0.03 1.8 (295)
1.3 (213)
1.4
0.2 (33)
2.3
0.1 (16)
2.1
1.2 (197)
1.8
0.6 (98)
1.9
0.3 (49)
1.5
patients even after stratifying patients based on age. Figure 2 highlights the mortality in the patients stratified based on age.
DISCUSION
Table 3 highlights the outcomes in the study population. Morbidly obese patients were more likely to have a longer hospital stay (OR, 1.2; 95% CI, 1.1Y1.3) and a longer intensive care unit of stay (OR, 1.15; 95% CI, 1.09Y1.2). The overall mortality rate was 2.6% (851 of 32,780). Mortality was higher in the morbidly obese patients (3% [491 of 16,390] vs. 2.2% [360 of 16,390]; OR, 1.4, 95% CI: 1.1Y1.5). Mortality rate remained higher in the morbidly obese
Obesity is a disease state associated with anatomic and physiologic alterations. This study highlights the adverse effects of morbid obesity on outcomes in critically injured trauma patients. Despite similarities in age, sex, mechanism of injury, admission vital signs, and injury severity, morbidly obese patients were more likely to have in-hospital complications and a higher mortality rate when compared with nonobese patients. Our study is the largest national study to assess the impact of morbid obesity in blunt trauma patients and validate the adverse effects of morbid obesity in trauma. BMI is the most accepted and readily applied method for stratifying patients based on weight. The National Institutes of Health and World Health Organization both define obesity as a BMI of 30 kg/m2 or greater and morbidly obese as BMI of 40 kg/m2 or greater, which is the standard definition used in most publications.9,10 However, there is an inconsistency among authors defining the optimal BMI cutoff to assess the impact of obesity in trauma patients. Choban et al.11 arbitrarily classified obese patients as having BMI of 31 kg/m2 or greater, while Byrnes et al.8 identified obese patients using a BMI score of 35 kg/m2 or greater. The use of unconventional BMI cutoff
Figure 1. Age-stratified mortality in morbidly obese and nonobese patients.
Figure 2. Age-stratified complications in morbidly obese and nonobese patients.
Urinary tract infection, % (n) (1.2Y3.1) Thromboembolic complications DVT, % (n) (1.5Y2.5) PE, % (n) (1.1Y2.2)
0.01
1.2 (197) 0.001 0.4 (65) 0.04
DVT, deep venous thrombosis; PE, pulmonary embolism.
178
* 2014 Lippincott Williams & Wilkins
Copyright © 2013 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 76, Number 1
for defining obesity has caused misleading results and raised questions about the generalizability of the results. In our study, we used BMI of 40 kg/m2 greater to define morbid obesity as reported by the NTDB. Using a BMI of 40 kg/m2 greater, we were able to demonstrate the adverse effects of obesity on outcomes in a national sample of matched trauma patients. The impact of obesity on mortality after blunt traumatic injury has been previously assessed by several studies.6Y8,11,12 However, the true impact of morbid obesity on mortality in patients with blunt trauma still remains unknown owing to the wide variability in the results of previously published studies. An increase in mortality among obese patients ranging from 1.5-folds to 6-folds has been previously reported.7,8,11,12 Our study adds to the growing literature demonstrating the adverse effect of morbid obesity in trauma patients. In this study, morbidly obese patients were 1.5 times more likely to die after a blunt injury compared with a similar cohort of matched nonobese patients. This wide variability in the results among published literature can be contributed to the retrospective nature, unmatched and mixed-patient population, and small sample sizes of these studies. In the current study, patients were matched for age, mechanism of injury, and injury severity, and patients with isolated traumatic brain injury were excluded to make our study population homogeneous. We believe that understanding the altered physiologic response to injury and the interplay of inflammatory mechanism in obese individuals would help manage obese trauma patients better and also help us assess the true impact of obesity on mortality after blunt traumatic injury. Obesity has been linked to an increased risk of morbidity in nonsurgical and nontrauma patients. The same is true in trauma patients where studies have highlighted the impact of obesity on in-hospital complications. Neville et al.6 and Brown et al.7 demonstrated that obese patients had a higher likelihood of developing in-hospital complications compared with nonobese patients. The current study demonstrates that morbidly obese patients were two times more likely to have an in-hospital complication compared with a similar cohort of nonobese patients. ARDS and pneumonia were the most common complications among morbidly obese patients. Similarly, Brown et al.7 also demonstrated higher incidence of ARDS in obese patients in comparison with nonobese patients. Even after stratifying based on age, morbidly obese individuals were more likely to have in-hospital complications. Studies have demonstrated that the proinflammatory milieu of cytokines and adipokines along with altered monocyte and macrophage functions in obese patients is responsible for the higher incidence of complications in this cohort of patients.9 We believe that future studies should focus on understanding the molecular and biochemical impact of obesity to help improve outcomes in obese trauma patients. The results of our study should be interpreted with the acknowledgement of the limitations of the NTDB. We were able only to assess the data points recorded in the NTDB and were not able to control for the missing data in the NTDB. We defined morbid obesity based on the reported comorbidity of
Ditillo et al.
obesity in the NTDB. We were not able to calculate the actual BMI of each patient because this is not reported in the NTDB. Despite these limitations, our study is the largest series demonstrating the adverse effect of morbid obesity in blunt trauma patients.
CONCLUSION In a cohort of matched patients, morbid obesity is a risk factor for the development of in-hospital complications and mortality after blunt traumatic injury. The results of our study call for attention through focused injury prevention efforts. Future studies are needed to help define the consequences of morbid obesity that influence outcomes. AUTHORSHIP M.D., V.P., P.R, R.S.F., K.D., and B.J. designed this study. M.D., V.P., H.A., B.B., and B.J. searched the literature. M.D., V.P., H.A., B.B., R.S.F., K.D. and B.J. abstracted the data. M.D., V.P., H.A., R.S.F., and B.J. analyzed the data. All other authors participated in the data interpretation and manuscript preparation.
DISCLOSURE The authors declare no conflicts of interest.
REFERENCES 1. Fryar C, Carroll M, Ogden C. Prevalence of Overweight, Obesity, and Extreme Obesity Among Adults: United States, Trends 1960Y1962 Through 2009Y2010. Center of Disease Control and Prevention. 2. Flegal KM, Carroll MD, Ogden CL, Johnson CL. Prevalence and trends in obesity among US adults, 1999Y2000. JAMA. 2002;288(14):1723Y1727. 3. Hubert HB, Feinleib M, McNamara PM, Castelli W.P. Obesity as an independent risk factor for cardiovascular disease: a 26-year follow-up of participants in the Framingham Heart Study. Circulation. 1983;67: 968Y977. 4. Boulanger BR, Milzman DP, Rodriguez A. Obesity. Crit Care Clin. 1994;10:613Y622. 5. Kisat M, Villegas CV, Onguti S, Zafar S.N, Latif A, David T, Haut E.R, Schneider E.B, Lipsett P.A, Zafar H, Haider A. Predictors of sepsis in moderately severely injured patients: an analysis of the National Trauma Data Bank. Surg Infect (Larchmt). 2013;14(1):62Y68. 6. Neville A, Brown CV, Weng J, Demetriades D, Velmahos G. Obesity is an independent risk factor for mortality in severely injured blunt trauma patients. Arch Surg. 2004;139:983Y987. 7. Brown CV, Neville A, Rhee P, Salim A, Velmahos G, Demetriades D. The impact of obesity on the outcomes of 1,153 critically injured blunt trauma patients. J Trauma. 2005;59:1048Y1051. 8. Byrnes MC, Mcdaniel MD, Moore MB, Helmer SD, Smith RS. The effect of obesity on outcomes among injured patients. J Trauma. 2005; 58:232Y237. 9. Winfield R, Bochicchio G. The critically injured obese patient: a review and a look ahead. J Am Coll Surg. 2013;216(6):1193Y1206. 10. Brown C, Velmahos G. The consequence of obesity on trauma, emergency surgery, and surgical critical care. World J Emerg Surg. 2006;1:27. 11. Choban PS, Weireter LJ Jr, Maynes C. Obesity and increased mortality in blunt trauma. J Trauma. 1991;31(9):1253Y1257. 12. Arbabi S, Wahl WL, Hemmila MR, Kohoyda-Inglis C. The cushion effect. J Trauma. 2003;54(6):1090Y1093.
* 2014 Lippincott Williams & Wilkins
Copyright © 2013 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
179
Morbid obesity predisposes trauma patients to worse outcomes: A National Trauma Data Bank analysis Michael Ditillo, DO, Viraj Pandit, MD, Peter Rhee, MD, Hassan Aziz, MD, Steven Hadeed, MPH, Bishwajit Bhattacharya, MD, Randall S. Friese, MD, Kimberly Davis, MD, and Bellal Joseph, MD, Tucson, Arizona
One third of US adults are obese. The impact of obesity on outcomes after blunt traumatic injury has been studied with discrepant results. The aim of our study was to evaluate outcomes in morbidly obese patients after blunt trauma. We hypothesized that morbidly obese patients have adverse outcomes as compared with nonobese patients after blunt traumatic injury. METHODS: We performed a retrospective analysis of all blunt trauma patients (Q18 years) using the National Trauma Data Bank for years 2007 to 2010. Patients with recorded comorbidity of morbid obesity (body mass index Q 40) were identified. Patients transferred, dead on arrival, and with isolated traumatic brain injury were excluded. Propensity score matching was used to match morbidly obese patients to nonYmorbidly obese patients (body mass index G 40) in a 1:1 ratio based on age, sex, Injury Severity Score (ISS), Glasgow Coma Scale (GCS), and systolic blood pressure on presentation. The primary outcome was mortality, and the secondary outcome was hospital complications. RESULTS: A total of 32,780 patients (morbidly obese, 16,390; nonobese, 16,390) were included in the study. Morbidly obese patients were more likely to have in-hospital complications (odds ratio [OR], 1.8, 95% confidence interval [CI], 1.6Y1.9), longer hospital stay (OR, 1.2; 95% CI, 1.1Y1.3), and longer intensive care unit stay (OR, 1.15; 95% CI, 1.09Y1.2). The overall mortality rate was 2.8% (n = 851). Mortality was higher in morbidly obese patients compared with the nonobese patients (3.0 vs. 2.2; OR, 1.4; 95% CI, 1.1Y1.5). CONCLUSION: In a cohort of matched patients, morbid obesity is a risk factor for the development of in-hospital complications and mortality after blunt traumatic injury. The results of our study call for attention through focused injury prevention efforts. Future studies are needed to help define the consequences of obesity that influence outcomes. (J Trauma Acute Care Surg. 2014;76: 176Y179. Copyright * 2014 by Lippincott Williams & Wilkins) LEVEL OF EVIDENCE: Prognostic study, level III. KEY WORDS: Obesity; trauma; adverse outcomes in obese patients; mortality in obese patients; obesity and trauma. BACKGROUND:
T
he percentage of obese individuals in the United States has been increasing at an alarming rate. According to a 2012 census by the Center of Disease Control and Prevention, 28.9% of the adult US population was categorized as obese (body mass index [BMI] Q 30).1 Obesity is a complex disease state, which is commonly associated with cardiovascular, metabolic, and endocrine abnormalities.2 The adverse impact of obesity has been extensively studied in the general population; however, with the increasing burden of trauma, the definite impact of morbid obesity on trauma patients still remains unclear. Obesity has been largely considered a public health issue, and the primary focus has been on its prevention and treatment
Submitted: August 2, 2013, Revised: September 21, 2013, Accepted: September 23, 2013. From the Division of Trauma, Critical Care, Burns and Emergency Surgery (V.P., P.R., H.A., S.H., R.S.F., B.J.), Department of Surgery, University of Arizona, Tucson, Arizona; and Division of Trauma, Critical Care, Burns and Emergency Surgery (M.D., B.B., K.D.), Department of Surgery, Yale-New Haven Medical Center, New Haven, Connecticut. This study was presented as a poster at the 72nd annual meeting of the American Association for the Surgery of Trauma, September 18Y21, 2013, in San Francisco, California. Address for reprints: Bellal Joseph, MD, Division of Trauma, Critical Care, Burns and Emergency Surgery, Department of Surgery, University of Arizona, 1501 N. Campbell Ave, Room.5411, PO Box 245063, Tucson, AZ 85724; email: [email protected]. DOI: 10.1097/TA.0b013e3182ab0d7c
176
of associated medical conditions.3 Although obesity is a critical topic, few studies have been published assessing the impact of obesity on outcomes in critically ill and injured surgical patients.4,5 Studies have demonstrated higher rates of in-hospital complications and increased mortality among obese patients after blunt traumatic injury.4Y8 However, all previous studies are single-institution studies with small sample sizes. Furthermore, the lack of reliability in identifying obese patients and lack of a standard BMI cutoff to define obese patients make the generalization of the results from the previous studies difficult. The aim of our study was to evaluate outcomes in morbidly obese trauma patients after blunt trauma using the National Trauma Data Bank (NTDB). We hypothesized that morbidly obese patients have worse outcomes as compared with nonobese patients after blunt traumatic injury.
PATIENTS AND METHODS This is a 4-year (2007Y2010) retrospective analysis of the NTDB (version 7.2). The NTDB is the largest collection of trauma index cases, which is maintained by the American College of Surgeons (Chicago, IL). The NTDB contains data on more than 1.8 million patients, which are contributed by more than 900 trauma centers across the United States. J Trauma Acute Care Surg Volume 76, Number 1
Copyright © 2013 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 76, Number 1
Ditillo et al.
This study includes all patients with blunt traumatic injury and those 18 years and older. Patients transferred from another facility, patients dead on presentation, patients with burn injuries or isolated traumatic brain injury (head Abbreviated Injury Scale [AIS] score 9 3 and other body regions AIS score G 3), and patients with missing data of comorbidities were excluded from our study. We defined blunt mechanism of injury using E-codes. The following data points were abstracted from the NTDB database: demographics (age, sex, race, and ethnicity), vitals on presentation (heart rate [HR], systolic blood pressure [SBP], respiratory rate, and temperature), mechanism of injury, Glasgow Coma Scale (GCS) score on presentation, comorbidities, in-hospital complications, hospital and intensive care unit length of stay, and in-hospital mortality. Patient’s injury characteristics were abstracted using the Injury Severity Score (ISS) and the AIS score. Morbidly obese patients were identified from the NTDB based on the recorded comorbid condition of obesity. The NTDB defines morbid obesity as a BMI of 40 or greater. Propensity scoring matched patients with recorded comorbidity of morbid obesity to nonobese patients in a 1:1 ratio based on the following factors: age, sex, race, SBP, HR, and GCS score on presentation, mechanism of injury, and ISS. Propensity matching is an analog to the process of randomization of a clinical trial that is commonly used in observational studies. The propensity score denotes the conditional probability of an individual to receive a certain treatment. A propensity score is generated for each patient based on all the confounding factors using a logistic regression model. In our study, morbidly obese patients were matched to nonobese patients based on their propensity scores within 0.00001 of the estimated score. We then quantified the accuracy of the model based on the area under the receiver operating characteristic curve. The primary outcome measure was in-hospital mortality. The secondary outcome measure was in-hospital complications. In-hospital complications are defined as infectious complications (sepsis, pneumonia, urinary tract infections), hematologic complications (deep venous thrombosis and disseminated intravascular coagulation), adult respiratory distress syndrome (ARDS), and decubitus ulcer. Data are reported as mean (SD) for continuous variables, median (range) for ordinal variables and as proportions for categorical variables. To explore for the differences between obese and nonobese patients, we performed Mann-Whitney U-test and Student’s t test for continuous variables and W2 test for categorical variables. Univariate analysis was performed to compare the outcomes between obese and nonobese patients. p e 0.05 was considered statistically significant. All statistical analyses were performed using SPSS (version 20, SPSS, Inc., Chicago, IL).
RESULTS A total of 1,699,373 patients with blunt injuries were identified, of whom 32,780 (morbidly obese, 16,390; nonobese, 16,390) were included in the study. The mean (SD) age was 53.35 (19.3) years, 73 % were male, mean (SD) SBP was 130 (27.3) mm Hg, and the median ISS was 9 (range, 4Y13) for all patients. There were no differences between the two groups
in demographics, vitals on presentation, and injury severity parameters. Table 1 highlights the demographics of the study populations. Morbidly obese patients were more likely to have inhospital complications (odds ratio [OR], 1.8; 95% confidence interval [CI], 1.6Y1.9). ARDS ( p = 0.001), urinary tract infections ( p = 0.01), and deep venous thromboembolism ( p = 0.001) were the common complications in the morbidly obese population. Table 2 has also been edited to add the proportion of patients in each group (obese and nonobese) who developed urinary tract infection. After stratifying patients based on age, morbidly obese patients were more likely to have in-hospital complications. Figure 1 highlights the complications in the patients stratified by age. TABLE 1. Patient Characteristic by Group Characteristic Demographics Age, mean (SD), y e25, % 26Y50, % 51Y75, % Q75, % Male, % Race, % White Black Hispanics Intoxication Vital parameters GCS score, median (range) 13Y15 9Y12 e8 ED SBP, mean (SD) Hypotensive (SBP e 90), % ED HR, mean (SD) Tachycardia (HR 9 90), % ED temperature, mean (SD) Injury parameters Mechanism of injury Fall, % MVC, % Head AIS, median (range) Head AIS score Q 3, % ISS, median (range) e8, % 9Y15, % 16Y24, % Q25, %
Morbidly-Obese (n = 16,390) 53.4 (17.3) 13.1 32.5 34 20.4 71
Nonobese (n = 16,390) 53.3 (21) 7.4 34 46.9 11.7 75
48.2 27.4 37.2 20.6 15 (13Y15)
p
45.7 23.6 39.1 18.7
0.8 V V V V 0.7 0.5 V V 0.4 0.8
15 (13Y15)
0.7
94.4 2.5 3.1 141 (26.3) 2.5
94.6 2.1 3.4 140 (26) 3.7
V V V 0.8 V
91.3 (19) 50.5
91.7 (19) 50.1
0.7 V
36.2 (0.7)
36.4 (0.6)
0.6
36.4 53.5 2 (3Y4)
32.7 51.8 2 (3Y4)
0.6 V V 0.7
22.9 9 (4Y13) 42.7 37.7 12.1 8.1
21
V
9 (4Y13) 42.1 38 12.1 7.8
0.5 V V V V
ED, emergency department.
* 2014 Lippincott Williams & Wilkins
Copyright © 2013 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
177
J Trauma Acute Care Surg Volume 76, Number 1
Ditillo et al.
TABLE 2. In-Hospital Complications
TABLE 3. Outcomes
Morbidly Obese Nonbese OR (n = 16,390) (n = 16,390) (95%CI) In-hospital complications, % (1.6Y1.9) ARDS, % (n) (1.5Y3.6) Decubitus Ulcer, % (n) (1.6Y2.9) Infectious complications Pneumonia, % (n) (1.2Y1.7) Surgical site infection, % (n) (1.6Y3.6) Sepsis, % (n) (1.3Y3.45)
9.3 (1,525) 0.01 2.1 (344) 0.001 0.8 (131) 0.01
5.8 (953)
1.8
1 (164)
2.4
0.4 (65)
Morbidly Obese Nonbese (n = 16,390) (n = 16,390)
p
2.1
Hospital LOS, mean (SD) ICU LOS, mean (SD) Ventilation days, mean (SD) Mortality, % (n)
OR (95%CI)
p
8.3 (10.8)
6.1 (8.3)
1.2 (1.1Y1.3)
0.03
7.5 (9.5)
5.8 (7.4)
1.15 (1.09Y1.2) 0.04
6.8 (5.4)
4.1 (6.3)
1.1 (0.9Y1.6)
0.1
3 (491)
2.2 (360)
1.4 (1.1Y1.5)
0.01
ICU, intensive care unit; LOS, length of stay.
1.7 (279) 0.02 0.4 (65) 0.01 0.3 (49) 0.03 1.8 (295)
1.3 (213)
1.4
0.2 (33)
2.3
0.1 (16)
2.1
1.2 (197)
1.8
0.6 (98)
1.9
0.3 (49)
1.5
patients even after stratifying patients based on age. Figure 2 highlights the mortality in the patients stratified based on age.
DISCUSION
Table 3 highlights the outcomes in the study population. Morbidly obese patients were more likely to have a longer hospital stay (OR, 1.2; 95% CI, 1.1Y1.3) and a longer intensive care unit of stay (OR, 1.15; 95% CI, 1.09Y1.2). The overall mortality rate was 2.6% (851 of 32,780). Mortality was higher in the morbidly obese patients (3% [491 of 16,390] vs. 2.2% [360 of 16,390]; OR, 1.4, 95% CI: 1.1Y1.5). Mortality rate remained higher in the morbidly obese
Obesity is a disease state associated with anatomic and physiologic alterations. This study highlights the adverse effects of morbid obesity on outcomes in critically injured trauma patients. Despite similarities in age, sex, mechanism of injury, admission vital signs, and injury severity, morbidly obese patients were more likely to have in-hospital complications and a higher mortality rate when compared with nonobese patients. Our study is the largest national study to assess the impact of morbid obesity in blunt trauma patients and validate the adverse effects of morbid obesity in trauma. BMI is the most accepted and readily applied method for stratifying patients based on weight. The National Institutes of Health and World Health Organization both define obesity as a BMI of 30 kg/m2 or greater and morbidly obese as BMI of 40 kg/m2 or greater, which is the standard definition used in most publications.9,10 However, there is an inconsistency among authors defining the optimal BMI cutoff to assess the impact of obesity in trauma patients. Choban et al.11 arbitrarily classified obese patients as having BMI of 31 kg/m2 or greater, while Byrnes et al.8 identified obese patients using a BMI score of 35 kg/m2 or greater. The use of unconventional BMI cutoff
Figure 1. Age-stratified mortality in morbidly obese and nonobese patients.
Figure 2. Age-stratified complications in morbidly obese and nonobese patients.
Urinary tract infection, % (n) (1.2Y3.1) Thromboembolic complications DVT, % (n) (1.5Y2.5) PE, % (n) (1.1Y2.2)
0.01
1.2 (197) 0.001 0.4 (65) 0.04
DVT, deep venous thrombosis; PE, pulmonary embolism.
178
* 2014 Lippincott Williams & Wilkins
Copyright © 2013 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 76, Number 1
for defining obesity has caused misleading results and raised questions about the generalizability of the results. In our study, we used BMI of 40 kg/m2 greater to define morbid obesity as reported by the NTDB. Using a BMI of 40 kg/m2 greater, we were able to demonstrate the adverse effects of obesity on outcomes in a national sample of matched trauma patients. The impact of obesity on mortality after blunt traumatic injury has been previously assessed by several studies.6Y8,11,12 However, the true impact of morbid obesity on mortality in patients with blunt trauma still remains unknown owing to the wide variability in the results of previously published studies. An increase in mortality among obese patients ranging from 1.5-folds to 6-folds has been previously reported.7,8,11,12 Our study adds to the growing literature demonstrating the adverse effect of morbid obesity in trauma patients. In this study, morbidly obese patients were 1.5 times more likely to die after a blunt injury compared with a similar cohort of matched nonobese patients. This wide variability in the results among published literature can be contributed to the retrospective nature, unmatched and mixed-patient population, and small sample sizes of these studies. In the current study, patients were matched for age, mechanism of injury, and injury severity, and patients with isolated traumatic brain injury were excluded to make our study population homogeneous. We believe that understanding the altered physiologic response to injury and the interplay of inflammatory mechanism in obese individuals would help manage obese trauma patients better and also help us assess the true impact of obesity on mortality after blunt traumatic injury. Obesity has been linked to an increased risk of morbidity in nonsurgical and nontrauma patients. The same is true in trauma patients where studies have highlighted the impact of obesity on in-hospital complications. Neville et al.6 and Brown et al.7 demonstrated that obese patients had a higher likelihood of developing in-hospital complications compared with nonobese patients. The current study demonstrates that morbidly obese patients were two times more likely to have an in-hospital complication compared with a similar cohort of nonobese patients. ARDS and pneumonia were the most common complications among morbidly obese patients. Similarly, Brown et al.7 also demonstrated higher incidence of ARDS in obese patients in comparison with nonobese patients. Even after stratifying based on age, morbidly obese individuals were more likely to have in-hospital complications. Studies have demonstrated that the proinflammatory milieu of cytokines and adipokines along with altered monocyte and macrophage functions in obese patients is responsible for the higher incidence of complications in this cohort of patients.9 We believe that future studies should focus on understanding the molecular and biochemical impact of obesity to help improve outcomes in obese trauma patients. The results of our study should be interpreted with the acknowledgement of the limitations of the NTDB. We were able only to assess the data points recorded in the NTDB and were not able to control for the missing data in the NTDB. We defined morbid obesity based on the reported comorbidity of
Ditillo et al.
obesity in the NTDB. We were not able to calculate the actual BMI of each patient because this is not reported in the NTDB. Despite these limitations, our study is the largest series demonstrating the adverse effect of morbid obesity in blunt trauma patients.
CONCLUSION In a cohort of matched patients, morbid obesity is a risk factor for the development of in-hospital complications and mortality after blunt traumatic injury. The results of our study call for attention through focused injury prevention efforts. Future studies are needed to help define the consequences of morbid obesity that influence outcomes. AUTHORSHIP M.D., V.P., P.R, R.S.F., K.D., and B.J. designed this study. M.D., V.P., H.A., B.B., and B.J. searched the literature. M.D., V.P., H.A., B.B., R.S.F., K.D. and B.J. abstracted the data. M.D., V.P., H.A., R.S.F., and B.J. analyzed the data. All other authors participated in the data interpretation and manuscript preparation.
DISCLOSURE The authors declare no conflicts of interest.
REFERENCES 1. Fryar C, Carroll M, Ogden C. Prevalence of Overweight, Obesity, and Extreme Obesity Among Adults: United States, Trends 1960Y1962 Through 2009Y2010. Center of Disease Control and Prevention. 2. Flegal KM, Carroll MD, Ogden CL, Johnson CL. Prevalence and trends in obesity among US adults, 1999Y2000. JAMA. 2002;288(14):1723Y1727. 3. Hubert HB, Feinleib M, McNamara PM, Castelli W.P. Obesity as an independent risk factor for cardiovascular disease: a 26-year follow-up of participants in the Framingham Heart Study. Circulation. 1983;67: 968Y977. 4. Boulanger BR, Milzman DP, Rodriguez A. Obesity. Crit Care Clin. 1994;10:613Y622. 5. Kisat M, Villegas CV, Onguti S, Zafar S.N, Latif A, David T, Haut E.R, Schneider E.B, Lipsett P.A, Zafar H, Haider A. Predictors of sepsis in moderately severely injured patients: an analysis of the National Trauma Data Bank. Surg Infect (Larchmt). 2013;14(1):62Y68. 6. Neville A, Brown CV, Weng J, Demetriades D, Velmahos G. Obesity is an independent risk factor for mortality in severely injured blunt trauma patients. Arch Surg. 2004;139:983Y987. 7. Brown CV, Neville A, Rhee P, Salim A, Velmahos G, Demetriades D. The impact of obesity on the outcomes of 1,153 critically injured blunt trauma patients. J Trauma. 2005;59:1048Y1051. 8. Byrnes MC, Mcdaniel MD, Moore MB, Helmer SD, Smith RS. The effect of obesity on outcomes among injured patients. J Trauma. 2005; 58:232Y237. 9. Winfield R, Bochicchio G. The critically injured obese patient: a review and a look ahead. J Am Coll Surg. 2013;216(6):1193Y1206. 10. Brown C, Velmahos G. The consequence of obesity on trauma, emergency surgery, and surgical critical care. World J Emerg Surg. 2006;1:27. 11. Choban PS, Weireter LJ Jr, Maynes C. Obesity and increased mortality in blunt trauma. J Trauma. 1991;31(9):1253Y1257. 12. Arbabi S, Wahl WL, Hemmila MR, Kohoyda-Inglis C. The cushion effect. J Trauma. 2003;54(6):1090Y1093.
* 2014 Lippincott Williams & Wilkins
Copyright © 2013 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
179
