Does obesity affect outcomes of adult burn patients?

j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 8 ( 2 0 1 5 ) 4 5 0 e4 5 5

Available online at www.sciencedirect.com

ScienceDirect journal homepage: www.JournalofSurgicalResearch.com

Association for Academic Surgery

Does obesity affect outcomes of adult burn patients? Juliet J. Ray, MD, Shevonne S. Satahoo, MD, Jonathan P. Meizoso, MD, Casey J. Allen, MD, Laura F. Teisch, BS, Kenneth G. Proctor, PhD, Louis R. Pizano, MD, MBA, FACS, Nicholas Namias, MD, MBA, FACS, and Carl I. Schulman, MD, PhD, MSPH, FACS* Divisions of Trauma and Surgical Critical Care, DeWitt Daughtry Family Department of Surgery, University of Miami Miller School of Medicine, Miami, Florida

article info

abstract

Article history:

Background: Obesity negatively affects outcomes after trauma and surgery; results after

Received 8 December 2014

burns are more limited and controversial. The purpose of this study was to determine the

Received in revised form

effect of obesity on clinical and economic outcomes after thermal injury.

9 March 2015

Methods: The National Inpatient Sample was queried for adults from 2005e2009 with

Accepted 18 March 2015

International Classification of Diseases-9 codes for burn injury. Demographics and clinical

Available online 25 March 2015

outcomes of obese and nonobese cohorts were compared. Univariate and multivariate analysis using logistic regression models were performed. Data are expressed as median

Keywords:

(interquartile range) or mean
standard deviation and compared at P < 0.05.

Obesity

Results: In 14,602 patients, 3.3% were obese (body mass index 30 kg/m2). The rate of

Overweight

obesity increased significantly by year (P < 0.001). Univariate analysis revealed significant

BMI

differences between obese and nonobese patients in incidence of wound infection (7.2%

Thermal injury

versus 5.0%), urinary tract infection (7.2% versus 4.6%), deep vein thrombosis in total body

TBSA

surface area (TBSA) 10% (3.1% versus 1.1%), pulmonary embolism in TBSA 10% (2.3%

Outcomes

versus 0.6%), length of stay [6 d (8) versus 5 d (9)], and hospital costs ($10,122.12 [$18,074.72]

DVT

versus $7892.07 [$17,191.96]) (all P < 0.05). Death occurred less frequently in the obese group

PE

(1.9% versus 4%, P ¼ 0.021). Significant predictors of grouped adverse events (urinary tract

Mortality

infection, wound infection, deep vein thrombosis, and pulmonary embolism) on multi-

Morbidity

variate analysis include obesity, TBSA 20%, age, and black race (all P  0.05). Conclusions: Obesity is an independent predictor of adverse events after burn injury; however, obesity is associated with decreased mortality. Our findings highlight the potential clinical and economic impact of the obesity epidemic on burn patients nationwide. ª 2015 Elsevier Inc. All rights reserved.

Portions of this work were presented at the Southern Regional Burn Conference, November 2014 and at the Academic Surgical Congress, February 2015. * Corresponding author. Divisions of Trauma and Surgical Critical Care, DeWitt Daughtry Family Department of Surgery, University of Miami Miller School of Medicine, Ryder Trauma Center, 1800 NW 10th Ave, Suite T 215 (D40), Miami, FL 33136. Tel.: þ1 305 585 1178; fax: þ1 305 326 7065. E-mail address: [email protected] (C.I. Schulman). 0022-4804/$ e see front matter ª 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jss.2015.03.049

j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 8 ( 2 0 1 5 ) 4 5 0 e4 5 5

1.

Introduction

Obesity is a global epidemic that is projected to worsen over the next decade [1]. Affecting over one-third of the U.S. population, this represents a major public health threat with significant clinical and economic implications in medical and surgical populations [2e6]. Physiologic mechanisms, specifically inflammatory and immune mediator regulation, are altered by excess adipose tissue. These altered physiologic mechanisms lead to challenges in managing resuscitation requirements due to the cardiac ramifications of excess weight; respiratory support due to increased ventilation-perfusion mismatch; and thromboprophylaxis due to a potentially amplified prothrombotic state [7]. The impact of obesity in the burn population presents conflicting results [8e11]. Mortality in a cohort of obese patients from the National Burn Repository was 2.6 times higher than that in nonobese patients, and obese patients were 4.1 times more likely to have a length of stay (LOS) of 7 d [12]. A recent study showed extremely obese patients had LOS almost double that of nonobese patients and a mortality of 36.4% [13]. In contrast, by a study in pediatrics reported no difference in mortality between obese and nonobese patients (11% versus 8%) [10]. Ghanem et al. [9] evaluated 95 patients with burns >15% total body surface area (TBSA) and stratified obesity into “moderate,” “severe,” and “morbid” based on the World Health Organization classifications of body mass index (BMI). A BMI of 35, representing severe and morbid obesity, was a “tilt point” for higher than predicted mortality. Another study stratified adult patients by level of obesity and showed improved survival in the mild obesity group [14]. Obesity has also been associated with higher incidence of sepsis in adult, [11] but not pediatric burns [10]. A 20-y review of morbidly obese burn patients found a 43% incidence of fatal pulmonary embolism (PE) [8]. The purpose of this study was to determine the effect of obesity (BMI 30 kg/m2) on clinical and economic outcomes after thermal injury in a large national sample. The main objective was to identify whether obesity is an independent predictor of adverse events.

2.

Methods

2.1.

Data source and study population

This retrospective cohort study was exempt from the institutional review board approval as the data set is publically available and deidentified. The Nationwide Inpatient Sample (NIS), Healthcare Cost and Utilization Projects, and Agency for Healthcare Research and Quality database were chosen as the source population. The NIS is the largest publicly available all-payer inpatient heath care database in the United States, containing more than 7 million hospitalizations yearly [15]. This large representative sample allows for greater external validity and generalizability of results.

451

The NIS was queried for all adult patients (age 18 y) from 2005e2009 with International Classification of Diseases-9 (ICD-9) codes for burn injury (941e946.5, 948e949.5), excluding those with isolated injuries to the internal organs or eyes (n ¼ 33,638). This sample was further selected for those with an “emergency,” “urgent,” or “trauma center” admission yielding a total of 27,166 patients. Cases were excluded if no TBSA burn information was available (n ¼ 11,690) or if burn depth was not indicated (n ¼ 874), resulting in a final sample of 14,602 patients.

2.2.

Definition of variables

NIS variables were used to identify various patient characteristics including age, sex, race, LOS, discharge disposition, hospital costs, and income quartile. Disease severity was coded using the NIS category of All Patient Refined- Diagnosis Related Groups severity subclass. This classification takes into account the interaction between the principal diagnosis, presence of operating room and nonoperating room procedures, and secondary diagnoses in a process consisting of three phases [16]. The scoring system is as follows: 1 ¼ minor loss of function, 2 ¼ moderate loss of function, 3 ¼ major loss of function, and 4 ¼ extreme loss of function. Categories 3 þ 4 were grouped to form the classification of “high disease severity” in our study. Obesity was defined as patients with BMI 30 kg/m2 and identified both by ICD-9 codes for obesity and morbid obesity (278.00, 278.01; n ¼ 392) and by the specific code as an NIS variable (n ¼ 83) for a total of 475 obese patients in our sample. Burn depth was identified by ICD-9 codes, and patients were classified based on highest burn depth type regardless of TBSA of each burn degree.

2.3.

Assignment of comorbidity index

An adapted clinical comorbidity index was used based on methods previously outlined by Deyo et al. [17], which modified the original comorbidity index developed by Charlson et al. [18]. We searched each of the 15 diagnosis groups in the NIS for the 3e5 character ICD-9 codes identified by Deyo et al. along with the corresponding comorbidity measures already coded in the NIS. Each patient was then assigned two scores as follows: a weighted score (Charlson Comorbidity Index) based on the adapted score by Deyo et al. and an unweighted comorbidity index where 1 point was assigned to each of 17 possible comorbidities.

2.4.

Outcomes

Clinical outcomes were identified based on ICD-9 codes and included deep vein thrombosis (DVT), PE, urinary tract infection (UTI), wound infection (WI), and need for mechanical ventilation (Table 1). A composite category designated as grouped adverse events (GAE) was created to account for common morbidities affecting obese patients that we assessed using ICD-9 codes in our data set. The GAE consisted of UTI, WI, DVT, and PE. The NIS variable of death during hospitalization was used to determine mortality.

452

j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 8 ( 2 0 1 5 ) 4 5 0 e4 5 5

Table 1 e ICD-9 codes for clinical outcomes.

Table 2 e Study sample by TBSA, degree burn, and race.

Clinical outcome

Demographic

DVT Acute venous embolism and thrombosis of deep vessels of lower extremity Acute venous embolism and thrombosis of other specified veins PE PE and infarction UTI Acute pyelonephritis Acute cystitis Infection and inflammatory reaction due to indwelling urinary catheter Wound infection Posttraumatic wound infection not elsewhere classified Infected postoperative seroma Need for mechanical ventilation Other continuous invasive mechanical ventilation

2.5.

ICD-9 codes

P value

No 453.4, 453.40e453.42

453.82, 453.84e453.86

415.1, 415.11, 415.13, 415.19 590.1, 590.10e590.11 595.0 996.64

958.3

TBSA 0%e9% 10%e19% 20%e29% 30%e39% 40%e100% Degree First Second Third Race White Black Hispanic

Yes

n

%

n

%

9514 2718 919 361 615

67.3 19.2 6.5 2.6 14.4

347 70 32 13 13

73.1 14.7 6.7 2.7 2.7

600 7466 6061

4.2 52.8 42.9

20 260 195

4.2 54.7 41.1

6370 1582 1040

70.8 17.6 11.6

248 61 34

72.3 17.8 9.9

0.214

0.712

0.64

998.51 96.7, 96.70e96.72

Statistical analysis

The two cohorts (obese versus nonobese) were compared in terms of demographics and outcomes. Statistical analyses were performed using SPSS version 21 (IBM Corporation; Armonk, NY). Data are reported as mean
standard deviation or median (interquartile range). Significance was assessed at P < 0.05. Continuous variables were compared with Student t-test for parametric data and ManneWhitney U test for nonparametric data. Categorical data were compared with chisquared or Fisher exact test as appropriate. A binary logistic regression was performed with GAE as the dependent variable and age, black race, TBSA 20%, and obesity as covariates. This regression was repeated with the addition of the unweighted comorbidity index as a covariate to control for the disparity of comorbidities between cohorts.

3.

Obese

Results

In 14,602 patients, 3.3% were obese (n ¼ 475). The majority of the population had TBSA burns between 0 and 9% (n ¼ 9861; 67.3% of nonobese patients, 73.1% of obese patients). Seconddegree burns were most prevalent (n ¼ 7726), and most patients were classified as white (n ¼ 6618). There were no significant differences between the obese and nonobese patients in terms of these characteristics (Table 2). The rate of obesity increased significantly over the study period (2005: 1.7%, 2006: 2.0%, 2007: 2.8%, 2008: 4.6%, 2009: 5.2%, P < 0.001). On univariate analysis, significant differences were noted in the incidence of wound infection (7.2% versus 5.0%), UTI (7.2% versus 4.6%), DVT in TBSA burns 10% (3.1% versus 1.1%), PE in TBSA burns 10% (2.3% versus 0.6%), high disease severity (91.8% versus 73.5%), LOS (6 d [8] versus 5

d [9]), hospital costs ($10,122.12 [$18,074.72] versus $7892.07 [$17,191.96]), and discharge to home (57.7% versus 66.6%) (all P < 0.05). Interestingly, inpatient death occurred less frequently in the obese group (1.9% versus 4%, P ¼ 0.021). When excluding patients with small burns (TBSA <20%), this mortality disparity persisted with less inpatient deaths in the obese group (10.3% versus 22.7%, P ¼ 0.025) (Table 3). There was no significant difference in terms of need for mechanical ventilation between cohorts. Predictors of GAE included obesity (15.2% versus 10.1%), TBSA 20% (14.6% versus 9.6%), age (53
20 y versus 45
18 y), and black race (13.0% versus 9.9%; all P < 0.001). These remained significant independent predictors of GAE on multivariate analysis using a logistic regression model (area under receiver operator curve ¼ 0.630). Odd ratios were calculated (Table 4). Patients with GAE were more likely to be obese, have TBSA burns 20%, be of older age, and of black race (all P  0.05). The data were then reanalyzed with comorbidity adjustment. There was no significant difference between the obese and nonobese groups in terms of mean Charlson Comorbidity Index on univariate analysis (1.88
1.30 versus 1.86
1.44, P ¼ 0.204). There was, however, a statistically significant difference in the unweighted comorbidity index between obese and nonobese cohorts (0.91
0.96 versus 0.38
0.74, P < 0.001). Therefore, the decision was made to include the unweighted comorbidity index in the predictive model for GAE. A binary logistic regression was performed with GAE as the dependent variable and age, black race, TBSA 20%, obesity, and the unweighted comorbidity index as covariates. All five of these covariates remained significant predictors of GAE on this multivariate model (area under receiver operator curve ¼ 0.635; Table 5).

4.

Discussion

The major new finding is that obesity (BMI 30 kg/m2) is an independent predictor of GAE, even after controlling for TBSA

453

j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 8 ( 2 0 1 5 ) 4 5 0 e4 5 5

Table 3 e Significant differences observed between obese and nonobese patients. Outcome

Obese

Wound infection, % UTI, % DVT in TBSA 10%, % PE in TBSA 10%, % Discharged to home, % Disease severity, % LOS Hospital cost, $ Died in hospital (all patients), % Died in hospital (TBSA 20), %

No

Yes

5.00 4.60 1.10 0.60 66.60 73.50 5 d (9) 7892.07 (17,191.96) 4.00 22.7

7.20 7.20 3.10 2.30 57.70 91.80 6 d (8) 10,122.12 (18,074.72) 1.90 10.3

burns 20%, age, and black race, in burn patients. The effect of obesity approaches that of TBSA burns 20%. In addition to higher incidence of UTI, wound infection, DVT, and PE, obese patients experienced longer LOS and higher hospital costs, as would be expected. However, mortality was lower in our population, which is consistent with findings of other studies [10,14]. This trend persisted when small burns were excluded from our analysis. These results support the need for further research regarding DVT screening and weight-based thromboprophylaxis [8], weight-based antibiotic adjustment [19], wound care [20], and urinary catheter management in obese burn patients [21,22]. The extensive list of negative health outcomes associated with obesity has placed this significant problem at the forefront of public health. The link between obesity and chronic health conditions, such as diabetes and heart disease, is well established. Recently, studies have looked at obesity as a negative risk factor for morbidity and mortality in critically ill postsurgical patients [23] and trauma patients [24e27]. Research, however, is still evolving in the burn population. It is difficult to explain the contradictory finding that obesity is associated with increased GAE but less mortality. Jeschke et al. [14] found that mild obesity was associated with the best survival after severe burn injury. This “obesity paradox” has been described in other populations of patients including those on hemodialysis [28] and those with heart failure [29]. It would be reasonable to assume that morbidly obese patients would have worse outcomes because of comorbid conditions but that mild obesity may be protective in countering the hypermetabolic response after severe burns as previously hypothesized [14]. Unfortunately, without the ability to stratify our data by level of obesity due to limitations of the NIS database, we are unable to support this hypothesis.

Table 4 e Significant contributors to adverse events in burn patients (model without unweighted comorbidity index as covariate). Outcome measure TBSA 20 Obesity Age Black race

Adjusted odds ratio (95% confidence interval) 1.613 1.509 1.021 1.369

P value

(1.402e1.856) (1.164e1.955) (1.018e1.024) (1.171e1.602)

0.04 0.009 0.039 0.013 <0.001 <0.001 <0.001 0.002 0.021 0.025

The major limitation in our study is the classification of obesity in the NIS data. Information was only available to indicate “obese” or “not obese;” however, no BMI calculations were presented and, therefore, subanalyses could not be performed on mild versus moderate versus severe obesity. Patients with ICD-9 codes for obesity were added when available but again no level of BMI information was present. It is likely that the NIS and ICD-9 codes undercoded obese patients as even at its highest prevalence, obesity was only noted in 5.2% of patients, which does not approach the reported national prevalence of 34.9% during the same period [30]. This introduces a potential misclassification bias. Furthermore, our prevalence of obesity increased significantly by year, likely due to improved coding, whereas the national prevalence remained relatively stable over the same period [31]. The second most important limitation lies in the measurement of BMI itself. Although BMI is an excellent tool to screen for obesity, other factors such as fat distribution, genetics, and fitness level may affect this calculation, and therefore BMI cannot be assumed to quantify excess adiposity directly [32]. Regardless of these limitations, the large sample size and reliable coding of clinical outcomes strengthen our results. Another limitation involves the method of adjustment for comorbidity index. The Charlson index was developed to predict 1-y mortality from comorbidity data obtained by medical chart review, and weights were applied to each of the 19 comorbidities based on adjusted relative risks [18]. Various groups have attempted to adapt this system to ICD-9 codes, with the method outlined by Deyo et al. [17] being one of the most frequently cited. Using this scheme in administrative databases is subject to many limitations and there are conflicting opinions as to which method is most appropriate

Table 5 e Significant contributors to adverse events in burn patients (model with unweighted comorbidity index as covariate). Outcome measure TBSA 20 Obesity Age Black race Unweighted Comorbidity Index

Adjusted odds ratio (95% confidence interval) 1.655 1.392 1.018 1.344 1.77

(1.437e1.905) (1.071e1.809) (1.015e1.022) (1.149e1.573) (1.101e1.259)

454

j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 8 ( 2 0 1 5 ) 4 5 0 e4 5 5

[17,33e35]. The primary problem with adapting this index to large databases is the assignment of ICD-9 codes to comorbidities, as variability exists [35,36]. Also, the NIS records a limited number of diagnoses so a patient with many diagnoses recorded in an acute setting may not have relevant comorbid conditions represented. Furthermore, through identification of ICD-9 codes in the NIS, we are unable to discriminate a temporal relationship and therefore cannot truly say if a diagnosis is a comorbid condition from before hospitalization or a complication of the patient’s current condition. For these reasons, we chose to analyze our data with and without the comorbidity adjustment and present both results in the article. It is important to consider the potential limitation of our data set for detecting our primary and secondary outcomes. As stated previously, only 15 diagnoses per patient are reported in the NIS. Depending on the amount of comorbidities and acute condition codes assigned to the patient. important outcomes may not be reflected. Furthermore, 46.2% of eligible patients had to be excluded because of missing data for TBSA or burn degree, which illuminates another major limitation of the large retrospective database review. Obesity affects all aspects of burn management from the inflammatory cascade to patient care issues in the intensive care unit. With obesity afflicting more than one-third of our nation, special attention must be paid to this population in all facets of medical care. Specific therapeutic strategies and protocols must be considered to help mitigate the excess risk posed by obesity in the thermally injured patient.

5.

Conclusions

In summary, obesity was associated with increased WI, UTI, DVT in TBSA burns 10%, PE in TBSA burns 10%, increased hospital cost, increased LOS, decreased likelihood of being discharged home, and higher disease severity. However, mortality was reduced. Obesity is an independent predictor of adverse events along with TBSA burns 20%, age, and black race in the burn population. The effect of obesity approaches that of TBSA burns 20%. Our findings highlight the potential, clinical, and economic impact of the obesity epidemic on burn patients nationwide, although further research is needed regarding the mechanisms behind these phenomena and how to mitigate the negative consequences.

Acknowledgment The authors acknowledge Katherina Julien for her assistance with creating tables and charts and Chad Thorson, MD, MSPH, for his assistance with statistical methods. Authors’ contributions: J.J.R., S.S.S., J.P.M., C.J.A., L.F.T., and C.I.S. participated in the experimental design, collection, analysis and interpretation of data and drafted the article. J.J.R., S.S.S., J.P.M., C.J.A., L.F.T., K.G.P., L.R.P., N.N. and C.I.S. read and revised the article, figures, and tables. C.I.S. had overall responsibility for the study; including conception and

experimental design; analysis and interpretation of data; drafting and revision of the article and tables; statistical expertise and evaluation; and supervision. There was no external support or funding for this study.

Disclosure None of the authors have declared conflicts of interest.

references

[1] World Health Organization. Obesity and overweight. Retrieved October 17 2014, from http://www.who.int/ mediacentre/factsheets/fs311/en/. [2] Flegal KM, Carroll MD, Kit BK, Ogden CL. Prevalence of obesity and trends in the distribution of body mass index among US adults, 1999-2010. JAMA 2012;307:491. [3] Peeters A, Barendregt JJ, Willekens F, et al. Obesity in adulthood and its consequences for life expectancy: a life-table analysis. Ann Intern Med 2003;138:24. [4] Calle EE, Thun MJ, Petrelli JM, Rodriguez C, Heath CW Jr. Body-mass index and mortality in a prospective cohort of U.S. adults. N Engl J Med 1999;341:1097. [5] Kiraly L, Hurt RT, Van Way CW 3rd. The outcomes of obese patients in critical care. JPEN J Parenter Enteral Nutr 2011; 35(5 Suppl l):29S. [6] Osborne Z, Rowitz B, Moore H, et al. Obesity in trauma: outcomes and disposition trends. Am J Surg 2014;207:387. discussion 391e382. [7] Goutos I, Sadideen H, Pandya AA, Ghosh SJ. Obesity and burns. J Burn Care Res 2012;33:471. [8] Sheridan RL, Rue LW 3rd, McManus WF, Pruitt BA Jr. Burns in morbidly obese patients. J Trauma 1992;33:818. [9] Ghanem AM, Sen S, Philp B, Dziewulski P, Shelley OP. Body mass index (BMI) and mortality in patients with severe burns: is there a “tilt point” at which obesity influences outcome? Burns 2011;37:208. [10] Kraft R, Herndon DN, Williams FN, Al-Mousawi AM, Finnerty CC, Jeschke MG. The effect of obesity on adverse outcomes and metabolism in pediatric burn patients. Int J Obes 2012;36:485. [11] Gottschlich MM, Mayes T, Khoury JC, Warden GD. Significance of obesity on nutritional, immunologic, hormonal, and clinical outcome parameters in burns. J Am Diet Assoc 1993;93:1261. [12] Carpenter AM, Hollett LP, Jeng JC, Wu J, Turner DG, Jordan MH. How long a shadow does epidemic obesity cast in the burn unit? A dietitian’s analysis of the strengths and weaknesses of the available data in the National Burn Repository. J Burn Care Res 2008;29:97. [13] Liodaki E, Senyaman O, Stollwerck PL, et al. Obese patients in a burn care unit: a major challenge. Burns 2014;40:1738. [14] Jeschke MG, Finnerty CC, Emdad F, et al. Mild obesity is protective after severe burn injury. Ann Surg 2013;258:1119. [15] HCUP Nationwide Inpatient Sample (NIS). Healthcare Cost and Utilization Project (HCUP). Rockville, MD: Agency for Healthcare Research and Quality, http://www.hcup-us.ahrq. gov/nisoverview.jsp; 2005-2009. [16] All patient refined diagnosis related groups (APR-DRGs) classification system: methodology overview. Wallingford, CT: 3M Health Information Systems; 2008. [17] Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol 1992;45:613.

j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 8 ( 2 0 1 5 ) 4 5 0 e4 5 5

[18] Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40: 373. [19] Polso AK, Lassiter JL, Nagel JL. Impact of hospital guideline for weight-based antimicrobial dosing in morbidly obese adults and comprehensive literature review. J Clin Pharm Ther 2014;39:584. [20] Wilson JA, Clark JJ. Obesity: impediment to wound healing. Crit Care Nurs Q 2003;26:119. [21] Christ-Libertin C, Black S, Latacki T, Bair T. Evidence-based prevent catheter-associated urinary tract infections guidelines and burn-injured patients: a pilot study. J Burn Care Res 2015;36:e1. [22] Semins MJ, Shore AD, Makary MA, Weiner J, Matlaga BR. The impact of obesity on urinary tract infection risk. Urology 2012;79:266. [23] Nasraway SA Jr, Albert M, Donnelly AM, Ruthazer R, Shikora SA, Saltzman E. Morbid obesity is an independent determinant of death among surgical critically ill patients. Crit Care Med 2006;34:964. quiz 971. [24] Bochicchio GV, Joshi M, Bochicchio K, Nehman S, Tracy JK, Scalea TM. Impact of obesity in the critically ill trauma patient: a prospective study. J Am Coll Surg 2006;203:533. [25] Christmas AB, Reynolds J, Wilson AK, et al. Morbid obesity impacts mortality in blunt trauma. Am Surg 2007;73:1122. [26] Ciesla DJ, Moore EE, Johnson JL, Burch JM, Cothren CC, Sauaia A. Obesity increases risk of organ failure after severe trauma. J Am Coll Surgeons 2006;203:539. [27] Serrano PE, Khuder SA, Fath JJ. Obesity as a risk factor for nosocomial infections in trauma patients. J Am Coll Surgeons 2010;211:61. [28] Fleischmann E, Teal N, Dudley J, May W, Bower JD, Salahudeen AK. Influence of excess weight on mortality and

[29]

[30]

[31]

[32]

[33]

[34]

[35]

[36]

455

hospital stay in 1346 hemodialysis patients. Kidney Int 1999; 55:1560. De Schutter A, Lavie CJ, Milani RV. The impact of obesity on risk factors and prevalence and prognosis of coronary heart disease-the obesity paradox. Prog Cardiovasc Dis 2014;56: 401. Ogden CL, Carroll MD, Kit BK, Flegal KM. Prevalence of childhood and adult obesity in the United States, 2011-2012. JAMA 2014;311:806. Centers for disease control and prevention. Prevalence of overweight, obesity, and extreme obesity among adults: United States, 1960e1962 through 2011e2012; retrieved October 17 2014 from http://www.cdc.gov/nchs/data/hestat/ obesity_adult_11_12/obesity_adult_11_12.pdf. Centers for disease control and rrevention. About adult BMI. Retrieved October 17 2014, from http://www.cdc.gov/ healthyweight/assessing/bmi/adult_bmi/index.html). Needham DM, Scales DC, Laupacis A, Pronovost PJ. A systematic review of the Charlson comorbidity index using Canadian administrative databases: a perspective on risk adjustment in critical care research. J Crit Care 2005;20:12. Yurkovich M, Avina-Zubieta JA, Thomas J, Gorenchtein M, Lacaille D. A systematic review identifies valid comorbidity indices derived from administrative health data. J Clin Epidemiol 2015;68:3. Romano PS, Roos LL, Jollis JG. Adapting a clinical comorbidity index for use with ICD-9-CM administrative data: differing perspectives. J Clin Epidemiol 1993;46:1075. discussion 1081e1090. Cleves MA, Sanchez N, Draheim M. Evaluation of two competing methods for calculating Charlson’s comorbidity index when analyzing short-term mortality using administrative data. J Clin Epidemiol Aug 1997;50:903.