Article, Traumatology

Denver ED Trauma Organ Failure Score predicts healthcare resource utilization in adult trauma patients

a b s t r a c t

Background: Early identification of trauma patients who need specialized Healthcare resources may facilitate goal-directed resuscitation and effective secondary triage.

Objective: To estimate associations between Denver Emergency Department (ED) Trauma Organ Failure (TOF) Score and healthcare resource utilization.

Methods: Retrospective study of adult trauma patients at Denver Health Medical Center. The outcome was re- source utilization including: intensive care unit (ICU) length of stay (LOS), hospital LOS, procedures, and costs. multivariable regression analyses were used to estimate associations between moderate- or high-risk patients, as determined by the Denver ED TOF Score, and healthcare resource utilization.

Results: We included 3000 patients with a median age of 42 (IQR 27-56) years, 71% male, median injury severity score 9 (IQR 5-16), and 83% blunt mechanism. Among the cohort, 1379 patients (46%) were admitted to the ICU and 122 (4%) died. The adjusted relative risk for high- and moderate-risk as compared to low risk for number of procedures performed was 2.31 (95% CI 2.07-2.57) and 1.80 (95% CI 1.59-2.03) respectively; ICU LOS was 2.87

(95% CI 2.70-3.05) and 1.71 (95% CI 1.60-1.83) respectively; hospital LOS was 3.33 (95% CI 3.21-3.45) and

1.97 (95% CI 1.90-2.05) respectively. The adjusted Geometric mean for high-, moderate-, and low-risk for costs was $48,881 (95% CI $43,799-$54,552), $27,890 (95% CI $25,460-$30,551), and $12,983 (95% CI $12,493-

$13,492), respectively.

Conclusions: The Denver ED TOF Score predicts healthcare resource utilization, and is a useful bedside tool to identify patients early after injury that are likely to require significant healthcare resources and specialized trauma care.

(C) 2018

Introduction

Resource utilization in healthcare is becoming an increasingly im- portant issue with the national focus on healthcare coverage and costs [1]. Improvements in trauma and critical care have led to an increasing rate of multiply-injured trauma patients surviving the early post- traumatic period [2,3]. During their convalescence, trauma patients often require expensive hospital-based healthcare services and re- sources [4,5]. Early recognition of trauma patients who will need

? Supported, in part, by the National Institute of General Medical Sciences (Dr. Vogel [F32 GM099344]), the Agency for Healthcare Research and Quality (Dr. Vogel [K08 HS023901]), and National Institute of Allergy and Infectious Diseases (Dr. Haukoos [R01 AI106057]).

* Corresponding author at: Department of Emergency Medicine, Denver Health Medical Center, 777 Bannock Street, Mail Code 0108, Denver, CO 80204, USA.

E-mail address: [email protected] (J.A. Vogel).

Specialized care and who are at risk for worse outcomes is especially im- portant to improve healthcare resource allocation.

Several scoring systems currently exist to grade the severity of trau- matic injuries and to stratify trauma patients into mortality risk groups [6]. However, the data necessary to compute many of these scores is not readily available early in the post-injury period; it may take several hours for all of the injuries to be identified and some scores necessitate data that are not available until hospital discharge. Identification of a Prediction tool that uses information easily obtained at the bedside early in the post-injury course to identify patients in need of specialized intensive care would be beneficial to facilitate targeted resuscitation and triage of injured patients to higher levels of trauma care to improve resource allocation and outcomes.

We recently derived and validated the Denver Emergency Depart- ment (ED) Trauma Organ Failure (TOF) Score, which uses data obtained within 4 h of injury to predict multiple organ failure (MOF) within

https://doi.org/10.1016/j.ajem.2018.08.073

0735-6757/(C) 2018

seven days of hospitalization in adult trauma patients (Fig. 1) [7,8]. The Denver ED TOF Score has not previously been evaluated to ascertain its ability to predict healthcare resource utilization. The objective of this study was to estimate the associations between different Denver TOF Score risk categories (i.e., low-, moderate- and high-risk) and healthcare resource utilization. We hypothesized that intensive care resource utili- zation (defined by mechanical ventilation, invasive procedures, length of intensive care unit and hospital stay, and Total costs) would increase incrementally among adult trauma patients categorized into low- (Den- ver TOF Score 0-1), moderate- (Score 2-3), and high-risk (Score >= 4) groups, respectively.

Methods

Study design and setting

This was a cohort study from Denver Health Medical Center in Den- ver, Colorado. Denver Health Medical Center is a 477-bed, urban safety- net hospital and the Rocky Mountain Regional Level 1 Trauma Center, the only academic Level 1 trauma center in Colorado [9]. The ED is the primary receiving facility for all trauma patients, and has an approxi- mate annual adult census of 60,000 visits, of which approximately 2000 are included in the trauma registry, and approximately 500 of which are classified as major trauma (defined by an Injury Severity Score [ISS] N 15). The study was approved by the Colorado Multiple In- stitutional Review Board.

Study population

Consecutive trauma patients (>=18 years of age) who presented to Denver Health Medical Center from June 23, 2010, through December 26, 2012, and who were entered in the trauma registry were included in this study. The trauma registry contains trauma patients who are ad- mitted to the ED observation unit for at least 12 h, admitted to the hos- pital, die in the ED, or are transferred from outside hospitals for further trauma care. Patients were excluded from this study if they: (1) were b 18 years of age; (2) died in the ED; or (3) were transferred from an- other hospital.

Data collection

Patients included in the trauma registry are identified daily from the ED log by a team of personnel specifically trained in the oversight of the trauma registry and in the acquisition of data. Data are systematically abstracted from the patients’ medical records and entered into an elec- tronic database. To maintain quality assurance, an internal review of

approximately 20% of the trauma registry records is conducted on a reg- ular basis using a standardized protocol. Each review is completed by registry staff members who did not perform the initial chart abstraction. Any data discrepancies identified in this review are presented to the trauma registry committee and corrections are made through a consen- sus process.

Data recorded in the trauma registry and obtained for purposes of this study included: (1) Denver TOF Score variables (age, intubation, and initial systolic blood pressure, hematocrit, blood urea nitrogen, and white blood cell count); (2) gender; (3) clinical care characteristics (invasive procedures, operative interventions, and the duration of me- chanical ventilation); (4) ICU admission and length of ICU stay;

(5) length of hospital stay; (6) mortality during hospitalization;

(7) patient’s five-digit household zip code (to assess socioeconomic sta- tus using median Household income); (8) payer source (commercial, Medicaid, Medicare, self); and (9) total hospital costs. Laboratory data obtained within 4 h of arrival were obtained from the hospital’s com- puterized laboratory reporting system (MiSYS healthcare systems, Ra- leigh, NC) and linked to patient data obtained from the trauma registry. If more than one laboratory value was available within 4 h of hospital arrival, the most extreme value was utilized. total hospital charges were obtained from the patient records and converted to hospi- tal charges using the Denver Health Medical Center cost-to-charge ratio. Costs were adjusted for inflation to 2012 costs using the Consumer Price Index for all consumers for 2010 and 2011 at 0.053 and 0.021 respectively.

The medical records of all patients were systematically reviewed in a blinded fashion by trained physician abstractors to obtain specific mea- sures of healthcare resource utilization not included in the trauma reg- istry, including the following: (1) clinical care characteristics (arrival vital signs and the type and volume of blood products transfused during hospitalization); (2) specific procedures performed during hospitaliza- tion (Arterial line and Central venous line); and (3) comorbidities (asthma, coronary artery disease, congestive heart failure, peripheral vascular occlusive disease, cerebrovascular disease, dementia, chronic obstructive pulmonary disease, asthma, liver disease including cirrhosis, diabetes, renal disease [dialysis dependent or dialysis eligible], leuke- mia, lymphoma, solid tumor, metastatic cancer, and Acquired Immuno- deficiency Syndrome).

The Blinded physician abstractors were trained to use a closed-

response data collection instrument to systematically abstract each endpoint using standardized medical record abstraction methodology [10]. Ten percent of the charts were re-abstracted to verify reliability of the abstraction process. Inter-rater reliability was assessed to deter- mine the overall agreement and the median kappa was 0.98 (IQR 0.84-1.00).

Predictor

Emergent intubation? Hematocrit < 20%

Score

3

2

Outcome measures

The outcomes for the study included several measures of intensive care and hospital resource utilization as follows: total hospital costs, tra- cheostomy tube, mechanical ventilation duration, central venous cathe- ter placement, arterial line placement, parenteral nutrition, quantity of

years

1

Hematocrit 20% 1

ED systolic blood pressure < 90 mmHg 1

1

1

?Emergent intubation defined as intubation in the prehospital or emergency department settings.

ED TOF Score: low-

(Score 0- – (Score 2- – .

Fig. 1. The Denver Emergency Department (ED) Trauma Organ Failure (TOF) Score for prediction of multiple organ failure in adult trauma patients [7,8].

blood product transfusion, hospital and ICU length of stay, and operative intervention.

Data management and statistical analyses

Data were transferred electronically from the trauma registry and MiSYS into separate electronic spreadsheets (Microsoft Excel, Microsoft Corporation, Redmond, WA). Abstracted data were recorded in an elec- tronic closed-response data collection instrument (REDCap). REDCap is a secure, web-based application designed to support data capture for re- search studies that provides online validated data entry, audit trails for tracking data manipulation, and export procedures for data downloads to statistical packages [11].

Each electronic file was then transferred into native SAS format. All analyses were performed using SAS Version 9.3 (SAS Institute, Inc., Cary, NC) or Stata Version 12.0 (Stata Corporation, College Station, TX). Continuous data are reported as medians with interquartile ranges (IQRs) and categorical data are reported as percentages with 95% confi- dence intervals (CIs).

Multivariable regression analyses were used to estimate the inde- pendent associations between patients categorized as moderate- or high-risk for MOF (with low-risk serving as the reference), as deter- mined by the Denver ED TOF Score, and intensive care resource utiliza- tion using the following conceptual model:

Outcome = f [moderate risk + high risk + age + comorbidities

+SES status + payer source]

Because no composite summary measure of intensive care resource utilization exists, we developed separate regression models a priori for the following outcomes: total hospital costs (linear regression); trache- ostomy tube (logistic regression); mechanical ventilation duration (Poisson); Central venous catheter placement (logistic); arterial line placement (logistic); parenteral nutrition (logistic); quantity (in mL) of blood product transfusion (linear); hospital and ICU length of stay (Poisson); and operative intervention defined as any operative inter- vention by any surgical specialty during hospitalization (logistic). In ad- dition, we developed a model that incorporated all procedures (arterial line, central venous catheter, mechanical ventilation, Gastrostomy tube, and tracheostomy) as a Composite outcome. We included age, the number of comorbidities [12,13], SES status as median household in- come for the zip code, and payer source as multi-selection (self-pay, government-sponsored, Medicare, and other, with commercial as the reference) in each model.

Automated variable selection methods were not used to select vari- ables for inclusion in any of the model. Instead we performed complete model analysis with variables being included based upon our a priori knowledge of known or hypothesized relationships between multiple organ failure and healthcare resource utilization.

Sample size estimation

We anticipated including nine independent variables in each model. Of all the dependent variables, we anticipated some of the procedures being assessed to have the lowest prevalence (approximately 3%) and therefore used this statistic to provide a conservative estimate of the total number of patients required for the study. As a general rule, multi- variable logistic regression analyses require an event to predictor ratio of 10:1 to prevent over-fitting and to thus allow for valid inferences [14]. We therefore estimated requiring a minimum of 90 patients who had the procedure performed, and therefore a total minimum of 3000 patients for inclusion.

Results

During the study period, 3000 consecutive trauma patients met criteria for inclusion and therefore represent our study sample. The me- dian age was 42 (IQR 27-56) years; 71% were male, 83% experienced a blunt injury mechanism, and the median ISS was 9 (IQR 5-16). A com- plete description of the demographics and characteristics of the study sample is presented in Table 1. Of the 3000 patients, the median hospital LOS was 4 (IQR 2-8) days, with 1379 (46%) admitted to the ICU with a median ICU length of stay (LOS) of 2 (IQR 1-5) days. Excluding patients with short ICU stays (e.g., those simply observed for 24 h then trans- ferred to a ward), 651 (22%) had an ICU LOS N 48 h with a median ICU LOS of 5 days (IQR 3-10). Among the 3000 participants, 122 (4%) died. A complete description of the outcomes for the study sample is presented in Table 2.

Table 1

Demographics and characteristics of the study sample (n = 3000).

Variable

No.

(% or IQR)

Missing

No.

(%)

Total (n)

3000

Median age, years

42

(27-56)

0

(0)

Male gender

2140

71

3

(0)

Race

27

(0)

White

2444

82

Black

307

10

American Indian

42

1

Asian

50

2

Pacific Islander

2

0

Other

128

4

Hispanic ethnicity

1002

33

Mechanism

Blunt

2499

83

1

(0)

injury type

2

(0)

Motor vehicle crash

637

21

motorcycle crash

208

7

Assault

346

12

Auto-pedestrian crash

236

8

Bicycle crash

119

4

Gunshot

163

5

Stabbing

229

8

Fall

827

28

Other

233

8

Median injury severity score

9

(5-16)

0

(0)

Abbreviations: ED, emergency department; IQR, interquartile range; MOF, multiple organ failure.

Trauma patients identified as high risk using the Denver ED TOF Score had significantly increased total volume of blood products trans- fused and hospital costs (Table 3). As the level of risk as identified by the Denver ED TOF Score increased, total hospital costs increased (Fig. 2). Patients at high or moderate risk using the Denver ED TOF Score had increased total procedures, ICU LOS, and hospital LOS (Table 4). There was no significant difference in duration of mechanical ventilation among patients identified as high versus low risk using the Denver ED TOF Score (Table 4). Patients identified as high risk by the Denver ED TOF Score had increased interventions during hospitalization including arterial line, central venous catheter, parenteral nutrition, gastrostomy tube, tracheostomy, and operations (Table 5). The associa- tion between patients who were high-risk determined by the Denver ED TOF Score and outcomes for the study were statistically significant

Table 2

Outcomes of the study sample (n = 3000).

Variable No. % or (IQR)

Total (n) 3000

Procedures

Arterial line

474

16

Central venous catheter

351

12

Mechanical ventilation

562

19

Gastrostomy tube

81

3

Tracheostomy

193

6

Median number of procedures performeda

2

(1-4)

Parenteral nutrition

46

6

Renal replacement therapy

6

0

Received blood products

603

20

Median volume of blood products transfused, mL

1213

(500-3206)

Operative interventionb

1746

58

Intensive care unit

Admission

1379

46

Median length of stay

2

(1-5)

Median hospital length of stay, days

4

(2-8)

In-hospital mortality

122

4

Abbreviations: ED, emergency department; IQR, interquartile range.

a Includes arterial line, central venous catheter, mechanical ventilation, percutaneous endoscopic gastrostomy tube, and tracheostomy.

b Defined as operative intervention by any surgeon during hospitalization.

Table 3 Adjusted associations between total Healthcare costs and volume of blood products trans- fused and risk as determined by the Denver Emergency Department (ED) Trauma Organ Failure (TOF) Score.a,b

Variable Geometric mean 95% CI

Table 4

Adjusted associations between healthcare resource utilization and risk as determined by the Denver Emergency Department (ED) Trauma Organ Failure (TOF) Score.a,b

Healthcare resource utilization measure Relative risk (95% CI) Duration of mechanical ventilation

Total volume of blood products transfusedc

High versus low risk

1.00

(0.92, 1.10)

High versus low risk 2561

2189-2996

Moderate versus low risk

0.68

(0.62, 0.97)

Moderate versus low risk 1345

1122-1611

Total procedures during hospitalizationc

Total hospital costs

High versus low risk

2.31

(2.07, 2.57)

High risk $48,881

$43,799-$54,551

Moderate versus low risk

1.80

(1.59, 2.03)

Moderate risk $27,890

$25,460-$30,551

Intensive care unit length of stay

Low risk $12,983

$12,493-$13,492

High versus low risk

2.87

(2.70, 3.05)

Abbreviations: confidence interval, CI; emergency department, ED; Trauma Organ Failure, Moderate versus low risk 1.71 (1.60, 1.83)

TOF. Hospital length of stay

a High risk is defined as a Denver ED TOF score of >=4; moderate risk is defined as a Denver ED TOF score of 2-3; and low risk is defined as a Denver ED TOF score of 0-1.

b Each model was adjusted for age, number of comorbidities, Median household income of home zip code as a measure of socioeconomic status, and payer source with commercial as the reference.

c Volume of total blood products transfused is reported in milliliters.

for all outcomes with the exception of duration of mechanical ventilation.

Discussion

This study was conducted to determine the associations between the Denver ED TOF Score risk categories, which use objective clinical data readily available within 4 h of injury, and healthcare resource utili- zation in trauma care. We demonstrated that the Denver ED TOF Score is associated with several measures of healthcare resource utilization in- cluding arterial line, central venous catheter, gastrostomy tube, hospital length of stay, intensive care unit length of stay, operative intervention, parenteral nutrition, total hospital costs, total procedures, tracheos- tomy, and volume of blood products transfused. The use of the Denver ED TOF Score early in the post-injury period may serve as a useful tool to identify patients in need of specialized healthcare resources and higher levels of trauma care. Early identification of these at-risk patients may subsequently facilitate improved Healthcare outcomes.

Several trauma and intensive care unit scoring systems have been used to evaluate outcomes other than mortality risk [15-25]. The ISS has been used to evaluate hospital stay, ICU stay, time ventilated, com- plications, and emergency surgery [15,16]. The ICISS has outperformed the ISS and TRISS for prediction of hospital charges and hospital length of stay [17]. The ISS, TRISS, ICISS and ASCOT are all valuable tools that

High versus low risk

3.33

(3.21, 3.45)

Moderate versus low risk

1.97

(1.90, 2.05)

Abbreviations: confidence interval, CI; emergency department, ED; Trauma Organ Failure, TOF.

a High risk is defined as a Denver ED TOF score of >=4; moderate risk is defined as a Denver ED TOF score of 2-3; and low risk is defined as a Denver ED TOF score of 0-1.

b Each model was adjusted for age, number of comorbidities, median household income of home zip code as a measure of socioeconomic status, and payer source with commercial as the reference.

c Procedures included arterial line, central venous catheter, mechanical ventilation, percutaneous endoscopic gastrostomy tube, and tracheostomy.

have been used to predict outcomes after trauma but unfortunately re- quire coding of the healthcare information in AIS nomenclature that may not be readily available at the bedside early in the post-injury pe- riod [17-20]. Studies have demonstrated that the RTS, a physiologic score which can be applied at the bedside, does not predict mortality as well as the anatomic scores [18]. The APACHE score has been demon- strated to predict mortality and healthcare outcomes but has limited ap- plicability clinically at the bedside early after injury as the score necessitates complex calculations [19]. Scores that require extensive calculation or data from the hospitalization are highly valuable for retro- spective analyses but their role is limited in informing resource use pro- spectively and in evaluating the severity of illness and the potential need for secondary triage early in the post-injury period [18].

The Denver ED TOF Score is easily applied at the bedside in the first few hours of the acute post-injury phase. In this study we found the as- sociations between high risk (as determined by the Denver ED TOF Score) and healthcare resource utilization outcomes were statistically significant for all outcome measures with the exception of the duration of mechanical ventilation. The duration of mechanical ventilation

Image of Fig. 2

Fig. 2. Association between total hospital costs and the Denver Emergency Department Trauma Organ Failure Score.

Table 5 Adjusted associations between measures of healthcare resource utilization and risk as deter- mined by the Denver Emergency Department (ED) Trauma Organ Failure (TOF) score.a,b

Measure of healthcare resource utilization

Odds ratio

(95% CI)

Arterial line

High versus low risk

4.75

(3.87, 5.83)

Moderate versus low risk

1.14

(0.95, 1.36)

Central venous catheter

High versus low risk

6.38

(5.16, 7.88)

Moderate versus low risk

1.13

(0.93, 1.38)

Parenteral nutrition

High versus low risk 2.27 (1.47, 3.62)

Moderate versus low risk 1.04 (0.63, 1.72) Gastrostomy tube

High versus low risk

4.00

(2.91, 5.50)

Moderate versus low risk

1.70

(1.21, 2.39)

Tracheostomy

High versus low risk

20.98

(14.18, 31.03)

Moderate versus low risk Operative interventionc

High versus low risk

9.19

1.90

(6.19, 13.65)

(1.53, 2.38)

Moderate versus low risk

0.77

(0.65, 0.92)

Abbreviations: confidence interval, CI; emergency department, ED; Trauma Organ Failure, TOF.

a High risk is defined as a Denver ED TOF score of >=4; moderate risk is defined as a Denver ED TOF score of 2-3; and low risk is defined as a Denver ED TOF score of 0-1. b Each model was adjusted for age, number of comorbidities, median household income of home zip code as a measure of socioeconomic status, and payer source with commercial

as the reference.

c Defined as operative intervention by any surgeon during hospitalization.

among high and low risk trauma patients was not significantly different. One potential explanation for the lack of association between duration of mechanical ventilation and high risk status is that these patients may have died sooner after injury, thereby limiting the duration of the post-injury mechanical ventilation period.

This study demonstrates that the Denver ED TOF Score could serve as a useful clinical prediction instrument for early identification of those patients at risk for necessitating specialized intensive care re- sources thereby facilitating aggressive, goal-directed resuscitation and secondary triage of these critically ill patients to highly-specialized facil- ities. The Transfer of patients at high risk for in-hospital mortality and specialized care needs to a higher level of trauma care with a cadre of trauma specialists may improve resource allocation and trauma out- comes thereby reducing healthcare costs associated with such patients. Before implementing the Denver ED TOF Score into clinical practice, additional investigations of the performance of the score in diverse set- tings are indicated. A trial of use of the score in a regional fashion to identify those patients in need of specialized healthcare resources and who may benefit from transfer to a higher level of care as compared to current regional triage criteria may be helpful. This would include an evaluation of the outcomes for patients who were transferred based upon the Denver ED TOF score and a comparison of the score to physician gestalt for the need for transfer to a higher level of trauma care. These studies would further the knowledge of the utility and appli-

cation of the Denver ED TOF Score in clinical practice.

The primary limitation of this study is that the study was conducted at a single urban Level I Trauma center. While we have a diverse trauma population, additional multi-center studies are needed to evaluate the performance of the Denver ED TOF Score in both urban and rural healthcare settings with heterogeneous trauma populations. We in- cluded patients from the Denver Health Trauma Registry and excluded patients who died in the ED and who were discharged home. It is possi- ble that some of the patients discharged home, and therefore not in- cluded in the trauma registry, had a subsequent decline in their condition and developed a later need for hospital-based procedures; these data would not have been captured by this study. It is possible that bias may have occurred as a result of the data abstracted from the

medical record by the physician abstractors for the study. However, the potential for this type of bias to occur was significantly mitigated by the use of a rigorous training and abstraction process and use of phy- sician abstractors blinded to the purpose of our study. We also com- pleted re-abstraction of 10% of the charts and compared these results to verify agreement of the abstraction process which demonstrated ex- cellent inter-rater reliability.

Conclusions

The Denver ED TOF Score predict healthcare resource utilization in an adult trauma population at a single center. The Denver ED TOF Score may be a useful tool to identify patients early in the post-injury phase who are at high risk for needing significant healthcare resources and specialized trauma care. Early identification of these patients may facilitate targeted resuscitation and secondary triage to improve re- source allocation and outcomes in trauma.

Acknowledgements

Author contributions

Dr. Vogel had full access to all of the data in the study and takes re- sponsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Vogel, Haukoos

Acquisition of data: Vogel, Sungar, Ryan, Murphy, Loar, Boatright, Adams

Analysis and interpretation of data: Vogel, Sungar, Ryan, Murphy, Loar, Boatright, Adams, Haukoos

Drafting of the manuscript: Vogel

Critical revision of the manuscript for important intellectual content: Vogel, Sungar, Ryan, Murphy, Loar, Boatright, Adams, Haukoos Statistical analysis: Vogel

Obtained funding: Vogel, Haukoos

Administrative, technical, or material support: Vogel, Haukoos

Study supervision: Vogel, Haukoos

Additional contributions

The authors would like to acknowledge and thank Emily Hopkins, MSPH, who assisted with the data for this study, and Craig Gravitz, RN, EMT-P, who assisted with obtaining the trauma registry data for this study.

Financial disclosures

Dr. Vogel received support from the National Institute of General Medical Sciences and Agency for Healthcare Research and Quality to conduct this research project. Dr. Haukoos received support from the National Institute of Allergy and Infectious Diseases.

References

  1. Goettler CE, Waibel BH, Goodwin J, Watkine F, Toschlog EA, Sagraves SG, et al. Trauma intensive care unit survival: how good is an educated guess? J Trauma 2010;68:1279-88.
  2. Frohlich M, Lefering R, Probst C, Paffrath T, Schneider MM, Maegele M, et al. Epidemiol- ogy and risk factors of multiple organ failure after multiple trauma: an analysis of 31,154 patients from the Trauma Register DGU. J Trauma Acute Care Surg 2014;76: 921-8.
  3. Ruchholtz S, Lefering R, Paffrath T, Oestern HJ, Neugebauer E, Nast-Kolb D, et al. Re- duction in mortality of severely injured trauma patients in Germant. Dtsch Arztebl Int 2008;105:225-31.
  4. O’Keefe GE, Maier RV, Diehr P, Grossman D, Jurkovich GJ, Conrad D. The complica- tion of trauma and their Associated costs in a level I trauma center. Arch Surg 1997;132:920-4.
  5. Hemmila MR, Jakubus JL, Wahl WL, Arabi S, Henderson WG, Khuri SF, et al. Detecting the blind spot: complications in the trauma registry and trauma quality improve- ment. Surgery 2007;142:439-48.
  6. Moore FA, Sauaia A, Moore EE, Haenel JB, Burch JM, Lezotte DC. Postinjury multiple organ failure a bimodal phenomenon. J Trauma 1996;40:501-10.
  7. Vogel JA, Liao MM, Hopkins E, Seleno N, Byyny RL, Moore EE, et al. Prediction of postinjury multiple-organ failure in the emergency department: Development of the Denver Emergency Department Trauma Organ Failure Score. J Trauma Acute Care Surg 2014;76:140-5.
  8. Vogel JA, Newgard CD, Holmes JF, Diercks DB, Arens AM, Boatright D, et al. Validation of the Denver Emergency Department Trauma Organ Failure Score. J Am Coll Surg 2016;222:73-82.
  9. Gabow P, Eisert S, Wright R. Denver Health a model for the integration of a public hospital and community health centers. Ann Intern Med 2003;138:143-50.
  10. Gilbert EH, Lowenstein SR, Koziol-McLain J, Barta DC, Steiner J. Chart reviews in emergency medicine research: Where are the methods? Ann Emerg Med 1996;27 (3):305-8.
  11. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap) a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform 2009;42: 377-81.
  12. Moore L, Lavoie A, Le Sage N, Bergeron E, Emond M, Liberman M, et al. Using infor- mation on preexisting conditions to predict mortality from traumatic injury. Ann Emerg Med 2008;52:356-64.
  13. Haukoos JS. Advanced modeling strategies in emergency medicine research: is the pendulum swinging? Ann Emerg Med 2008;52:365-7.
  14. Hosmer DW, Lemeshow S. Applied Logistic Regression. NY, New York: Wiley; 2000.
  15. Copes WS, Champion Hr, Sacco WJ, Lawnick MM, Keast SL, Bain LW. The injury se- verity score revisited. J Trauma 1988;28:68-77.
  16. Lavoie A, Moore L, LeSage N, Liberman M, Sampalis JS. The injury severity score or the New Injury Severity Score for predicting intensive care unit admission and hos- pital length of stay? Injury, Int J Care Injured 2005;36:477-83.
  17. Ruteledge R, Osler T, Emery S, Krumhout-Schiro S. The end of the Injury Severity Score and the Trauma and Injury Severity Score: an International Classification of Diseases, Ninth Revision-based prediction tool outperforms both ISS and TRISS as predictors of trauma patient survival, hospital charges, and hospital length of stay. J Trauma 1998;44:41-9.
  18. Gabbe BJ, Cameron PA, Wolfe R. TRISS does it get better than this? Acad Emerg Med 2003;11:181-6.
  19. Champion HR, Sacco WJ, Hunt TK. Trauma injury severity scoring to predict mortal- ity. World J Surg 1983;7:4-11.
  20. Baker SP, O’Neill B, Haddon W, Long WB. The injury severity score: a method for de- scribing patients with multiple injuries and evaluating emergency care. J Trauma 1974;14:187-96.
  21. Balogh A, Offner PJ, Moore EE, Biffl WL. NISS predicts postinjury multiple organ fail- ure better than the ISS. J Trauma 2000;48:624-7.
  22. Osler T, Baker SP, Long W. A modification of the Injury Severity Score that both im- proves accuracy and simplifies scoring, 43; 1997; 922-6.
  23. Brenneman FD, Boulanger BR, McLellan BA, Redelmeier DA. Measuring injury sever- ity: time for a change? J Trauma 1998;44:580-2.
  24. Lavoie A, Moore L, LeSage N, Liberman M, Sampalis JS. The new injury severity score: a more accurate predictor of in-hospital mortality than the injury severity score. J Trauma 2004;56:1312-20.
  25. Vassar MJ, Lewis FR, Chambers JA, Mullins RJ, O’Brien PE, Weigelt JA, et al. Prediction of outcome in intensive care unit trauma patients: a multicenter study of APACHE, TRISS, and a 24-hour intensive care unit point system. J Trauma 1999;47:324-9.

Leave a Reply

Your email address will not be published. Required fields are marked *