AAST 2013 PLENARY PAPER

Epidemiology and risk factors of multiple-organ failure

after multiple trauma: An analysis of 31,154 patients
from the TraumaRegister DGU
Matthias Frohlich, MD, Rolf Lefering, PhD, Christian Probst, MD,
Thomas Paffrath, MD, Marco M. Schneider, MD, Marc Maegele, MD,
Samir G. Sakka, MD, Bertil Bouillon, MD, Arasch Wafaisade, MD,
and Committee on Emergency Medicine, Intensive Care and Trauma Management
of the German Trauma Society (Sektion NIS), Cologne, Germany

In the severely injured who survive the early posttraumatic phase, multiple-organ failure (MOF) is the main cause of morbidity
and mortality. An enhanced prediction of MOF might inuence individual monitoring and therapy of severely injured patients.
METHODS:
We performed a retrospective analysis of a nationwide prospective database, the TraumaRegister DGU of the German Trauma
Society. Patients with complete data sets (2002Y2011) and a relevant trauma load (Injury Severity Score [ISS] Q 16), who were
admitted to an intensive care unit, were included.
RESULTS:
Of a total of 31,154 patients enclosed in this study, 10,201 (32.7%) developed an MOF according to the Sequential Organ
Failure Assessment score. During the study period, mortality of all patients decreased from 18.1% in 2002 to 15.3% in 2011
( p G 0.001). Meanwhile, MOF occurred signicantly more often (24.6% in 2002 vs. 31.5% in 2011, p G 0.001), but mortality
of MOF patients decreased (42.6% vs. 33.3%, p G 0.001). MOF patients who died survived 2 days less (11 days in 2002 vs.
8.9 days in 2011, p G 0.001). Independent risk factors for the development of MOF following severe trauma were age, ISS,
head Abbreviated Injury Scale (AIS) score of 3 or higher, thoracic AIS score of 3 or higher, male sex, Glasgow Coma Scale
(GCS) score of 8 or less, mass transfusion, base excess of less than j3, systolic blood pressure less than 90 mm Hg at
admission, and coagulopathy.
CONCLUSION:
Over one decade, we observed an ongoing decrease of mortality after multiple trauma, accompanied by decreasing mortality
in the subgroup with MOF. However, incidence of MOF in the severely injured increased signicantly. Thus, MOF after
multiple trauma remains a challenge in intensive care. The risk factors from multivariate analysis could be instrumental in
anticipating the early development of MOF. Furthermore, a reliable prediction model might be supportive for patient enrolment
in trauma studies, in which MOF marks the primary end point. (J Trauma Acute Care Surg. 2014;76: 921Y928. Copyright * 2014
by Lippincott Williams & Wilkins)
LEVEL OF EVIDENCE: Epidemiologic study, level III.
KEY WORDS:
Multiple-organ failure; multiple trauma; epidemiology.
BACKGROUND:

uring the last years, an increasing rate of multiple trauma

patients has survived the early posttraumatic course because of improved trauma and critical care.1 Nevertheless,
multiple-organ failure (MOF) is still considered the main cause
of late postinjury mortality and intensive care unit (ICU) resource use,2 as reected in several single-center studies. Ciesla
et al.3 reported an overall incidence of MOF of 25% in highrisk polytrauma patients in a 12-year prospective study.
Submitted: September 21, 2011, Revised: November 7, 2011, Accepted: November 9, 2011.
From the Department of Trauma and Orthopedic Surgery (M.F., C.P., T.P.,
M.M.S., M.M., B.B., A.W.), Institute for Research in Operative Medicine
(IFOM) (R.L.), and Department of Anaesthesiology and Intensive Care
Medicine (S.G.S.), Cologne-Merheim Medical Center (CMMC), University
of Witten/Herdecke, Cologne, Germany.
This study was presented at the 72nd annual meeting of the American Association for
the Surgery of Trauma, September 18Y21, 2013, in San Francisco, California.
Address for reprints: Matthias Frohlich, MD, Department of Trauma and Orthopedic Surgery, University of Witten/Herdecke, Cologne-Merheim Medical
Center (CMMC), Ostmerheimerstr. 200, D-51109 Cologne, Germany; email:
froehlichm@kliniken-koeln.de.
DOI: 10.1097/TA.0000000000000199

According to further studies, MOF incidence in trauma patients

decreased slightly4Y6 or remained constant7 during the last years.
Furthermore, postinjury MOF caused signicant increased ICU
length of stay5,8 and caused approximately 51% of late trauma
deaths.3 However, multicenter epidemiologic data on incidence
and mortality are scarce; one prospective cohort study on seven
US trauma centers excluding relevant brain injury has even
implicated an early onset of MOF.9
In addition, independent risk factors for posttraumatic
MOF have been discussed recently. Different authors have described general patient data such as age, male sex or injury severity, as well as early therapy such as the amount of packed red
blood cells (pRBC) as independent predictors.3,10 Dewar et al.11
recently described further laboratory analyses such as platelet
count, bilirubin, and creatinine for a prediction model. However, injury pattern or anatomic localizations of injuries have
not yet been analyzed.
In the context of decreasing mortality from multiple
trauma, we hypothesized an increase in MOF incidence. Based
on a large multicenter database, the aims of the current study

J Trauma Acute Care Surg

Volume 76, Number 4

Copyright 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

921

J Trauma Acute Care Surg

Volume 76, Number 4

Frohlich et al.

were (1) to assess the potential changes in the incidence and

outcome of MOF after multiple trauma in Germany during
one decade (2002Y2011) and (2) to evaluate independent risk
factors for posttraumatic MOF.

PATIENTS AND METHODS

The TraumaRegister DGU of the German
Trauma Society
The TraumaRegister DGU (TR-DGU) of the German
Trauma Society (DGU) was founded in 1993 as an initiative
for prospective, standardized, and anonymous documentation
of severely injured patients.
Initially, participation was voluntary and free of costs.
Because of the development of trauma networks by the DGU
(TraumaNetzwerk DGU) during the recent years, participation
in the TR-DGU has become mandatory for certied trauma
centers in the context of quality management. Fees from the
participating hospitals fund the registry, and data quality is reviewed during regular audits.12 The number of documented
cases has consequently increased during the last years. An
annual amount of data sets of more than 25,000 patients from
more than 600 trauma centers of all levels have been entered
recently. Thus, the majority of the severely injured is covered
by the TR-DGU. Until the end of 2011, a total of 93,024 cases
were registered. In 2002, the online version of the registry was
introduced, replacing paper form data collection. Further details
regarding the TR-DGU have been described previously.13Y15
The TR-DGU is approved by the review board of the
DGU and is in compliance with the institutional requirements
of its members.

Study Population
For the present study, data sets of patients with multiple
injuries entered into the TR-DGU between 2002 and 2011 were
analyzed. Inclusion criteria were a relevant trauma load (Injury
Severity Score [ISS] Q 16), admission to an ICU, and complete
data sets regarding MOF. Therefore, a total of 31,154 patients
were included for further analysis.

Denitions of Organ Failure and Sepsis

In the data report form, the occurrences of organ failure
and sepsis during the hospitalization are documented dichotomously for each patient as either yes or no. Organ failure
was assessed using the Sequential Organ Failure Assessment
(SOFA) score to determine organ function status.16 Organ systems monitored and recorded for failure were the lungs, cardiovascular and central nervous system (CNS), kidney, liver,
and coagulation system. The SOFA score is composed of
scores from six organ systems, graded from 0 to 4 according
to the degree of dysfunction or failure.16,17 A score of 3 or
higher for one of the organ systems was dened as a failure
of this organ. In both the initial description and the evaluation
of the SOFA score by Vincent et al.,16,17 there is no statement
on when to dene multiple organ failure. Thus, in our database, MOF was dened as organ failure for at least two of
the listed organs or systems. Sepsis was assessed according to
the criteria of Bone et al.18 According to our previous work,
922

coagulopathy was dened by the presence of abnormal coagulation parameters upon emergency department (ED) arrival of
the patient, that is, prothrombin time ratio (Quicks value) of
less than 70% and/or platelet count of less than 100,000/mL.19
The occurrences of organ failure and sepsis are assessed on
a daily basis during the complete ICU stay. However, the day
of onset is not documented in the trauma registry.

Statistical Analysis
Data are presented as mean (SD) for continuous variables
or percentages for categorical variables. Categorical variables
were analyzed using the W2 test. The Mann-Whitney U-test was
applied for comparison of continuous variables in MOF versus
no-MOF patients.
To identify independent risk factors for the development
of MOF, we selected potential predictors from the literature
and from the clinical experience of our research group. These
variables were composed of demographic characteristics (age,
sex), severity of injury, early physiology, and early treatment.
Subsequently, a multivariate analysis was performed using stepwise logistic regression with MOF as the dependent variable.
Where appropriate, mortality and other incidence rates
are presented with 95% condence intervals.
For all statistical analyses, a probability of less than 0.05
was considered to be statistically signicant. All data were analyzed by using IBM SPSS 20 (IBM Corporation, Chicago, IL).

RESULTS
Demographics
In total, 31,154 severely injured patients were identied
for further analysis. Patients had a mean age of 45 (21) years,
were predominantly male (73%), and were injured relevantly
with a mean (SD) ISS of 28 (12), with 96% sustaining blunt
trauma. During the entire period of observation, 10,201 patients (32.7%) developed MOF.
With a mean (SD) age of 48 (22) years, MOF patients were
older compared with non-MOF patients (44 [21], p G 0.001).
Moreover, the MOF group presented a signicantly higher
trauma load (ISS, 33.4 [13.4] vs. 25.8 [9.7]; p G 0.001), and
more patients had severe head injuries as reected by head
Abbreviated Injury Scale (AIS) score of 3 or higher (69.6% vs.
53.5%, p G 0.001) with consecutively compromised neurologic status at the scene (Glasgow Coma Scale [GCS] score e 8,
53.1% vs. 24.1%, p G 0.001).
Furthermore, the occurrence of MOF during posttraumatic
hospitalization was associated with more deranged physiologic and laboratory values during the initial trauma resuscitation. Patients who developed MOF during the later hospital
sequelae had received signicantly more pRBCs (3.2 [7] vs.
1 [3.5]) and fresh frozen plasma (2.5 [6.5] vs. 0.7 [2.9]) between arrival in the ED and ICU admission compared with nonMOF patients. In 11.6%, massive transfusion was required
( p G 0.001). MOF patients stayed longer in the ICU (18.9
[12.2] days vs. 9.2 [10.4] days for the non-MOF patients,
p G 0.001) and in the hospital (28.5 [28.7] days vs. 25.5
[26.2] days, p G 0.001).
* 2014 Lippincott Williams & Wilkins

Copyright 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

J Trauma Acute Care Surg

Volume 76, Number 4

Frohlich et al.

Detailed characteristics for patients with versus without

MOF are summarized in Table 1.
Of all enclosed patients, 52% showed at least one organ
failure. The most commonly affected systems were central nervous, cardiovascular, and respiratory systems. While the coagulation system failed in 31.5% of the patients who developed
MOF, only 2.5% of the non-MOF patients showed a deranged
coagulation. Detailed characteristics are summarized in Table 2.
Nevertheless, since 2008, a slight increase in failure of the
coagulation system could be observed.

Incidence and Mortality 2002 to 2011

During the entire study period, 43.1% of the MOF patients died within 30 days after trauma as compared with 7.5%
in the non-MOF patients ( p G 0.001). Among these patients,

patients who developed MOF died after a mean (SD) of 9.3

(15) days, while nally deceased non-MOF patients survived
only a mean (SD) of 6.6 (12) days.
Detailed analyses of the single years during the study
period showed that mortality of all enclosed patients decreased
from 18.1% in 2002 to 15.3% in 2011 (Fig. 1, p G 0.001).
In 2002, 47.2% of multiple trauma patients experienced any
organ failure as compared with 51.1% in 2011 ( p G 0.001).
Accordingly, MOF occurred signicantly more often (24.6%
in 2002 vs. 31.5% in 2011, p G 0.001, Fig. 2). However,
mortality decreased signicantly in the subgroup of MOF
patients (42.6% vs. 33.3%, p G 0.001). Nevertheless, patients
with MOF who died survived 2 days less (11 days in 2002 vs.
8.9 days in 2011, p G 0.001), translating into an average reduction of 0.244 survived days per year (regression coefcient, j0.244; p = 0.001).

TABLE 1. Basic Demographic/Clinical Characteristics at Admission and Outcome of Multiple Trauma Patients With and Without
MOF (2002Y2011)
All Enclosed Patients, 31,154 MOF, 10,201 (32.7%) No MOF, 20,953 (67.3%)
Age, mean (SD), y
Age less than 60 y, %
Male, %
Blunt/penetrating trauma, %
ISS, mean (SD)
New ISS (NISS), mean (SD)
Head AIS score Q 3, %
Thoracic AIS score Q 3, %
Abdominal AIS score Q 3, %
Extremities AIS score Q 3, %
Isolated traumatic brain injury, %
Combined traumatic brain injury, %
SBP at scene, mean (SD), mm Hg
Heart rate at scene, mean (SD), beats/min
GCS score at scene, mean (SD)
GCS score at scene e 8, %
Intravenously administered uids prehospital, mean (SD), mL
SBP at ED, mean (SD), mm Hg
Heart rate at ED, mean (SD), beats/min
Hemoglobin, mean (SD), g/dL
Prothrombin time, mean (SD), Quick%
Base excess, mean (SD), mmol/L
Preexisting medical condition, %
pRBC transfusion,* %
Massive transfusion* (Q10 pRBCs), %
pRBC units,* mean (SD), n
FFP units,* mean (SD), n
Organ failure, %
Sepsis, %
ICU LOS, mean (SD), d
In-hospital LOS, mean (SD), d
Ventilator days, mean (SD), d
30-d mortality, %
In-hospital mortality overall, %

45 (21.3)
72.0
73.1
96.1 / 3.9
28.3 (11.6)
34.7 (11.6)
58.8
53.0
19.8
34.2
14.8
51.7
120.1 (34.3)
92 (24)
10.6 (4.7)
33.8
1,228 (910)
123 (29)
90 (21)
11.7 (2.7)
78 (22)
j3.1 (4.7)
14.2
24.0
5.6
1.7 (5.0)
1.3 (4.5)
52.3
10.4
12.2 (13.8)
26.5 (27.1)
7.8 (12.0)
6.5
16.2

48 (22)
65.7
74.3
96.0/4.0
33.4 (13.4)
41.6 (15.1)
69.6
59.9
22.6
36.6
15.0
60.3
113 (39)
94 (29)
8.5 (4.9)
53.1
1,420 (992)
116 (33)
93 (24)
10.9 (2.9)
70 (24)
j4.3 (5.3)
18.7
37.7
11.6
3.2 (7.0)
2.5 (6.5)
100
24.1
18.2 (17.6)
28.5 (28.7)
13.9 (14.9)
13.1
34.1

44 (21)
75.1
72.6
96.2/3.8
25.8 (9.7)
31.4 (12.1)
53.5
53.0
18.4
33.0
14.7
47.5
123 (31)
92 (22)
11.6 (4.3)
24.1
1,133 (851)
126 (27)
89 (19)
12.1 (2.5)
82 (20)
j2.3 (4.0)
12.0
17.4
2.7
1.0 (3.5)
0.7 (2.9)
29.0
3.8
9.2 (10.4)
25.6 (26.2)
4.8 (8.3)
3.3
7.5

p
G 0.001
G0.001
G0.001
0.30
G0.001
G0.001
G0.001
G0.001
G0.001
G0.001
G0.001
G0.001
G0.001
G0.001
G0.001
G0.001
G0.001
G0.001
G0.001
G0.001
G0.001
G0.001
G0.001
G0.001
G0.001
G0.001
G0.001
G0.001
G0.001
G0.001
G0.001
G0.001
G0.001
G0.001

*Blood products transfused between ED arrival and ICU admission.

FFP, fresh frozen plasma; LOS, length of stay.

* 2014 Lippincott Williams & Wilkins

Copyright 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

923

J Trauma Acute Care Surg

Volume 76, Number 4

Frohlich et al.

TABLE 2. Organ Systems Failed During Hospital Course

(n = 31,154)
n

All Enclosed MOF, 16,283

No MOF,
Patients, 31,154
(52.3%)
14,871 (47.7%)

Single organ failure, %

Organ failure
Lung, %
Coagulation
system, %
Liver, %
Heart/circulation,%
CNS, %
Kidney

52.3

100

29.0

G0.001

26.1
12.0

64.7
31.5

7.4
2.5

G0.001
G0.001

3.2
32.5
29.5
5.4

9.5
82.1
96.0
15.5

0.2
8.3
10.3
0.4

G0.001
G0.001
G0.001
G0.001

Basic Characteristics 2002 to 2011

During the study period, we observed an increasing age
of the study population. In 2002 to 2003, the mean age was
40.5 years compared with 46.7 years in 2010 to 2011. In contrast, trauma patients, who developed MOF, were constantly
approximately 5 years older ( p G 0.0001, Table 3). The rate
of male patients and the injury severity marked by ISS remained constant. While the sepsis rate remained unchanged
in non-MOF patients (approximately 4%), signicantly less
MOF patients developed sepsis (30% in 2002Y2003 vs. 21%
in 2010Y2011, p G 0.0001).

Predictors of MOF From Multivariate Analysis

A total of 10 variables were entered into the stepwise
logistic regression. The following independent risk factors for
the development of MOF after multiple trauma were calculated:
male sex, age, ISS, head AIS score of 3 or higher, thoracic AIS
score of 3 or higher, GCS score of 8 or lower, coagulopathy,
systolic blood pressure (SBP) at ED admission of 90 mm Hg or
less, base excess of less than j3 at ED admission, and number

Figure 1. Mortality in all trauma patients (n = 31,154)

and in the subgroups of trauma patients who developed no,
single-, or multiple-organ failure from 2002 to 2011.
924

Figure 2. Incidence of any organ failure and MOF in multiple

trauma patients (n = 31,154, 2002Y2011).

of red blood cell units transfused. The nal model is presented

in Table 4. For the model, Nagelkerkes r2 was 0.278.

DISCUSSION
The presented retrospective analysis of a nationwide prospective database, the TR-DGU, displays the incidence, mortality,
and risk factors of MOF in 31,154 multiple trauma patients over
one decade. To our knowledge, the presented study is the largest
analysis on postinjury MOF to date.
Our main ndings were as follows. First, the incidence
of posttraumatic MOF increased from 26% in 2002 to 33%
in 2011. Second, MOF-related mortality decreased from 43%
in 2002 to 34% in 2011. Third, various independent risk factors
were derived as strong predictors of posttraumatic MOF.

Epidemiology
During the entire study period, we observed an incidence
of MOF of 33% in patients with ISS of 16 or higher. This
nding resembles previously published single-center data reporting MOF rates of 25% to 40% among trauma populations.3,7,9,20 A recently published study with data from 2005
to 2010 by Dewar el al.11 described a lower incidence of 15%
and an MOF-related mortality of 24%. The authors discussed
if the difference from previous studies might be caused by
modern resuscitation strategies or a changing population.11
Compared with the study of Dewar et al.,11 our cohort showed
very similar age (48 years vs. 47 years) and trauma severity
(ISS, 33 vs. 32). However, during the study period, we observed
an increasing age in trauma patients and especially the subgroup of MOF patients, reecting the general demographic
changes of the aging German population.
In this context, the TR-DGU uses the SOFA score to
determine organ failure, while other analyses have applied other
scores, such as the Denver MOF score,3,11,21 the Marshall score,5
or a modied MOF score based on the criteria of Goris et al.,7,22
which might explain differences in calculated incidences.
As described earlier, the SOFA score takes into account
the function of the CNS, which might represent a bias in trauma
populations. In our cohort, MOF patients had isolated traumatic
brain injury in 15% and combined traumatic brain injury in
* 2014 Lippincott Williams & Wilkins

Copyright 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

J Trauma Acute Care Surg

Volume 76, Number 4

Frohlich et al.

TABLE 3. Basic Characteristics, Injury Pattern, and Outcome of the Subgroup of Patients Who Developed MOF During the Hospital
Course, Stratified According to Their Respective Trauma Years (n = 10,201)
MOF Patients
n
Age, mean, y
Male, %
ISS, points
Sepsis, %
In-hospital mortality, %

2002Y2003

2004Y2005

2006Y2007

2008Y2009

2010Y2011

861
45.9
76
33
30
43

1217
45.3
73
32
26
36

2093
46.62
75
34
27
31

2613
49.54
74
33
23
35

3417
50.47
74
33
21
33

60%. During the ICU stay, assessment of CNS function might

be biased by ventilated and sedated patients who required
mechanical ventilation because of extracranial injuries (e.g.,
thoracic trauma).
Since an increasing number of patients survived severe
trauma, this higher rate of survivors potentially develops MOF,
explaining the increase in MOF incidence in our analysis. During
the entire decade, deceased MOF patients survived nearly 3 days
longer than deceased non-MOF patients. This is in line with a
recently published review that describes MOF next to sepsis
as the predominant cause of death in trauma patients who die
later than 1 week after trauma.23 Therefore, early trauma deaths
are caused by CNS injury and exsanguination.23 Presumably,
these early victims do not survive long enough to develop MOF.
Furthermore, in 2011, MOF patients died 2 days earlier
than in 2002. In synopsis, with the increased MOF incidence
and decreased mortality during this period, there are severe
courses of MOF that cannot be intercepted even by improved
trauma care. Improvements in the treatment of the critical ill,
however, were described and implied in clinical praxis during
the observed period. For example, lung protective ventilation
and24 the use of intensive insulin therapy25 or cortisol replacement therapy for acute adrenal insufciency26 have been
shown to reduce mortality and improve outcome in critically
ill patients.

MOF Risk Factors

The present study showed a variety of independent risk
factors of MOF (Table 4). As previously described, predictors
may be xed, such as age and injury pattern/severity, or modiable such as received blood products or uncorrected lactate
levels.2 These clinical predictors with a high predictive value
in our analysis as GCS, transfused pRBCs, and base decit
are available in nearly all situations of acute trauma care. Especially, the base decit was recently discussed to identify the
presence of hypovolemic shock and to risk-stratify patients in
need of early blood product transfusion.27 Irrespective of the
occurrence of MOF, base excess is one of the most important
predictors of mortality after severe trauma.28 Accompanied by
the reliable prediction of MOF, this could help identify MOF
patients at an early stage and might inuence treatment.
In a recent study, Minei et al.9 analyzed prospectively
during a 4-year period the incidence and implications of posttraumatic MOF in approximately 1,000 patients of seven US
trauma centers. Similar to our data, the ndings of the authors
described an overall MOF incidence of 29.4%. However,
one major difference to our study was the exclusion of head

p
G0.001

G0.001
G0.001

injuries.9 In their publication, Minei et al. did not explain the

reason for the exclusion of traumatic brain injury. However,
during our study design, we felt that because of the incidence
of more than 50% of traumatic brain injury in patients with
multiple injuries, the inclusion would image the clinical practice
more realistically. They found 11 independent predictors applying an independent model as well, including male sex, ISS,
blood units, severe thoracic trauma, and base decit, identical
to our predictive model. Another major predictor in that study
was the use of vasopressors, which is in accordance to the risk
factor SBP e 90 mm Hg in our calculation, since vasopressors are not documented in our database.9
Finally, a reliable prediction model might be supportive for patient recruitment and inclusion in trauma studies, in
which MOF marks the primary end point. Recently, Holcomb

TABLE 4. Independent Risk Factors for the Development of

MOF After Multiple Trauma, Derived From a Multivariate
Analysis Using a Stepwise Logistic Regression Model With
Forward Variable Selection
Variable
Entered
Male sex
ISS, point
pRBC units*
0
1 to 9
Q10
Head AIS score Q 3
Thoracic AIS
score Q 3
Coagulopathy
GCS score e 8
at scene
Age, y
SBP e 90 mm Hg
Base excess
j3 to j6
j6 to j15
less than j15
Constant

Regression
Adjusted Odds Ratio,
Coefcient A eA (95% Condence Interval)

0.298
0.033

1.347 (1.61Y2.03)
1.033 (1.030Y1.037)

G0.001
G0.001

0.578
1.077
0.416
0.097

1.782 (1.638Y1.939)
2.935 (2.531Y3.405)
1.515 (1.395Y1.646)
1.102 (1.022Y1.188)

G0.001
G0.001
G0.001
G0.001

0.478
0.854

1.614 (1.499Y1.737)
2.356 (2.185Y2.541)

G0.001
G0.001

0.016
0.408

1.016 (1.015Y1.018)
1.504 (1.336Y1.657)

G0.001
G0.001

0.331
0.448
0.931
j3.767

1.392 (1.276Y1.519)
1.565 (1.414Y1.731)
2.536 (1.881Y3.420)

G0.001
G0.001
G0.001

*pRBCs transfused between ED arrival and ICU admission.

The logistic regression model was started with 12 variables. The following variables
were excluded by the model: Abdominal AIS score of 3 or higher; Extremities AIS score of
3 or higher.

* 2014 Lippincott Williams & Wilkins

Copyright 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

925

J Trauma Acute Care Surg

Volume 76, Number 4

Frohlich et al.

et al.29 have highlighted that 30-day mortality represents a

problematic end point in emergency trauma research. The authors emphasize that new, clinically relevant end points should
be considered, which should take into account the current
trauma epidemiology.
The risk factors represent formal conditions that are applicable during the initial phase of trauma care to stratify and
include/exclude patients for clinical trials and in quantifying
outcome and success of such trials.

Limitations
For all participating hospitals, regular audits are conducted and sample tests are performed to ensure data quality.
However, the validity of the documentation is not controlled
by external monitoring as in prospective trials.30 Furthermore,
the applied conventional statistical analyses are problematic
in large populations. Because of the large sample size, generally small p values were calculated. Therefore, interpretation
has to consider the clinical importance of the observed differences. The day of onset of both organ failure and sepsis is not
documented in the trauma registry. Furthermore, the cause of
death is not documented in the TR-DGU; therefore, MOFrelated mortality was indicated in this study.
The current analysis includes a European population
where the majority experiences blunt trauma, which might differ
from cohorts with a higher percentage of penetrating injury.

CONCLUSION
During a study period of 10 years, we observed an ongoing decrease of mortality after multiple trauma in a population of 31,154 patients. In the subgroup with MOF, mortality
decreased likewise. However, incidence of MOF in severely
injured increased signicantly. Thus, MOF after multiple trauma
remains a challenge in intensive care.
The risk factors from multivariate analysis could be instrumental in anticipating the early development of MOF. A
correspondent clinical score will be the objective of further
studies. Furthermore, a reliable prediction model might be supportive for patient enrolment in trauma studies, in which MOF
marks the primary end point.
AUTHORSHIP
M.F., R.L., A.W., and B.B. designed this study. R.L. analyzed the data,
which M.F., A.W., M.M.S., and B.B. interpreted. M.F. and A.W. wrote the
manuscript, which C.P., T.P., M.M.S., M.M., and S.G.S. critically reviewed.

DISCLOSURE
The authors declare no conflicts of interest.

REFERENCES
1. Ruchholtz S, Lefering R, Paffrath T, Oestern HJ, Neugebauer E, Nast-Kolb
D, Pape HC, Bouillon B. Reduction in mortality of severely injured patients in Germany. Dtsch Arztebl Int. 2008;105(13):225Y231.
2. Dewar D, Moore FA, Moore EE, Balogh Z. Postinjury multiple organ
failure. Injury. 2009;40(9):912Y918.
3. Ciesla DJ, Moore EE, Johnson JL, Burch JM, Cothren CC, Sauaia A. A
12-year prospective study of postinjury multiple organ failure: has anything
changed? Arch Surg. 2005;140(5):432Y438; discussion 438Y440.

926

4. Aldrian S, Koenig F, Weninger P, Vecsei V, Nau T. Characteristics of

polytrauma patients between 1992 and 2002: what is changing? Injury.
2007;38(9):1059Y1064.
5. Durham RM, Moran JJ, Mazuski JE, Shapiro MJ, Baue AE, Flint LM.
Multiple organ failure in trauma patients. J Trauma. 2003;55(4):608Y616.
6. Kahl JE, Calvo RY, Sise MJ, Sise CB, Thorndike JF, Shackford SR. The
changing nature of death on the trauma service. J Trauma Acute Care Surg.
2013;75(2):195Y201.
7. Nast-Kolb D, Aufmkolk M, Rucholtz S, Obertacke U, Waydhas C. Multiple organ failure still a major cause of morbidity but not mortality in blunt
multiple trauma. J Trauma. 2001;51(5):835Y841; discussion 841Y842.
8. Laudi S, Donaubauer B, Busch T, Kerner T, Bercker S, Bail H, Feldheiser
A, Haas N, Kaisers U. Low incidence of multiple organ failure after major
trauma. Injury. 2007;38(9):1052Y1058.
9. Minei JP, Cuschieri J, Sperry J, Moore EE, West MA, Harbrecht BG,
OKeefe GE, Cohen MJ, Moldawer LL, Tompkins RG, et al. The changing
pattern and implications of multiple organ failure after blunt injury with
hemorrhagic shock. Crit Care Med. 2012;40(4):1129Y1135.
10. Sauaia A, Moore FA, Moore EE, Haenel JB, Read RA, Lezotte DC. Early
predictors of postinjury multiple organ failure. Arch Surg. 1994;129(1):
39Y45.
11. Dewar DC, Tarrant SM, King KL, Balogh ZJ. Changes in the epidemiology
and prediction of multiple-organ failure after injury. J Trauma Acute Care
Surg. 2013;74(3):774Y779.
12. Frink M, Kuhne C, Debus F, Pries A, Ruchholtz S. The TraumaNetzwerk
DGU project. Goals, conception, and successes achieved [in German].
Unfallchirurg. 2013;116(1):61Y71; quiz 72Y73.
13. Trauma register of the German Society of Trauma Surgery. Scoring
study committee of the German Society of Trauma Surgery. Unfallchirurg.
1994;97(4):230Y237.
14. Brockamp T, Nienaber U, Mutschler M, Wafaisade A, Peiniger S, Lefering
R, Bouillon B, Maegele M. Predicting on-going hemorrhage and transfusion requirement after severe trauma: a validation of six scoring systems
and algorithms on the TraumaRegister DGU\. Crit Care. 2012;16(4):
R129.
15. Wafaisade A, Lefering R, Bouillon B, Sakka SG, Thamm OC, Paffrath T,
Neugebauer E, Maegele M. Epidemiology and risk factors of sepsis after
multiple trauma: an analysis of 29,829 patients from the Trauma Registry
of the German Society for Trauma Surgery. Crit Care Med. 2011;39(4):
621Y628.
16. Vincent JL, de Mendonca A, Cantraine F, Moreno R, Takala J, Suter PM,
Sprung CL, Colardyn F, Blecher S. Use of the SOFA score to assess the
incidence of organ dysfunction/failure in intensive care units: results of
a multicenter, prospective study. Working group on sepsis-related problems of the European Society of Intensive Care Medicine. Crit Care
Med. 1998;26(11):1793Y1800.
17. Vincent JL, Moreno R, Takala J, Willatts S, De Mendonca A, Bruining H,
Reinhart CK, Suter PM, Thijs LG. The SOFA (Sepsis-related Organ
Failure Assessment) score to describe organ dysfunction/failure. On behalf
of the Working Group on Sepsis-Related Problems of the European
Society of Intensive Care Medicine. Intensive Care Med. 1996;22(7):
707Y710.
18. Bone RC. Lets agree on terminology: denitions of sepsis. Crit Care Med.
1991;19(7):973Y976.
19. Wafaisade A, Wutzler S, Lefering R, Tjardes T, Banerjee M, Paffrath T,
Bouillon B, Maegele M. Drivers of acute coagulopathy after severe trauma:
a multivariate analysis of 1987 patients. Emerg Med J. 2010;27(12):
934Y939.
20. Brattstrom O, Granath F, Rossi P, Oldner A. Early predictors of morbidity
and mortality in trauma patients treated in the intensive care unit. Acta
Anaesthesiol Scand. 2010;54(8):1007Y1017.
21. Dewar DC, Mackay P, Balogh Z. Epidemiology of post-injury multiple
organ failure in an Australian trauma system. ANZ J Surg. 2009;79(6):
431Y436.
22. Goris RJ, te Boekhorst TP, Nuytinck JK, Gimbrere JS. Multiple-organ
failure. Generalized autodestructive inammation? Arch Surg. 1985;120(10):
1109Y1115.

* 2014 Lippincott Williams & Wilkins

Copyright 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

J Trauma Acute Care Surg

Volume 76, Number 4

Frohlich et al.

23. Pfeifer R, Tarkin IS, Rocos B, Pape HC. Patterns of mortality and causes of
death in polytrauma patientsVhas anything changed? Injury. 2009;40(9):
907Y911.
24. Hirvela ER. Advances in the management of acute respiratory distress
syndrome: protective ventilation. Arch Surg. 2000;135(2):126Y135.
25. van den Berghe G, Wouters P, Weekers F, Verwaest C, Bruyninckx F,
Schetz M, Vlasselaers D, Ferdinande P, Lauwers P, Bouillon R. Intensive
insulin therapy in critically ill patients. N Engl J Med. 2001;345(19):
1359Y1367.
26. Cooper MS, Stewart PM. Corticosteroid insufciency in acutely ill patients. N Engl J Med. 2003;348(8):727Y734.
27. Mutschler M, Nienaber U, Brockamp T, Wafaisade A, Fabian T, Paffrath T,
Bouillon B, Maegele M. Renaissance of base decit for the initial
assessment of trauma patients: a base decit-based classication for
hypovolemic shock developed on data from 16,305 patients derived from
the TraumaRegister DGU\. Crit Care. 2013;17(2):R42.
28. Rixen D, Raum M, Bouillon B, Neugebauer E. Base excess as prognostic
indicator in patients with polytrauma [in German]. Anasthesiol Intensivmed
Notfallmed Schmerzther. 2002;37(6):347Y349.
29. Holcomb JB, Weiskopf R, Champion H, Gould SA, Sauer RM, Brasel K,
Bochicchio G, Bulger E, Cotton BA, Davis D, et al. Challenges to
effective research in acute trauma resuscitation: consent and endpoints.
Shock. 2011;35(2):107Y113.
30. Lefering R, Paffrath T. Reality of care based on the data from the Trauma
Registry of the German Society of Trauma Surgery [in German]. Unfallchirurg.
2012;115(1):30Y32.

DISCUSSION
Dr. Saman Arbabi (Seattle, Washington): Thank you. I
want to thank the AAST for the opportunity to discuss this
well-presented paper. Dr. Frohlich and colleagues performed a
retrospective analysis of the German National Trauma Registry. They focused their analysis to trauma patients with ISS of
16 or higher for the last ten years.
They demonstrated a slight decrease in overall mortality,
from 18% to 15%. They also demonstrated an increase in
multiple organ failure, MOF, from 25% to 32%. However, the
mortality for MOF patients decreased signicantly. I have one
comment and a question.
You suggested that the increase in multiple organ failure
may be due to a decrease in mortality, where the surviving
patients may be sicker with MOF. However, there was only a
3% decrease in your mortality. And even if all these patients
developed MOF, it still does not explain the increased rate of
MOF observed in your study.
Considering that your mortality for the MOF group
decreased over the time period, it appears that the MOF patients in later years had less severe burden of disease. I wonder
if the observed increased rate in MOF was not a real increase.
Since the organ failure in your data registry was yes/no,
maybe over the years there was an increased emphasis in the
recognition of organ failure. Please comment.
And my question, assuming that you can predict MOF,
how would you change the treatment of the patients at high risk to
develop MOF? Do you have a specic treatment to prevent it?
Again, I enjoyed the paper and thank you very much.
Dr. Zsolt Balogh (Newcastle, Australia): Impressive
data. My comment is all about how you dene the problem,
namely MOF. You used SOFA score, which is a very sensitive
score, not so specic to trauma, actually never been properly

evaluated for trauma. Because of this, you have a very, very

high incidence of MOF that we dont see anymore. I wonder
how your data would look if you re-evaluated with leaving out
at least the CNS or gut or platelet score from the SOFA score.
My understanding is that in Germany many of these patients
get prehospital intubation and arrive to the ICU straight away
and score high on the CNS part of the SOFA score, while in the
Denver score we leave out the head injury patients straight
away because they just die differently and they have different
complications.
I think its worthwhile to look at that. This could change
your prediction model big time because we dont see any more
that shock is a signicant independent predictor these days like
it was 15 years ago. Your data is exactly as it was 15 years ago
as age, injury severity, and shock severity are still high up
amongst the predictors, while these days we see that these
predictors are not strong any more in our MOF database.
If you could comment on the scoring, specically, that
would be good. Thanks.
Dr. Matthias N. Frohlich (Cologne, Germany): Dr.
Arbabi, Dr. Balogh, thank you for your insightful comments.
With the recognition of organ failure with the SOFA
score, its with all the intubated patients the difculty to assess
the CNS malfunction.
And assessment for the trauma registry, these patients are
left out or to explain it more in detail, we used in that situation
only ve of the six points of the SOFA score to dene the
multiple organ failure.
For the further decrease of mortality after multiple organ
failure, we did not nd any further explanation. Its a good
thought. And we have to look at an explanations and further
studies. But so far this study was an epidemiologic assessment
of that registry and we do not have any solutions what we can
do with the patients at risk to develop a multiple organ failure.
With the prediction model we can raise awareness. But
its true that it does not affect the treatment so far.
Thank you very much.

EDITORIAL CRITIQUE
When studies collide
This important study from the German group conrms
previous evidence that multiple organ failure (MOF) remains a
challenge in severely injured patients. The investigators report
a signicant increase in MOF incidence but a decrease in casefatality rate. The reader may recall a recently published Glue
Grant-based study, presented at the last 2013 AAST meeting,
which also concluded that postinjury MOF is still resourceintensive, morbid and lethal, but reported different temporal
trends: in the Glue Grant-based dataset the MOF incidence
decreased over time while case-fatality rate remained stable.
We are left with two apparently disparate conclusions. How to
reconcile them so the messages can be appropriately translated
to our clinical practice and/or advance our research agenda?
In order to better understand these differences, we should
rst invoke the PICO (Problem/Population, Interventions,
Comparator, and Outcomes) framework to determine to which
population each study specically applies and also to ensure

* 2014 Lippincott Williams & Wilkins

Copyright 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

927

J Trauma Acute Care Surg

Volume 76, Number 4

Frohlich et al.

that the Outcomes (i.e., postinjury MOF) are indeed the same.
Once we do this, the differences are glaring. The entry criteria
for the two studies were different: the patients enrolled in the
Glue Grant study were more severely injured (i.e., blunt torso
trauma with hemorrhagic shock; all required at least 1 pRBC/
12 hours) than the German group (24% required 1 PRBC
between ED arrival and ICU admission). The German study
population included a large proportion of victims of traumatic
brain injury (TBI: 58.8% had AIS Head >=3), while the Glue
Grant study specically excluded TBI patients. Second, the two
studies used different denitions of the Outcome postinjury
MOF. The German investigators employed the SOFA score,
which assess the dysfunction of six organ systems including
the central nervous system (CNS). In contrast, the Glue Grant
investigators used the Denver MOF score, which does not
assess the CNS, and a modied version of the Marshall Multiple Organ Dysfunction Score without its CNS component.

928

Thus, the two studies apply to different populations and to

somewhat different outcomes.
In addition, the two studies employed distinct statistical
treatment of potential confounding of the temporal trends (age,
injury severity, obesity, comorbid conditions, resuscitation,
etc.). Finally, the large size of the German dataset compared to
the Glue Grant has well-known implications for statistical
signicance. There are several other dissimilarities between
two studies, and I invite the reader to be meticulous about them
so the translation to practice and future research is adequately
done. All in all, the two studies are very relevant and advance
our knowledge about this still resource-intensive, morbid and
lethal condition.
Angela Sauaia, MD, PhD
Denver, Colorado

* 2014 Lippincott Williams & Wilkins

Copyright 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.