Injury, Int. J.

Care Injured 45 (2014) 13121319

Contents lists available at ScienceDirect

Injury
journal homepage: www.elsevier.com/locate/injury

Advancing age and trauma: Triage destination compliance and

mortality in Victoria, Australia
Shelley Cox a,b,*, Chris Morrison a,b, Peter Cameron b, Karen Smith a,b
a
b

Research & Evaluation Department, Strategy, Research & Innovation Division, Ambulance Victoria, Australia
Department of Epidemiology & Preventive Medicine, Monash University/The Alfred Hospital, Australia

A R T I C L E I N F O

A B S T R A C T

Article history:
Accepted 20 February 2014

Objective: To describe the association between increasing age, pre-hospital triage destination
compliance, and patient outcomes for adult trauma patients.
Methods: A retrospective data review was conducted of adult trauma patients attended by Ambulance
Victoria (AV) between 2007 and 2011. AV pre-hospital data was matched to Victorian State Trauma
Registry (VSTR) hospital data. Inclusion criteria were adult patients sustaining a traumatic mechanism of
injury. Patients sustaining secondary traumatic injuries from non-traumatic causes were excluded. The
primary outcomes were destination compliance and in-hospital mortality. These outcomes were
evaluated using multivariable logistic regression.
Results: There were 326,035 adult trauma patients from 2007 to 2011, and 18.7% met the AV prehospital trauma triage criteria. The VSTR classied 7461 patients as conrmed major trauma (40.9% > 55
years). Whilst the trauma triage criteria have high sensitivity (95.8%) and a low under-triage rate (4.2%),
the adjusted odds of destination compliance for older trauma patients were between 23.7% and 41.4%
lower compared to younger patients. The odds of death increased 8% for each year above age 55 years
(OR: 1.08; 95% CI: 1.07, 1.09).
Conclusions: Despite effective pre-hospital trauma triage criteria, older trauma patients are less likely to
be transported to a major trauma service and have poorer outcomes than younger adult trauma patients.
It is likely that the benet of access to denitive trauma care may vary across age groups according to
trauma cause, patient history, comorbidities and expected patient outcome. Further research is required
to explore how the Victorian trauma system can be optimised to meet the needs of a rapidly ageing
population.
Crown Copyright 2014 Published by Elsevier Ltd. All rights reserved.

Keywords:
Pre-hospital
Trauma
Triage
Destination compliance
Mortality
Advancing age
Major trauma service
Denitive care

Introduction
Consistent with other developed countries, the population in
Australia is ageing rapidly [1]. The Australian Bureau of Statistics
(ABS) has predicted that by the year 2056, 25% of the Victorian
population will be aged 65 years [2]. With an increased
proportion of older people in the community, comes an increase
in the number of older trauma patients attended by emergency
medical services and triaged to hospital emergency departments
[3]. It is unclear whether current clinical management protocols

* Corresponding author at: Research & Evaluation Department, Strategy,

Research & Innovation Division, Ambulance Victoria, 31 Joseph Street, Blackburn
North, 3130, Victoria, Australia. Tel.: +61 3 9896 6089.
E-mail address: Shelley.Cox@ambulance.vic.gov.au (S. Cox).
http://dx.doi.org/10.1016/j.injury.2014.02.028
00201383/Crown Copyright 2014 Published by Elsevier Ltd. All rights reserved.

are optimal for the care of this patient group [3]. Understanding the
relationship between age and trauma is critical to improving the
management and outcomes of older trauma patients [4].
Trauma triage is predicated on the idea of getting the right
patient to the right hospital as quickly as possible [5,6]. It is
widely accepted that morbidity and mortality can be reduced by
effective triage of trauma patients to specialised trauma hospitals
[79]. There is evidence to suggest that the adjusted risk of
death is 25% lower when care is provided at a specialised trauma
service [10].
Compared to younger patients, older trauma patients are
known to have poorer outcomes and require far less forceful
mechanisms to produce serious injuries [1114]. Age related
physiological changes, together with pre-existing medical conditions and medications (e.g. anticoagulants, antiplatelets)
can further complicate traumatic injuries and result in worse

S. Cox et al. / Injury, Int. J. Care Injured 45 (2014) 13121319

outcomes for older patients [15,16]. With advancing age, outcomes

for trauma patients have been shown to deteriorate, with each one
year increase in age resulting in an increase in mortality of 6%
[11,17].
Despite the known benets of denitive trauma care, studies
have shown that older trauma patients are less likely to be triaged
to specialised trauma services [1826]. Proposed reasons for
hospital destination non-compliance for older trauma patients
include age-inappropriate pre-hospital trauma triage criteria
[19,20,23,27,28], proximity to specialised trauma services [24],
patient/family destination preference [29], paramedic discretionary decision making [30] and patient acuity [23]. While numerous
international studies have investigated factors associated with the
triage of older trauma patients, the drivers of triage destination
compliance have not been investigated in the Australian prehospital environment.
With an ageing population it is important to have in place a
trauma system that is optimised to care for the needs of older
trauma patients. Identifying potential factors that may contribute
to poor outcomes for older patients is a critical rst step in
determining how the trauma system may be optimised to meet the
demands of a rapidly ageing Victorian population. This descriptive
study aims to:

1313

The Victorian state adult pre-hospital trauma triage criteria [35]

have been described previously [6]. The current triage criteria
consist of physiological, anatomical, mechanistic and logistic
elements [35]. When a trauma patient meets one or more of these
criteria, and ambulance transport time is less than 30 min, the
patient should be transported to a MTS, or next highest level of care
within 30 min [6,33]. Based on geographic location, the majority of
trauma scenes within metropolitan Melbourne are considered to
be within 30 min of the MTSs. This does not take into account
possible confounders like trafc congestion, road works, time of
day or day of week. No regional scene locations are considered to
be within 30 min of the MTSs by road, however the same triage
criteria apply to rural cases, and air transport can be arranged for
more severe trauma cases.
Patients aged 16 years who experienced a traumatic mechanism between January 1st 2007 and December 31st 2011 were
included in this study. Data were extracted for blunt, penetrating
and burns related trauma. Injured patients were only excluded if
injuries were secondary to a non-traumatic cause. Patients who
meet one or more of in-hospital death, ISS > 12, ICU admission
with mechanical ventilation >24 h or urgent surgery are dened as
conrmed major trauma by the VSTR [36].
Methods of measurement, outcome variables and covariates

1. Determine the clinical utility of the Victorian State adult prehospital trauma triage criteria overall, and for older and younger
trauma patients.
2. Determine whether the clinical utility of the pre-hospital
trauma triage criteria matches actual system performance with
respect to trauma triage destination compliance.
3. Investigate whether destination non-compliance leads to poorer
outcomes for older trauma patients.

A retrospective review of data of all adult trauma patients

attended by AV during a ve-year time period (20072011) was
conducted.

Pre-hospital data were captured via VACIS, an in-eld electronic

data capture application that is linked to the AV data warehouse.
Hospital trauma data were sourced electronically from the VSTR.1
Pre-hospital and hospital data were matched via probabilistic
linkage using LinkageWiz Software (LinkageWiz Inc, Adelaide, SA).
Record linkage was manually reviewed to ensure a 100%
match rate.
Destination compliance is dened as access to the highest
level trauma service within 30 min transport time, for patients
who meet the pre-hospital trauma triage criteria. Mortality is
dened as in-hospital death. Covariates included age, gender,
trauma cause, injury severity, paramedic type, vital signs,
mechanism of injury, pain score  3, signicant comorbidity,
inter-hospital transfer, air ambulance transport, transport time,
injury count, paramedic judgement of major trauma and scene
location region.

Study setting and selection of participants

Statistical analyses

The study was conducted in Victoria, a state of Australia that

covers 227,590 km2 [31]. During the study period Victoria had a
population of approximately 5.4 million people, 4 million of whom
lived in metropolitan Melbourne [32].
Established in 2000, the Victorian State Trauma System (VSTS)
coordinates all pre-hospital paramedic and acute medical
services across the state of Victoria. The VSTS works to reduce
mortality and morbidity by matching severely injured patients
with an appropriate level of care in a timely manner [5]. Victoria
follows a two-tiered emergency response model, which has been
described previously [5,6,33]. There are two adult hospitals and
one paediatric hospital designated as major trauma services
(MTSs), which are the equivalent of international level-1 trauma
centres [5]. Several metropolitan and regional hospitals have
lower levels of designation. These hospitals are hierarchical and
provide lower levels of trauma care delivery prior to patient
transfers to MTSs [5,6,34]. Within the VSTS patients can also
be transferred from rst receiving hospital to one of three
specialised metropolitan neurological or spinal hospitals designated to treat isolated head or spinal injuries. This is an interhospital transfer protocol, which is not currently indicated in the
pre-hospital trauma triage criteria.

Analyses were performed using SPSS Version 20.0 (SPSS

Inc, Chicago. IL). Statistical signicance was set at p < 0.05.
Summary statistics were used to describe the characteristics of
trauma patients. Chi square tests for categorical variables were
performed to compare proportions across older and younger
patient groups. Categorical data is summarised as counts and
percentages.
Sensitivity (a/(a + c)), specicity (d/(b + d)), accuracy (a + d/
a + b + c + d),2 under-triage (1 sensitivity) and over-triage
(1 specicity) were calculated for the current Victorian adult
pre-hospital trauma triage criteria.3 Over and under-triage values
were based on the number of trauma patients identied by trauma
triage criteria, not on triage destination.
Logistic regression analyses were performed to determine
univariate associations and to construct two multivariable

Methods
Study design

1
Rural data for road transported conrmed major trauma patients were not
available at the time of the study.
2
a = true positive; b = false positive; c = false negative; d = true negative.
3
Patients with insufcient information for classication according to the prehospital trauma triage criteria were excluded from calculations of diagnostic
statistics.

S. Cox et al. / Injury, Int. J. Care Injured 45 (2014) 13121319

1314

Table 1
Summary of current pre-hospital trauma triage criteria for adult conrmed major trauma patients.
Pre-hospital triage criteria

Age > 55 years (n = 3054)

Age  55 years (n = 4407)

p-value

Physiological vital signs

Pulse rate > 124
Respiratory rate <12 or >24
Systolic blood pressure < 90 mmHg
Conscious state: GCS < 13
Oxygen saturation < 90%
Specic injuries
Penetrating wound
Burns > 10%/respiratory tract
2 long bone fractures
Blunt abdomen
Amputation/limb threat
Head injury
Isolated Head Injury
Fractures open major
Blunt chest
Fractured pelvis
Suspected spinal cord injury
Signicant crush injury
Mechanism of injury (MOI)
Prolonged extrication
Ejected from vehicle
Motor/cyclist collision > 30 kph
Fall from height > 3 m
Struck by falling object > 3 m
Explosion
Motor vehicle collision > 60 kph
Pedestrian impact
Patients meeting adult pre-hospital trauma triage criteria
Vital signs only
Injury only
MOI only
Aberrant vital signs + injury
Aberrant vital signs + MOI
Injury + MOI
Aberrant vital signs + injury + MOI
Patients not meeting adult pre-hospital trauma triage criteria
Patients that do not meet pre-hospital trauma triage criteria
Patients with insufcient pre-hospital data to be classied
Total

163
395
234
574
261

(5.3)
(12.9)
(7.7)
(18.8)
(8.5)

609
904
565
1125
658

(13.8)
(20.5)
(12.8)
(25.5)
(14.9)

0.001
0.001
0.001
0.001
0.381

53
103
448
207
50
1698
1023
71
699
128
488
11

(1.9)
(3.4)
(4.7)
(6.8)
(1.6)
(55.6)
(33.5)
(2.3)
(22.9)
(4.2)
(16.0)
(0.5)

313
253
795
658
81
2327
1046
263
1292
227
779
45

(7.1)
(6.0)
(8.5)
(15.6)
(1.9)
(54.8)
(23.7)
(6.3)
(30.7)
(5.4)
(18.5)
(1.3)

0.001
0.001
0.001
0.001
0.360
0.498
0.001
0.001
0.001
0.019
0.005
0.001

60
16
248
157
2
9
379
216

(2.0)
(0.5)
(8.1)
(6.4)
(0.1)
(0.3)
(12.4)
(7.1)

189
125
918
204
8
29
1061
274

(4.3)
(2.8)
(20.8)
(4.6)
(0.2)
(0.7)
(24.1)
(6.2)

0.001
0.001
0.001
0.545
0.350
0.060
0.001
0.146

178
1312
211
480
74
290
266

(5.8)
(43.0)
(6.9)
(15.7)
(2.4)
(9.5)
(8.7)

394
2055
21
1547
7
30
31

(8.9)
(46.6)
(0.5)
(35.1)
(0.2)
(0.7)
(0.7)

0.001
0.002
0.001
0.001
0.001
0.001
0.001

132 (4.3)
111(3.6)

173 (3.9)
149 (3.4)

3054 (100.0)

4407 (100.0)

0.395
0.557

Missing data for 1 vital sign: 4.2%; 2 vital signs: 1.0%; 3 vital signs: 0.4% and 4 vital signs: 6.0%. The pre-hospital trauma triage criteria were unable to be used to classify
just 3.5% of conrmed major trauma patients due to missing data (N = 260).

models. The primary outcomes of interest were destination

compliance (model 1) and mortality (model 2). To avoid issues of
non-linearity both models were constructed using segmented
partitioning [37,38] (or piecewise regression) for the continuous
variables transport time (model 1) and age and injury severity
(model 2). Breakpoints were pre-determined by the adult prehospital trauma triage criteria (transport time < 30 min and
age > 55 years) and the conrmed major trauma criterion
(ISS > 12) [36]. The spline method employed did not allow for
discontinuities at the breakpoints [38]. Odds ratios (OR) and 95%
condence intervals (95% CI) were reported to describe the
relationships between covariates and outcome variables. Goodness of t was assessed using the HosmerLemeshow statistic.
The receiver operating characteristic (ROC) curve was used
to evaluate the statistical discrimination of the regression
models [39].
To minimise case exclusion from the multivariable regression
analyses, missing values were imputed for Glasgow Coma Scale
(GCS) score, pulse rate, systolic blood pressure, respiratory rate
and ISS. The pattern of missing data was not missing completely
at random (MCAR) (Littles MCAR test: x2 = 410.1, p < 0.001).
Accordingly multiple imputation using a fully conditional

specication (FCS) procedure was deemed appropriate. FCS is

described in detail elsewhere [40]. Five data sets were imputed
and pooled regression estimates reported.
Results
Characteristics of all adult trauma patients
Between January 1st 2007 and December 31st 2011, Ambulance
Victoria (AV)4 attended 326,0355 adult trauma patients
aged  16 years. Older trauma patients (age > 55 years) were
more likely to be female (60.5% vs. 39.5% younger patients)
and sustain a fall from standing height (84.0%). Trafc related
incidents were common among younger adult patients (39.5%).
4
On July 1st 2008 the Metropolitan Ambulance Service, Rural Ambulance
Victoria and Alexandra and District Ambulance Service merged to form Ambulance
Victoria.
5
All pre-hospital road car data was captured electronically via VACIS and an
integrated Data Warehouse. This gure includes electronic metropolitan data from
1st January 2007 to 31st December 2011, and rural electronic data from 1st July
2008 to 31st December 2011. Electronic rural data for the time period 1st January
2007 to 30th June 2008 is not available.

S. Cox et al. / Injury, Int. J. Care Injured 45 (2014) 13121319

1315

major trauma patients. The mechanistic criteria identied

a larger proportion of older conrmed major trauma patients
(Table 1).
Clinical effectiveness of trauma triage criteria
A summary of the current adult pre-hospital trauma triage
criteria for conrmed major trauma patients is presented in
Table 1. These analyses are based on the number of patients
identied by the trauma triage criteria, regardless of transport
destination. The pre-hospital trauma triage criteria have an overall
sensitivity of 95.8%, a specicity of 79.2% and an overall accuracy of
79.6%. The over-triage rate is 20.8% and the under-triage rate is
4.2%. For older patients (age > 55 years), sensitivity is 95.5%,
specicity 80.1% and overall accuracy 80.4%. Over-triage for older
patients is 19.9% and under-triage is 4.5%. For younger adult
trauma patients (age  55 years), sensitivity is 95.9%, specicity
78.1% and overall accuracy 78.7%. Over-triage for younger patients
is 21.9% and under-triage is 4.1%.
Trauma triage destination compliance

Fig. 1. Destination compliance by age for patients identied by the adult prehospital trauma triage criteria (A); the association between age and mortality (B).

Comorbidities were documented on the pre-hospital patient care

records for 77.1% older and 23.0% younger trauma patients.
Hypertension (32.6%), diabetes (12.7%) and dementia (9.8%) were
common among older trauma patients. The southern metropolitan
region of Melbourne was the most common locality for trauma
incidents (22.0% older vs. 21.4% younger trauma patients). Private
residences (54.5%) and nursing homes or supported accommodation facilities (17.9%) were common scene locations for older
trauma patients.
Pre-hospital triage criteria characteristics of trauma patients
Of the 326,035 adult trauma patients attended by AV,
there were 60,751 trauma patients who met the Victorian
State adult pre-hospital trauma triage criteria and were transported to a hospital destination. Of these patients, the VSTR
identied 7461 adult conrmed major trauma patients, specically, 3054 (40.9%) older and 4407 (59.1%) younger patients.
Greater proportions of younger patients experienced vital
signs outside the normal range and serious injuries, compared
to older patients (Table 1). Blunt head trauma was the most
common injury for both patient groups, with isolated head
injuries sustained by a larger proportion of older conrmed

There was a steady decline in destination compliance with

age for all trauma patients who met the pre-hospital trauma
triage criteria (Fig. 1A). Of the older conrmed major trauma
patients who met the trauma triage criteria, 66.9% were
transported to a MTS, compared to 87.6% of younger adults
(p < 0.001). Of the older conrmed major trauma patients who
met the triage criteria but were not transported to a MTS, 28.2%
recorded vital signs outside the normal range, 10.5% had a
signicant mechanism of injury and 82.3% had one or more
serious injuries identied in the pre-hospital setting. Sixty-two
per cent of these injuries were identied by paramedics as blunt
head injuries.
The results of a univariate logistic regression analysis showed
that for every one-year increase in age, the odds of being
transported to a MTS decreased by 2 per cent (OR 0.982, 95% CI:
0.982, 0.983). When age was categorised according to the current
pre-hospital trauma triage criteria (age > 55 years), the unadjusted odds of being transported to a MTS for older trauma patients
was 56.9% lower (OR 0.431; 95% CI: 0.416, 0.446) than for younger
adult trauma patients.
Table 2 presents a multivariable logistic regression model for
the odds of destination compliance, for patients who met the prehospital trauma triage criteria (model 1, N = 60,751). The area
under the ROC curve for this model was 0.852, and the Hosmer
Lemeshow goodness-of-t statistic was signicant (p = 0.001),
however this is not unusual with a sample of this size. The tted
model correctly classied 79.4% of cases. After adjusting for
potential confounders, the odds of destination compliance for
older trauma patients were between 23.7% and 41.1% lower,
compared to trauma patients aged 16 to 25 years. This model was
replicated for subgroups of older trauma patients, namely patients
aged >55 years, >65 years and >75 years. Table 3 presents the
ndings of these regression analyses, which despite decreased
statistical power, are comparable to the original destination
compliance model.
Destination compliance and patient outcomes
For older conrmed major trauma patients, a larger proportion
of in-hospital deaths occurred at non-MTSs (32.1% vs. MTSs,
20.2%), while the opposite was true for younger adult patients
(non-MTSs, 4.3% vs. MTSs 6.8%). In contrast, larger proportions of
older patients who met the major trauma criteria for ISS > 12, ICU
admission or urgent surgery were transported to MTSs compared

S. Cox et al. / Injury, Int. J. Care Injured 45 (2014) 13121319

1316

Table 2
Multivariable logistic regression model of destination compliance for trauma patients who meet the pre-hospital
trauma triage criteria (model 1).
Unadjusted odds ratio
[95% CI] (p-value)

Adjusted odds ratio

[95% CI] (p-value)

Age category
1625 years
2635 years
3645 years
4655 years
5665 years
6675 years
7685 years
86+ years
Gender (female)

(Reference)
1.09 [1.02, 1.15]
0.95 [0.89, 1.01]
0.82 [0.76, 0.87]
0.62 [0.58, 0.66]
0.49 [0.46, 0.52]
0.35 [0.33, 0.37]
0.28 [0.26, 0.31]
0.53 [0.52, 0.55]

(Reference)
1.03 [0.95, 1.12]
0.90 [0.83, 0.97]
0.85 [0.78, 0.93]
0.76 [0.69, 0.83]
0.68 [0.62, 0.75]
0.58 [0.54, 0.64]
0.62 [0.56, 0.68]
0.74 [0.70, 0.77]

Trauma cause
Trafc relateda
Pedestrian collision
Fall
Assault
Penetrating injury
Struck by object
Other
Injury severity (ISS > 12)
Paramedic type (MICA)
Signicant comorbidity
Inter-hospital transfer
AAV transport
Transport time (30 min)
Transport time (>30 min)
Pre-hospital injury count
Hospital AIS injury count
Paramedic judgementb

(Reference)
2.14 [1.95, 2.33] (0.001)
0.40 [0.39, 0.42] (0.001)
0.71 [0.67, 0.76] (0.001)
3.24 [2.82, 3.73] (0.001)
0.54 [0.49, 0.60] (0.001)
0.58 [0.55, 0.63] (0.001)
10.9 [10.21, 11.69] (0.001)
4.72 [4.49, 4.96] (0.001)
0.46 [0.44, 0.47] (0.001)
0.51 [0.42, 0.61] (0.001)
33.9 [28.07, 40.88] (0.001)
1.05 [1.05, 1.06] (0.001)
1.01 [1.01, 1.02] (0.001)
2.18 [2.12, 2.24] (0.001)
2.43 [2.36, 2.49] (0.001)
4.53 [4.33, 4.75] (0.001)

(Reference)
2.35 [2.23, 2.50] (0.001)
0.66 [0.62, 0.69] (0.001)
0.66 [0.61, 0.72] (0.001)
1.83 [1.52, 2.20] (0.001)
0.45 [0.25, 0.48] (0.001)
0.57 [0.53, 0.62] (0.001)
2.80 [2.40, 3.25] (0.001)
2.11 [1.97, 2.27] (0.001)
0.78 [0.74, 0.82] (0.001)
0.31 [0.24, 0.41] (0.001)
31.82 [24.5, 41.3] (0.001)
1.04 [1.04, 1.05] (0.001)
1.02 [1.02, 1.02] (0.001)
1.33 [1.29, 1.38] (0.001)
1.47 [1.39, 1.55] (0.001)
2.71 [2.54, 2.89] (0.001)

AV region
Southern metropolitan
Eastern metropolitan
Northern metropolitan
Western metropolitan
Barwon South West
Gippsland
Grampians
Hume
Loddon Mallee
Unspecied

(Reference)
0.52 [0.49, 0.54] (0.001)
1.27 [1.21, 1.34] (0.001)
2.01 [1.91, 2.11] (0.001)
0.07 [0.06, 0.09] (0.001)
0.19 [0.16, 0.22] (0.001)
0.26 [0.22, 0.30] (0.001)
0.17 [0.15, 0.20] (0.001)
0.24 [0.21, 0.27] (0.001)
3.11 [2.79, 3.47] (0.001)

(Reference)
0.45 [0.42, 0.48] (0.001)
1.32 [1.24, 1.40] (0.001)
2.20 [2.07, 2.33] (0.001)
0.02 [0.02, 0.03] (0.001)
0.04 [0.03, 0.05] (0.001)
0.10 [0.08, 0.13] (0.001)
0.03 [0.02, 0.04] (0.001)
0.11 [0.09, 0.13] (0.001)
0.65 [0.54, 0.78] (0.001)

Variable

(0.006)
(0.104)
(0.001)
(0.001)
(0.001)
(0.001)
(0.001)
(0.001)

(0.480)
(0.015)
(0.001)
(0.001)
(0.001)
(0.001)
(0.001)
(0.001)

MICA: mobile intensive care ambulance, AAV: Air Ambulance Victoria.

a
Trafc related: motor vehicle collision, motorcycle collision, bicycle collision.
b
Paramedic judgement: Paramedic documentation of any one of major trauma, major blunt, or major
penetrating on the pre-hospital patient care record.

to non-MTSs. A similar pattern was observed for younger adult

patients.
There were 1034 (13.9%) conrmed major trauma patients
who died in hospital and 72.5% of these patients were aged > 55
years. Fig. 1B shows a steady increase in mortality with age. Of the
older conrmed major trauma patients who died in hospital,
74.4% had sustained a fall, 72.8% had one or more comorbidities
recorded by paramedics, 51.9% were transported to a MTS and
73.2% of these patients were aged over 75 years. One-quarter of
older patients who died in hospital met the VSTR major trauma
criteria on death alone and 72% of these patients were aged over
55 years and were transported to a non-MTS. The results of a
univariate logistic regression analysis showed that with each oneyear increase in age, the odds of in-hospital death increased by 4%
(OR 1.04; 95% CI: 1.03, 1.04). When age was dichotomised
according to the current pre-hospital trauma triage criteria
(age > 55 years), the unadjusted odds of in-hospital death for
older trauma patients was 4.72 times higher (95% CI: 4.07, 5.46)
than for younger adult trauma patients.
Table 4 presents a multivariable logistic regression model for the
odds of mortality, for conrmed major trauma patients (model 2,
N = 7461). The area under the ROC curve for this model was 0.89,
and the HosmerLemeshow statistic suggested a good model t

(p = 0.064). The tted model correctly classied 89.1% of cases. After

adjusting for potential confounders, the odds of death increased 8%
for each year above age 55 years (OR: 1.08; 95% CI: 1.07, 1.09).
Limitations
The study was limited by the inability to directly control for
destination compliance using estimated drive time and distance
from scene to MTS. A gross measure of compliance was used in the
logistic regression modelling, however further research is necessary using sophisticated GIS modelling to map drive time and
distance from scene to MTSs.
Patients with insufcient information for classication by the
pre-hospital trauma triage criteria were excluded from sensitivity
and under-triage calculations. If sufcient information was
available to nd that all the excluded patients legitimately did
not meet the triage criteria, the sensitivity would still be 92.4%
with an under-triage rate of 7.6%.
Despite the sophistication of VACIS and an integrated data
warehouse, it is possible that the trauma lters did not identify all
trauma patients. This does not affect sensitivity because all
conrmed major trauma patients were sourced from the VSTR,
however it is possible that specicity may be lower than reported.

S. Cox et al. / Injury, Int. J. Care Injured 45 (2014) 13121319

1317

Table 3
Multivariable logistic regression model of destination compliance for trauma patients who meet the pre-hospital
trauma triage criteria, by older age categories.
Variable

Age > 55 years

Age > 65 years

Age > 75 years

Gender (female)

0.81 [0.76, 0.86]

0.86 [0.80, 0.93]

0.96 [0.87, 1.05]

Trauma cause
Trafc relateda
Pedestrian collision
Fall
Assault
Penetrating injury
Struck by object
Other
ISS > 12
MICA paramedic(s)
Comorbidity
AAV transport
Transport time (30 min)
Transport time (>30 min)
Pre-hospital injury count
Hospital AIS injury count
Paramedic judgementb

(Reference)
2.25 [1.94, 2.62]
0.55 [0.51, 0.59]
0.62 [0.45, 0.88]
3.22 [1.69, 6.13]
0.44 [0.34, 0.57]
0.70 [0.59, 0.83]
1.98 [1.65, 2.38]
1.78 [1.59, 1.99]
0.73 [0.68, 0.79]
19.33 [12.9, 28.96]
1.06 [1.05, 1.06]
1.01 [1.01, 1.02]
1.31 [1.24, 1.38]
1.40 [1.31, 1.50]
2.87 [2.56, 3.22]

(Reference)
2.17 [1.80, 2.61]
0.52 [0.47, 0.57]
0.56 [0.32, 0.97]
4.22 [1.75, 10.18]
0.42 [0.28, 0.62]
0.75 [0.61, 0.93]
2.02 [1.65, 2.48]
1.66 [1.45, 1.91]
0.78 [0.70, 0.86]
8.98 [5.53, 14.58]
1.06 [1.05, 1.06]
1.01 [1.01, 1.02]
1.26 [1.18, 1.34]
1.42 [1.32, 1.54]
2.97 [2.58, 3.42]

(Reference)
1.91 [1.45, 2.46]
0.46 [0.40, 0.52]
0.86 [0.40, 1.86]
4.91 [1.31, 18.40]
0.47 [0.27, 0.81]
0.74 [0.57, 0.97]
2.15 [1.70, 2.73]
1.70 [1.42, 2.05]
0.83 [0.73, 0.95]
6.34 [3.25, 12.37]
1.06 [1.05, 1.07]
1.01 [1.00, 1.02]
1.26 [1.17, 1.36]
1.41 [1.28, 1.55]
2.64 [2.20, 3.17]

AV region
Southern metro
Eastern metro
Northern metro
Western metro
Barwon South West
Gippsland
Grampians
Hume
Loddon Mallee
Unspecied

(Reference)
0.36 [0.33, 0.40]
1.17 [1.07, 1.27]
1.89 [1.73, 2.06]
0.02 [0.01, 0.04]
0.05 [0.03, 0.07]
0.08 [0.06, 0.12]
0.02 [0.02, 0.04]
0.13 [0.10, 0.17]
0.56 [0.44, 0.72]

(Reference)
0.31 [0.28, 0.35]
1.09 [0.99, 1.21]
1.70 [1.53, 1.88]
0.03 [0.02, 0.05]
0.04 [0.03, 0.08]
0.07 [0.04, 0.12]
0.03 [0.02, 0.05]
0.12 [0.09, 0.17]
0.65 [0.49, 0.86]

(Reference)
0.28 [0.24, 0.32]
1.04 [0.92, 1.17]
1.56 [1.37, 1.76]
0.01 [0.01, 0.04]
0.03 [0.01, 0.07]
0.04 [0.02, 0.09]
0.02 [0.01, 0.05]
0.12 [0.08, 0.18]
0.56 [0.40, 0.80]

MICA: mobile intensive care ambulance, AAV: Air Ambulance Victoria.

a
Trafc related: motor vehicle collision, motorcycle collision, bicycle collision.
b
Paramedic judgement: Paramedic documentation of any one of major trauma, major blunt, or major
penetrating on the pre-hospital patient care record.
Table 4
Multivariable logistic regression model of predictors of in-hospital death for conrmed major trauma patients
(model 2).
Variable

Unadjusted odds ratio

[95% CI] (p-value)

Adjusted odds ratio

[95% CI] (p-value)

Age  55 years
Age > 55 years
ISS  12
ISS > 12
Gender (female)
Major trauma service
Paramedic type (MICA)
Underlying comorbidity

1.05
1.07
0.91
1.05
1.71
0.46
1.02
2.61

1.06]
1.08]
0.93]
1.05]
1.96]
0.53]
1.16]
2.98]

(0.001)
(0.001)
(0.001)
(0.001)
(0.001)
(0.001)
(0.801)
(0.001)

1.01 [1.00, 1.02] (0.185)

1.08 [1.07, 1.09] (0.001)
1.01 [0.89, 1.12] (0.944)
1.10 [1.09, 1.12] (0.001)
0.93 [0.78, 1.12] (0.441)
0.79 [0.64, 0.96] (0.018)
1.76[1.43, 2.17] (0.001)
1.27 [1.08, 1.58] (0.032)

Trauma cause
Trafc relateda
Pedestrian collision
Fall
Assault
Penetrating injury
Struck by object
Other
Aberrant vital signs
Mechanism of injury
Pain score  3b

(Reference)
2.84 [2.16, 3.72]
4.35 [3.68, 5.13]
0.84 [0.53, 1.28]
1.12 [0.71, 1.76]
1.33 [0.75, 2.34]
2.92 [2.19, 3.90]
2.78 [2.30, 3.35]
1.03 [0.84, 1.26]
0.41 [0.37, 0.46]

(0.001)
(0.001)
(0.452)
(0.635)
(0.328)
(0.001)
(0.001)
(0.768)
(0.001)

(Reference)
1.73 [1.28, 2.53]
2.14 [1.63, 2.81]
1.67 [1.00, 2.79]
1.11 [0.63, 1.94]
1.24 [0.54, 2.88]
2.82 [1.95, 4.08]
2.37 [1.78, 3.16]
0.81 [0.60, 1.11]
0.51 [0.44, 0.58]

[1.05,
[1.07,
[0.89,
[1.04,
[1.49,
[0.40,
[0.89,
[2.28,

(0.001)
(0.001)
(0.050)
(0.713)
(0.611)
(0.001)
(0.001)
(0.189)
(0.001)

MICA: mobile intensive care ambulance.

a
Trafc related: motor vehicle collision, motorcycle collision, bicycle collision.
b
Ambulance Victoria scores pain on a scale ranging from 0 to 10, where 0 = no pain, 2 = mild pain, 5 = moderate
pain and 10 = severe pain.

Discussion
This study has shown that the Victorian state adult prehospital trauma triage criteria have high sensitivity and are
effective at identifying both older and younger adult trauma
patients, regardless of well described differences in trauma

proles. Despite this, the ndings of this study have shown that
the clinical utility of the criteria does not match actual system
performance for older trauma patients. Consistent with international literature [1826] older trauma patients in Victoria are less
likely to be transported to a MTS for denitive care. After
controlling for trauma cause, injury severity, air transport,

1318

S. Cox et al. / Injury, Int. J. Care Injured 45 (2014) 13121319

transport duration, paramedic qualications, secondary transfer

to specialised care, triage criteria and paramedic judgement of
serious trauma, there was a steady decline in destination
compliance for older trauma patients identied by the prehospital trauma triage criteria. Older Victorian trauma patients
were also shown to have increased odds of mortality compared to
younger patients, with the odds of death 25% lower for patients
transported to a major trauma service.
While a larger proportion of older patients who died in hospital
were transported to non-MTS hospitals, the opposite was true for the
injury severity, ICU admission and urgent surgery major trauma
criteria. Larger proportions of older patients who met these criteria
were transported to MTSs compared to non-MTSs. Taken together
these ndings suggest that it is likely that paramedic subjective
judgement plays a role in in-eld triage decision making. These
ndings suggest that paramedics correctly identify severe trauma
based on injury severity, the need for intensive care and urgent
surgery. The nding that 72% of non-destination compliant older
patients who died in hospital and met the major trauma criteria on
death alone suggests that there is something different about these
patients that paramedics identify and base their triage decisions on.
Paramedic judgement is likely to consider the prospect of active
trauma care for older patients with respect to age, injury severity,
specic comorbidities (e.g. dementia), and patient prognosis.
Historically there has been a tendency in pre-hospital care to
triage older trauma patients who are considered low acuity to
non-MTSs, even though they may meet trauma triage criteria
[41,42]. In the current study, over eighty per cent of older
conrmed major trauma patients who met the current triage
criteria, but were not transported to a MTS had one or more
injuries identied in the pre-hospital setting. More than sixty per
cent of these injuries were assessed as blunt head injuries by
paramedics, and subsequent in-hospital assessment classied
over eighty-six per cent of these injuries as serious6 or greater in
severity. Almost thirty per cent of these patients ultimately died in
hospital. Previous reports have shown that even minor head
trauma without any loss of consciousness in older trauma patients
can result in severe intracranial injury, even when patients are
neurologically intact in the pre-hospital setting and upon arrival
at hospital [15,43]. The risk of complications is further increased
for older head trauma patients if they have one or more
comorbidities [15].
Fifty-two per cent of older conrmed major trauma patients
who died in hospital were transported to a MTS, and 74.2% of these
patients experienced a fall. In Australia and internationally [41,44]
the high mortality of low falls has been underestimated due to the
seemingly benign nature of the trauma [41]. Consistent with the
current study, prior studies [15,16,41,44] have reported that low
falls can be associated with severe injuries and worse outcomes. As
such, patients who experience low falls may require more focused
assessment in the pre-hospital setting to determine whether
denitive trauma care at a MTS is appropriate.
In summary, older trauma patients experience more falls and
have poorer outcomes than younger adult trauma patients. Despite
this, for older patients the level of destination compliance dened
by the current pre-hospital trauma triage criteria, declines steadily
with advancing age. Even though these triage criteria are effective
at identifying both older and younger adult trauma patients, their
use by paramedics does not match their clinical utility. The
ndings of the current study suggest it is likely that paramedic
judgement plays a role in pre-hospital triage decisions. Apart from
unusual circumstances (e.g. multi casualties), optimal pre-hospital
trauma triage criteria should reduce the need for discretionary
decision making.
6

Abbreviated Injury Scale (AIS) score > 2.

Future research is required to determine whether MTS

intervention is the best option for all older major trauma patients.
The benet of access to denitive trauma care in an MTS may vary
across age groups according to trauma cause, patient history,
comorbidities (e.g. dementia) and expected outcome for each
patient. Research should focus on a thorough review of the current
pre-hospital trauma triage criteria to determine whether these
criteria are optimised to care for the needs of older trauma
patients. This review should also include geospatial mapping
analyses to determine how time and distance from scene to
denitive trauma care impacts on destination compliance.
Paramedics may well be providing optimal care to older trauma
patients even though they are not strictly following the clinical
practice guidelines for pre-hospital trauma triage. This study has
provided a baseline from which we can further explore how the
trauma system can be optimised to meet the demands of a rapidly
ageing population.
Funding
None to declare.
Conict of interest statement
None of the authors has any conict of interest with this
research project.
References
[1] Christensen K, Doblhammer G, Rau R, Vaupel JW. Ageing populations: the
challenges ahead. The Lancet 2009;374:1196208.
[2] Australian Bureau of Statistics. Population projections, Australia, 20062101;
2008.
[3] Mitra B, Cameron PA. Optimising management of the elderly trauma patient.
Injury 2012;43:9735.
[4] Mitra B, Cameron PA, Gabbe BJ, Rosenfeld JV, Kavar B. Management and
hospital outcome of the severely head injured elderly patient. ANZ Journal
of Surgery 2008;78:58892.
[5] Gabbe B, Simpson P, Sutherland A, Wolfe R, Fitzgerald M, Judson R, et al.
Improved functional outcomes for major trauma patients in a regionalized,
Inclusive Trauma System. Annals of Surgery 2012;255:100915.
[6] Cox S, Currell A, Harriss L, Barger B, Cameron P, Smith K. Evaluation of the
Victorian state adult pre-hospital trauma triage criteria. Injury 2012;43:573
81.
[7] American College of Emergency Physicians. Guidelines for trauma care systems. American College of Emergency Physicians. Annals of Emergency Medicine 1993;22:1079100.
[8] OConnor RE. Trauma triage: concepts in prehospital trauma care. Prehospital
Emergency Care 2006;10:30710.
[9] Trauma & Emergency Services Ministerial Taskforce. Review of trauma and
emergency services in Victoria, 1999: Report of the Ministerial Taskforce on
trauma and emergency services and the Department of Human Services
Working Party on emergency and trauma services. Melbourne, Victoria:
Department of Human Services; 1999.
[10] MacKenzie EJ, Rivara FP, Jurkovich GJ, Nathens AB, Frey KP, Egleston BL, et al. A
national evaluation of the effect of trauma-center care on mortality. New
England Journal of Medicine 2006;354:36678.
[11] ONeill S, Brady RR, Kerssens JJ, Parks RW. Mortality associated with traumatic
injuries in the elderly: a population based study. Archives of Gerontology &
Geriatrics 2012;54:e42630.
[12] Pracht EE, Langland-Orban B, Flint L. Survival advantage for elderly trauma
patients treated in a designated trauma center. Journal of Trauma-Injury
Infection & Critical Care 2011;71:6977.
[13] Aldrian S, Nau T, Koenig F, Vecsei V. Geriatric polytrauma. Wiener Klinische
Wochenschrift 2005;117:1459.
[14] Callaway DW, Wolfe R. Geriatric trauma. Emergency Medicine Clinics of North
America 2007;25:83760.
[15] Aschkenasy MT, Rothenhaus TC. Trauma and falls in the elderly. Emergency
Medicine Clinics of North America 2006;24:41332.
[16] Pudelek B. Geriatric trauma: special needs for a special population. Clinical
Issues 2002;13:6172.
[17] Grossman MD, Miller D, Scaff DW, Arcona S. When is an elder old? Effect of
preexisting conditions on mortality in geriatric trauma,. Journal of TraumaInjury Infection & Critical Care 2002;52:2426.
[18] Demetriades D, Sava J, Alo K, Newton E, Velmahos GC, Murray JA, et al. Old age
as a criterion for trauma team activation. Journal of Trauma-Injury Infection &
Critical Care 2001;51:7546. discussion 756757.

S. Cox et al. / Injury, Int. J. Care Injured 45 (2014) 13121319

[19] Phillips SP, Rond PCI, Kelly SMM, Swartz PDM. The failure of triage criteria to
identify geriatric patients with trauma: results from the Florida trauma triage
study. Journal of Trauma 1996;40:27883.
[20] Zimmer-Gembeck MJ, Southard PA, Hedges JR, Mullins RJ, Rowland D, Stone JV,
et al. Triage in an established trauma system. Journal of Trauma-Injury
Infection & Critical Care 1995;39:9228.
[21] Chang DC, Bass RR, Cornwell EE, Mackenzie EJ. Undertriage of elderly trauma
patients to state-designated trauma centers. Archives of Surgery
2008;143:77681. discussion 782.
[22] Fitzharris M, Stevenson M, Middleton P, Sinclair G. Adherence with the prehospital triage protocol in the transport of injured patients in an urban setting.
Injury 2011;43:136876.
[23] Ma MH, MacKenzie EJ, Alcorta R, Kelen GD. Compliance with prehospital triage
protocols for major trauma patients. Journal of Trauma-Injury Infection &
Critical Care 1999;46:16875.
[24] Hsia R, Wang E, Saynina O, Wise P, Auerbach A. Factors associated with trauma
center use for elderly patients with trauma: a statewide analysis, 19992008.
Archives of Surgery 2011;146:58592.
[25] Nakamura Y, Daya M, Bulger EM, Schreiber M, Mackersie R, Hsia RY, et al.
Evaluating age in the eld triage of injured persons. Annals of Emergency
Medicine
2012;60:33545.
http://dx.doi.org/10.1016/j.annemergmed.2012.04.006 [Epub 2012 May 24].
[26] Staudenmayer KL, Hsia RY, Mann NC, Spain DA, Newgard CD. Triage of elderly
trauma patients: a population-based perspective. Journal of the American
College of Surgeons 2013;217:56976. http://dx.doi.org/10.1016/j.jamcollsurg.2013.06.017.
[27] Baez A, Lane P, Sorondo B. System compliance with out-of-hospital trauma
triage criteria. Journal of Trauma 2003;54:34451.
[28] Lane P, Sorondo B, Kelly JJ. Geriatric trauma patientsare they receiving
trauma center care? Academic Emergency Medicine 2003;10:24450.
[29] Newgard CD, Mann NC, Hsia RY, Bulger EM, Ma OJ, Staudenmayer K, et al.
Patient choice in the selection of hospitals by 9-1-1 emergency medical
services providers in trauma systems. Academic Emergency Medicine
2013;20:9119. http://dx.doi.org/10.1111/acem.12213.

1319

[30] Newgard CD, Nelson MJ, Kampp M, Saha S, Zive D, Schmidt T, et al. Out-ofhospital decision making and factors inuencing the regional distribution of
injured patients in a trauma system. Journal of Trauma 2011;70:134553.
[31] Australian Bureau of Statistics. Year book Australia 2007; 2007.
[32] Australian Bureau of Statistics. Regional population growth, Australia, 2008
2009; 2009.
[33] Cox S, Smith K, Currell A, Harriss L, Barger B, Cameron P. Differentiation of
conrmed major trauma patients and potential major trauma patients using
pre-hospital trauma triage criteria. Injury 2011;42:88995.
[34] State Government of Victoria DoH. Victorian state trauma system: Metropolitan Melbourne trauma services; 2010.
[35] Department of Health. Adult pre hospital major trauma criteria 2010; 2011.
[36] Victorian State Trauma Outcomes Registry. Victorian state trauma registry
summary report. Melbourne: Department of Health; 2011.
[37] Von Eye A, Schuster C. Regression analysis for social sciences. Academic Press;
1998.
[38] Marsh L, Cormier C. Spline regression models. Sage University papers series on
quantitative applications in the social sciences, 07-137. Thousand Oaks, CA:
Sage; 2001.
[39] Hosmer D, Lemeshow S. Applied logistic regression. 2nd ed. New York: John
Wiley & Sons, Inc; 2000.
[40] van Stef B. Multiple imputation of discrete and continuous data by fully conditional specication. Statistical Methods in Medical Research 2007;16:21942.
[41] Helling TS, Watkins M, Evans LL, Nelson PW, Shook JW, Van Way CW. Low falls:
an underappreciated mechanism of injury. Journal of Trauma-Injury Infection
& Critical Care 1999;46:4536.
[42] Helling TS, Nelson PW, Moore BT, Kintigh D, Lainhart K. Is trauma centre care
helpful for less severely injured patients? Injury 2005;36:12937.
[43] Mack LR, Chan SB, Silva JC, Hogan TM. The use of head computed tomography
in elderly patients sustaining minor head trauma. Journal of Emergency
Medicine 2003;24:15762.
[44] Bergeron E, Clement J, Lavoie A, Ratte S, Bamvita J-M, Aumont F, et al. A simple
fall in the elderly: not so simple. Journal of Trauma-Injury Infection & Critical
Care 2006;60:26873.