Resuscitation 83 (2012) 1259–1264

Contents lists available at SciVerse ScienceDirect

Resuscitation
journal homepage: www.elsevier.com/locate/resuscitation

Clinical paper

Prehospital Trauma Life Support (PHTLS) training of ambulance caregivers and

impact on survival of trauma victims夽
Jakob Johansson a,b,∗ , Hans Blomberg a,b , Bodil Svennblad c , Lisa Wernroth c , Håkan Melhus e ,
Liisa Byberg a,c , Karl Michaëlsson c,d , Rolf Karlsten a , Rolf Gedeborg a,c
a
Department of Surgical Sciences, Anaesthesiology and Intensive Care, Uppsala University, Uppsala, Sweden
b
Centre of Emergency Medicine, Uppsala University Hospital, Uppsala, Sweden
c
Uppsala Clinical Research Center, Uppsala University, Uppsala, Sweden
d
Department of Surgical Sciences, Orthopedics, Uppsala University, Uppsala, Sweden
e
Department of Medical Sciences, Uppsala University, Uppsala, Sweden

a r t i c l e i n f o a b s t r a c t

Article history: Background: The Prehospital Trauma Life Support (PHTLS) course has been widely implemented and
Received 21 November 2011 approximately half a million prehospital caregivers in over 50 countries have taken this course. Still, the
Received in revised form 23 January 2012 effect on injury outcome remains to be established. The objective of this study was to investigate the
Accepted 10 February 2012
association between PHTLS training of ambulance crew members and the mortality in trauma patients.
Methods: A population-based observational study of 2830 injured patients, who either died or were hospi-
Keywords:
talized for more than 24 h, was performed during gradual implementation of PHTLS in Uppsala County in
Trauma
Sweden between 1998 and 2004. Prehospital patient records were linked to hospital-discharge records,
Advanced life support
Prehospital Trauma Life Support
cause-of-death records, and information on PHTLS training and the educational level of ambulance crews.
Survival The main outcome measure was death, on scene or in hospital.
Results: Adjusting for multiple potential confounders, PHTLS training appeared to be associated with a
reduction in mortality, but the precision of this estimate was poor (odds ratio, 0.71; 95% confidence
interval, 0.42–1.19). The mortality risk was 4.7% (36/763) without PHTLS training and 4.5% (94/2067)
with PHTLS training. The predicted absolute risk reduction is estimated to correspond to 0.5 lives saved
annually per 100,000 population with PHTLS fully implemented.
Conclusions: PHTLS training of ambulance crew members may be associated with reduced mortality in
trauma patients, but the precision in this estimate was low due to the overall low mortality. While there
may be a relative risk reduction, the predicted absolute risk reduction in this population was low.
© 2012 Elsevier Ireland Ltd. All rights reserved.

1. Introduction
Trauma is the leading cause of death among persons below 60
years of age.1 It is a well-established belief that optimal treatment
in the early phase after trauma has a major impact on mortality.2
Therefore, over the years, raising the educational level among pre-
hospital caregivers and the implementation of specific educational
programs that target trauma care have been two widely adopted
strategies aimed at improving the outcome for trauma victims.
夽 A Spanish translated version of the summary of this article appears as Appendix
in the final online version at doi:10.1016/j.resuscitation.2012.02.018.
∗ Corresponding author at: Department of Surgical Sciences – Anaesthesiology &
Intensive Care, Uppsala University Hospital, S-751 85 Uppsala, Sweden.
Tel.: +46 18 6110000; fax: +46 18 559357.
E-mail address: jakob.johansson@surgsci.uu.se (J. Johansson).
This strategy has high face validity, but the underlying evidence is
poor.3
In the late 1970s, the American College of Surgeons Com-
mittee on Trauma developed the Advanced Trauma Life Support
(ATLS) course for physicians.4,5 Although implemented world-
wide, there is still no strong evidence that ATLS lowers mortality
in trauma victims.6,7 According to a recent study, ATLS-training
might even impair outcome.8 The Prehospital Trauma Life Support
(PHTLS) program was introduced in 1983 to integrate prehospital
trauma care with the ATLS program.9 PHTLS has been recog-
nized as one of the leading educational programs for prehospital
emergency trauma care and approximately half a million pre-
hospital caregivers in over 50 countries have taken this course.9
However, the scientific support for improved patient outcome
is limited, and there is no evidence to recommend advanced
life-support (ALS) training for ambulance crews.10–12 Substantial
effort and money are put into the PHTLS training program and

0300-9572/$ – see front matter © 2012 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.resuscitation.2012.02.018
1260 J. Johansson et al. / Resuscitation 83 (2012) 1259–1264
there is an obvious need to evaluate the possible benefits for
patients.
The aim of this study was to investigate the association between
PHTLS training of ambulance crew members and the outcome in
trauma patients.
2. Methods
2.1. Source population
Uppsala County is an administrative region for health care in
Sweden, with a population of 302,564 and a population density of
43 inhabitants/km2 in 2004. There are two hospitals in the county;
Uppsala University Hospital, a 1100-bed tertiary care facility, and
Enköping Hospital, a small local hospital handling only minor
trauma.
2.2. Emergency Medical Service (EMS)-system
The ambulance staff consisted of registered nurses and emer-
gency medical technician (EMT) equivalents (nursing assistants
with special ambulance training). During the study period the num-
ber of registered nurses employed as ambulance crew members
increased rapidly. Prior to and during the study period all ambu-
lance crew members were, as part of their regular skill practice,
annually trained and authorized to provide ALS in medical emer-
gencies and trauma, with the exception of endotracheal intubation
(only performed in cardiac arrest) and chest drainage.
2.3. Study population
All primary incident hospital admissions for injury in the county
of Uppsala from 1998 through 2004 were extracted from the
Swedish National Patient Registry (NPR).13 Injury deaths, where
the underlying cause of death was V01-Y36 according to ICD-10,14
were extracted from the Cause of Death Registry (CDR). These
codes include all external causes of injury except those referring
to “Complications of medical and surgical care”, “Sequelae of exter-
nal causes of morbidity and mortality”, and “Supplementary factors
related to causes of morbidity and mortality classified elsewhere”.
These datasets were combined using the unique personal identifi-
cation number for person-based linkage.15 Admissions resulting in
a hospital stay of one day or less and discharge alive were excluded
to reduce the number of minor injuries included.16 Dispatch infor-
mation, prehospital time intervals, medical information, and the
identity of the ambulance crew members were added from prehos-
pital electronic patient records.17 For each individual ambulance
crew member the following information was collected: years of
employment in health care services and in ambulance services; the
date for exam as a registered nurse and the date for PHTLS train-
ing. The final dataset consisted of 2830 injury events with complete
data (Fig. 1). The study was approved by the regional Human Ethics
Committee.
2.4. Exposure
The PHTLS program is a standardized curriculum for prehospital
caregivers. The core component is a 16-h course with a mixture of
lectures and interactive skill stations.18 If at least one ambulance
crew member was PHTLS certified (had passed the final exam-
ination in PHTLS), the injured patient was considered exposed.
Exposure status was not dependent on whether the certified crew
member was in charge of the crew or not. The time elapsed since
PHTLS certification was considered in a sensitivity analysis. No
refresher courses were performed during the study period.
There were no major changes in the standard of pre-hospital
trauma care or equipment used during the study period. All
ambulance crew members in the organisation were trained and
authorized to perform trauma care according to international stan-
dard treatment, except from endotracheal intubation. A PHTLS
certification did not change the authorization to use any equipment
or perform any specific intervention.
2.5. Outcome
Prehospital injury deaths were defined as autopsied injury
deaths not associated with a hospital admission.19 Only prehos-
pital deaths with documented involvement of ambulance services
were included. Hospital death was defined as death during a hos-
pital stay with a principal discharge diagnosis indicating injury.
The composite outcome of either prehospital or hospital death was
studied.
2.6. Possible confounders
The International Classification of disease Injury Severity Score
(ICISS) was calculated based on diagnosis-specific survival prob-
abilities for individual injury ICD-10 codes.20–22 For descriptive
purposes, ICISS was categorized as critical (0–0.219), severe
(0.220–0.354), serious (0.355–0.664), moderate (0.665–0.940), or
minor (0.941–1.0). Injury severity estimates based on the Revised
Trauma Score (RTS) were also available.23
Each victim’s injuries were categorized according to the injury
mortality diagnosis matrix for ICD-10.24 Five categories of major
injury regions (head, spinal cord, thorax, abdomen, and pelvis) were
defined from the matrix.
Causes of injury were classified according to the matrix devel-
oped by the National Center for Health Statistics, Centers for Disease
Control, USA.25 Data was collapsed to three categories: traffic
injuries, falls, and other injuries.
The basic educational level (nurse or EMT equivalent) and
employment time of the ambulance crew members were also con-
sidered as potential confounders. Exposure status was determined
from the highest educational level and the longest work experience
in the ambulance crew. To account for a possible period effect (such
as changes in trauma care over time during the study period), the
calendar year of the injury was included in the multivariable mod-
els. The patients’ age and sex were also considered as potential
confounders.
2.7. Statistics
Multivariable logistic regression analysis was used to model the
composite outcome of prehospital or hospital death using all the
possible confounders described above. Collinearity was assessed
from variance inflation factors. Both age and calendar year of injury
were entered as linear effects in the models based on inspec-
tion of a logit plot of each variable. ICISS was used after logit
transformation.26 Effect-measure modification was evaluated by
including product terms between the exposure variable and injury
severity (ICISS) and year, respectively. The multivariable logistic
regression model was also used to calculate the predicted mortal-
ity for each injury event. The difference in mean predicted mortality
between the PHTLS group and the non-PHTLS group estimated the
absolute risk reduction.
Population-averaged models using generalized estimation
equations (GEE) were used to handle correlations from ambulance
crews appearing multiple times in the dataset and patients appear-
ing multiple times in the study population.27,28 The number of
lives potentially saved annually was estimated by the proportion
of eligible patients corresponding to the estimated absolute risk
 J. Johansson et al. / Resuscitation 83 (2012) 1259–1264 1261

5235 severe injuries (hospitalized > 1 day or dead) with documented participation of EMS

483 without a valid personal

identification number in EMS record

Information on educational level 1129 with incomplete information to

among ambulance personnel added determine crew PHTLS certification
status

3623 severe injuries (hospitalized > 1 day or dead) with documented participation of EMS and
information on ambulance crew’s PHTLS certification status

793 without complete information on

both crew members’ identity and
PHTLS certification status and
covariates

2830 incident injuries (hospitalized > 1 day or dead) with documented participation of EMS and
information on identity and PHTLS certification status for both members of the ambulance
crew and all covariates in the multivariable model

Fig. 1. Flow diagram for selection of the study population. EMS: Emergency Medical Service.

reduction. All analyses were done as complete subjects analyses.
As a sensitivity analysis, multiple imputation using predictive mean
matching (aregImpute) that fits separate flexible additive imputa-
tion models, was also applied.29,30 Missing data for PHTLS status
and employment time were imputed based on mortality, PHTLS sta-
tus, educational level (nurse/EMT), calendar year, and employment
time. Ten bootstrap samples were generated.
The statistical packages SAS version 9 (SAS Institute Inc., Cary,
NC, USA) and R version 2.9.2 (R Foundation for Statistical Comput-
ing, Vienna, Austria) were used for data management and statistical
analyses.
3. Results
3.1. Cohort characteristics
During the 7-year study period, EMSs responded to 2830 injury
events with complete data for the analyses (Fig. 1). The proportion
of injury events handled by ambulance crew members with PHTLS
training increased over time (Fig. 2). There were no major differ-
ences in patient characteristics; however, ambulance crews in the
PHTLS group had more years of employment in ambulance services,
compared to the non-PHTLS group (Table 1). There was no appar-
ent overall difference in response time, on-scene time, or transport
time between the two groups.
0.41–1.25), with a corresponding predicted absolute risk reduc-
tion of 0.2%. There was no indication of serious collinearity in the
regression model, with the highest variance inflation factor being
1.78.
This analysis, however, violates the assumption of independent
observations since the same ambulance crews were involved in
many cases of injury and also the same patients occurred multi-
ple times in the study population. We analytically considered such
possible correlations (supplementary data: Table 4). Results from
these analyses were very similar to those from using standard logis-
tic regression. The choice of variance structure did not affect the
result and, in general, the correlations were minimal (<0.1).
3.3. Subgroup analyses
A product term for PHTLS and injury severity in the logistic
regression model indicated the possible presence of effect-measure
modification (likelihood ratio test P = 0.08), but analysis stratified
for injury severity did not indicate any substantial influence on
the estimated OR for PHTLS (Table 2). Exclusion of injuries caused
by falls strengthened the estimated protective effect of PHTLS to
OR = 0.54 (Table 2) but with poor precision (95% CI, 0.13–2.41).
A short interval from PHTLS training resulted in a stronger pro-
tective association with mortality (Table 2). The precision in these
estimates was poor.

3.2. Relative and absolute mortality risk 3.4. Characteristics of excluded patients

The mortality was 4.7% (36/763) without PHTLS training and There were no major differences in age, sex, cause of injury,
4.5% (94/2067) with PHTLS training. The crude (unadjusted) odds injury severity, major organs injured, or mortality when charac-
ratio (OR) for mortality was 0.96 [95% confidence interval (CI), teristics among the excluded cases were compared to the study
0.66–1.44]. The adjusted OR for mortality was 0.71 (95% CI, population (supplementary data: Table 3). In electronic EMS patient
1262 J. Johansson et al. / Resuscitation 83 (2012) 1259–1264

B. PHTLS trained present A. Highest educational level present

100

100
80

80
Proportion of injury events (%)
Proportion of injury events (%)
60

60
40

40
At least one with PHTLS Registered nurse
No one with PHTLS EMT equivalent
20

20
0

0
1998 2000 2002 2004 1998 2000 2002 2004

Fig. 2. Change during the study period in the distribution of highest educational level in the ambulance crews (left panel) and the proportion of injury events where at least
one PHTLS-trained individual was present (right panel). EMT equivalent, Emergency Medical Technician equivalent (nursing assistants with special ambulance training).

Table 1
Baseline characteristics of the Study Cohort.

N At least one member of the ambulance No member of the ambulance crew

crew with PHTLS certification with PHTLS certification
(N = 2067) (N = 763)

Age, % (N) 2830

0–14 2 (39) 1 (9)
15–24 4 (93) 5 (37)
25–44 11 (219) 10 (78)
45–64 18 (362) 11 (86)
65–74 11 (220) 11 (85)
≥75 55 (1134) 61 (468)
Sex, % (N) 2830
Male 37 (759) 38 (292)
ICISS stratified, % (N) 2830
Critical 1 (23) 0 (2)
Severe 1 (20) 1 (7)
Serious 5 (96) 4 (30)
Moderate 48 (1000) 50 (379)
Minor 45 (928) 45 (345)
Head injury, % (N) 2830 9 (194) 9 (71)
Spinal cord injury, % (N) 2830 0 (6) 0 (0)
Thoracic injury, % (N) 2830 6 (122) 6 (42)
Abdominal injury, % (N) 2830 1 (27) 1 (7)
Pelvic injury, % (N) 2830 6 (118) 6 (46)
No. of major injury regions, % (N) 2830
None 80 (1653) 80 (612)
One 18 (372) 18 (140)
Two 2 (33) 1 (7)
Three or more 0 (9) 1 (4)
Cause of injury, % (N) 2830
Fall 76 (1570) 80 (607)
Traffic 11 (234) 9 (71)
Other 13 (263) 11 (85)
Highest crew member education, % (N) 2830
EMT equivalent 32 (660) 75 (574)
Registered nurse 68 (1407) 24 (189)
Shortest interval from PHTLS training in the
ambulance crew (years), median (IQR) 2067 1.4 (0.7–2.3) NA
Maximum years of individual experience in EMS,
median (IQR) 2830 18 (15–22) 16 (12–19)
Response time minutes, median (IQR) 2563 9 (5–15) 9 (5–15)
On-scene time minutes, median (IQR) 2497 11 (8–16) 10 (7–14)
Transport time minutes, median (IQR) 2059 11 (7–25) 12 (7–24)
Prehospital mortality, % (N) 2830 1% (20) 0% (2)
Hospital mortality, % (N) 2830 4% (74) 4% (34)

ICISS, International Classification of disease Injury Severity Score; IQR, interquartile range; EMS, Emergency Medical Service; EMT, Emergency Medical Technician.
 J. Johansson et al. / Resuscitation 83 (2012) 1259–1264
Table 2
Complementary analysesa , using subgroups and graded exposures, of the association between PHTLS certification and mortality.
Subgroup Number of patients
(no PHTLS/PHTLS)
Traffic and other injury, excluding falls 156/497
Effect-measure modification by injury severity
More severe injury (ICISS < 0.665) 39/139
Less severe injury (ICISS ≥ 0.665) 724/1928
Length of the interval since PHTLS certification 763/2067
(shortest in the ambulance crew)
≤1 year
>1 year
OR, odds ratio; CI, confidence interval; ICISS, International Classification of disease Injury Severity Score.
a
Multivariable logistic regression model also included age, sex, injury severity (ICISS), cause of injury (all transport/fall/other), study year, head injury, spinal cord injury,
thoracic injury, abdominal injury, pelvic injury, and years of employment in ambulance service.
care reports without a valid personal identification number for the
patient, the median RTS was 8, identical to the median RTS among
those with a complete personal identification number. Multiple
imputation of the major offending variables – PHTLS status and
employment time – did not substantially change the estimated
association or the precision in the estimate (supplementary data:
Table 4).
4. Discussion
This population-based observational study indicates that PHTLS
training of ambulance crew members may be associated with
approximately 30% relative reduction of mortality in trauma vic-
tims. However, with the low overall mortality in this population,
the precision of this estimate was low making the interpretation
of this possible association difficult. The predicted absolute risk
reduction is estimated to correspond to 0.5 lives saved annually
per 100,000 population with PHTLS fully implemented. Despite
decades of widespread implementation of educational programs
such as ATLS and PHTLS, there is scant scientific evidence for benefi-
cial effects in trauma outcome.6,7,12 Today, a substantial proportion
of ambulance crews include an ALS-trained member.31 While there
are studies on conditions other than trauma indicating that ALS
training and higher educational levels among ambulance crews
improves outcomes,32,33 there is little support for such an asso-
ciation in trauma care.34–36
Three conditions provided a rare opportunity to perform a com-
prehensive observational study of the association between PHTLS
training and outcome: (a) the introduction of prehospital electronic
medical records, (b) the subsequent gradual implementation of
PHTLS, and (c) exact record linkage of patient injury data in national
healthcare databases. We could associate each injury case with
the educational level of the ambulance crew and adjust for each
crew’s years of experience and basic educational level. We further
controlled for injury cause, type and severity, and also the year
of the study in which the exposure/incident occurred in order to
account for a possible period effect. The population-based design
and good control of confounding – and especially a potential period
effect – support the findings in this study. Exclusion of observations
with incomplete data was unfortunate; however, the comparable
characteristics of these patients to the study population, the small
impact on the estimated association from imputation of missing
values, and the absence of apparent effect-measure modification
from injury severity did not raise any grave concern of a selection
bias.
A particular challenge for this type of study is that the same
ambulance crews appear several times in the data, and similarly
a patient may also present repeated times. Ignoring such correla-
tions within the data may lead to incorrect statistical inference and
1263
Number of deaths Crude (unadjusted) Adjusted OR (95%
(no PHTLS/PHTLS) OR (95% CI) CI)
7/29 1.32 (0.60–3.33) 0.54 (0.13–2.41)
9/35 1.12 (0.50–2.71) 0.80 (0.20–3.44)
27/59 0.81 (0.52–1.31) 0.77 (0.28–2.04)
36/94
0.75 (0.44–1.24) 0.66 (0.36–1.20)
1.09 (0.72–1.66) 0.76 (0.42–1.41)
biased estimates of variance. However, methods that do not uti-
lize all available data may be inefficient. Analytically accounting
for these showed that their impact on the estimated associations
was minimal.
Despite good control of several important confounders, limited
causal inference can be made from a single observational study.
Largely due to the relatively low overall mortality in our study, the
precision of our estimates was low.
We can only speculate why PHTLS training appeared to be asso-
ciated with reduced mortality after trauma. According to the PHTLS
concept, endotracheal intubation is the preferred method of air-
way control, although several studies dispute the benefits of this
procedure.36–38 In our study, no member of the ambulance crew
was authorized to perform endotracheal intubation on trauma
patients, irrespective of PHTLS training. Thus, acquisition of this
specific airway management skill could not explain the lowered
mortality seen in our study. In fact, in this EMS system, ALS inter-
ventions related to PHTLS had already been implemented before
PHTLS education, with the exception of endotracheal intubation.
Therefore, the observed association between PHTLS education and
reduced mortality was most likely not due to the application of
individual ALS interventions.
It has been demonstrated that PHTLS training improves adher-
ence to established priorities and management of the trauma victim
in a structured approach, which could possibly shorten time to
definitive care.12,39 In our study though, response time, on-scene
time, and transport time did not appear influenced by PHTLS train-
ing. The impact of the structured approach, including assessment
and setting priorities that the PHTLS concept is focused on, is dif-
ficult to measure and could possibly explain the association with
reduced mortality.
The percentage of ambulance crews trained in PHTLS went from
approximately 15% to almost 90% in a relatively short time period.
This focused educational effort could possibly create a Hawthorne-
like effect, making the PHTLS-trained personnel particularly prone
to perform better in general, regardless of the content of the
course. However, the implementation of PHTLS in the organiza-
tion could possibly also have influenced ambulance crew members
not trained in PHTLS in a positive direction, thus also improving
outcome in the control group.
Concerning the ability to generalize our results to other set-
tings, it is notable that the PHTLS course is highly standardized.
However, the basic educational level of prehospital crews varies
between organizations and therefore other results could possibly
be achieved in a similar study in another setting such as a paramedic
or EMT-based system. The study population represents the popu-
lation admitted to a general hospital, but is not representative of a
selected trauma-center population. This is reflected by the female
dominance and large number of falls causing minor or moderate
injuries.
1264 J. Johansson et al. / Resuscitation 83 (2012) 1259–1264
Excluding falls from the study population, i.e. mainly elderly
patients with hip fractures, as expected strengthened the asso-
ciation between PHTLS and reduced mortality. In the subgroup
of traffic crash victims there was a signal of stronger association
between PHTLS training and reduced mortality. It is therefore possi-
ble that both the relative and especially the absolute risk reduction
might be higher in a population with higher prevalence of multiply
injured patients. The relation with injury severity is also of high
interest, but likely complex. One may expect to find the strongest
effect of PHTLS on injuries with intermediate severity. With low
injury severity mortality is very low and with high injury severity
mortality might be inevitable. However, due to the limited mortal-
ity in the study population we are unable to analytically pursue this
in detail. The presented estimates from all the subgroup analyses
are highly uncertain and must be interpreted with great caution.
5. Conclusion
This population-based study of lethal or hospitalized injuries
indicates that PHTLS training of ambulance crew members may
be associated with reduced mortality in trauma patients, but
the precision in this estimate was low due to the overall low
mortality. Although our study indicates a relative risk reduc-
tion, the predicted absolute risk reduction in this population was
low.
Conflicts of interest
Dr. Blomberg reported that he has been an instructor in PHTLS
since 2002. No other conflicts of interest were reported.
Acknowledgments
This project was funded by Uppsala University and the Laerdal
Foundation for Acute Medicine. No funding organization or sponsor
was involved in the design and conduct of the study.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.resuscitation.2012.02.018.
References
1. Krug EG, Sharma GK, Lozano R. The global burden of injuries. Am J Public Health
2000;90:523–6.
2. Trunkey DD, Lim Jr RC, Blaisdell FW. Traumatic injury. A health care crisis. West
J Med 1974;120:92–4.
3. Lerner EB, Moscati RM. The golden hour: scientific fact or medical “urban leg-
end”? Acad Emerg Med 2001;8:758–60.
4. American College of Surgeons. Atls advanced trauma life support. Lippincott
Williams & Wilkins; 2006.
5. Collicott PE, Hughes I. Training in advanced trauma life support. JAMA
1980;243:1156–9.
6. Soreide K. Three decades (1978–2008) of advanced trauma life support (atls)
practice revised and evidence revisited. Scand J Trauma Resusc Emerg Med
2008;16:19.
7. Jayaraman S, Sethi D. Advanced trauma life support training for hospital staff.
Cochrane Database Syst Rev 2009:CD004173.
8. Drimousis PG, Theodorou D, Toutouzas K, et al. Advanced trauma life support
certified physicians in a non trauma system setting: is it enough? Resuscitation
2011;82:180–4.
9. Salomone JP, Pons PT. Phtls prehospital trauma life support; 2007.
10. Ali J, Adam RU, Gana TJ, Williams JI. Trauma patient outcome after the pre-
hospital trauma life support program. J Trauma 1997;42:1018–21 [discussion
1021–1012].
11. Arreola-Risa C, Mock C, Herrera-Escamilla AJ, Contreras I, Vargas J. Cost-
effectiveness and benefit of alternatives to improve training for prehospital
trauma care in mexico. Prehosp Disaster Med 2004;19:318–25.
12. Jayaraman S, Sethi D. Advanced trauma life support training for ambulance
crews. Cochrane Database Syst Rev 2010:CD003109.
13. Gedeborg R, Engquist H, Berglund L, Michaelsson K. Identification of incident
injuries in hospital discharge registers. Epidemiology 2008;19:860–7.
14. World Health Organization. International statistical classification of diseases
and health related problems icd-10. Geneva: World Health Organization; 2005.
15. Ludvigsson JF, Otterblad-Olausson P, Pettersson BU, Ekbom A. The swedish
personal identity number: possibilities and pitfalls in healthcare and medical
research. Eur J Epidemiol 2009;24:659–67.
16. Cryer C, Langley JD. Studies need to make explicit the theoretical and case defi-
nitions of injury. Inj Prev 2008;14:74–7.
17. Karlsten R, Sjoqvist BA. Telemedicine and decision support in emergency ambu-
lances in uppsala. J Telemed Telecare 2000;6:1–7.
18. National association of emergency medical technicians (naemt).
<http://www.naemt.org/education/PHTLS/phtls a.aspx>.
19. Gedeborg R, Chen L, Thiblin I, et al. Prehospital injury deaths—strengthening
the case for prevention: Nationwide cohort study. J Trauma;
doi:10.1097/TA.0b013e3182288272, in press.
20. Clark DE, Winchell RJ. Risk adjustment for injured patients using administrative
data. J Trauma 2004;57:130–40 [discussion 140].
21. Osler T, Rutledge R, Deis J, Bedrick E. Iciss: an international classification of
disease-9 based injury severity score. J Trauma 1996;41:380–6 [discussion
386–388].
22. Stephenson S, Henley G, Harrison JE, Langley JD. Diagnosis based injury severity
scaling: investigation of a method using australian and new zealand hospitali-
sations. Inj Prev 2004;10:379–83.
23. Champion HR, Sacco WJ, Copes WS, Gann DS, Gennarelli TA, Flanagan ME. A
revision of the trauma score. J Trauma 1989;29:623–9.
24. Fingerhut LA, Warner M. The icd-10 injury mortality diagnosis matrix. Inj Prev
2006;12:24–9.
25. Recommended framework for presenting injury mortality data. MMWR
Recomm Rep 1997;46:1–30.
26. Moore L, Lavoie A, Bergeron E, Emond M. Modeling probability-based injury
severity scores in logistic regression models: the logit transformation should be
used. J Trauma 2007;62:601–5.
27. Carey V, Zeger SL, Diggle P. Modelling multivariate binary data with alternating
logistic regressions. Biometrica 1993;80:517–26.
28. Miglioretti DL, Heagerty PJ. Marginal modeling of nonnested multilevel data
using standard software. Am J Epidemiol 2007;165:453–63.
29. Harrell Jr FE. Regression modeling strategies. New York: Springer; 2001.
30. Marshall A, Altman DG, Holder RL. Comparison of imputation methods for han-
dling missing covariate data when fitting a cox proportional hazards model: a
resampling study. BMC Med Res Methodol 2010;10:112.
31. Bulger EM, Nathens AB, Rivara FP, MacKenzie E, Sabath DR, Jurkovich GJ. National
variability in out-of-hospital treatment after traumatic injury. Ann Emerg Med
2007;49:293–301.
32. Woodall J, McCarthy M, Johnston T, Tippett V, Bonham R. Impact of advanced
cardiac life support-skilled paramedics on survival from out-of-hospital cardiac
arrest in a statewide emergency medical service. Emerg Med J 2007;24:134–8.
33. Stiell IG, Spaite DW, Field B, et al. Advanced life support for out-of-hospital
respiratory distress. N Engl J Med 2007;356:2156–64.
34. Haas B, Nathens AB. Pro/con debate: is the scoop and run approach the best
approach to trauma services organization? Crit Care 2008;12:224.
35. Isenberg DL, Bissell R. Does advanced life support provide benefits to patients?:
a literature review. Prehosp Disaster Med 2005;20:265–70.
36. Stiell IG, Nesbitt LP, Pickett W, et al. The opals major trauma study: impact of
advanced life-support on survival and morbidity. CMAJ 2008;178:1141–52.
37. Eckstein M, Chan L, Schneir A, Palmer R. Effect of prehospital advanced life
support on outcomes of major trauma patients. J Trauma 2000;48:643–8.
38. Bulger EM, Maier RV. Prehospital care of the injured: what’s new. Surg Clin North
Am 2007;87:37–53, vi.
39. Ali J, Adam R, Josa D, et al. Effect of basic prehospital trauma life support program
on cognitive and trauma management skills. World J Surg 1998;22:1192–6.