Injury, Int. J. Care Injured 41 (2010) 10–20

Review

A systematic review of the costs and benefits of helicopter emergencymedical services§

Colman B. Taylor a,*, Mark Stevenson a,b, Stephen Jan a,b, Paul M. Middleton d,Michael Fitzharris e,f, John A. Myburgh a,c

a The George Institute for International Health, Sydney, NSW, Australiab University of Sydney, Sydney Medical School, Sydney, NSW, Australiac University of New South Wales, Faculty of Medicine, Sydney, NSW, Australiad Ambulance Research Institute, Ambulance Service of NSW, Sydney, NSW, Australiae Accident Research Centre, Monash South Africa, Johannesburg, South Africaf Accident Research Centre, Monash University, Melbourne, Australia

Contents

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

Categorisation of studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

Synthesis of results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

A R T I C L E I N F O

Article history:

Accepted 21 September 2009

Keywords:

Trauma

Injury

Critical care

Helicopter emergency medical services

Transport

Cost–effectiveness

A B S T R A C T

Introduction: Helicopter emergency medical services (HEMS) are popular in first world health systems

despite inconsistent evidence in the scientific literature to support their use. The aim of the current study

was to perform a systematic review of economic evaluations of HEMS, in order to determine the

economic cost of HEMS and the associated patient-centered benefits.

Method: A systematic review was performed of studies that provided a cost estimate of HEMS. The

inclusion criteria consisted of English language articles that estimated both the costs and outcomes of a

HEMS and fulfilled pre-specified criteria in relation to a cost analysis, cost–minimisation, cost–

effectiveness or cost–benefit evaluation. Identified studies were synthesised according to the patient

diagnosis (trauma, non-trauma or non-specific) and the type of HEMS transport under review (primary

scene retrieval or secondary inter-facility transport). All costs were converted to US dollars and indexed

for inflation.

Results: Fifteen studies met the inclusion criteria. Among all studies the annual cost of HEMS ranged

from $115,777 to $5,571,578. Five studies showed HEMS to be a more expensive transport alternative

without an associated benefit while eight studies provided cost–effectiveness ratios of $3292 and $2227

per life year saved for trauma, $3258 per life saved and $7138 and $12,022 per quality adjusted life year

for non-trauma and $30,365 and $91,478 per beneficial mission for non-specific patient populations. One

study also evaluated the cost of HEMS to societal benefit, producing a ratio of 1:6.

Interpretation: The cost and effectiveness of HEMS varied considerably between studies. Despite

generally being more expensive than ground transport, a number of studies found HEMS to be cost-

effective. However, given the variation in the intervention design, context and study methods between

studies it was not possible to assess the cost–effectiveness of HEMS in general. Given the variation

inherent in the health systems in which HEMS operate, synthesis and extrapolation of study findings

across differing health environments is difficult. To address economic and clinical evidence in relation to

HEMS, future research that is tailored to account for local system factors is required.

� 2009 Elsevier Ltd. All rights reserved.

Contents lists available at ScienceDirect

Injury

journa l homepage: www.e lsevier .com/ locate / in jury

§ Previous presentations: Combined Australasian Trauma Society and Trauma Association of Canada ASM, Auckland, New Zealand, February 2009.

* Corresponding author at: P.O. Box M201, Missenden Rd, Camperdown, NSW 2050, Australia. Tel.: +61 2 9657 0352; fax: +61 2 9657 0301.

E-mail address: ctaylor@george.org.au (C.B. Taylor).

0020–1383/$ – see front matter � 2009 Elsevier Ltd. All rights reserved.

doi:10.1016/j.injury.2009.09.030

C.B. Taylor et al. / Injury, Int. J. Care Injured 41 (2010) 10–20 11

Search results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

Trauma patients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

Primary transport . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

Secondary (inter-facility) transport . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

Primary and secondary transport. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

Non-trauma patients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

Non-specific patients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

Differences in interventions and service provision . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

Differences in context. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

Variation in study method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

Introduction

Out-of-hospital emergency care has developed from basicmilitary models into highly sophisticated and integrated systemsof care. In the military environment, the first documented case ofhelicopter transport of an injured patient was in Burma duringWorld War II.8 This experience was expanded upon in the Koreanand Vietnam wars with the use of formal helicopter emergencymedical services (HEMS).14 The success of HEMS in the military ledto their subsequent introduction into the civilian environmentwhere they are now used for primary scene retrievals andsecondary inter-facility transports in modern health systems.

A number of studies examining the effectiveness of HEMScompared to ground based services have been undertaken.30–33

Despite inconsistent results, some studies have shown HEMS toimprove patient mortality in comparison to ground based servicesor ‘usual care’.2,3,12,34 Given the added expense of operating andmaintaining HEMS in a health system, an important step in thejustification for their continued use is to relate their costs to anybenefits they provide. A previous review17 highlighted a paucity ofquality economic evaluations in the emergency care environment.As an intervention, HEMS has unique costs and consequences forthe health system in which they operate. Hence, a targeted reviewthat provides a detailed synthesis of the current evidence isrequired. To date, there has been no comprehensive evaluation ofthe economics of HEMS.

The aim of this study therefore is to undertake a systematicreview of the literature to determine the economic costs of HEMSand the relationship of these costs to the patient-centered benefit.In doing so we consider the external validity of the currentevidence so that thought can be given to the broader operatingscenarios of HEMS.

Methods

A systematic review of the literature was conducted of articleswhich provided an economic evaluation of a HEMS. The inclusion

Table 1Categories and definitions for four types of economic studies.

Type of article Definition

Cost analysis The study provides an estimate of

Cost–minimisation An economic evaluation in which

is made solely in terms of costs

Cost–effectiveness An economic evaluation comparin

measured and compared in terms

such as life years gained. The resu

Cost–benefit An economic evaluation in which

is deemed ‘value for money’ if ben

criteria consisted of English language articles that estimated bothcosts and outcomes of a HEMS and fulfilled pre-specified criteria inrelation to a cost analysis, cost–minimisation, cost–effectivenessor cost–benefit evaluation (Table 1).

A search was conducted of Medline, EMBASE, Cochrane Libraryand Google Scholar using the following search string: ‘‘airambulance’’/‘‘air medical transport’’ OR ‘‘helicopter’’ OR ‘‘emer-gency medical service’’/‘‘emergency health service’’ OR ‘‘emer-gency medical technician’’/‘‘emergency medicine’’ OR ‘‘emergencycare’’ combined using ‘‘and’’ with ‘‘economics, medical’’/‘‘healtheconomics’’ OR ‘‘costs and cost analysis’’/‘‘costs’’ OR ‘‘cost–benefitanalysis’’ OR ‘‘cost–effectiveness analysis’’ OR ‘‘cost–minimisationanalysis’’ OR ‘‘cost utility analysis’’.

A hand-search of relevant journals was also performed. Thisincluded Injury, Air Medical Journal, Annals of Emergency Medicine,Emergency Medicine Journal and Pre-hospital and Emergency Care.

Study review and selection was undertaken by a singleinvestigator (CT). The initial set of abstracts was reviewed andthose considered unrelated to either the cost or the cost and benefitof HEMS were excluded. Editorials and commentaries were alsoexcluded after they were reviewed for relevant references. The fulltext of articles from the selected abstracts were also reviewed andconsidered valid only if criteria was met as to one of four pre-specified types of economic evaluations (Table 1). The referencesections of all full text articles were reviewed to identify furtherarticles of relevance.

A tool to abstract health economic information was developedfrom an established guide designed to assess health economicliterature.10 Components recorded from each study included theauthor and year of publication, type of economic evaluation,setting, competing alternative(s) to HEMS, methods, main healtheconomic outcome, economic estimate, main patient-based out-comes (including mortality, morbidity hospital length of stay, etc.),viewpoint adopted for economic evaluation and whether uncer-tainties were varied in cost and benefit estimates. For studiesproviding an estimate of the annual cost of a HEMS, furtherinformation was extracted relating to the characteristics of the

the cost in relation to a pre-specified performance outcome with methodology

the effectiveness of alternatives compared is equivalent and thus comparison

g the costs and outcomes of alternative interventions. Patient outcomes are

of quality adjusted life years (QALYS) or some other patient relevant measure

lts are presented in terms of an incremental cost–effectiveness ratio

the costs and benefits to society are valued in monetary terms. An intervention

efits exceed costs

C.B. Taylor et al. / Injury, Int. J. Care Injured 41 (2010) 10–2012

service (e.g. staffing, hours of operation, type of helicopter) and thecomponents included in the cost analysis (e.g. personnel, opera-tional and capital costs).

Categorisation of studies

Selected articles were categorised according to the primarypatient diagnosis reviewed, i.e. trauma, or non-trauma. Articlesthat included multiple patient diagnoses were categorised as non-specific. Within patient categories, articles were classifiedaccording to the HEMS transport type evaluated. This includedprimary scene retrieval, secondary inter-facility transport or both.

Synthesis of results

All economic information was converted to US dollars usinghistorical exchange rates22 as at the first of January in the year ofpublication. Amounts were then adjusted to 2008 values,according to the US Bureau of Labor Statistics35 to account forinflation. Annual cost estimates of HEMS and the comparator (ifapplicable) along with cost–effectiveness or cost–benefit ratioswere then synthesised according to the primary patient diagnosisand the HEMS transport type (primary, secondary or both).

Results

Search results

The literature search derived an initial set of 183 abstracts(Fig. 1). From this set, 62 articles received a full text review.

Fig. 1. Review flow and initial categorisation b

Scanning the reference lists of the selected full texts revealed anadditional 4 relevant articles. Of these 66 articles, 38 articles wereidentified that included an economic assessment of HEMS. A totalof 23 articles were excluded due to not meeting pre-specifiedcriteria for a cost analysis, cost–minimisation, cost–effectivenessor cost–benefit evaluation. In total 15 articles (from 6 countries)met the inclusion criteria and received a full structured review. Thedetails of these studies are described in Table 2. A total of 7articles5,6,13,16,21,26,27 included an estimate of the annual cost ofHEMS, and received further review. Table 3 shows the character-istics of these services and the components included in the costestimates.

Costs are reported as 2008 equivalent amounts in USD withoriginal estimates in brackets.

Trauma patients

Of the fifteen studies identified, seven5,9,13,20,21,27,28 concen-trated specifically on the trauma patient population. Twostudies21,27 included a small group of non-trauma patientsalthough both predominantly focused on trauma and as such,were included in the trauma category.

Primary transport

Of the eight trauma studies, three cost–minimisation stu-dies5,21,27 evaluated HEMS for primary patient retrieval. All threestudies were undertaken in the UK and used ground transport asthe comparator. The annual cost of HEMS from one studyevaluating three services27 ranged from $115,777 (£54,503) to

y patient diagnosis and type of transport.

Table 2Characteristics of studies included in review.

Author (year) Type of

economic

evaluation

Setting Comparator Methods Main health

economic outcome

Economic estimate Main patient

outcome

Viewpoint

adopted

Uncertainties

varied

Trauma

Nicholl (1994)21 Cost–

minimisation

Cornwall,

UK

Ground

units

D: Prospective Total operating cost

of HEMS and ground

per annum (GBP)

HEMS: £594,666;

ground: £97,805

N/A Stated as

sociological

Assumptions

and costing

methods

N: 178 (HEMS), 180 (ground)

T: 1 year

M: HEMS was compared to a

group of ground based controls

through performance times

Brazier (1996)5 Cost–

minimisation

London,

UK

Ground

units

D: Prospective Incremental cost of

HEMS per annum

(GBP)

£1.97M

(£1.52M–£2.6M)

No difference

residual in

disability

Appears to

be societal

Helicopter

residual value

and excess

hospital costs

N: 337 (HEMS), 466 (ground)

T: 2 years

M: Comparison of residual disability

in accidentally injured survivors,

transported by air and ground

Snooks (1996)27 Cost–

minimisation

UK Ground

units

D: Prospective Total operating cost

per annum (GBP)

Sussex: £54,503; No difference

in mortality

or general

health status

Appears to

be societal

None

N: 900 (HEMS), 982 (ground) Wiltshire: £151,200

M: Processes of care and costs were

compared across 4 locations and

health outcomes were compared in 2

locations only (Cornwall and London)

Cornwall £594,666;

London £1,146,891

Thomas (1990)28 Cost–

minimisation

UT, US Fixed

wing

D: Retrospective Cost per mile (USD) Fixed wing: $6;

HEMS: $24

No difference

in mortality

or discharge

disability

Appears to

be service

provider

None

N: 660 (HEMS), 266 (fixed wing)

T: �8 year period

M: Patients transported by HEMS

were compared to patients

transported by fixed wing for

several outcome measures

Gearhart (1997)13 Cost–

effectiveness

US N/A D: Review of literature Cost per life year

saved (USD)

$2454 per life year

saved ($1400–

$9677)

Median 5 extra

survivors per

100 transports

(1.1–12.2)

Stated as

service

provider

Survival,

transport cots,

hospital costs,

remaining life

expectancy

N: 4078 (13 sets of data)

M: Previous TRISS articles were

elicited from a MEDLINE search.

Cost data was obtained from a

single university based helicopter

service

De Wing (2000)9 Cost–

minimisation

CA, US Ground

units

P: Burns Sum of all

transportation

charges

(N = 47) (USD)

Ground: $80,646;

HEMS: $595,780

No difference

in LOS, days on

ventilator or

mortality

Appears to

be service

provider

None

C.B

.T

ay

lor

eta

l./Inju

ry,

Int.

J.C

are

Inju

red4

1(2

01

0)

10

–2

01

3

Table 2 (Continued )

Author (year) Type of

economic

evaluation

Setting Comparator Methods Main health

economic outcome

Economic estimate Main patient

outcome

Viewpoint

adopted

Uncertainties

varied

D: Retrospective

N: 47 (HEMS), 47 (ground)

T: 2 years

M: Burn injured (�30% total body

surface area) transported by HEMS

were compared to matched

controls transported by ground

Mitchell (2007)20 Cost–

effectiveness

Nova Scotia,

CAN

Ground

units

D: Retrospective Cost per life year

saved (CAN)

$2500 per life

year saved

6.4 extra survivors

per 100 transports

for HEMS

Appears to

be service

provider

None

N: 237 (HEMS), 554 (ground)

T: 4 years

M: Actual mortality was compared

to expected mortality for patients

transported by air or ground using

TRISS

Non-trauma

Low (1988)19 Cost–

effectiveness

US N/A P: Obstetric Cost per life saved

(USD)

$1790 per

life saved

HEMS transport

allows 20%

improvement in

survival

Appears to

be service

provider

None

N: 357 (HEMS), 18 (ground) 88

(fixed wing)

D: Prospective

M: Charts were reviewed following

transport of obstetric patients in

33 transport services. HEMS benefit

was established from previous data

Silbergleit

(2003)26

Cost–

effectiveness

MI, US N/A P: Stroke Cost per additional

good outcome;

cost per QALY (USD)

$35,000 per

additional good

outcome; $6100

per QALY

N/A Stated as

broad health

care system

Costs of

transport,

treatment

and long term

care; patient

eligibility

D: Economic model

M: HEMS transport of stroke patients

followed by treatment with IV or

IA thrombolysis was modelled.

Effectiveness was described through

the frequency of good outcome at

3 months

Bardach (2004)1 Cost–

effectiveness

CA, US N/A P: Stroke Cost per QALY (USD) $10,548 per

QALY

N/A Stated as

societal

Cut points to

define low/high

volume hospitals;

absolute mortality

difference; risk

of death in transfer;

costs

D: Economic model

C.B

.T

ay

lor

eta

l./Inju

ry,

Int.

J.C

are

Inju

red4

1(2

01

0)

10

–2

01

4

M: HEMS transport of stroke patients

from a low volume to a high volume

hospital was modelled. Patient

outcome was modelled over the

lifetime in 1 of 3 health states

Selmer (2005)24 Cost–

effectiveness

Norway N/A P: Cardiac Lifetime health care

costs related to

cardiovascular

disease (EUR)

HEMS transport

resulted in gain

of 0.7 life years

and a s5250

saving in lifetime

care costs

N/A Stated as the

provider

(the health

care system)

Model parameters;

age at intervention;

Monte–Carlo

simulations

D: Economic model

M: HEMS transport of patients with

STEMI followed by treatment with

percutaneous coronary intervention

(PCI) was compared to patients

receiving thrombolytic therapy

without transport. Patient outcome

was modelled over the lifetime in

1 of 5 health states

Non-specific

Bruhn (1993)6 Cost analysis MA, US Ground

units

D: Economic model Annual cost for

mandated

performance

outcome (USD)

HEMS: $1.7M;

ground: $3.8M

N/A Appears to

be service

provider

None

M: An economic model was

developed to assess the costs of a

helicopter vs. a network of

advanced care ground ambulances

to achieve a 30 min response time

Lechleuthner

(1994)16

Cost analysis GER Ground

units

D: Economic model Annual cost for

mandated

performance

outcome in three

locations (DM)

Additional HEMS:

12.2–38.2M;

additional ground

units: 15.1–41.2M

N/A Appears to

be service

provider

Number of

ground and

HEMS units

and cost ratio

of ground: HEMS

M: The addition of helicopters or

multiple advanced care ground

units was modelled in three

different geographic locations under

different cost ratios to achieve a

15 min response time

Kurola (2002)15 Cost–

effectiveness

Varkaus,

FIN

N/A D: Retrospective Cost per life saving

or beneficial

mission (EUR)

s28,444 per life

saving or beneficial

mission

HEMS attendance

was ‘life saving or

beneficial’ in 22%

of patients

Stated as

service

provider

None

N: 206

T: 1 year

M: HEMS missions were reviewed

and divided into subgroups as to

whether the patient was judged

to benefit

Elvik (2002)11 Cost–

effectiveness

NOR N/A D: Review of literature Cost per life saved;

cost per beneficial

mission (NOK)

2,952,273 per life

saved; 683,684

per life saving or

beneficial mission

HEMS saves 6% of

patients attended

per year and

improves 20%

Appears to

be societal

C.B

.T

ay

lor

eta

l./Inju

ry,

Int.

J.C

are

Inju

red4

1(2

01

0)

10

–2

01

5

Ta

ble

2(C

on

tin

ued

)

Au

tho

r(y

ea

r)T

yp

eo

f

eco

no

mic

ev

alu

ati

on

Se

ttin

gC

om

pa

rato

rM

eth

od

sM

ain

he

alt

h

eco

no

mic

ou

tco

me

Eco

no

mic

est

ima

teM

ain

pa

tie

nt

ou

tco

me

Vie

wp

oin

t

ad

op

ted

Un

cert

ain

tie

s

va

rie

d

M:

Pre

vio

us

lite

ratu

rew

as

use

dto

est

ima

teth

ee

ffe

cts

of

HE

MS

on

surv

iva

la

nd

qu

ali

tyo

fli

fe

Elv

ik(2

00

2)1

1C

ost

–b

en

efi

tN

OR

N/A

D:

Re

vie

wo

fli

tera

ture

Co

st–

be

ne

fit

rati

o

for

HE

MS

(NO

K)

Be

ne

fit:

76

2.1

M;

cost

:1

29

.9M

(ra

tio

:1

:5.8

7)

As

ab

ov

eA

pp

ea

rsto

be

soci

eta

l

Pe

rce

nta

ge

of

life

sav

ing

mis

sio

ns

an

d

pe

rce

nta

ge

of

pa

tie

nts

wit

h

imp

rov

em

en

t

inq

ua

lity

of

life

M:

Th

ea

nn

ua

lco

sto

fH

EM

Sw

as

com

pa

red

toth

eco

stsa

vin

gs

of

pre

ve

nti

ng

44

de

ath

sa

nd

imp

rov

ing

the

liv

es

of

14

6p

eo

ple

(ex

tra

po

late

dfr

om

the

cost

of

a

roa

dtr

affi

cd

ea

th)

P:

pa

tie

nt;

D:

de

sig

n;

N:

nu

mb

ers

;T

:ti

me

pe

rio

d;

M:

me

tho

ds.

C.B. Taylor et al. / Injury, Int. J. Care Injured 41 (2010) 10–2016

$2,436,178 (£1,146,861), demonstrating a 21-fold variation(Table 3). Potential savings in ground ambulance costs due toHEMS operations were estimated in one study21 at $210,221(£97,805) per annum. Among all studies, none were able to relatethe higher cost of HEMS to an incremental patient benefit such asreduced mortality and/or morbidity (Table 2).

Secondary (inter-facility) transport

One cost–minimisation study28 from the US evaluated thesecondary transport of trauma patients by HEMS and fixed wing inrelation to patient benefit expressed as hospital length of stay,mortality and discharge disability score. Costs were estimated on aper mile basis and were calculated at $40 ($24) per mile for HEMSwith fixed wing transport estimated to be four times less, at $10($6) per mile. Despite the higher costs of HEMS, no incrementalpatient benefit was found (Table 2).

Primary and secondary transport

Two studies9,13 from the US and one20 from Canada evaluatedboth primary and secondary HEMS transport of trauma. Groundtransport was used as a comparator in two evaluations.9,20 Costcalculations for HEMS were not directly comparable and includedan annual cost estimate ($3,023,568 [$2,253,952])13 and anestimate for the sum of transportation charges, which could beextrapolated into a cost per patient transport ($15,849 [$12,676]).9

The latter study (classified as cost–minimisation) also estimatedthe sum of charges for ground transport ($2145 per transport[$1716]), finding HEMS to be seven times more expensive withoutconferring additional patient benefit as measured by length ofhospital stay and mortality. In contrast, two cost–effectivenessstudies13,20 found that HEMS prevented expected deaths followingtrauma, producing similar cost–effectiveness ratios of $329213

($2454) and $222720 ($2500) per life year saved, respectively(Table 2).

Non-trauma patients

Three studies1,19,26 from the US and one study from Norway24

focused specifically on non-trauma patients, including patientdiagnosis categories of obstetric, cardiac and stroke. Among thesestudies, one study26 provided an annual cost estimate for HEMS at$5,571,578 ($4,761,524) (Table 3). All four studies modelled thecost of HEMS transport in combination with upgraded clinicaltreatment, with three studies1,19,26 producing cost–effectivenessratios and one study24 comparing life years gained and lifetimetreatment costs with non-transported patients. For stroke patients,two studies found HEMS transport to cost $713826 ($6100) and$12,0221 ($10,548) per quality adjusted life year, respectively withthe former study also producing a ratio of $40,954 ($35,000) peradditional good outcome (minimal or no disability). For cardiacpatients (patients with ST-elevated MI), helicopter transport toreceive percutaneous coronary intervention (PCI) was found toresult in 0.7 life year gain and a reduction in lifetime care costs of$7854 (s5250) compared to non-transported patients.24 Finally,one study evaluating transport of obstetric patients19 found HEMStransport to cost $3258 ($1790) per life saved for neonates(Table 2).

Non-specific patients

Four studies6,11,15,16 from four countries (US, Germany, Finlandand Norway) included non-specific patient diagnoses. For thesestudies, the economic cost of HEMS was estimated in a variety ofways and only two studies6,16 used ground transport as a

Table 3Characteristics of HEMS and components included in annual cost calculations for studies including an annual cost estimate of a HEMS.

Study Characteristics of HEMS Components included in cost estimate Cost estimate per annum 2008 equivalent value

per annum (USD)

Bruhn et al.6 Helicopter: BK 117 A3 Office utilities $1,686,500 $2,512,862

Staffed by paramedic and nurse Medical supplies

Operates in the state of Massachusetts Other supplies

Operates 24/7 Maintenance

Used as an addition to ground based

services (30 min response time)

Unexpected costs

Targets emergency trauma and critically

ill patients

Vehicle fuel

Training

Lease of helicopter

Personnel (nurse, paramedic, pilot,

clinical operations director, office staff)

Lechleuthner et al.16 Cologne, Sanderbusch and Traunstein Personnel (2.5 pilots, 2.5 paramedics,

2.5 physicians)

1,575,000–2,500,000 DM $1,319,113–$2,093,829

Staffed by physician and paramedic Maintenance/technical (gas, inspection,

leasing, depreciation)

Operates in daylight hours (Cologne and

Traunstein) and 24/7 (Sanderbusch)

Single pilot

Used in addition to ground based services

to achieve a 15 min response time

Nichol et al.21 Helicopter: Stretched Bolkow 105 Capital and running costs for the helicopter £594,666 $1,278,173

Staffed by paramedics Staff and training costs

Operates in rural environment Costs of hangarage and developing appropriate

landing sites

Operates 8 h/day Costs of equipping the aircraft

Single pilot

Used as first responder

Targets trauma and medical emergencies

Undertakes primary and secondary missions

Snooks et al.27 Cornwall Pre-hospital capital and operating costs of the

helicopter

Cornwall: £594,666 Cornwall: $1,278,173

See Nicholl21 Special landing and patient transfer facilities

(where appropriate)

London: £1,146,861 London: $2,436,178

London Crewing costs Sussex: £54,503 Sussex: $115,777

See Brazier5 Consequences for the ambulance service concerned Wiltshire: £151,200 Wiltshire: $321,182

Sussex Impact of helicopter on hospital services

Helicopter: Bolkow 105

Staffed by paramedic and police observer

Operates in rural/urban environment

Operates daylight and evenings

Single pilot

Targets primarily trauma

Undertakes primary missions only

Joint police and ambulance service venture

Wiltshire

Not available

Brazier et al.5 Helicopter: Aerospatiale Dauphin 365N Helicopter £1,146,861 $2,436,178

Staffed by a physician and a paramedic Landing deck

Operates in the London metropolitan area Staff (including pilots, crew, dispatchers and

management)

Operates in daylight hours only Other costs (insurance, fuel, fees, etc.)

C.B

.T

ay

lor

eta

l./Inju

ry,

Int.

J.C

are

Inju

red4

1(2

01

0)

10

–2

01

7

Ta

ble

3(C

on

tin

ued

)

Stu

dy

Ch

ara

cte

rist

ics

of

HE

MS

Co

mp

on

en

tsin

clu

de

din

cost

est

ima

teC

ost

est

ima

tep

er

an

nu

m2

00

8e

qu

iva

len

tv

alu

e

pe

ra

nn

um

(US

D)

Tw

op

ilo

tsM

ain

ten

an

ce

Use

da

sa

na

dd

itio

nto

gro

un

db

ase

dre

spo

nse

Ta

rge

tstr

au

ma

on

ly

Un

de

rta

ke

sp

rim

ary

an

dse

con

da

rym

issi

on

s

Ge

arh

art

et

al.

13

He

lico

pte

r:D

au

ph

inA

S-3

65

N2

Pe

rso

nn

el

cost

s(s

ala

rie

s,e

mp

loy

ee

be

ne

fits

,

tra

inin

ge

xp

en

ses)

$2

,25

3,9

52

$3

,02

3,5

68

Sta

ffe

db

yp

ara

me

dic

an

da

nu

rse

Ca

pit

al

cost

s(e

qu

ipm

en

tp

urc

ha

ses

an

d

de

pre

cia

tio

no

fth

eh

an

ga

rfa

cili

tya

nd

the

he

lico

pte

r)

Op

era

tes

inru

ral

an

dsu

bu

rba

na

rea

Op

era

tio

na

lco

sts

(fu

el,

ma

inte

na

nce

,p

urc

ha

sed

serv

ice

s,

lea

seo

fre

pla

cem

en

ta

ircr

aft

,a

dm

inis

tra

tio

n)

Sin

gle

pil

ot

Un

de

rta

ke

sp

rim

ary

an

dse

con

da

rym

issi

on

s

Sil

be

rgle

ite

ta

l.2

6N

ot

av

ail

ab

leP

ers

on

ne

l$

4,7

61

,52

4$

5,5

71

,57

8

Ca

pit

al

ex

pe

nse

s

Op

era

tio

ns

Ad

min

istr

ati

on

Insu

ran

ce

Me

dic

al

sup

pli

es

C.B. Taylor et al. / Injury, Int. J. Care Injured 41 (2010) 10–2018

comparator. Both studies were classified as cost analysis studies,and modelled the annual cost of a HEMS in comparison to multipleground transport units, finding HEMS to be operationally cheaperfor a given geographic area. The annual cost of HEMS in thesestudies ranged from $1,319,11316 (1,575,000 DM) to $2,512,8626

($1,686,500), compared to ground transport which ranged from$2,522,64616 (3,012,000 DM) to $5,667,9076 ($3,804,000) for thesame geographical coverage. In contrast, two cost–effectivenessstudies using differing study methods11,15 found HEMS to be ofincremental benefit above usual care and produced estimates of$30,36515 (s28,444) and $91,47811 (683,684 NOK) per life savingor beneficial mission. In the latter study, also classified as a cost–benefit study,11 the cost of HEMS was related to the societalbenefits (of saving and improving lives), producing a 1:6 cost tobenefit ratio.

Discussion

Numerous studies have attempted to assess if HEMS improvespatient outcomes in trauma compared to ground based services or‘usual care’.33 In non-trauma patients, studies have focused on thesafety and feasibility of transporting patients between facilitiesusing HEMS.32 Due to a lack of Level 1 evidence (randomised trials),much of the research has been unable to demonstrate that HEMSimproves patient outcomes. Despite this, authors have commentedthat the weight of observational evidence supports an associationbetween HEMS and a reduction in patient mortality in certaintrauma patients.29 The potential benefits of HEMS in traumainclude a reduction in the time to definitive care as well as theprovision of advanced pre-hospital interventions by highly trainedstaff. Outside of trauma, HEMS provide stable and expedienttransport between facilities. Given the constant evolution ofemergency services, there is a continual need for further studies toassess the advantages of HEMS in both trauma and non-traumapatient populations.

As well as the benefit provided by HEMS, another importantquestion is the relative cost and its association with any benefitprovided. From studies in this review, the annual cost of HEMSranged from $115,777 to $5,571,578 per annum, incorporating a48-fold variation. In terms of the cost–effectiveness of HEMS, eightstudies reported variant cost–effectiveness ratios including $3292and $2227 per life year saved, $7138 and $12,022 per qualityadjusted life year, $40,954 per ‘additional good outcome’, $3258per life saved and $30,365 and $91,478 per life saving or beneficialmission. In contrast, five studies showed HEMS to be a moreexpensive transport alternative compared to ground and fixedwing transport without an incremental patient benefit.

Having provided a synthesis of the current economic evidenceregarding HEMS, a discussion point in this investigation was toconsider the external validity of this evidence and whether theinformation can be useful in guiding service provision in otherhealth systems and jurisdictions. It is clear that overall there isconsiderable variation in the study findings included in thisreview. Three key considerations were identified as being integralto this variation: first, differences in the nature of the interventionand service provision; second, differences in the context in whichthe study was conducted; and third, differences in study methodand cost and benefit calculation. These three considerations andtheir implications for HEMS are discussed below.

Differences in interventions and service provision

An underlying fundamental reason for the differences in annualcosts of HEMS was the difference in intervention delivery betweenthe services. Snooks et al.27 highlighted the economic conse-quences of this variation for services operating in the same

C.B. Taylor et al. / Injury, Int. J. Care Injured 41 (2010) 10–20 19

country, showing a 21-fold variation in annual cost. Across studieswhich included an estimate of annual cost, differences betweenservices included capital costs such as the type of helicopter,operational costs such as fuel (and hence geographical areacovered) and personnel costs such as the number of pilots and theuse of a physician or paramedic. The discrepancy in cost due to thevariation in HEMS delivery is important justification as to theproblem of generalising results in this area.

Patient population represents a key element of differencebetween the studies. For studies classified as examining traumapopulations, two13,20 demonstrated a mortality benefit fromHEMS, while, four studies5,9,27,28 showed HEMS to be considerablymore expensive than ground transport without an incrementalpatient benefit. These divergent findings are consistent throughoutthe trauma literature.30,31,33 Put simply, in order to achieve patientbenefit for trauma, HEMS needs to either reduce time to definitivecare for the patient, or bring more appropriate care than thatprovided by ground services. Currently, the literature remainsmixed as to the time advantages of HEMS4,7 as well as the benefit ofadvanced treatment in the pre-hospital environment.18,25 Onestudy in this review13 relating the cost of HEMS to benefit derivedfrom previous literature, did highlight the potential of HEMS to becost-effective for trauma. Given the cost implications of thedifferent HEMS delivery, there is an imperative to test this findingwithin differing jurisdictions using consistent analytical methods.

Despite many HEMS now transporting a majority of non-trauma patients between hospitals, only four studies1,19,24,26 wereidentified which had reviewed HEMS transport for a non-traumapopulation. All four studies demonstrated that the cost of this formof transport was justifiable with respect to the incremental patientbenefit. In these patient populations, HEMS provides expedienttransport for time-critical conditions (such as stroke), critical carelevel transport for inter-facility transfers and can also reduce theneed for sophisticated and costly hospital services in rurallocations due to the ability to cover large distances with anappropriately trained crew. If HEMS continues to operate in theseroles future economic evaluations must account for this workloadas well as assessing the benefit of HEMS transport to both thepatient and the broader health system.

Differences in context

Context plays a crucial role in the economic evaluation of HEMS,more so than with most other health care interventions. Thisincludes elements such as the health system funding model,system design and physical geography in which HEMS operate.Within different health systems, the financial implications ofHEMS are variable. For example in a user-pays health care model,HEMS are often an integral part of the financial sustainability of aninstitution by broadening the patient catchment area. In contrast,within a socialised model of health care service, HEMS are requiredto be financially acceptable within the broader system and‘compete’ against the many services funded by the single provider.

Different health system designs also impact on the financialcost of HEMS and its comparator. Despite a commonly heldperception that HEMS is more expensive than ground transport,two studies6,16 in this review showed HEMS to be operationallycheaper than multiple ground units. This contradiction is seldommentioned in the literature and is a consequence of the variation inthe emergency medicine system design between jurisdictions.Across systems a spectrum of designs exist, ranging from a singleemergency responder to a more complex tiered system incorpor-ating multiple emergency response ‘layers’. For instance in therural environment HEMS represents a mechanism to cover largedistances in an expedient fashion and are at times used as theprimary responder. In urban contexts, HEMS can provide a second,

more advanced ‘layer’ on top of regular ground based services. Inpractice, HEMS operate across this spectrum adding weight to theinherent difficulty of generalising findings from economicevaluations of HEMS.

Variation in study method

With respect to study method, a previous review by Lerneret al.17 of the economic literature in the emergency medicineenvironment highlighted a lack of high-quality studies with theauthors suggesting a conceptual framework was needed to guideresearchers. Studies in the present review contained methodolo-gical weaknesses, best exemplified by few providing a full costingof HEMS against a specified comparator while others used patientcharges as a proxy for actual costs.

The viewpoint adopted in a health economic analysis is centralto the validity of results as it determines who the costs andoutcomes are applicable to, which generally includes either thepatient, the institution or the broader society.10 Many studies inthis review adopted a broad or poorly defined viewpoint, such thatthey either did not specify the viewpoint adopted or theyconsidered some of the costs and consequences of HEMS forsociety, but not all. For HEMS, the viewpoint adopted can haveconsiderable implications for its cost–effectiveness and futureevaluations need to consider a narrower, or well defined andconsistent focus. The lack of consistent viewpoints for studies inthis review serves to highlight the difficulty of using study findingsto base decisions about the appropriate use of HEMS in differenthealth care systems.

Limitations

A review of this nature has several limitations including therestrictions of the search, the number of published evaluations andthe complexity in synthesising results. The search strategy used inthis review was limited to published English language articles,which may have led to the omission of important evaluationsundertaken in non-English speaking countries or evaluationsundertaken by government or private institutions. Only fifteenarticles were identified in this review and the variation in method,patient population, service structure and context made synthesisof study outcomes necessarily complex. As such, synthesisincluded a degree of subjectivity and the study elements reportedmay not provide a true reflection of the individual studiesthemselves. In addition, all monetary information was convertedinto US dollars as at the 1st of January in the year of publication andthen adjusted for inflation to 2008 equivalent values. In doing so, itis accepted that the values obtained were somewhat crude giventhe variation in costing utilised, the potential variation in costbetween the data collection period and year of publication, andfinally the higher inflation of medical goods and services.

Conclusion

Despite an established role within health systems and theapparent willingness of the public to pay for helicopter basedsystems,23 HEMS are expensive to implement and run. Thisexpense may be justified if HEMS are cheaper than ground services,provides improved patient outcomes or expedites the transfer ofpatients that cannot be transported by any other means. Thisreview highlighted the considerable variation in HEMS serviceoperations across health systems and the subsequent difficulties ingeneralising findings. Across the studies, wide variation was notedin the nature of HEMS interventions, the context in which HEMSoperates (e.g. rural or urban) and the method used to arrive at costsand benefits. Under specific assumptions and circumstances,

C.B. Taylor et al. / Injury, Int. J. Care Injured 41 (2010) 10–2020

HEMS has been shown to be economically justifiable for trauma,however equally, other studies have reported findings to thecontrary. In non-trauma populations such as stroke, myocardialinfarction and obstetrics this review highlighted emergingevidence as to the cost–effectiveness of using HEMS in thesepopulations. Given such findings and the widespread use of HEMS,the high costs to the health system and the variation in HEMSdelivery, there is an imperative to evaluate HEMS across their fulloperational workload including both trauma and non-traumapatients using consistent economic cost–benefit calculationmethods. The difficulties inherent in synthesising and extrapolat-ing the findings of studies conducted in different settings meansthat future studies need to completely account for these contextualdeterminants of cost–effectiveness and tailor evaluations togenerate locally relevant economic and clinical evidence. Theresults of such evaluations can then be used to directly addresslocally relevant health service delivery policy.

Conflicts of interest

The authors have no conflicts of interest to report.

References

1. Bardach NS, Olson SJ, Elkins JS, et al. Regionalization of treatment for subar-achnoid hemorrhage: a cost–utility analysis. Circulation 2004;109:2207–12.

2. Baxt WG, Moody P. The impact of advanced prehospital emergency care on themortality of severely brain-injured patients. Journal of Trauma-Injury Infection& Critical Care 1987;27:365–9.

3. Biewener A, Aschenbrenner U, Rammelt S, et al. Impact of helicopter transportand hospital level on mortality of polytrauma patients. Journal of Trauma-Injury Infection & Critical Care 2004;56:94–8.

4. Biewener A, Aschenbrenner U, Sauerland S, et al. Impact of rescue method andthe destination clinic on mortality in polytrauma. A status report. Unfallchirurg2005;108:370–7.

5. Brazier J, Nicholl J, Snooks H. The cost and effectiveness of the London Heli-copter Emergency Medical Service. Journal of Health Services & Research Policy1996;1:232–7.

6. Bruhn JD, Williams KA, Aghababian R. True costs of air medical vs. groundambulance systems. Air Medical Journal 1993;12:262–8.

7. Carr BG, Caplan JM, Pryor JP, Branas CC. A meta-analysis of prehospital caretimes for trauma. Prehospital Emergency Care 2006;10:198–206.

8. Carter G, Couch R, O’Brien DJ. The evolution of air transport systems: a pictorialreview. Journal of Emergency Medicine 1988;6:499–504.

9. De Wing MD, Curry T, Stephenson E, et al. Cost-effective use of helicopters forthe transportation of patients with burn injuries. Journal of Burn Care &Rehabilitation 2000;21:535–40.

10. Drummond MF, Richardson WS, O’Brien BJ, et al. Users’ guides to the medicalliterature. XIII. How to use an article on economic analysis of clinical practice. A.Are the results of the study valid? Evidence-Based Medicine Working Group.JAMA 1997;277:1552–7.

11. Elvik R. Cost–benefit analysis of ambulance and rescue helicopters in Norway:reflections on assigning a monetary value to saving a human life. AppliedHealth Economics & Health Policy 2002;1:55–63.

12. Frankema SPG, Ringburg AN, Steyerberg EW, et al. Beneficial effect of helicopteremergency medical services on survival of severely injured patients. BritishJournal of Surgery 2004;91:1520–6.

13. Gearhart PA, Wuerz R, Localio AR. Cost–effectiveness analysis of heli-copter EMS for trauma patients. Annals of Emergency Medicine 1997;30:500–6.

14. Hankins DG. Air medical transport of trauma patients. Prehospital EmergencyCare 2006;10:324–7.

15. Kurola J, Wangel M, Uusaro A, Ruokonen E. Paramedic helicopter emergencyservice in rural Finland—do benefits justify the cost? Acta AnaesthesiologicaScandinavica 2002;46:779–84.

16. Lechleuthner A, Koestler W, Voigt M, Laufenberg P. Helicopters as part of aregional EMS system—a cost–effectiveness analysis for three EMS regions inGermany. European Journal of Emergency Medicine 1994;1:159–66.

17. Lerner EB, Maio RF, Garrison HG, et al. Economic value of out-of-hospitalemergency care: a structured literature review. Annals of Emergency Medicine2006;47:515–24.

18. Liberman M, Mulder D, Sampalis J. Advanced or basic life support for trauma:meta-analysis and critical review of the literature. Journal of Trauma-InjuryInfection & Critical Care 2000;49:584–99.

19. Low RB, Martin D, Brown C. Emergency air transport of pregnant patients: thenational experience. Journal of Emergency Medicine 1988;6:41–8.

20. Mitchell AD, Tallon JM, Sealy B. Air versus ground transport of major traumapatients to a tertiary trauma centre: a province-wide comparison using TRISSanalysis. Canadian Journal of Surgery 2007;50:129–33.

21. Nicholl JP, Beeby NR, Brazier JE. A comparison of the costs and performance ofan emergency helicopter and land ambulances in a rural area. Injury1994;25:145–53.

22. OANDA. FXHistory—historical currency exchange rates; 2009.23. Ringburg AN, Buljac M, Stolk EA, et al. Willingness to pay for lives saved by

helicopter emergency medical services. Prehospital Emergency Care 2009;13:37–43.

24. Selmer R, Halvorsen S, Myhre KI, et al. Cost–effectiveness of primary percu-taneous coronary intervention versus thrombolytic therapy for acute myo-cardial infarction. Scandinavian Cardiovascular Journal 2005;39:276–85 [seecomment].

25. Sethi D, Kwan I, Kelly AM, Roberts I. Advanced trauma life support training forambulance crews. Cochrane Database of Systematic Reviews 2008;2.

26. Silbergleit R, Scott PA, Lowell MJ, Silbergleit R. Cost–effectiveness of helicoptertransport of stroke patients for thrombolysis. Academic Emergency Medicine2003;10:966–72.

27. Snooks HA, Nicholl JP, Brazier JE, Lees-Mlanga S. The costs and benefits ofhelicopter emergency ambulance services in England and Wales. Journal ofPublic Health Medicine 1996;18:67–77.

28. Thomas F, Wisham J, Clemmer TP, et al. Outcome, transport times, and costs ofpatients evacuated by helicopter versus fixed-wing aircraft. Western Journal ofMedicine 1990;153:40–3.

29. Thomas S, Biddinger P. Helicopter trauma transport: an overview of recentoutcomes and triage literature. Current Opinion in Anaesthesiology 2003;16:153–8.

30. Thomas SH. Helicopter emergency medical services transport outcomes litera-ture: annotated review of articles published 2000–2003. Prehospital Emer-gency Care 2004;8:322–33.

31. Thomas SH. Helicopter EMS transport outcomes literature: annotated reviewof articles published 2004–2006. Prehospital Emergency Care 2007;11:477–88.

32. Thomas SH, Cheema F, Cumming M, et al. Nontrauma helicopter emergencymedical services transport: annotated review of selected outcomes-relatedliterature. Prehospital Emergency Care 2002;6:242–55.

33. Thomas SH, Cheema F, Wedel SK, Thomson D. Trauma helicopter emergencymedical services transport: annotated review of selected outcomes-relatedliterature. Prehospital Emergency Care 2002;6:359–71.

34. Thomas SH, Harrison TH, Buras WR, et al. Helicopter transport and blunt traumamortality: a multicenter trial. Journal of Trauma-Injury Infection & Critical Care2002;52:136–45.

35. United States Department of Labor Bureau of Labor Statistics. Inflation calcu-lator; 2009.