economic evaluations in the diagnosis and management of ... · adherence to brain trauma foundation...

Avai lable onl ine at www.sc iencedirect .com

V A L U E I N H E A L T H 1 8 ( 2 0 1 5 ) 7 2 1 – 7 3 4

1098-3015$36.00 – s

Published by Elsevie

http://dx.doi.org/10

E-mail: aziz.alali

* Address correspoToronto, ON, Canad

journal homepage: www.elsevier .com/ locate / jva l

Economic Evaluations in the Diagnosis and Management ofTraumatic Brain Injury: A Systematic Review and Analysisof QualityAziz S. Alali, MD, PhD1,2,3,*, Kirsteen Burton, MD, MSc, MBA2,4, Robert A. Fowler, MDCM, MSc1,2,5,6,David M.J. Naimark, MD, MSc2,7, Damon C. Scales, MD, PhD1,2,5,6, Todd G. Mainprize, MD8,Avery B. Nathens, MD, PhD, MPH1,2,9,10

1Sunnybrook Research Institute, Sunnybrook Health Sciences Center, Toronto, ON, Canada; 2Institute of Health Policy, Managementand Evaluation, University of Toronto, Toronto, ON, Canada; 3Division of Neurosurgery, University of Ottawa, Ottawa, ON, Canada;4Department of Medical Imaging, University of Toronto, Toronto, ON, Canada; 5Department of Critical Care, Sunnybrook HealthSciences Center, Toronto, ON, Canada; 6Interdepartmental Division of Critical Care, University of Toronto, Toronto, ON, Canada;7Department of Medicine, University of Toronto, Toronto, ON, Canada; 8Division of Neurosurgery, Sunnybrook Health SciencesCenter, University of Toronto, Toronto, ON, Canada; 9Keenan Research Centre of the Li Ka Shing Knowledge Institute, St Michael’sHospital, Toronto, ON, Canada; 10Department of Surgery, Sunnybrook Health Sciences Center, University of Toronto, Toronto, ON,Canada

A B S T R A C T

Background: Economic evaluations provide a unique opportunity toidentify the optimal strategies for the diagnosis and management oftraumatic brain injury (TBI), for which uncertainty is common and theeconomic burden is substantial. Objective: The objective of this studywas to systematically review and examine the quality of contempo-rary economic evaluations in the diagnosis and management of TBI.Methods: Two reviewers independently searched MEDLINE, EMBASE,Cochrane Central Register of Controlled Trials, NHS Economic Evalua-tion Database, Health Technology Assessment Database, EconLit, andthe Tufts CEA Registry for comparative economic evaluations pub-lished from 2000 onward (last updated on August 30, 2013). Data onmethods, results, and quality were abstracted in duplicate. The resultswere summarized quantitatively and qualitatively. Results: Of 3539citations, 24 economic evaluations met our inclusion criteria. Ninewere cost-utility, five were cost-effectiveness, three were cost-mini-mization, and seven were cost-consequences analyses. Only six

ee front matter Copyright & 2015, International S

r Inc.

.1016/j.jval.2015.04.012

@mail.utoronto.ca.

ndence to: Aziz S. Alali, Institute of Health Policy,a M4N 3M5.

studies were of high quality. Current evidence from high-qualitystudies suggests the economic attractiveness of the following strat-egies: a low medical threshold for computed tomography (CT) scan-ning of asymptomatic infants with possible inflicted TBI, selective CTscanning of adults with mild TBI as per the Canadian CT Head Rule,management of severe TBI according to the Brain Trauma Foundationguidelines, management of TBI in dedicated neurocritical care units,and early transfer of patients with TBI with nonsurgical lesions toneuroscience centers. Conclusions: Threshold-guided CT scanning,adherence to Brain Trauma Foundation guidelines, and care forpatients with TBI, including those with nonsurgical lesions, in speci-alized settings appear to be economically attractive strategies.Keywords: cost-benefit analysis, cost-effectiveness, cost utility,economic evaluation, systematic review, traumatic brain injury.

Copyright & 2015, International Society for Pharmacoeconomics andOutcomes Research (ISPOR). Published by Elsevier Inc.

Introduction

Traumatic brain injury (TBI) presents a major economic, social,and health challenge worldwide [1]. According to the WorldHealth Organization, TBI will surpass many diseases as the majorcause of death and disability by 2020 [2]. In the United States,53,000 individuals die annually from TBIs, and at least 5.3 millionAmericans are currently living with long-term disabilities directlyattributable to TBI [3,4]. In 2010, the estimated overall burden ofTBI on the US economy was approximately $76.5 billion, withcosts for disability and lost productivity outweighing those for

acute medical care and rehabilitation [5,6]. Despite the profoundconsequences of TBI, there is substantial uncertainty surround-ing the costs and benefits of several alternative diagnostic andtherapeutic strategies from the societal and health care systemperspectives [7,8].

Economic evaluations are tools to weigh the relative costs andbenefits of alternative courses of action [9]. They provide meansto support decision making in a cost-constrained environment ofhealth care and can be used to explore areas of uncertainty inboth costs and benefits. Without considering economic data, theallocation of resources may not lead to optimal value for health

ociety for Pharmacoeconomics and Outcomes Research (ISPOR).

Management and Evaluation, 2075 Bayview Avenue, Room D-574,

http://dx.doi.org/10.1016/j.jval.2015.04.012http://dx.doi.org/10.1016/j.jval.2015.04.012http://dx.doi.org/10.1016/j.jval.2015.04.012http://crossmark.crossref.org/dialog/?doi=10.1016/j.jval.2015.04.012&domain=pdfhttp://crossmark.crossref.org/dialog/?doi=10.1016/j.jval.2015.04.012&domain=pdfhttp://crossmark.crossref.org/dialog/?doi=10.1016/j.jval.2015.04.012&domain=pdfmailto:[email protected]://dx.doi.org/10.1016/j.jval.2015.04.012

V A L U E I N H E A L T H 1 8 ( 2 0 1 5 ) 7 2 1 – 7 3 4722

care spending. Therefore, economic evaluations provide a uniqueopportunity to identify the optimal strategies for the diagnosisand management of TBI, for which uncertainty is common andthe consequential economic burden on patients and health caresystem is substantial [1,7,10]. Inadequately informed, designed,or poorly executed economic evaluations, however, may ill-inform resource allocation and lead to poor health policy deci-sions [9].

In this context, we systematically reviewed and evaluated thequality of economic evaluations in the diagnosis and manage-ment of TBI.

Methods

Search Methods

We duplicate searched the following databases: MEDLINE,EMBASE, NHS Economic Evaluation Database, Health TechnologyAssessment Database, EconLit, Tufts CEA Registry, and CochraneCentral Register of Controlled Trials. The search retrieved articlespublished between January 1, 2000, and August 30, 2013. Wefollowed published guidelines on EMBASE searches for economicstudies [11]. We also searched the reference lists of relevantstudies, and the Web sites of health technology assessmentagencies, including the Canadian Agency for Drugs and Tech-nologies in Health and the National Institute for Health ResearchHealth Technology Assessment program. The search was notrestricted by language. We did not consider conference abstractsor unpublished data. The full text of any cited article that wasconsidered potentially relevant was retrieved. The search strat-egies are described in detail in the Appendix in SupplementaryMaterials found at http://dx.doi.org/10.1016/j.jval.2015.04.012.

Study Selection

Type of studiesWe included published comparative economic evaluations. Thesearch was limited to the literature published from 2000 onward,to focus on contemporary studies that are potentially relevant tothe current economic environment. Furthermore, older studiesare likely to derive their input values from outdated clinicaleffectiveness and cost-estimation data. We included both singlestudy–based (i.e., clinical data were derived from a single study)and model-based (i.e., incorporating clinical data from varioussources) economic evaluations.

Type of patientsEconomic evaluations related to patients of any age who hadmild, moderate, and/or severe TBI were included. Studies thatincluded general trauma patients or those with acquired braininjury were included only if a subgroup analysis on patients withTBI was reported.

Type of comparatorsWe included economic evaluations of any diagnostic modality,medical, or surgical intervention aimed at the diagnosis and/ormanagement of patients with TBI.

We scanned the abstracts of every record retrieved to deter-mine which studies should be further assessed. If it was clearfrom the abstract that the article was irrelevant, we rejected therecord. The full texts of the remaining articles were evaluated.Two reviewers (A.A. and K.B.) independently assessed and deter-mined the eligibility of each study. Disagreements were resolvedby discussion and consensus.

Data Extraction

Data abstraction forms were created to collect the relevant datafrom the included studies (see Appendix in SupplementalMaterial). Two review authors (A.A. and K.B.) independentlyextracted the data on target population, comparators, methods,outcomes, and results using a data extraction form (see Appendixin Supplemental Material). Any discrepancy in data extractionwas resolved by discussion and consensus. We divided theeconomic evaluations into five groups as defined by Drummondet al. [9]: cost-utility analyses, cost-effectiveness analyses,cost-benefit analyses, cost-minimization analyses, and cost-consequences analyses. To facilitate comparison, costs wereconverted, if price date and currency were specified, to 2012international dollars (Geary-Khamis dollars) using the WorldBank’s purchasing power parity conversion factors, after adjust-ing for temporal changes in country-specific gross domesticproduct [12,13]. The international dollar (Int.$) is a hypotheticalunit of currency that has the same purchasing power parity (i.e.,same ability to buy the same amount of goods and services) thatUS dollar had in the United States at a given point in time [14].

Assessment of Methodological Quality

Two review authors (A.A. and K.B.) independently assessed thequality of each included study using the 24-item Consolidated HealthEconomic Evaluation Reporting Standards checklist [15,16]. Qualityscores were assigned on the basis of this checklist (range 0–24 points).Any disagreement was resolved by discussion and consensus.

The results of all economic evaluations that met our eligibilitycriteria were summarized qualitatively and quantitatively(Tables 1–3). Studies of high quality (quality score 4 20/24 [480%])were discussed in more detail.

Results

Results of the Search

Our search strategy yielded a total of 4079 potentially relevantcitations; 540 duplicates were excluded, and 3465 were excludedafter scanning the abstracts because they did not meet ourinclusion criteria (Fig. 1). A total of 74 citations were retrievedfor detailed evaluation of full-text articles. Fifty of these articleswere excluded because they did not meet our inclusion criteria(further details are available in the Appendix in SupplementalMaterial). Common reasons for exclusion were lack of compara-tor, combining data from patients with TBI and other causes ofacquired brain injury (e.g., stroke) in the analysis, and reportinghospital charges (rather than costs) only.

Description of Included Studies

Twenty-four economic evaluations met the inclusion criteria.Nine studies were cost-utility analyses [17–25], five were cost-effectiveness analyses [26–30], three were cost-minimizationanalyses [31–33], and seven were cost-consequences analyses[34–40]. There were no cost-benefit analyses that met our inclu-sion criteria.

Tables 1, 2, and 3 summarize the characteristics and findingsof economic evaluations of diagnostic approaches, managementstrategies, and alternative structures of care, respectively, forpatients with TBI. Further details are provided in the Appendix inSupplemental Material. Seven studies examined diagnostic strat-egies for mild TBI in children and/or adults [20–22,26,31,33,40],two compared different screening strategies for suspected bluntcerebrovascular injury [24,27], two evaluated management strat-egies for patients with mild TBI [20,23], and four studies

http://dx.doi.org/10.1016/j.jval.2015.04.012

Table 1 – Economic evaluations of diagnostic strategies.

Study Design Population Setting Perspective(currency)

Comparators Time horizon Incremental costsand outcomes

Qualityscore

Af Geijerstamet al. [31]

Model-basedcost-minimizationanalysis

Patients with mildTBI; age was notspecified

Swedish HealthCare System

Not specified(1998 poundsterling)

CT scan and homecare strategy vs.in-hospitalobservation

2 d 36% reduction in costwith CT strategy(£300 vs. £470 [Int.$612 vs. 959])

15/24

Campbellet al. [26]

Model-basedcost-effectivenessanalysis

Asymptomatic 5-wk-old infant witheither 1)unexplained scalpbruising or 2) ahistory of anapparent life-threatening event(ALTE)

ED. Location wasnot specified

Medical payerand societal(2005 USdollars)

CT head onpresentation vs.discharge fromhospital with noCT scan

Until age 52 wkor death

From the medicalpayer perspective,the CT strategysaves $3880 (Int.$4500) for everysevere/fatal inflictedTBI averted in thescalp-bruisingscenario while CTcosts $72,744 (Int.$84,368) per caseaverted in the ALTEscenario. From thesocietal perspective,CT costs $132,701(Int.$153,907) and$209,328 (Int.$242,779) per caseaverted for bruisingand ALTE scenarios,respectively

22/24

Kaye et al.[27]

Model-basedcost-effectivenessanalysis

Adults withsuspected bluntcerebrovascularinjury (based onpredefinedcriteria)

Not specified Institutional andsocietal (2008US dollars)

No screening, duplexultrasound,magneticresonanceangiography,catheterangiography, andCT angiography(CTA)

Not specified From the societalperspective, CTAwas the dominantstrategy ($3727 [Int.$3983] per patientwith 1% stroke rate).From theinstitutionalperspective, duplexultrasound and CTAwere the most cost-effective screeningmodalities, with acost of $8940 (Int.$9554) and $10,670(Int.$11,402) perstroke prevented,respectively

17/24

continued on next page

VALUE

INH

EALTH

18

(2015)721–734

723

Table 1 – continued



Qualityscore

Norlund et al.[40]

Single study–based cost-consequencesanalysis (amulticenterRCT; n ¼ 2602)

Patients with mildTBI (aged Z6 y)

Acute hospitalsin Sweden

Societal (2003euros)

Immediate CTduring triage foradmission vs.observation inhospital

3 mo Clinical outcomeswere similar forboth groups. Thetotal average costwas €718 (Int.$1034)per patient for theCT group vs. €914(Int.$1316) perpatient for theobservation group, adifference of €196(Int.$282)

19/24

Pandor et al.[20]

Model-basedcost-utilityanalysis

Patients with mildTBI aged 1, 10, 40and 75 y,presenting to theED

EDs in the UnitedKingdom

The NHS and thePSS in Englandand Wales(poundsterling; pricedate notspecified)

Diagnostic strategiesfor mild TBI inchildren andadults (CT all,selective CTscanning usingmultiple decisionrules, no CT)

Lifetime 1. For children aged 1 or10 y, the CHALICE ruledominates the otherstrategies by virtue ofgaining more QALYs(mean 22.41) withlower costs (mean£3567)

2. For adults aged 40 y,the CCHRhm strategyhad an ICER of £3879/QALY whencompared with theScandinavianstrategy. All otherstrategies were eitherdominated orextendedlydominated. For adultsaged 75 y, theCCHRhm strategy hadan ICER of £10,397/QALY whencompared with theScandinavianstrategy. All otherstrategies weredominated


VALUE

INH

EALTH

18

(2015)721–734

724

Ruan et al.[33]

Model-basedcost-minimizationanalysis

Adults with mild TBI ED. Location wasnot specified

Hospital (USdollars; pricedate notspecified)

Screening withserum S100B,followed by CThead only if theS100B is abnormal(40.1 μg/L) vs.ordering CT headat the discretionof the EDphysician basedon presentingsymptoms

2 d S-100B as a screen testwould costhospitals onaverage $281 perpatient, whereasselective CTscanning based onphysician discretionwould cost onaverage $160 perpatient

16/24

Smits et al.[21]

Single study–based cost-utility analysis(prospectivemulticentercohort study;n ¼ 3181)

Adults with mild TBI Dutch universityhospital

Societal (2006 USdollars)

Diagnostic strategiesfor mild TBI inadults (CT all,selective CTscanning usingmultiple decisionrules, no CT)

Lifetime All CT strategies werealmost equallyeffective (22.464QALYs). The CCHRwas the least costly(lifetime cost $8800[Int.$9889]) and thusthe most cost-effective

23/24

Stein et al.[22]

Model-basedcost-utilityanalysis

Adults with mild TBI Not specified Health carepayer* (2005 USdollars)

Diagnostic strategiesfor mild TBI inadults (CT all,selective CT as perCCHR, observationin ED, admission,skull x-ray)

Lifetime Use of the CCHR andthe CT All strategiesyields the greatestnumber of QALYs(28.85 QALY). TheCCHR strategy hadthe lowest expectedtotal cost ($1668[Int.$1935]),followed by the CTAll strategy ($1888[Int.$2190])

19/24

Wang et al.[24]

Single study–based cost-utility analysis(retrospectivesingle-centercohort study;n ¼ 222)

Adults withsuspected bluntcerebrovascularinjury (based onpredefinedcriteria)

Level 1 traumacenter, AnnArbor, MI

Not specified (USdollars; pricedate notspecified)

CTA vs. catheter-based digitalsubtractionangiography (DSA)

30 y The ICER for DSA is$3199/QALY

13/24

CCHR, Canadian CT Head Rule; CCHRh, Canadian CT Head Rule high risk; CCHRhm, Canadian CT Head Rule high or medium risk; CHALICE, Children’s Head injury Algorithm for the predictionof Important Clinical Events; ED, emergency department; ICER, incremental cost-effectiveness ratio; Int.$, 2012 international dollar (Geary-Khamis dollar); NHS, National Health Service; PSS,personal social services; QALY, quality-adjusted life-year; RCT, randomized controlled trial; TBI, traumatic brain injury.* The study was described as of societal perspective. However, productivity loss and indirect health care expenses were not taken into account.

VALUE

INH

EALTH

18

(2015)721–734

725

Table 2 – Economic evaluations of management strategies.


Comparators Time horizon Incremental costs andoutcomes

Qualityscore

Cotton et al.[17]

Model-based cost-utility analysis

Patients with TBI; agewas not specified

Level I traumacenter. Locationwas notspecified


Phenytoin vs.levetiracetam forposttraumaticseizure prophylaxis

Not specified Levetiracetam strategy (23.2QALY, $480) wasdominated by thephenytoin strategy (23.6QALY, $37.5)

13/24

Faul et al. [30] Model-based cost-effectivenessanalysis

Patients with severeTBI (age Z12 y)

Nonfederal short-stay hospital inthe UnitedStates

Not specified (2002US dollars)

33% compliance ratewith BTF guidelinesfor the treatment ofsevere TBI (currentstate) vs. increasingthe compliance rateto 80% (widespreadadoption of BTFguidelines)

Not specified 3607 lives and $4.08 billion(Int.$5.13 billion) would besaved every year withwidespread adoption ofBTF guidelines

15/24

Ho et al. [19] Single study–basedcost-utility analysis(retrospectivesingle-center caseseries; n ¼ 168)

Adult patients withTBI who needdecompressivecraniectomy (DC) asa life-savingprocedure includingthose withbilaterallyuncreative pupils

Two neurotraumacenters inWesternAustralia

Societal (2011–2012 US dollars)

DC (unilateral orbilateral) as a life-saving procedurevs. withholdingsurgery andwithdrawal of lifesupport

Lifetime For patients with 480% riskof an unfavorableoutcome, the ICER for DCwas $682,000/QALY (Int.$698,368/QALY). Forpatients with r80%predicted risk of anunfavorable outcome, theICER was $140,000/QALY(Int.$143,360/QALY)

16/24

Ibrahim et al.[34]

Model-based cost-consequencesanalysis

Adults with severe TBI Theneurointensivecare unit atHospitalUniversiti Sains,Malaysia

Not specified (2002US dollars)

Intracranial pressuremonitoring alone(BNM) vs.multimodalitymonitoring

6 mo Mean total provider cost was$8023.78 (Int.$10,090) and$9022.63 (Int.$11,336) inthe BNM and M3 groups,respectively (P ¼ 0.49).There was a statisticallysignificant difference inthe mean Barthel Indexscore (P ¼ 0.031) in favorof the M3 strategy.Differences in outcomepersisted after adjustedanalysis

15/24

(M3) strategy(transcranialDopplerultrasonography,jugular venousoximetery, cerebraloxygen monitoring)

Joseph et al.[35]

Single study–basedcost-consequencesanalysis(retrospectivesingle-center cohortstudy; n ¼ 85)

Adults with TBI andinduced or acquiredcoagulopathy

Level 1 traumacenter, Arizona


Prothrombin complexconcentrate (PCC)vs. recombinantfactor VII (rFVIIa)

Not specified The overall mortality ratewas lower in the patientgroup that received PCC(47%) compared with themortality rate in thepatient groups receivingfactor VIIa (67%; P ¼ 0.03).Mean hospital cost wassignificantly higher in therFVIIa-treated group thanin patients who receivedPCC ($62,491 vs. $31,694;P o 0.01)

9/24


VALUE

INH

EALTH

18

(2015)721–734

726

Pandor et al.[20]

Model-based cost-utility analysis

Adults with mild TBI EDs in the UnitedKingdom

The NHS and thePSS in Englandand Wales(pound sterling;price date notspecified)

Admission strategiesfor adults with mildTBI

Lifetime 1. Hospital admission forpatients with a normal CTscan compared withdischarge home with aresponsible adult has an ICERof £39 million/QALY

2. Hospital admission forpatients with a normal CTscan compared withdischarge home without aresponsible adult has anICER of £2.5 million/QALY

3. The admission strategy forthose with anonneurosurgical lesioncosts approximately £340less and gains 0.004 QALYscompared with dischargehome and thereforedominates discharge home

23/24

Pieracci et al.[32]

Model-based cost-minimizationanalysis


Not specified Acute careinstitution (2011US dollars)

Phenytoin vs.levetiracetam forposttraumaticseizure prophylaxis

Acutehospitalizationperiod

Phenytoin was more costsaving with a mean costper patient of $151.24 (Int.$155) when comparedwith $411.85 (Int.$422) forlevetiracetam

14/24

Schoenberget al. [37]

Model-based cost-consequencesanalysis

Patients whosustained moderateto severe TBI atleast 1 y ago; agewas not specified

RehabilitationHospital,Oklahoma


Internet-basedcognitiverehabilitationteletherapyprogram(intervention) vs.face-to-faceoutpatient speech–language therapy(control)

Not specified There was no differencebetween the two groups inreturn to independentliving, work, and school.The mean total cost perpatient in the interventiongroup was notsignificantly differentfrom that in the controlgroup ($3672 vs. $2610;P ¼ 0.23)

10/24

Stein et al. [23] Model-based cost-utility analysis

Adults with mild TBIwho have anintracranialabnormality onCT scan thatdoes not needsurgicalintervention

Not specified Health care payer*

(US dollars;price date notspecified)

Routine follow-up CTscanning vs.repeating the CTscan only in case ofclinicaldeterioration

Lifetime The ICER for routine CTstrategy is $12,670/QALY

18/24

Stein et al.[38]

Single study–basedcost-consequencesanalysis(retrospectivesingle-center cohortstudy;n ¼ 179)

Patients with TBI andinduced or acquiredcoagulopathy; agewas not specified

Level I traumacenter,Maryland


Recombinant factorVIIa (rFVIIa) vs.conventionaltherapy forcoagulopathy (freshfrozen plasma andvitamin K) that doesnot include rFV11a.Dosing regimen wasnot specified

Not specified Mortality, ICU LOS, andhospital LOS were thesame in both groups, butamong survivors,discharge GCS score andRLAS were higher inconventionally treatedpatients. Total hospitalcost was similar for thetwo groups

13/24


VALUE

INH

EALTH

18

(2015)721–734

727

Table

2–co

ntinued

Stu

dy

Des

ign

Population

Setting

Perspec

tive

(curren

cy)

Com

parators

Tim

ehorizon

Increm

entalco

stsan

dou

tcom

esQuality

score

Whitmore

etal.[25]

Model-bas

edco

st-

utility

analys

isAdultswithse

vere

TBI

Notsp

ecified

Societal

(201

1US

dollars)

“Agg

ressive

trea

tmen

t”defi

ned

asfollowingth

eBTF

guidelines

inat

leas

t50

%ofth

eca

ses;

“Routine

care”defi

ned

asfollowingth

eBTF

guidelines

infewer

than

50%

ofth

eca

ses;

and“C

om

fort

care,”

which

invo

lves

asingle

day

ofag

gres

sive

care

followed

byca

reon

am

edical-

surgical

floor

Lifetim

e“A

ggress

ivetrea

tmen

t”dom

inates

oth

erstrategies

byvirtueof

gainingm

ore

QALY

s(m

ean

11.7)with

lower

costs(m

ean

$1,263

,899

[Int.$1

,294

,233

])

22/24

BTF,

Brain

TraumaFo

undation;ED

,em

erge

ncy

dep

artm

ent;GCS,

Glasg

owCom

aSc

ale;

ICER

,increm

entalco

st-effec

tive

nes

sratio;

ICU,intensive

care

unit;Int.$,

2012

intern

ational

dollar(Gea

ry-

Kham

isdollar);L

OS,

lengthof

stay

;NHS,

National

Hea

lthSe

rvice;

PSS,

perso

nal

social

services

;QALY

,quality-ad

justed

life-yea

r;RLA

S,Ran

choLo

sAmigos

Cog

nitiveSc

ale;

TBI,trau

matic

braininjury.

*Auth

ors

stated

that

thestudywas

undertake

nfrom

theso

cietal

persp

ective

.Howev

er,pro

ductivityloss

andindirec

thea

lth

care

expen

seswerenottake

ninto

acco

unt.

V A L U E I N H E A L T H 1 8 ( 2 0 1 5 ) 7 2 1 – 7 3 4728

examined severe TBI management strategies [19,25,30,34]. Twostudies compared management options for coagulopathy inpatients with TBI [35,38], and two studies compared alternativeregimens for posttraumatic seizure prophylaxis [17,32].

Prehospital and in-hospital alternative structures of care forTBI were examined by four economic evaluations [28,29,36,39].Only one study [37] examined the economic impact of differentrehabilitation pathways for patients with moderate to severe TBI.

Thirteen economic evaluations were single study–based. Spe-cifically, the input values were derived from a randomizedcontrolled trial (RCT) in two studies [39,40], a subgroup analysisof an RCT in one study [28], a prospective cohort in two studies[18,21], and a retrospective single-center cohort/case series ineight studies [19,24,29,34–38]. The other studies (n ¼ 11) weremodel-based, in which the input values were derived frommultiple observational studies (none of the source studies wasan RCT) [17,20–23,25–27,30–33].

Two studies were published during the period 2000 to 2005[31,36], whereas the remaining 22 were published from 2006 toAugust 2013 [17–30,32–35,37–40].

Quality Assessment of Included Studies

The quality scores of the studies ranged between 9 and 23 out of atotal of 24 points (see Tables 1–3 and Appendix in SupplementalMaterial). Most of the studies were of low to moderate quality(n ¼ 18; scores between 9 and 19). Only six studies were consideredto be of high quality (score 4 20/24 [480%]) [18,20,21,25,26,29].

Common methodological shortcomings included failure to spec-ify a study perspective (n¼ 12) [17,22–24,30,31,34–39], clinical setting/location (n ¼ 8) [17,22,23,25–27,32,33], time horizon of the analysis(n ¼ 8) [17,27,30,35–39], price date (n ¼ 8) [17,20,23,24,33,35,37,38],and/or the funding source (n ¼ 12) [17,19,22–24,27,29,30,32,35–37].Only seven studies discounted future outcomes/costs [18,20–23,25,29], whereas none of the studies that used discrete-time,health state transition models (n ¼ 14) addressed the issue of half-cycle correction during analysis [17,20–27,29–33]. Twelve studies didnot provide a statement on potential conflict of interest [17,19,22,23,27,28,30,31,34,36,38,39]. Only eight studies considered a lifetimehorizon [18–23,25,29].

Among cost-utility studies (n ¼ 9), four studies [19,22,23,25]derived their utility scores for the Glasgow Outcome Scale (GOS)categories from Aoki et al. [41]. Two studies [20,21] used utilityscores provided by Smits et al. [21] cohort. Harrison et al. [18]prospectively measured the quality of life of patients with TBI at6 months postinjury using the EuroQol five-dimensional ques-tionnaire (the questionnaire was completed by 49% and 48% ofthe survivors in the intensive care unit cohort and the transfercohort, respectively), and the results were converted to utilityscores to include in their cost-utility analysis. Two studiesmentioned that their utility estimates were obtained from theliterature, but the source was not specified [17,24]. None of theincluded studies considered utility scores for short-term healthstates (i.e., health states during the first 6 months postinjury, e.g.,utility score for acute hospitalization period or disutility of asurgical intervention).

Contrary to findings from previous studies [42,43], most of thecost-utility studies (n ¼ 7) assumed that functional outcomes werefixed after the first 6 to 18 months following TBI [17,19,21–23,25].

Results of High-Quality Studies

Economic evaluations of diagnostic strategiesOne study estimated the cost-effectiveness of a policy of com-puted tomography (CT) scanning of the head for inflicted TBI(i.e., brain damage due to violence caused by another person)in selected asymptomatic infants seen in the emergency

Table 3 – Economic evaluations of structure of care.



Qualityscore

Harrison et al.[18]


Adults with TBIadmitted to ICU

Adult ICUs in theUnitedKingdom

The NHS andthe PSS inEngland andWales (2010–2011 poundsterling)

Management in adedicatedneurocritical careunit comparedwith a combinedneuro/generalcritical care unitfor adult patientswith TBIpresenting at aneurosciencecenter

Lifetime Dedicated neurocriticalcare units had highermean lifetime QALYs(9.99 vs. 9.49) atsmall additionalmean costs (£34,909vs. 31,007), with anICER ofapproximately£14,000/QALY (Int.$21,509/QALY)

22/24

Harrison et al.[18]


Adults with TBI whoinitially present ata nonneurosciencecenter and do notrequire surgery forevacuation of amass lesion

Neurosciencecenters in theUnitedKingdom

The NHS andthe PSS inEngland andWales (2010–2011 poundsterling)

“Early” (within 18 hof hospitalpresentation)transfer to aneurosciencecenter comparedwith “no or late”(after 24 h)transfer

Lifetime The “Early” transfergroup has highermean lifetime QALYs(11.55 vs. 7.19) at anadditional mean cost(£36,422 vs. 16,105)with an ICER of£11,000/QALY (Int.$16,900/QALY)

22/24

Macnab et al.[36]

Single study–based cost-consequencesanalysis(retrospectivesingle-centercohort study;n ¼ 43)

Children with TBItransported by airambulance to atertiary carefacility fromanother facility

Provincial airambulanceservice,Vancouver, BC,Canada

Not specified(1988 USdollars)

Transportation bytrained pediatrictransportpersonnel vs.transportation bypersonnel (fromthe referringhospital) withoutspecializedtransport trainingas escorts

Not specified There were 13 adverseevents in six patients(55%) in theuntrained transportgroup and 5 adverseevents in fourpatients (12%) in thetrained transportgroup (P o 0.05).The calculatedadditional cost ofcare resulting fromthe adverse eventsoccurring duringtransport byuntrained team is$12,300 (Int.$21,277)per patienttransported

9/24


VALUE

INH

EALTH

18

(2015)721–734

729

Table 3 – continued



Qualityscore

Saltzherret al. [28]

Single study–based cost-effectivenessanalysis(subgroupanalysis of anRCT; n ¼ 120)

Adults with severeTBI

Level 1 traumacenters,Amsterdam,TheNetherlands

Health careprovider(2007 euros)

Mobile 4-slice CTscanner in thetrauma room(intervention) vs.64-slice CTscanner in theradiologydepartment(control)

1 y No significantdifference in meannoninstitutionalizeddays alive or costs

15/24

Taylor et al.[29]

Single study–based cost-effectivenessanalysis(retrospectivesingle-centercohort study; n¼ 1067)

Adult with TBI(defined as headAIS score of Z3)

Major traumacenter,Australia

Health careprovider(2010Australiandollars)

Direct transfer fromthe scene to LevelI trauma centervia Physician-staffed HelicopterEmergencyMedical Services(HEMS) vs. groundambulance orindirect transport(via a referralhospital)

Lifetime The mean lifeexpectancy for theHEMS strategy washigher (15.3 vs. 14.6y). The incrementalcost of the HEMSstrategy is Au$49,159 (Int.$36,296)per life saved

20/24

Wong et al.[39]

Single study–based cost-consequencesanalysis(single-centerRCT; n ¼ 179)


Conductedbetween adistrict generalhospital and atertiaryneurosurgicalcenter, HongKong, China

Not specified(2001 HongKong dollars)

Three modes ofneurosurgicalconsultation:

Lifetime No significantdifference betweenthe differentconsultationmethods in GOS at 6mo. Cost per patientwas similar for allconsultationmethods

11/24

1. Telephoneconsultation

2. Teleradiology3. Video consultation

AIS, Abbreviated Injury Scale; GOS, Glasgow Outcome Scale; ICER, incremental cost-effectiveness ratio; ICU, intensive care unit; Int.$, 2012 international dollar (Geary-Khamis dollar); NHS,National Health Service; PSS, personal social services; QALY, quality-adjusted life-year; RCT, randomized controlled trial; TBI, traumatic brain injury.

VALUE

INH

EALTH

18

(2015)721–734

730

Fig. 1 – Flow chart of study selection process.

V A L U E I N H E A L T H 1 8 ( 2 0 1 5 ) 7 2 1 – 7 3 4 731

department [26]. From a medical payer perspective, the CTstrategy was less costly in the unexplained scalp-bruising sce-nario (US $3880 [Int.$4,500] saved for every severe/fatal TBIaverted), but cost US $72,744 (Int.$84,368) per case averted inthe apparent life-threatening event scenario [26]. When a societalperspective was adopted (i.e., adding the costs of child protectiveservices to the overall cost calculation), the CT scanning strategycost US $132,701 (Int.$153,907) and US $209,328 (Int.$242,779) percase averted for bruising and apparent life-threatening eventscenarios, respectively [26].

Two high-quality cost-utility studies compared diagnosticstrategies for mild TBI in two health care systems (from theUnited Kingdom National Health Service [UK NHS] perspectiveand from the Dutch societal perspective) [20,21]. Both found thatusing the widely validated Canadian CT Head Rule is likely themost cost-effective diagnostic imaging strategy in adults (Table 1)[20,21]. In children, Pandor et al. [20] found that the Children’sHead injury Algorithm for the prediction of Important ClinicalEvents (CHALICE) was the dominant strategy at all willingness-to-pay threshold values, from the UK NHS perspective. In addi-tion, the authors found that hospital admission dominated (i.e.,less costly and more effective) discharge home for adults withmild TBI and nonsurgical lesions, but the admission strategycosts £39 million/quality-adjusted life-year (QALY) for clinicallynormal patients with a normal head CT [20].

Economic evaluations of management strategiesIn a cost-utility analysis of compliance with the Brain TraumaFoundation (BTF) guidelines for the management of severe TBI,“aggressive care” (defined as a compliance rate of 450%) foradults with severe TBI was found to be the dominant strategy (i.e., less costly and more effective) from a US societal perspectivewhen compared with “routine care” (i.e., compliance with BTFguidelines o50%) and “comfort care” (i.e., 1 day of aggressive carein the intensive care unit followed by transfer to a medical-surgical floor) [25]. The findings were consistent when thepatient’s age was assumed to be 40 or 60 years (instead of thebase-case age of 20 years) [25]. For age 80 years, the incrementalcost-effectiveness ratio for aggressive versus routine care was US$88,507 (Int.$90,631)/QALY [25].

Economic evaluations of structure of careUsing data from a large prospective cohort study (n ¼ 2975),Harrison et al. [18] found that managing patients with TBI in a

dedicated neurocritical care unit compared with a general criticalcare unit resulted in higher mean QALYs at a small additionalcost, with an incremental cost-effectiveness ratio of £14,000 (Int.$21,509)/QALY from the UK NHS perspective. There was consid-erable statistical uncertainty; however, the probability that dedi-cated compared with general intensive care units are more cost-effective was approximately 60% on the cost-effectivenessacceptability curve at a willingness-to-pay threshold of £20,000/QALY [18]. In addition, the authors found that “early” transfer(within 18 hours of presentation) of patients with TBI withnonsurgical lesions to a neuroscience center may be more cost-effective compared with no or “late” transfers (after 24 hours) atan incremental cost-effectiveness ratio of £11,000 (Int.$16,900)/QALY, with a probability close to 100% on the cost-effectivenessacceptability curve at a willingness-to-pay threshold of £20,000[18]. Subgroup analyses suggested that early transfer had a verylow probability (�20%) of being cost-effective for patients olderthan 70 years and a low probability (�60%) of being cost-effectivefor patients without major extracranial injury [18]. Early transferwas relatively cost-effective for patients within different TBIseverity subgroups, particularly for severe TBI [18].

Based on a retrospective single-center cohort study (n ¼ 1067)from Australia, direct transfer of adults with significant TBI(defined as a head Abbreviated Injury Scale score of 43) fromthe scene to a Level I trauma center by Physician-staffed Heli-copter Emergency Medical Services compared with ground ambu-lance or indirect transport (via a referral hospital) resulted in ahigher mean life expectancy (15.3 vs. 14.6) at an incremental costof Au $49,159 (Int.$36,296) per life saved [29].

Discussion

We conducted a comprehensive systematic review of the liter-ature to identify published economic analyses of diagnostic andmanagement approaches for patients with TBI. Current evidencefrom high-quality studies supports the economic attractivenessof a low medical threshold for CT scanning of asymptomaticinfants with possible inflicted TBI, the utilization of the CanadianCT Head Rule in adults and the CHALICE rule in children as thediagnostic strategies for mild TBI, the management of adults withsevere TBI according to the BTF guidelines, the management ofTBI in dedicated neurocritical care units instead of combinedcritical care units, and the early transfer of patients with TBI withnonsurgical lesions to neuroscience centers.

Although these findings were demonstrated in high-qualityeconomic evaluations, certain methodological deficienciesremained in each study. In estimating the cost-effectiveness ofCT scanning for inflicted TBI from a societal perspective, theshort time horizon of cost calculation (52 weeks) may have led tounderestimating the long-term impact of missed TBI in infants,and therefore may have biased the results against the CTscanning strategy [26]. In studies of alternative diagnostic strat-egies for mild TBI, the CHALICE rule has limited validationoutside its derivation cohort, and the results were based onsensitivity and specificity estimates obtained from the derivationcohort [20]. In addition, the generalizability of these studies’findings to patients with mild TBI who have induced or acquiredcoagulopathy is uncertain because the source of outcome datawas studies that excluded these patients [20]. In the economicevaluation of different rates of compliance with BTF guidelines[25], the probability estimates of long-term outcomes for eachstrategy were based on pooled data from multiple observationalstudies conducted between 1970 and 2009 [44]. This may havebiased the results in favor of the “aggressive strategy” (i.e., highcompliance rate) because its outcome data were derived frommore recent studies (as opposed to data from older studies for the

V A L U E I N H E A L T H 1 8 ( 2 0 1 5 ) 7 2 1 – 7 3 4732

“routine care” and “comfort care” strategies), which reflect morecontemporary care and perhaps better overall outcomes inde-pendent of compliance with the BTF guidelines. Although theauthors of economic evaluations of alternative structures of TBIcare adjusted for differences in case-mix in the original cohorts,there is still a high risk of unmeasured confounding on eachstudy’s findings [18,29]. Moreover, these studies did not accountfor the implementation and maintenance costs of alternativestructures of care (e.g., cost of establishing and maintaining aPhysician-staffed Helicopter Emergency Medical Services Pro-gram), which may have overestimated the economic attractive-ness of these alternative settings of care [18,29]. Finally, becausethese studies were conducted in specific health care settings,their results may not be generalizable to other health caresystems with different economic conditions.

We used an extensive and well-defined search strategy thatencompassed multiple electronic databases and the gray liter-ature to identify relevant economic evaluations. Furthermore,two reviewers independently applied the inclusion criteria andassessed the quality of the included studies using a structuredappraisal tool developed by an international and multidiscipli-nary team of experts in economic evaluation methodology[15,16]. In addition to the structured quality assessment ofcontemporary economic evaluations using a generic instrument,we highlighted a number of methodological aspects specific tothe population with TBI, which can help to improve future studydesign.

The number of economic evaluations in the diagnosis andmanagement of TBI has increased over the past decade. Most ofthe studies, however, are still of low to moderate quality. Only 3of 24 studies derived clinical effectiveness estimates from RCTs[28,39,40], with the remainder based on retrospective observatio-nal data. Although the functional, cognitive, social, and economiceffects of TBI are lifelong, most of these economic evaluations didnot use a lifetime horizon when comparing costs and conse-quences of alternative strategies. Among the studies that con-sidered the lifetime horizon, most assumed that functionaloutcomes were fixed 6 to 18 months following TBI. This assump-tion runs contrary to findings from previous follow-up studies,which suggest that functional outcomes of patients with TBIevolve considerably beyond the first few years postinjury [42].

Because of the substantial and lifelong impact of TBI, eco-nomic evaluations of TBI diagnostic or management strategiesshould incorporate a lifetime horizon when evaluating costs andconsequences. Although previous studies have attempted tocharacterize the long-term natural history of TBI, these studieswere of limited quality because of the small sample sizes, highdropout rates, short follow-up times, and/or the combination ofpatients with heterogeneous injuries [42,43,45,46]. Mild, moder-ate, and severe TBI differ greatly in expected recovery andoutcomes, and may present as different diseases [47]. Therefore,the duration, natural history, and life course manifestations ofmild, moderate, and severe TBI should be characterized in largelongitudinal studies that evaluate each TBI category separately.Better characterization of TBI natural history can help avoidmaking strong assumptions about the trajectory of recovery ofpatients with TBI in future economic evaluations.

Previous cost-utility studies were limited by the scarce num-ber of utility-measurement studies that examined the long-termhealth states specific to the TBI population [18,21]. Among cost-utility analyses (n ¼ 9), four studies [19,22,23,25] derived theirutility scores for the GOS categories from the study by Aoki et al.[41]. The latter investigators used the standard gamble techniqueto estimate utility scores for long-term health states amongpatients with subarachnoid hemorrhage, using a conveniencesample of health care workers (n ¼ 140) at a hospital in Japan [41].Two studies [20,21] used GOS utility scores provided by Smits

et al. [21] cohort, in which long-term GOS outcomes and theirutility scores (using the EuroQol five-dimensional questionnaire)were measured in a subset of patients with mild TBI (n ¼ 87;which represents 28% of the original sample) from the CT in headinjury patients study [48,49]. None of the included studiesconsidered utility scores for short-term health states (i.e., healthstates during the first 6 months postinjury). Short-term healthstates should be accounted for when calculating quality-adjustedlife expectancy in the TBI population. Although accounting forsuch temporary states may not substantially affect the overallquality-adjusted life expectancy of patients with long life expect-ancy (e.g., young patients who have good functional recovery),the quality of life during these health states may significantlyaffect the quality-adjusted life expectancy calculation for olderpatients and for patients with poor functional recovery becausethese short-term states likely constitute a major portion of therelatively short life expectancy of these patients. Collectively,these two patient groups represent a large proportion of thepopulation with TBI. Furthermore, discounting future costs andconsequences tends to disproportionately weight acute andsubacute health states relative to long-term states, which alsohighlights the importance of accounting for short-term healthstates. It is plausible that certain long-term health states follow-ing TBI (e.g., persistent vegetative state) could be perceived asworse than death by patients, their families, caregivers, and/orsociety. Future utility-measurement studies should consider thispossibility and allow for negative utility scores (i.e., worse thanutility score for death) when measuring preferences for healthstates among patients with TBI.

To refine cost calculations in future studies, certain costelements require special consideration. The cost of care atnursing homes and assisted living represents a major cost driverin TBI economic evaluations, particularly for severely injuredpatients [25]. Future studies are indicated to highlight thevariability in this cost component between jurisdictions, varia-bility in the proportion of patients with TBI who require skillednursing care over time, and strategies to improve the efficiency oflong-term care programs for TBI survivors. In addition, familymembers of TBI survivors suffer a great deal of personal healthand financial difficulties [50,51]. To optimize economic evalua-tions that consider costs from a societal perspective, furtherinvestigations are needed to quantify the contribution of costsof medical care for families of TBI survivors and their lostproductivity to the overall economic burden of TBI on society.

To help prioritize research funding, future economic evalua-tions in the population with TBI may consider conductingexpected value of perfect information (EVPI) and partial EVPIanalyses to quantify the differential cost of multiple sources ofuncertainty in the evidence base [52]. EVPI is the expected benefitin monetary terms per patient of a hypothetical study that wouldeliminate all parameter uncertainty (or uncertainty in certainparameters of interest in the case of partial EVPI). More researchis justified if the expected benefit to future patients (estimated asthe product of EVPI per patient and the population that isexpected to benefit from future research) exceeds the cost ofproposed research [53]. Such analyses may help to focusfuture TBI research funding on uncertainty areas with thehighest value.

Lu et al. [54] recently reviewed economic evaluations in TBI.Our study, however, involved a more comprehensive search ofthe literature. In contrast to Lu et al., two reviewers conducteda more rigorous and detailed quality assessment of includedstudies using Consolidated Health Economic EvaluationReporting Standards guidelines. In addition, we provide a sum-mary of results from high-quality economic evaluations andhighlight methodological aspects specific to the populationwith TBI.

V A L U E I N H E A L T H 1 8 ( 2 0 1 5 ) 7 2 1 – 7 3 4 733

Conclusions

Threshold-guided CT scanning, adherence to BTF guidelines, andcare for patients in specialized settings appear to be economicallyattractive strategies among patients with TBI. Most of theeconomic evaluations to date, however, are of modest quality.To better inform future economic evaluations, further methodo-logically rigorous studies are needed to characterize the naturalhistory, quality of life, and utility estimates for short-term andlong-term health states among patients with TBI.

Acknowledgment

We thank Elizabeth Uleryk for her assistance in designing andperforming the literature search.

Source of financial support: No competing financial interestexists. This work was supported in part by funds from a CanadaResearch Chair Program in Systems of Trauma Care (to A.B.N.). R.A.F. was supported by a Heart and Stroke Foundation Clinician-Scientist Award. D.C.S. was supported by a New InvestigatorAward from the Canadian Institutes of Health Research and aFellowship in Translational Health Research from Physicians’Services Incorporated Foundation.

Supplemental Materials

Supplemental material accompanying this article can be found inthe online version as a hyperlink at http://dx.doi.org/10.1016/j.jval.2015.04.012 or, if a hard copy of article, at www.valueinhealthjournal.com/issues (select volume, issue, and article).

R E F E R E N C E S

[1] Maas AI, Stocchetti N, Bullock R. Moderate and severe traumatic braininjury in adults. Lancet Neurol 2008;7:728–41.

[2] Hyder AA, Wunderlich CA, Puvanachandra P, et al. The impact oftraumatic brain injuries: a global perspective. NeuroRehabilitation2007;22:341–53.

[3] Faul M, Xu L, Wald M, Coronado V. Traumatic Brain Injury in the UnitedStates: Emergency Department Visits, Hospitalizations, and Deaths.Atlanta, GA: Centers for Disease Control and Prevention, NationalCenter for Injury Prevention and Control, 2010.

[4] Langlois JA, Rutland-Brown W, Wald MM. The epidemiology andimpact of traumatic brain injury: a brief overview. J Head TraumaRehabil 2006;21:375–8.

[5] Rosenfeld JV, Maas AI, Bragge P, et al. Early management of severetraumatic brain injury. Lancet 2012;380:1088–98.

[6] Coronado VG, McGuire LC, Faul M, et al. Traumatic brain injuryepidemiology and public health issues. In: Zasler ND, Katz DI, Zafonte RD,et al., eds., Brain Injury Medicine: Principles and Practice (2nd ed.).New York: Demos Medical Publishing, 2012.

[7] Park E, Bell JD, Baker AJ. Traumatic brain injury: can the consequencesbe stopped? CMAJ 2008;178:1163–70.

[8] Stein SC, Georgoff P, Meghan S, et al. 150 years of treating severetraumatic brain injury: a systematic review of progress in mortality.J Neurotrauma 2010;27:1343–53.

[9] Drummond MF, Sculpher MJ, Torrance G, et al. Methods for theEconomic Evaluation of Health Care Programmes (3rd ed). New York:Oxford University Press, 2005.

[10] Chen A, Bushmeneva K, Zagorski B, et al. Direct cost associated withacquired brain injury in Ontario. BMC Neurol 2012;12:76.

[11] McKinlay RJ, Wilczynski NL, Haynes RB. Optimal search strategies fordetecting cost and economic studies in EMBASE. BMC. Health Serv Res2006;6:67.

[12] Tuominen R, Joelsson P, Tenovuo O. Treatment costs and productivitylosses caused by traumatic brain injuries. Brain Inj 2012;26:1697–701.

[13] Prang K-H, Ruseckaite R, Collie A. Healthcare and disability serviceutilization in the 5-year period following transport-related traumaticbrain injury. Brain Inj 2012;26:1611–20.

[14] United Nations Department of Economic and Social Development,Statistical Division. Handbook of the International ComparisonProgramme, Studies in Methods, Series F, No. 62. New York: UnitedNations Publication, 1992.

[15] Husereau D, Drummond M, Petrou S, et al. Consolidated HealthEconomic Evaluation Reporting Standards (CHEERS) statement. BMJ2013;346:f1049.

[16] Husereau D, Drummond M, Petrou S, et al. Consolidated HealthEconomic Evaluation Reporting Standards (CHEERS)–explanation andelaboration: a report of the ISPOR Health Economic EvaluationPublication Guidelines Good Reporting Practices Task Force. ValueHealth 2013;16:231–50.

[17] Cotton BA, Kao LS, Kozar R, Holcomb JB. Cost-utility analysis oflevetiracetam and phenytoin for posttraumatic seizure prophylaxis.J Trauma 2011;71:375–9.

[18] Harrison DA, Prabhu G, Grieve R, et al. Risk Adjustment In Neurocriticalcare (RAIN)—prospective validation of risk prediction models for adultpatients with acute traumatic brain injury to use to evaluate theoptimum location and comparative costs of neurocritical care: a cohortstudy. Health Technol Asses 2013;17:vii–viii, 1–350.

[19] Ho KM, Honeybul S, Lind CRP, et al. Cost-effectiveness ofdecompressive craniectomy as a lifesaving rescue procedure forpatients with severe traumatic brain injury. J Trauma 2011;71:1637–44.

[20] Pandor A, Goodacre S, Harnan S, et al. Diagnostic managementstrategies for adults and children with minor head injury: a systematicreview and an economic evaluation. Health Technol Asses2011;15:1–202.

[21] Smits M, Dippel DWJ, Nederkoorn PJ, et al. Minor head injury: CT-basedstrategies for management–a cost-effectiveness analysis. Radiology2010;254:532–40.

[22] Stein SC, Burnett MG, Glick HA. Indications for CT scanning in mildtraumatic brain injury: a cost-effectiveness study. J Trauma2006;61:558–66.

[23] Stein SC, Fabbri A, Servadei F. Routine serial computed tomographicscans in mild traumatic brain injury: when are they cost-effective? JTrauma 2008;65:66–72.

[24] Wang AC, Charters MA, Thawani JP, et al. Evaluating the use and utilityof noninvasive angiography in diagnosing traumatic bluntcerebrovascular injury. J Trauma Acute Care Surg 2012;72:1601–10.

[25] Whitmore RG, Thawani JP, Grady MS, et al. Is aggressive treatment oftraumatic brain injury cost-effective? J Neurosurg 2012;116:1106–13.

[26] Campbell KA, Berger RP, Ettaro L, Roberts MS. Cost-effectiveness ofhead computed tomography in infants with possible inflictedtraumatic brain injury. Pediatrics 2007;120:295–304.

[27] Kaye D, Brasel KJ, Neideen T, Weigelt JA. Screening for bluntcerebrovascular injuries is cost-effective. J Trauma 2011;70:1051–7.

[28] Saltzherr TP, Goslings JC, Bakker FC, et al. Cost-effectiveness of traumaCT in the trauma room versus the radiology department: the REACTtrial. Eur Radiol 2013;23:148–55.

[29] Taylor C, Jan S, Curtis K, et al. The cost-effectiveness of physicianstaffed Helicopter Emergency Medical Service (HEMS) transport to amajor trauma centre in NSW, Australia. Injury 2012;43:1843–9.

[30] Faul M, Wald MM, Rutland-BrownW, et al. Using a cost-benefit analysisto estimate outcomes of a clinical treatment guideline: testing theBrain Trauma Foundation guidelines for the treatment of severetraumatic brain injury. J Trauma 2007;63:1271–8.

[31] Af Geijerstam JL, Britton M, Marke LA. Mild head injury: observation orcomputed tomography? Economic aspects by literature review anddecision analysis. Emerg Med J 2004;21:54–8.

[32] Pieracci FM, Moore EE, Beauchamp K, et al. A cost-minimizationanalysis of phenytoin versus levetiracetam for early seizurepharmacoprophylaxis after traumatic brain injury. J Trauma AcuteCare Surg 2012;72:276–81.

[33] Ruan S, Noyes K, Bazarian JJ. The economic impact of S-100B as apre-head CT screening test on emergency department managementof adult patients with mild traumatic brain injury. J Neurotrauma2009;26:1655–64.

[34] Ibrahim MI, Abdullah M, Naing L, et al. Cost effectiveness analysis ofusing multiple neuromodalities in treating severe traumatic braininjury in a developing country like Malaysia. Asian J Surg 2007;30:261–6.

[35] Joseph B, Hadjizacharia P, Aziz H, et al. Prothrombin complexconcentrate: an effective therapy in reversing the coagulopathy oftraumatic brain injury. J Trauma Acute Care Surg 2013;74:248–53.

[36] Macnab AJ, Wensley DF, Sun C. Cost-benefit of trained transportteams: estimates for head-injured children. Prehosp Emerg Care2001;5:1–5.

[37] Schoenberg MR, Ruwe WD, Dawson K, et al. Comparison of functionaloutcomes and treatment cost between a computer-based cognitiverehabilitation teletherapy program and a face-to-face rehabilitationprogram. Prof Psychol-Res Pr 2008;39:169–75.

[38] Stein DM, Dutton RP, Kramer ME, Scalea TM. Reversal ofcoagulopathy in critically ill patients with traumatic brain injury:

http://dx.doi.org/10.1016/j.jval.2015.04.012http://dx.doi.org/10.1016/j.jval.2015.04.012www.valueinhealthjournal.com/issueswww.valueinhealthjournal.com/issueshttp://refhub.elsevier.com/S1098-3015(15)01983-X/sbref1http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref1http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref2http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref2http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref2http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref3http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref3http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref3http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref3http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref4http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref4http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref4http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref5http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref5http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref6http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref6http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref7http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref7http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref7http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref8http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref8http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref8http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref9http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref9http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref10http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref10http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref10http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref11http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref11http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref12http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref12http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref12http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref13http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref13http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref13http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref13http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref14http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref14http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref14http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref15http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref15http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref15http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref15http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref15http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref16http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref16http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref16http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref17http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref17http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref17http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref17http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref17http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref18http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref18http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref18http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref19http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref19http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref19http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref19http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref20http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref20http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref20http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref21http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref21http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref21http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref22http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref22http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref22http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref23http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref23http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref23http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref24http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref24http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref25http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref25http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref25http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref26http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref26http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref27http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref27http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref27http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref28http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref28http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref28http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref29http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref29http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref29http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref29http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref30http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref30http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref30http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref31http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref31http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref31http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref31http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref32http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref32http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref32http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref32http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref33http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref33http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref33http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref34http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref34http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref34http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref35http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref35http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref35http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref36http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref36http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref36http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref36http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref37http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref37

V A L U E I N H E A L T H 1 8 ( 2 0 1 5 ) 7 2 1 – 7 3 4734

recombinant factor VIIa is more cost-effective than plasma.J Trauma 2009;66:63–75.

[39] Wong HT, Poon WS, Jacobs P, et al. The comparative impact of videoconsultation on emergency neurosurgical referrals. Neurosurgery2006;59:607–13; discussion 607–13.

[40] Norlund A, Marke L-A, af Geijerstam J-L, et al. Immediate computedtomography or admission for observation after mild head injury: costcomparison in randomised controlled trial. BMJ 2006;333:469.

[41] Aoki N, Kitahara T, Fukui T, et al. Management of unrupturedintracranial aneurysm in Japan: a Markovian decision analysis withutility measurements based on the Glasgow Outcome Scale. Med DecisMaking 1998;18:357–64.

[42] Whitnall L, McMillan TM, Murray GD, Teasdale GM. Disability in youngpeople and adults after head injury: 5-7 year follow up of a prospectivecohort study. J Neurol Neurosurg Psychiatry 2006;77:640–5.

[43] Hammond FM, Grattan KD, Sasser H, et al. Five years after traumaticbrain injury: a study of individual outcomes and predictors of changein function. NeuroRehabilitation 2004;19:25–35.

[44] Stein SC, Georgoff P, Meghan S, et al. Relationship of aggressivemonitoring and treatment to improved outcomes in severe traumaticbrain injury. J Neurosurg 2010;112:1105–12.

[45] Selassie AW, McCarthy ML, Ferguson PL, et al. Risk ofposthospitalization mortality among persons with traumatic braininjury, South Carolina 1999–2001. J Head Trauma Rehabil2005;20:257–69.

[46] Corrigan JD, Harrison-Felix C, Bogner J, et al. Systematic bias intraumatic brain injury outcome studies because of loss to follow-up.Arch Phys Med Rehabil 2003;84:153–60.

[47] Ragnarsson KT. Traumatic brain injury research since the 1998 NIHConsensus Conference: accomplishments and unmet goals. J HeadTrauma Rehabil 2006;21:379–87.

[48] Smits M, Dippel DW, Steyerberg EW, et al. Predicting intracranialtraumatic findings on computed tomography in patients with minorhead injury: the CHIP prediction rule. Ann Intern Med 2007;146:397–405.

[49] Smits M, Hunink MG, van Rijssel DA, et al. Outcome after complicatedminor head injury. Am J Neuroradiol 2008;29:506–13.

[50] Kreutzer JS, Gervasio AH, Camplair PS. Primary caregivers’psychological status and family functioning after traumatic braininjury. Brain Inj 1994;8:197–210.

[51] Hashimoto K, Nakamura T, Wada I, et al. How great is willingness topay for recovery from sequelae after severe traumatic brain injury inJapan? J Rehabil Med 2006;38:141–3.

[52] Sculpher M, Claxton K. Establishing the cost-effectiveness of newpharmaceuticals under conditions of uncertainty—when is theresufficient evidence? Value Health 2005;8:433–46.

[53] Koerkamp BG, Weinstein MC, Stijnen T, et al. Uncertainty and patientheterogeneity in medical decision models. Med Decis Making2010;30:194–205.

[54] Lu J, Roe C, Aas E, et al. Traumatic brain injury: methodologicalapproaches to estimate health and economic outcomes. J Neurotrauma2013;30:1925–33.

http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref37http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref37http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref38http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref38http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref38http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref39http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref39http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref39http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref40http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref40http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref40http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref40http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref41http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref41http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref41http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref42http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref42http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref42http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref43http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref43http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref43http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref44http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref44http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref44http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref44http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref45http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref45http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref45http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref46http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref46http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref46http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref47http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref47http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref48http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref48http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref48http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref49http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref49http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref49http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref50http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref50http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref50http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref51http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref51http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref51http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref52http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref52http://refhub.elsevier.com/S1098-3015(15)01983-X/sbref52

Economic Evaluations in the Diagnosis and Management of Traumatic Brain Injury: A Systematic Review and Analysis of QualityIntroductionMethodsSearch MethodsStudy SelectionType of studiesType of patientsType of comparators

Data ExtractionAssessment of Methodological Quality

ResultsResults of the SearchDescription of Included StudiesQuality Assessment of Included StudiesResults of High-Quality StudiesEconomic evaluations of diagnostic strategiesEconomic evaluations of management strategiesEconomic evaluations of structure of care

DiscussionConclusionsAcknowledgmentSupplemental MaterialsReferences

economic evaluations in the diagnosis and management of ... · adherence to brain trauma foundation...

Documents