Choosing Between Unicompartmental and Total Knee Replacement

13 Pages·2017·0.55 MB·English

by  Edward Burn

PharmacoEconomicsOpen(2017)1:241–253
https://doi.org/10.1007/s41669-017-0017-4
SYSTEMATIC REVIEW
Choosing Between Unicompartmental and Total Knee
Replacement: What Can Economic Evaluations Tell Us?
A Systematic Review
Edward Burn1 • Alexander D. Liddle1,2 • Thomas W. Hamilton1 •
Sunil Pai3 • Hemant G. Pandit1,3 • David W. Murray1,3 • Rafael Pinedo-Villanueva1,4
Publishedonline:15March2017
(cid:2)TheAuthor(s)2017.ThisarticleispublishedwithopenaccessatSpringerlink.com
Abstract Results Twelvestudiessatisfiedtheinclusioncriteria.Five
Background and objective Patients with anteromedial were within-study analyses, while another was based on a
arthritis who require a knee replacement could receive literature review. The remaining six studies were model-
either a unicompartmental knee replacement (UKR) or a basedanalyses.Allstudieswereinformedbyobservational
total knee replacement (TKR). This review has been data. While methodological approaches varied, studies
undertaken to identify economic evaluations comparing generally had either limited follow-up, did not fully
UKRandTKR,evaluatetheapproachesthatweretakenin accountforbaselinedifferencesinpatientcharacteristicsor
the studies, assess the quality of reporting of these evalu- relied on previous research that did not. The quality of
ations,andconsiderwhattheycantellusabouttherelative reportingwasgenerallyadequateacrossstudies,exceptfor
value for money of the procedures. considerations of the settings to which evaluations applied
Methods AsearchofMEDLINE,EMBASEandtheCentre and the generalisability of the results to other decision-
for Reviews and Dissemination National Health Service making contexts. In the short-term, UKR was generally
Economic Evaluation Database was undertaken in January associatedwithbetterhealthoutcomesandlowercoststhan
2016toidentifyrelevantstudies.Studycharacteristicswere TKR. Initial cost savings associated with UKR seem to
described, the quality of reporting and methods assessed persist over patients’ lifetimes even after accounting for
using the Consolidated Health Economic Evaluation higher rates of revision. For older patients, initial health
Reporting Standards (CHEERS) checklist, and study find- improvements also appear to be maintained, making UKR
ings summarised. the dominant treatment choice. However, for younger
& EdwardBurn 1 NuffieldDepartmentofOrthopaedics,Rheumatologyand
[email protected] MusculoskeletalSciences,UniversityofOxford,Botnar
ResearchCentre,WindmillRoad,OxfordOX37LD,UK
AlexanderD.Liddle
[email protected] 2 UniversityCollegeLondonInstituteofOrthopaedicsand
MusculoskeletalSciences,RoyalNationalOrthopaedic
ThomasW.Hamilton
Hospital,Stanmore,MiddlesexHA74LP,UK
[email protected]
3 NuffieldOrthopaedicCentre,OxfordUniversityHospitals
SunilPai
NHSFoundationTrust,WindmillRoad,OxfordOX37LD,
[email protected]
UK
HemantG.Pandit
4 MRCLifecourseEpidemiologyUnit,Universityof
[email protected]
Southampton,SouthamptonGeneralHospital,Tremona
DavidW.Murray Road,SouthamptonSO166YD,UK
[email protected]
RafaelPinedo-Villanueva
[email protected]
242 E.Burnetal.
patients findings for health outcomes and overall cost of implant revision than TKR [6]. Although a primary
effectiveness are mixed, with the difference in health out- UKR can be expected to be cheaper than a TKR, given a
comes depending on the lifetime risk of revision and shorter hospital stay, the costs associated with revisions
patient outcomes following revision. and any other differences in future healthcare utilisation
Conclusions UKR appears to be less costly than TKR. For could outweigh any short-term cost saving.
older patients, UKR is also expected to lead to better health An economic evaluation can provide a means of
outcomes,makingitthedominantchoice;however,foryounger informingthechoicebetweenUKRandTKRbyproviding
patients health outcomes are more uncertain. Future research a comparative analysis of the alternative courses of action
should better account forbaseline differences in patientchar- intermsofboththeircostsandhealthoutcomes[8].Foran
acteristicsandconsiderhowtherelativevalueofUKRandTKR economic evaluation of UKR and TKR to be useful, it
variesdependingonpatientandsurgicalfactors. needs to use appropriate methods so that its results are
valid [8]. Given that evaluations address a question rele-
vanttoaplaceandsetting,[9]decisionmakersalsoneedto
be able to consider whether the results apply to their
Key Points for Decision Makers
decision-making context [8].
This review has been undertaken to identify economic
Twelve economic evaluations comparing
evaluations comparing UKRs and TKRs, examine the
unicompartmental knee replacement (UKR) with
approaches taken, assess the quality of these evaluations,
total knee replacement (TKR) were identified and
and consider what they can tell us about the relative value
analysed.
for money of the procedures.
Model-based analysis best captured the different
factors relevant to the choice between UKR and
TKR. Studies were limited by either small sample 2 Methods
sizes or not accounting for baseline differences in
patient characteristics. 2.1 Eligibility Criteria
UKR appears to offer a less costly alternative to
Economic evaluations including both UKR and TKR as
TKR, and also seems to lead to better health
treatment options for primary knee arthroplasty were eli-
outcomes for older patients. Uncertainty surrounds
gible for this review. While primary UKR and TKR were
the difference in health outcomes for younger
requiredtobespecified astreatment options, studies could
patients, which depends on a patient’s lifetime risk
also include additional treatment alternatives. No restric-
of revision, and health outcomes following a
tions were imposed on the study populations.
revision.
Onlyfulleconomic evaluations,inwhichboth thecosts
andhealthoutcomesofthealternativecoursesofactionare
estimated, [8] could be included. Any type of economic
evaluation could have been undertaken, hence while costs
1 Introduction were expected to be expressed in monetary terms, health
outcomescouldbemeasuredintermsofacommonunitof
Forpatientswithend-stageosteoarthritis(OA)oftheknee, clinical effect (a cost-effectiveness analysis), a generic
total knee replacement (TKR) provides a highly effective measure of health gain (a cost-utility analysis) or in mon-
treatment strategy associated with significant improve- etary units (a cost-benefit analysis).
ments in pain, function and quality of life [1, 2]. When No restriction was placed on the date ofpublication but
compared with nonsurgical treatments, TKR has been studies were required to be written in English. The review
foundtobehighlycosteffective[3,4].However,forthose was conducted as per the Preferred Reporting Items for
patients with OA predominantly in only one compartment Systematic Reviews and Meta-Analyses (PRISMA)
of the knee, unicompartmental knee replacement (UKR) guidelines and the review protocol was registered
offers an alternative approach, where only the diseased prospectively with the PROSPERO database (registration
cartilage in an isolated part of the joint is replaced [5]. number: CRD42015026664) [10].
The choice between UKR and TKR is not clear-cut.
While UKR is associated with a significantly reduced risk 2.2 Study Selection
of postoperative complications and mortality as well as
better functional and general health outcomes after SearcheswereundertakeninJanuary2016ofMEDLINEand
6 months, [6, 7] UKR is also associated with a higher rate EMBASE, using the OVID platform, and the Centre for
ChoosingBetweenUnicompartmentalandTotalKneeReplacement 243
Reviews and Dissemination National Health Service Eco- assessed.Thewayinwhichestimatesofeffectivenesswere
nomic Evaluation Database (NHS EED), which contains derived are examined, with potential sources of bias
critical overviews of economic evaluations. The specific identified.FactorsrelevanttothechoicebetweenUKRand
searchtermsusedaredetailedinthe‘‘Appendix’’.Twoofthe TKR included in the studies are considered. In particular,
authors (EB and SP) independently screened studies for attention is paid as to whether pain, function, or quality of
inclusion based on their titles, with only those clearly not life and risk of revision were assessed when summarising
eligible being excluded, and then based on abstracts, with health outcomes, and whether the cost of the primary
any discrepancies resolved by a third author (RPV). Two procedure and revision procedures were incorporated into
authors (EB and RPV) then screened the full texts of the the estimates of the overall costs.
remaining studies against the inclusion criteria and dis- Thequalityofreportingforeachstudyisassessedusing
crepancieswereresolvedthroughconsensusdiscussion. the Consolidated Health Economic Evaluation Reporting
Standards (CHEERS) checklist, which represents a set of
2.3 Analysis of Included Studies reporting standards for health economic studies [9]. For
each study, items on the CHEERS checklist were assessed
Economic evaluations were grouped by the type of ana- ashavingbeensatisfied,partiallysatisfied,notsatisfied,or
lytical frameworks used to inform decision making. A not applicable.
within-study analysisuses a single studyas the vehicle for
economic analysis, [11] with data collected within the
study used to estimate costs and health outcomes. A liter- 3 Results
ature review study identifies and compares estimates of
healthoutcomesandcostsfromdisparateresearch,whilea 3.1 Overview of Included Studies
model-based analysis defines a set of mathematical rela-
tionships to characterise the range of possible prognoses Twelve studies were included in the review [13–24]. A
and the impact of the alternative interventions [8]. flowchart of the screening process is provided in the Ap-
The key characteristics of the included studies are pendix, and the key characteristics of the included studies
summarised. The ages of the study populations and the and the decision-making factors that they considered are
interventionsconsideredarenoted,andthetimehorizonof detailed in Tables 1 and 2, respectively. Estimates of
analysis, over which costs and health outcomes where changesincostsandQALYsassociated withtheprovision
estimated, are recorded. The measures of overall health of UKR rather than TKR are presented on a cost-effec-
outcomes are detailed. The costing perspectives are also tiveness plane (Fig. 1).
identified, with studies considering costs from a patient,
hospital, payer, health system, government or societal 3.2 Within-Study Analyses
perspective, and only those costs of relevance to the given
perspective considered in an analysis. Five publications used a within-study analysis as their
Subsequently, findings relating to UKR and TKR are framework of analysis [15, 16, 19, 23, 24]. Each of these
compared.Estimatesofthedifferenceincostsandoutcomes studieswasobservationalandcomparedthecostsandhealth
between UKR and TKR are summarised. The incremental outcomesofpatientswhoreceivedUKRorTKR.Thestudy
cost-effectivenessratio(ICER),givenbythedifferenceincost populationsweresimilaracrossstudiesandappropriatefor
divided by the difference in health outcome, is detailed for thosereceivingthealternativeprocedures.Tobeacandidate
studies thatexpress healthoutcomesinterms ofquality-ad- foreitherUKRorTKRindividualsarerequiredtohaveOA
justedlife-years(QALYs),whichprovideagenericmeasure in one compartment of the knee, to have previously failed
ofhealthcapturing bothqualityand quantity oflife. Where nonsurgical treatment, and to have symptoms that had a
onealternativeisbothcostsavingandhealthimproving,itis substantial impact on their quality of life [25]; these were
consideredtobethedominanttreatmentoption.Inaddition, generallythecharacteristicsofthepatientsdescribedinthe
the difference in costs and QALYs associated with under- studies. While four studies imposed no age restriction for
takingUKRratherthanTKR are presented ona cost-effec- patientstobeincludedinthestudy,oneincludedonlythose
tiveness plane. For this figure, costs were transformed into olderthan50 yearsofage,[24]whileanotherassessedonly
2015Eurosbyfirstinflatingcostsintheoriginalcurrencyto thoseover60 yearsofage[16].
2015 prices using Consumer Price Index (CPI) indices and Twostudiesconsideredmultipleperi-andpostoperative
thenconvertedtoEurosusingofficialexchangerates[12]. healthoutcomesforUKRandTKRupto3 yearsfollowing
The methodological approaches of the studies are dis- the procedures [16, 24]. Outcomes considered included
cussed.Thespecificationofthedecisionproblem,interms blood loss during surgery, days required for independent
of the choice of study population and treatment options, is ambulation, knee flexion, and knee-specific patient-
244 E.Burnetal.
bICER NA NA NA TKRexpectedtocost65245,60382and4860perQALYgainedfromasocietal,patientandgovernmentperspective,respectively NA UKRtobecostsavingcomparedwithTKR,withthesamehealthoutcomes TKRtocost30300and63000perQALYgainedforthoseaged45and55years.UKRexpectedtobedominant65,75and85years TKRtocost12400perQALYgained UKRtobedominantforallagesubgroups UKRtobedominant
bCosts TKRhadalowercost(amountnotspecified) UKRledtoacostsavingof3100(Euros,yearnotspecified) UKRhadalowercost(amountnotspecified) UKRledtoacostsavingof1689,1564and125(USdollars,2008)fromasocietal,patientandgovernmentperspective,respectively UKRexpectedtoleadtoacostsavingof3300(Singaporedollars,yearnotspecified) UKRledtoacostsavingof800(Euros,yearnotspecified) UKRtoleadtoacostsavingof1000,1700,5300,6100and7000(USdollars,2012)forthoseaged45,55,65,75and85years,respectively UKRtoleadtoacostsavingof124(USdollars,2012) UKRtoleadtoacostsavingof1565,2327,2883and3220(Euros,2014)for\thoseaged55,55–65,[65–75and75years UKRtoleadtoacostsavingof200(USdollars,2005)
bHealthoutcomes UKRhada20percentagepointlowersurvivalrateafter15years UKRresultedinsuperiorresults UKRhada4percentagepointlowersurvivalrateafter10years UKRledto0.026fewerQALYS UKRresultedinsuperiorresults UKRandTKRassumedtoresultinequivalentoutcomes UKRtoleadto0.032and0.027fewerQALYsforthoseaged45and55years,but0.007,0.005and0.002moreQALYsforthoseaged65,75and85years UKRtoleadto0.01fewerQALYS UKRtoleadto0.07,0.05,0.06and0.05moreQALYsforthose\aged55,55–65,65–75and[75years UKRtoleadto0.05moreQALYS
ave andment
Costingperspecti Hospital Hospital Hospital Societal,patientgovern Hospital Hospital Societal Healthsystem Payer Payer
Measureofhealthbenefit Implantsurvival Multiplecoutcomes Implantsurvival QALYs Multiplecoutcomes QALYs QALYs QALYs QALYs QALYs
Timeahorizon 15years 3years 10years 2years Sixmonths Lifetime Lifetime Lifetime Lifetime Lifetime
Agelimitsforstudypopulation,years None [60 None None [50 None Cohortsaged45,55,65,75,and85 50–60 Cohorts\aged55,55–65,65–75,[75 78
includedstudies Interventionsconsidered UKRvs.TKR UKRvs.computer-assistedTKR UKRvs.TKR UKRvs.TKR UKRvs.TKR KineSpringvs.UKRvs.TKRvs.HTOvs.conservativenonsurgicaltreatment UKRvs.TKR UKRvs.TKRvs.HTO UKRvs.TKR UKRvs.TKR
e
h e e
Table1Overviewoft aAuthorCountry Within-studyanalyses KoskinenFinlandetal.[15] ManzottiItalyetal.[16] RobertssonSwedenetal.[19] Xieetal.Singapor[23] Yangetal.Singapor[24] Literaturereview MarcacciItalyetal.[17] Decisionmodelanalyses GhomrawiUSetal.[13] KonopkaUSetal.[14] PeersmanBelgiumetal.[18] Sloveretal.US[20]
ChoosingBetweenUnicompartmentalandTotalKneeReplacement 245
Y s reported outcome measures, such as the Knee Society
L es
QA ven score. In both studies, UKRs were matched with compa-
ost277per cost-effecti rToafKbmlReosTtoiKonnRt,sha.egInbeaoasnnisedossfteupxdrye[,1o36p4]e,rUawtKihviRelesawrintherrthieteimsosatethcvehereridsttyuw,dirytah,n53g04e
bICER UKRtocgained NA cremental UmfoKautnRcdhsiUnwgKerRmetemocahatcachnhiieesdmvewbwietehttree5r0opuTrotKcvoRidmse,edsbua[2tc4rno]o.ssdBaeoltlathmilsesaotusfudtriheeess
n
i
anincreaseddollars,1998) ostsavingofpounds,year ICERcement, cs[1tou6nAd,siin2edos4et]rhc,eeoadrnn.ssdIitndufedoaryudenddcdiotoimUonnpKly,aRrtetahdtkoeiqnhcugaoavsaleittshyaoooslfopfwipltiraefilermpcaaeosrrssysotpcptehircaaottnicevedTed,wKutriRhethes.
oleadtoof5(US edtoac(Britishpecified) neerepla r(eScFe)i-v3i6ngquUeKstiRonannadirTeK[2R3,].mSecaosruerserdecuosrindgedthpereSohpoerrta-tFivoermly
bCosts UKRtcost UKRl1761nots totalk ainnddiv6idaunadl’2s4QmAoLnYthssopvoesrtotpheera2tiyveealyrswfoelrleowusiendgtsouergsteirmya.tIet
R
K wasestimatedthatTKRwouldleadtoagreaternumberof
T
ore nt, QALYs than UKR. The discrepancy with this finding
sc me compared with those of the previous studies that consid-
more KQ ace eredcondition-specificmeasuresofoutcomeislikelytobe,
02 erT repl atleastinpart,becausethisstudydidnotappeartocontrol
bHealthoutcomes UKRtoleadto0.QALYS UKRledtoabett mpartmentalknee fTcbooeKrncRsboia.dsseteOrlseinandlveyiinndgtihtfhfeceiosrmceaonpnscaatesrlseydsboiefswtwatihtnehede,nTpaKtrghiRamoi.snaeW,ryrUeitcKhperRitovhciwinesdgasustUrufedKosyuRnwcdaoennrtode-
o sideringcostsfromagovernmentandsocietalperspective,
c
ni excluding the costs of revision is likely inappropriate.
ae u
Costingperspectiv Societal Hospital UKRomy, fnoaltlTioohwneainltgwaUrothKrreRompalanaidsntiTynKgreRsgtoiusvdteeierrss10[a1sa5sne,ds1s1e95d].yteOhaenrseruisssktiundogfydardetaivdfisroniomont
ot
e controlfordifferencesinthebaselinecharacteristicsofthose
Measureofhealthbenefit QALYs TKQ hightibialost omplications raaegscseoe,icgvieainntegddeUrwKaintRhdaaynegdarreToaKtfeoRrpre[is1rka5t]io,ofwnreh[1vil9ies]i.tohBneo,othhtohfweoreuvcneodrnUttrhKoelRsleedteosfbtoie-r
Timeahorizon Lifetime 1year HTOears, surgicalc m4whaptieleerscivenantrthiaeegdoestihpgeonrii,nfiUtcKalonRwtlwye.raIsnsueosrnvtieimv,UaaltKerdRattoewhaaasfvteeexrpa1e2c00teypdeeatrocrseh,na[tva1eg9ae]
y d
Agelimitsforstudypopulation,years None None adjustedlife- pitalstay,an purTeosKveiinRosti.folAornewsvoeiafsrirasoeunsUrurvKlatit,vReUasliKsarsamRtaesoharaeefrateeslttrbhre1aot5iwguhytecteefoanomrrsfweo[ua1mrr5daa]y.nthNdbaoesnnimxereittshvilmeieslaeieodssnismn,ogtohfariaess
y- os
Table1continued aAuthorCountryInterventionsconsidered SooHooUSUKRvs.TKRetal.[21] Willis-UKUKRvs.TKROwenetal.[22] TKQQALYsTotalKneeQuestionnaire,qualitNAratio,notapplicableaImpliedifnotexplicitlystatedbUKRcomparedwithTKRcIncludedfunctionaloutcomes,lengthofh issfalowtfaihtooatsnukeuhesuutdtmeidToecnnsslyncoyhtfiddstuhimeeeadtUettosdhtoheyetnaKaiesdnm[t(tbthR2sutacweheaea6ldoytseetit]irsitto.seenetiheSsmCssBvgbtsUeiiHooaeotsmchKtfetwnEtehced.arRdoeEtwecs6erstvRote[htad)hui1s.aSsisdta5gcgnisaTio]heesrvco.henmnasihfanetseheiattTgrnhetcleiaolcKrenkyceloltloraRdoyeirwlmdspiur,tftevioetwfpfwwreerrfenaeaeenrieiraelttrtteilhethebnig-idodnetvrahitnecrseahwtpseehaleppiodneitbrianhrysnocetwraecrtoiTmiiirndeenssaeKdrg,kteilvetulRoiwoiyacfrsfg,feuloiieosotnrr[msh.ee1nnrcaWvv9eptvimfiirr]aossiuaahnriitnotllooiehgellalsynydneesstl
246 E.Burnetal.
Table2 Decision-makingfactorsconsidered
Author Pain,function,oroverall Riskofrevision Costofprimary Costofrevisions
qualityoflife procedures
Within-studyanalysis
Koskinenetal.[15] 8 4 4 4
Manzottietal.[16] 4 8 4 8
Robertssonetal.[19] 8 4 4 4
Xieetal.[23] 4 8 4 8
Yangetal.[24] 4 8 4 8
Literaturereview
Marcaccietal.[17] 4 8 4 8
Decisionmodel
Ghomrawietal.[13] 4 4 4 4
Konopkaetal.[14] 4 4 4 4
Peersmanetal.[18] 4 4 4 4
Sloveretal.[20] 4 4 4 4
SooHooetal.[21] 4 4 4 4
Willis-Owenetal.[22] 4 4 4 4
Fig.1 Cost-effectivenessplane
withstudyfindings.Onlythose
studiesthatusedQALYsasa
healthoutcomeareincluded.
Thehorizontalaxisrepresents
thedifferenceinexpected
QALYsfollowingUKRand
TKR(DQALYs=UKR
QALYs-TKRQALYs);the
verticalaxisrepresentsthe
differenceinexpectedcosts(D
Costs=UKRcost–TKRcost).
Studyauthorandagegroup
consideredareinparentheses
well reported, with the outcomes and costs described. Hayward, CA, USA), an extra-articular device designed to
However, no study discussed why the analysis used was reduce the load on the knee joint [17], compared against
sufficient to inform an economic evaluation of UKR and UKR,TKR,hightibialosteotomy(HTO)andconservative
TKR. In addition, studies did not fully describe the setting nonsurgical treatment.Thepatientpopulation towhichthe
to which the study applied, or discuss the generalisability analysis applied was not described in detail. Nonetheless,
of their findings. thecomparisonofthesealternativecoursesofactioninthis
study is likely inappropriate. For example, candidates for
3.3 Literature Review Study UKR and TKR would be expected to have received and
failed nonsurgical treatments [27]. Consequently, compar-
Onestudywasinformedbyaliteraturereviewfromwhich ing the outcomes of these treatments is unlikely to be
estimatesofhealthoutcomesandcostswere identifiedand informativeasitdoesnotreflectthechoicefacedinreality.
compared [17]. Five alternative treatments were consid- In this study, UKR and TKR were grouped together as
ered, with the KineSpring implant system (Moximed, Inc., surgical procedures. It was assumed that both procedures
ChoosingBetweenUnicompartmentalandTotalKneeReplacement 247
would lead to the same health outcomes, in terms of which do allow time to be explicitly modelled. Across all
QALYs. The effect of revisions on health outcomes was of the Markov models, as time progresses patients could
not considered. With only the costs of the primary proce- remain either unrevised or have a revision. Costs and
dures considered in this analysis, as with similar within- health outcomes, in terms of QALYs, are estimated over
study analyses, UKR was estimated to be cost saving theremaininglifetimesofpatients.Twostudiesallowedfor
compared with TKR. one revision following UKR and TKR [13, 20], one
This study mostly partially satisfied items on the allowed for up to two revisions following both procedures
CHEERschecklist(seeSect.6).Thealternativetreatments [18],andnolimitwasplacedonthenumberofrevisionsin
consideredwerewell-described; however,therationalefor another study [14]. In all cases, UKR was expected to be
comparing them was not discussed. In addition, while the cost saving [13, 14, 18, 20]. Where health outcomes for
studies included in the analysis were reported, the process patients aged 65 years and over were estimated
by which they were identified was not described in detail. [13,18,20],UKRwasalsoexpectedtoleadtobetterhealth
outcomes than TKR, making it the dominant treatment
3.4 Model-Based Analyses option.However,findingsforhealthoutcomesand,inturn,
costeffectivenessweremixedforyoungerpatients.While,
Sixstudiesuseddecision-analyticmodelsastheframework in one study, UKR was expected to lead to a gain in
for analysis [13, 14, 18, 20–22]. Five of the studies com- QALYs compared with TKR for those under 65 years of
pared UKR with TKR, [13, 18, 20–22], while one study age [18], in another study TKR was expected to lead to
also included HTO as a further treatment option [14]. better health outcomes for those between 50 and 60 years
While two studies did not specify any age restriction ofage[14],whileanotherstudyestimatedthatTKRwould
[21,22],onestudyconsideredonlythoseagedbetween50 lead to better health outcomes for patients aged 45 and
and 60 years [14], and another was based on those aged 55 years [13]. The contrasting estimates appear to be dri-
78 years [20]. The remaining two studies estimated costs venbybothdifferencesinestimatesfortheriskofrevision
and health outcomes for a number of subgroups based on and,inparticular,theexpectedeffectofrevisiononquality
age [13, 18]. All of these model-based analyses incorpo- of life. While the study that found UKR to be health
rated each of the key decision-making factors relevant to improving assumed that a revision of UKR would lead to
the choice between UKR and TKR (see Table 2). quality of life equivalent to that following a primary TKR
Twostudiesuseddecisiontrees[21,22],inwhichbranches [13], the other studies expected that quality of life fol-
represent possible future treatment pathways. Neither study lowing revision of a UKR would either be equivalent to
imposedanyagerestrictiononthestudypopulationsthatthey that following revision of aTKR [13] orbetween that ofa
considered. In one study, a decision tree was only used to primary TKR and revision of a TKR [14].
estimate costs; with the costs of revisions in the year after Each of the models required estimates for risk of revi-
surgery incorporated, UKR was found to be cost saving sion following UKR and TKR. National arthroplasty reg-
compared with TKR. Meanwhile, health outcomes were istries were used in four of the studies [13, 18, 20, 22]. In
measuredpostoperativelyusingtheTotalKneeQuestionnaire each of these it appears that the rates of revision for those
(TKQ),with20UKRsmatchedwith20TKRsonageandsex receiving UKR and TKR were compared, with no adjust-
[22]. In line with the matched within-study analyses, UKR mentmadefordifferencesinbaselinecharacteristics.Inthe
wasfoundtoleadtobetterpostoperativescoresthanTKR.The two other studies, estimates were derived by a literature
otherstudyusedadecisiontreetoestimatecostsandhealth review [14, 21].
outcomes,intermsofQALYs,overtheremaininglifetimesof Theitemsonthe CHEERSchecklistweremostlyeither
patients.Astimeisnotexplicitlymodelledindecisiontrees,it partially or fully satisfied. The modelstructures were typi-
wasassumedthatimplantfailurewouldoccurforallpatients callydescribedindetail,butthewayinwhichmodelinputs
12 years following a UKR and 15 years following a TKR were estimated was generally not fully explained. Mean-
[21].This is likelya significantoversimplification, withthe while,althoughcostswerereportedindetail,theapproaches
risk of revision continuous over time and a proportion of usedtoestimatecosts,inparticularwhenprovidedbyhos-
patientswhoarelikelytoneverrequirearevision.UKRwas pitaladministrators,werenotfullydescribed.
estimated toleadtoa greater number ofQALYs thanTKR
[21].Inaddition,withallpatientsexpectedtorequirearevi-
sion, UKR was also expected to be marginally more costly 4 Discussion
thanTKR[21].UKRwasexpectedtobecosteffective,with
theestimatedhealthgainjustifyingtheadditionalcost. A large proportion of individuals who require knee
The four remaining model-based analyses were each replacement are suitable for either TKR or UKR. Sub-
informed by state-based Markov models [13, 14, 18, 20], stantial uncertainty exists around the identification of
248 E.Burnetal.
patients for whom UKR or TKR is most appropriate, and taking into account the higher risk of revision associated
significantvariationhasbeen observed intreatmentchoice with UKR. However, for younger patients with a greater
[28, 29]. While both operations have been practiced for lifetime risk of revision, findings are mixed for health
over 30 years, controversy remainsover which is the most outcomes and the cost effectiveness of the procedures.
suitableintervention.UKRisassociatedwithlowerratesof Differences in both the estimates of revision risk and the
early complications, morbidity and mortality, and superior consequencesassociatedwithrevisionsappeartodrivethis
patient-reported outcomes, but TKR is associated with a uncertainty. In particular, assumptions around quality of
significantly lower risk of revision [6]. Economic evalua- life following revision of a UKR appear to be key.
tions provide a method of reducing any unwarranted vari- Research findings are mixed as to how revision of a UKR
ation in surgical choice by providing a systematic compares with that of a primary TKR [31–33]. Additional
consideration of both the costs and health outcomes asso- research is required to better understand patients’ lifetime
ciated with each procedure. risk of revision and the consequences of revisions.
This review identified 12 economic evaluations that Aswellasage,anumberofotherpatientcharacteristics
have compared the costs and health outcomes of TKR and have been found to be associated with differences in out-
UKR.Fivewithin-studyanalysesandonestudybasedona comes following knee replacement, such as sex, weight,
literaturereviewprovideapartialconsiderationofthecosts andseverityofsymptoms[27,34].Surgicalfactorssuchas
and health outcomes associated with the procedures, with surgeon grade, their caseload, and the number of cases
their focus generally either on perioperative and early performed by the unit per year have also been found to be
postoperative outcomes or long-term revision rates. Six associated with implant survival [34, 35]. Differences in
decision-analytic models provided a broader consideration factorssuchasthesecouldalsoinfluencetherelativemerits
ofhealthoutcomesandcosts,incorporatingeachofthekey of the procedures, but none of the studies identified here
factors for decision making. In particular, four studies that considered how the cost effectiveness of UKR and TKR
used state-based Markov models were able to consider varies based on any factor other than age.
costs and health outcomes over the remaining lifetimes of Further research is required to establish the cost effec-
patients following UKR and TKR. tiveness of UKR and TKR, and how this varies depending
The studies included in this review differed in time on patient and surgical factors. National registries, in par-
horizons of analysis, study design, outcome measures and ticular,providearichsourceofinformation,capturingreal-
costing perspectives. While methodological approaches worldoutcomeswithrelativelylong-termfollow-up,which
varied, studies generally had either limited follow-up, did areofcriticalimportanceforthese procedures.If basedon
not fully account for baseline differences in patient char- such data, studies should utilise methods for addressing
acteristics, or were informed by on previous research that potential bias, such as regression analysis or matching
did not. The quality of reporting was generally adequate estimators, to better estimate treatment effects [36]. In
acrosstheincludedstudies,exceptforconsiderationsofthe addition to the numerous sources of observational data, a
settings to which evaluations applied and the generalis- large randomised controlled trial comparing UKR and
ability of the results to other decision-making contexts. TKR is currently underway [37]. While this study will
Thismakesitdifficultfordecisionmakerstoknowwhether provide valuable insight into the procedures by ascertain-
results apply to their setting. ingmorecomprehensivelytheircostsandhealthoutcomes,
In the short-term, based on the economic evaluations data from observational studies will also be needed to
considered in this study, UKR appears to be associated better understand the long-term effects of the procedures.
with better peri- and postoperative outcomes than TKR.
Better early outcomes for UKR have also been observed 4.1 Limitations of this Review
for routine practice in the UK [6, 7], and are even more
pronounced for UKRs performed by high-usage and high- A systematic approach was taken to identify studies that
volume surgeons [30]. Moreover, as would be expected have considered the costs and health outcomes associated
givenalowerlengthofstay,UKRisalsoestimatedtooffer with UKR and TKR; however, there does remain the
an immediate cost saving compared with TKR. possibility that relevant studies were not identified and
ThisinitialcostsavingassociatedwithUKRseemstobe included in the study. While MEDLINE, EMBASE and
maintained over patients’ lifetimes, even after accounting NHSEEDweresearched,searchesofadditionaldatabases,
for higher risk ofrevision. Forolderpatients, initial health such as the Health Economic Evaluations Database
improvements also appear to be maintained, even after (HEED) may have returned more results. The search filter
ChoosingBetweenUnicompartmentalandTotalKneeReplacement 249
used was designed specifically for this study and has not ofeffectivenessadjustedforbaselinedifferencesinpatient
previouslybeenvalidated.Inaddition,whilethetermsused characteristics. Further economic evaluations are required
were chosen to find appropriate results from both data- to better understand how the relative value of the proce-
bases, searching MEDLINE and EMBASE together using dures varies depending on patient and surgical factors.
the OVID platformmay have ledtomissing studies, given
the distinct characteristics of these databases. CompliancewithEthicalStandards
The descriptions of the included studies provide an
David Murray and Hemant Pandit receive research funding in the
objective overview of the economic evaluations that have
form of grants paid to the Nuffield Department of Orthopaedics,
been undertaken of UKR and TKR and their findings but RheumatologyandMusculoskeletalSciences(NDORMS),University
theanalysisofthemethodsusedandqualityofreportingis of Oxford, and the Oxford University Hospitals National Health
Service Trust from Zimmer Biomet and Stryker, both of which are
necessarily subjective. While an established checklist was
manufacturers of orthopaedic implants. David Murray receives roy-
used to consider the quality of reporting, categorising
altiesrelatedtotheOxfordUKR,andconsultancyfeesfromZimmer
studies as satisfying, partially satisfying, or not satisfying Biomet. Hemant Pandit receives consultancy fees from Zimmer
particularitemswasnecessarilybasedonthejudgementof Biomet. None of these companies were involved in the funding or
conduct of this study. Thomas Hamilton, Rafael Pinedo-Villanueva,
the authors of this study.
Alexander Liddle, Sunil Pai and Edward Burn have no potential
As well as comparing the approaches taken, the key
conflictsofinteresttodeclare.
findings of the studies have also been compared. Any
such comparison should be treated with caution due to Data availability statement Data sharing is not applicable to this
articleasnodatasetsweregeneratedduringthestudy.
methodological limitations of the studies and the wide
range of factors that limit the generalisability of results
Author’s contributions Edward Burn, Alexander Liddle, Thomas
across economic evaluations [38]. In particular, cost Hamilton, Hemant Pandit, David Murray, and Rafael Pinedo-Vil-
estimates can vary across studies due to differences in lanuevamadesubstantialcontributionstotheconceptionanddesign
of the study. Edward Burn, Sunil Pai and Rafael Pinedo-Villanueva
costing perspectives of the studies and in health systems.
identifiedstudiesforinclusioninthereview.EdwardBurnandRafael
In addition, converting costs from one currency to
Pinedo-Villanueva drafted the manuscript, with Alexander Liddle,
another adds further uncertainty [12]. However, a com- Thomas Hamilton, Sunil Pai, Hemant Pandit and David Murray
parison of findings was still felt to be merited so as to revising it for important intellectual content. All authors read and
approvedthefinalmanuscript.
provide a broad summary of the consistency in research
findings and to provide an indication of the effect of
Open Access This article is distributed under the terms of the
differences in methodological approaches on results.
Creative Commons Attribution-NonCommercial 4.0 International
Identifying such sources of variation across studies can License (http://creativecommons.org/licenses/by-nc/4.0/), which per-
help individual decision makers determine which studies mits any noncommercial use, distribution, and reproduction in any
medium,providedyougiveappropriatecredittotheoriginalauthor(s)
best apply to their particular settings, and can guide
andthesource,providealinktotheCreativeCommonslicense,and
future research [39].
indicateifchangesweremade.
5 Conclusions Appendix
TheeconomicevaluationsofUKRandTKRthathavebeen Search Terms
undertakenvary,withdifferencesinstudypopulationsand
methods of analysis. In the short-term, UKR appears to be Search terms for Medline and Embase via Ovid SP
both health improving and cost saving compared with
1. Exp Knee Joint/
TKR. This initial cost saving associated with UKR seems
2. Knee/
topersist,evenafteraccountingforhigherratesofrevision
3. Knee.tw.
over patients’ lifetimes. For older patients, UKR can also
4. Arthroplasty, Replacement, Knee/
be expected to lead to better overall health outcomes,
5. Knee Prosthesis/
making it the dominant treatment choice. However, for
6. (Knee arthroplast$ or knee replacement or knee
younger patients, findings are mixed, with differences in
prosthes$).tw.
estimates of the risk of revision and outcomes following
7. (Uka or unicompartmental knee arthroplast$).tw.
revision leading to substantial differences in estimates of
8. Unicompartmental.tw.
overall health outcomes.
9. Economics/
Toestimateallthecostsandhealthoutcomesassociated
10. Exp ‘‘costs and cost analysis’’/
with the choice between UKR and TKR, future research
11. Exp economics, hospital/
should incorporate long-term time horizons and estimates
250 E.Burnetal.
12. Economics, Medical/ Screening Results
13. Economics, Nursing/
14. Economics, Pharmaceutical/
15. (Economic$ or cost or costs or costly or costing or
price or prices or pricing or
pharmacoeconomic$).ti,ab.
16. (Expenditure$ not energy).ti,ab.
17. Value for money.ti,ab.
18. Budget$.ti,ab.
19. 1 or 2 or 3
20. 4 or 5 or 6 or 7 or 8
21. 9or10or11or12or13or14or15or16or17or18
22. 19 and 20 and 21
NHS EED, via The Cochrane Library
1. MeSHdescriptor:[Arthroplasty,Replacement,Knee]
explode all trees
2. MeSH descriptor: [Osteoarthritis, Knee] explode all
trees
3. MeSH descriptor: [Knee Joint] explode all trees
4. Knee/
5. Knee prosthesis
6. Knee arthroplasty
7. Knee replacement
8. Uka
9. Unicompartmental
10. 1 or 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9