REVIEW Open Access
Preference-based measures of health-related quality of life in congenital mobilityimpairment: a systematic review of validityand responsivenessNathan Bray1,2*, Llinos Haf Spencer1,2 and Rhiannon Tudor Edwards1,2
Abstract
Introduction: Mobility impairment is the leading cause of disability in the UK. Individuals with congenital mobilityimpairments have unique experiences of health, quality of life and adaptation. Preference-based outcomesmeasures are often used to help inform decisions about healthcare funding and prioritisation, however theapplicability and accuracy of these measures in the context of congenital mobility impairment is unclear. Inaccurateoutcome measures could potentially affect the care provided to these patient groups. The aim of this systematicreview was to examine the performance of preference-based outcome measures for the measurement of utilityvalues in various forms of congenital mobility impairment.
Methods: Ten databases were searched, including Science Direct, CINAHL and PubMed. Screening of reference listsand hand-searching were also undertaken. Descriptive and narrative syntheses were conducted to combine andanalyse the various findings. Results were grouped by condition. Outcome measure performance indicators wereadapted from COSMIN guidance and were grouped into three broad categories: validity, responsiveness andreliability. Screening, data extraction and quality appraisal were carried out by two independent reviewers.
Results: A total of 31 studies were considered eligible for inclusion in the systematic review. The vast majority ofstudies related to either cerebral palsy, spina bifida or childhood hydrocephalus. Other relevant conditions includedmuscular dystrophy, spinal muscular atrophy and congenital clubfoot. The most commonly used preference-basedoutcome measure was the HUI3. Reporting of performance properties predominantly centred around constructvalidity, through known group analyses and assessment of convergent validity between comparable measures anddifferent types of respondents. A small number of studies assessed responsiveness, but assessment of reliability wasnot reported. Increased clinical severity appears to be associated with decreased utility outcomes in congenitalmobility impairment, particularly in terms of gross motor function in cerebral palsy and lesion level in spina bifida.However, preference-based measures exhibit limited correlation with various other condition-specific and clinicallyrelevant outcome measures.
(Continued on next page)
© The Author(s). 2020 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License,which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you giveappropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate ifchanges were made. The images or other third party material in this article are included in the article's Creative Commonslicence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commonslicence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtainpermission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to thedata made available in this article, unless otherwise stated in a credit line to the data.
* Correspondence: [email protected] of Health Sciences, Fron Heulog, Bangor University, Gwynedd LL572EF, Wales, UK2Centre for Health Economics and Medicines Evaluation, Ardudwy, BangorUniversity, Gwynedd LL57 2PZ, Wales, UK
Bray et al. Health Economics Review (2020) 10:9 https://doi.org/10.1186/s13561-020-00270-3
(Continued from previous page)
Conclusion: Preference-based measures exhibit important issues and discrepancies relating to validity andresponsiveness in the context of congenital mobility impairment, thus care must be taken when utilising thesemeasures in conditions associated with congenital mobility impairments.
Keywords: Disability, Mobility impairment, Quality of life, Health-related quality of life, Patient reported outcomes,Preference-based outcome measures, Utilities, QALYs
IntroductionMobility impairment and assistive mobility technologyMobility impairment is the leading cause of disability inthe UK, accounting for 52% of reported disabilities [1].Mobility impairments arise from a vast array of differentdisabilities, conditions, injuries and illnesses. However,they can be classified broadly as either congenital (i.e.from birth) or acquired (i.e. occurring later in life).Whether a disability is present from birth or acquiredlater on in life significantly influences individual adapta-tion. For instance, individuals with congenital disabilitiesexhibit higher degrees of life satisfaction, self-identityand self-efficacy (related to their disability) than individ-uals who have had to adapt to acquired disability [2].Adaptation to disability is influenced by self-concept anddisability identity, which in turn are related to the onsetof disability [2].Common congenital conditions which can impact mo-
bility include cerebral palsy (CP) and spina bifida (SB).CP refers to a number of conditions caused by damageto the parts of the brain which control movement, bal-ance and posture, and can be caused either by abnormalbrain development or trauma. CP is symptomized byvarying degrees of permanent movement disorder, in-cluding poor coordination, muscle stiffness/weaknessand involuntary movements. SB affects the developmentof the spine and spinal cord before birth, and can resultin leg weakness and paralysis. There are three types ofSB: myelomeningocele, meningocele and SB occulta.Both congenital and acquired mobility impairments
may necessitate the use of assistive technology to allevi-ate impairments. Assistive technology refers to a widearray of products and services which enhance function-ing, participation and promote independence for peoplewho have disabilities. The Medicines and HealthcareProducts Regulatory Agency defines assistive technologyas any device “intended to compensate for or alleviate aninjury, handicap or illness or to replace a physical func-tion” [3]. Assistive technology, such as wheelchairs, arean “essential component for inclusive sustainable devel-opment” [4], and can enhance the fundamental freedomsand equality of opportunity for people with mobility im-pairments and other disabilities. The United Nations(UN) states that access to appropriate and affordable as-sistive technology is a basic human right [5]; the UN
Convention on the Rights of Persons with Disabilitieshas been ratified by 175 Member States, who are obli-gated to ensure that affordable assistive technology isavailable to all individuals in need. However, The WorldHealth Organization (WHO) estimates that only 10% ofpeople who need assistive technology have access to it[6], and there remain persistent challenges in the equit-able provision of assistive technology, particularly in de-veloping countries. One of the keys issues is in assessingthe costs and benefits of different assistive technologies,and developing evidence-based approaches to provisionwhich make best use of limited resources to maximisethe outcomes of people with disabilities.The National Health Service (NHS) in the UK spends
almost £200million per year on wheelchairs alone [7],thus there is an imperative to ensure that assistive mo-bility technologies (AMTs) such as wheelchairs andother mobility-enhancing interventions are provided inan evidence-based manner, utilising evidence of cost-effectiveness to guide service commissioning.
Economic evaluation and quality-adjusted life yearsMethods of economic evaluation are now routinely em-bedded in the evaluation of health technologies, andused to estimate the cost of incremental benefits associ-ated with new and alternative health interventions. Cost-utility analysis, specifically estimation of cost per quality-adjusted life years (QALYs), has become the predomin-ant form of economic evaluation for new health tech-nologies in the UK, in part due to the National Institutefor Health and Care Excellence’s (NICE) advocacy forthis approach [8]. The QALY framework has become in-creasingly influential in health policy as a theoreticallyuniversal and generic approach to measuring benefits viaa single common outcome.In order to calculate QALYs, health state utility values
are needed. These values are most commonly derivedfrom preference-based measures (PBMs) of health-related quality of life (HRQoL). HRQoL is a subjectiveand multi-dimensional construct defined as the per-ceived impact of health status on quality of life, includ-ing physical, psychological and social functioning. PBMsof HRQoL are used to assess the social desirability andutility values associated with different states of health.
Bray et al. Health Economics Review (2020) 10:9 Page 2 of 38
As the descriptive systems and value sets of genericPBMs are usually derived from adult samples of the gen-eral population, a common criticism is that their gener-icity limits relevance and sensitivity in certain conditions[9]. Moreover, in health states where quality of life takesprecedent over quantity of life (e.g. chronic illness, life-limiting conditions and disability), QALYs derived fromgeneric PBMs can devalue the effectiveness of an inter-vention [10].
Use of preference-based measures in the context ofmobility impairmentThe accuracy of a QALY estimate is subject to the sensi-tivity and applicability of the measurement tool used togenerate the utility data. PBMs have been found to be in-consistent in both congenital and acquired mobility im-pairments [11–13], furthermore different PBMs producesignificantly different results for AMT users [14, 15].Previous research shows that patients with congenital
mobility impairments do not necessarily consider mobil-ity to have a major impact on their HRQoL when suit-able adaptations (such as AMT) are available [16, 17].However, general population PBM value sets heavily im-pact estimation of HRQoL when ability to walk is af-fected. As an example, using the NICE approved UKvalue set for the EuroQoL five-dimension (three levelversion) (EQ-5D-3 L), the lowest possible mobility level(‘confined to bed’) has a disutility of − 0.664, meaningthat an individual who is unable to walk but is otherwisemobile using AMT can achieve a maximum utility valueof 0.336 (0 = death; 1 = perfect health), even if they haveno other HRQoL impacts. This raises the questions as towhether existing PBMs are a valid source of utility valuesin mobility impaired populations, particularly AMTusers.The validity of PBMs can be tested by comparing re-
sults across groups of patients and by comparing genericPBMs with condition-specific measures. For instance,the EQ-5D-3 L and the Health Assessment Question-naire (HAQ; an outcome measure for rheumatoid arth-ritis) both measure health status in significantly differentways [18], suggesting that the EQ-5D-3 L is lacking con-sideration of important health impacts associated withrheumatoid arthritis. These issues are partly due to theinsensitivity of the EQ-5D-3 L ‘mobility’ dimension toaccurately assess the varied impacts of rheumatoid arth-ritis on mobility [19]. Similarly, the limited level choiceson the EQ-5D-3 L have been found to cause some indi-viduals with mobility impairments to choose levelswhich are more or less severe than their actual state,such as using ‘I am confined to bed’ to substitute beingconfined ‘to an electric wheelchair’ [20]. The updatedfive level version of the EQ-5D (EQ-5D-5 L) is unlikelyto address this issue as the five level choices still focus
on walking and do not take account of alternativemethods of mobility.Even simple generic measures of health status, such as
the single question self-reported health (SRH) scale (i.e.“in general, would you say your health is excellent, verygood, good, fair, or poor?”), exhibit only limited correl-ation with PBMs such as the Health Utilities Index(HUI) 3 and Assessment of Quality of Life (AQoL) inthe context of SB [21] and CP [22]. Likewise, for individ-uals with spinal cord injuries, the wording of the 36-Item Short Form Health Survey (SF-36) (from which theShort-Form Six-Dimension (SF-6D) PBM is calculated)must be modified in order to maintain relevance [23].Considering the potential issues of using generic PBMs
in disability, and congenital mobility impairment specif-ically, it is apparent that there are a number of import-ant considerations when using PBMs to evaluate AMTinterventions and other mobility-enhancing interven-tions for people with congenital mobility impairments.The objective of this systematic review is therefore toexamine the measurement properties of generic utility-based PBMs in various forms of congenital mobility im-pairment. All evidence reporting (or inferring) the valid-ity, reliability and/or responsiveness of PBMs inconditions associated with congenital mobility impair-ment was collated and synthesised. Comparablecondition-specific reviews of PBM performance havebeen conducted in mobility impairments such asrheumatoid arthritis [24], CP [25] and multiple sclerosis[26], however PBM performance has not been systemat-ically summarised and collated across a range of con-genital mobility impairments to date.
MethodsThis systematic review followed the University of YorkCentre for Reviews and Dissemination (CRD) principlesfor conducting searches and extracting data [27]. Inter-net reference database searching was the main strategyfor gathering evidence. Databases included: CochraneCollaboration Register and Library, Science Direct,CINAHL, ASSIA, PsychINFO, PubMed and Web of Sci-ence. Screening of reference lists, hand-searching andtargeted searching via the CRD database (which coversthe Database of Abstracts of Reviews of Effects (DARE),the NHS Economic Evaluation Database (NHS EED) andthe Health Technology Association (HTA) database)were carried out in addition to the primary databasesearches. Due to limited translation resources, only stud-ies written or translated into English or Welsh were eli-gible for inclusion. Search results were managed usingthe online bibliographic management software Refworks(for storage of titles from the systematic searches) andMendeley (for referencing purposes). The systematic
Bray et al. Health Economics Review (2020) 10:9 Page 3 of 38
review protocol was registered on PROSPERO(CRD42018088932).
Search termsFor the purpose of this review, mobility impairment wasdefined as any congenital (i.e. present from or shortly afterbirth) condition, impairment, disability or illness whichcauses significant restrictions to mobility for 12months orlonger, and which necessitates the use of AMT, surgery orrehabilitation to maintain, facilitate or substitute ambula-tion, or to reduce complications related to mobility im-pairment. Acquired mobility impairments (i.e. not presentfrom birth) and short-term injuries, such as sprains oracute muscular injuries, were not included under this def-inition of mobility impairment.Search terms included a mixture of MeSH (Medical
Subject Heading) and non-MeSH words and phrases, di-vided into two groups: ‘population’ and ‘outcomes’ (seeTable 1).In order to identify studies referring to interventions ra-
ther than patient groups (e.g. studies examining ‘wheel-chair users’ more generally), the ‘population’ search termsalso covered relevant AMTs and mobility-enhancing in-terventions. The ‘outcomes’ search terms covered relevantPBM keywords, including specific outcome measures(such as the various versions of the EQ-5D and HUI mea-sures). An NHS posture and mobility service manager wasconsulted to refine the search terms. An example of asearch string is shown in Table 2.
Study eligibilityAny study reporting the performance of PBMs ofHRQoL in patient groups with congenital mobility im-pairments was eligible for inclusion. This included stud-ies reporting proxy outcomes. There was no restrictionon study type. Studies focussing solely on non-PBMs ofHRQoL or acquired mobility impairments were ex-cluded. Studies which included patient groups with vary-ing degrees of disability/disease severity were consideredfor inclusion if the majority of patients had a congenitalmobility impairment or if the data for patients with con-genital mobility impairments was reported separately(i.e. sub-group analysis).
ScreeningTwo researchers undertook each stage of the screeningprocess. For the initial screening process, all identifiedstudies were assessed for relevance based on their titleand descriptor terms, the remaining studies were thenassessed by their abstract. All studies considered relevantafter the initial screening process were then obtained infull. Both reviewers screened each study independently.A third researcher was consulted when there was dis-agreement about the inclusion of a specific study.
Quality appraisalAll relevant studies which met the initial inclusion cri-teria were critically appraised for methodological qualityby two researchers. Quality appraisal methods wereadapted from similar systematic reviews in other clinicalareas [28–30]. Quality appraisal was not used to excludestudies, but to illustrate the overall quality of researchconducted in this topic area. Quality appraisal focussedon six key areas:
� Whether tests of statistical significance were carriedout
� Differences between interventions and/or patientgroups (i.e. sub-group analysis)
� Clinical significance and relevance of results� Reporting of missing data� Response and completion rates� Explicit reporting of inclusion/exclusion criteria
Data extractionData extraction criteria included:
� Study characteristics: study type, country, number/composition of study groups, missing data
� Demographics: number of participants, age, gender,type/severity of mobility impairment
� Measures: Generic PBMs used, condition-specificmeasures used, other clinically relevant measuresused
� Outcomes: Mean utility scores, mean utility scoresfor relevant sub-groups, statistical significancebetween groups
� Performance: Known group analyses, convergentvalidity (correlation between outcomes and/orrespondent types), responsiveness, reliability,response/completion rates
Analysis of the performance of preference-basedmeasuresPBM performance indicators were adapted from COS-MIN measurement property guidance for health-relatedpatient-reported outcomes [31].
Assessment of validityIn this context validity refers to the extent to which aPBM can be considered to measure what it has been de-signed to measure (i.e. HRQoL), and whether it does soin a systematic manner. By establishing whether a spe-cific PBM is sufficiently valid in a particular patientgroup, greater confidence can be placed in the generateddata. We focussed predominantly on construct validityin this review.Construct validity was assessed in a number of ways.
Firstly, known-group analyses were used to assess
Bray et al. Health Economics Review (2020) 10:9 Page 4 of 38
whether specific PBMs were able to detect expected dif-ferences between different patient groups (i.e. variancedue to severity of illness). Secondly, convergent validitywas determined by examining correlation between com-parable outcomes, for instance between PBMs andcondition-specific measures with comparable constructs.Finally, convergent validity was further examined bylooking at correlation between respondents types (i.e.self-reported and proxy utility outcomes). In the interest
of uniformity, the strength of correlations was defined asabsent (r < 0.20), weak (r = 0.20 to 0.35), moderate (r =0.35 to 0.50) and strong (r ≥ 0.50) [32].
Assessment of reliabilityReliability refers to the replicability of results. Reliabilityis commonly assessed by examining test-retest resultsand inter-rater reliability of PBMs in defined unchangingpatient groups.
Table 1 Search terms and phrases
Population Outcomes
Assisted mobility Mobility scooter 15D
Assistive mobility Mobility technolog* AQoL
Brain damage* Motor dis* Assessment of Quality of Life
Brain injur* Motorised scooter Child health utilities
Buggy Neurodisability Child health utility
Caliper Neurological dis* CHU9D
Cane Neuromotor dis* CHU-9D
Cerebral palsy Neuromuscular dis* EQ 5D
Club foot Orthoti* EQ-5D
Clubfoot Osteogenesis imperfecta EuroQoL
Crutch* Paraly* Health utilities
Diplegi* Paraplegi* Health-utilities
Dysmelia Physical disab* HUI
Dystroph* Physical impair* HUI2
Electric chair Physically disab* HUI3
Electric powered indoor outdoor chair Physically impaired Preference based
Electric powered indoor/outdoor chair Power chair Preference-based
Electric scooter Powered chair QALY
Electrically powered indoor outdoor chair Pushchair Quality adjusted life year
Electrically powered indoor/outdoor chair Quadriplegi* Quality-adjusted life year
Electronically powered indoor outdoor chair Rollator Quality of well-being scale
Electronically powered indoor/outdoor chair Scooter QWB-SA
*encephal* Spina bifida Short Form Six Dimension
EPIOC Spinal muscular atrophy Short From 6 Dimension
Functional disab* Talipes SF6D
Handicap* Tetraplegi* SF-6D
Hemiplegi* Walk aid
Hydrocephalus Walk-aid
Knee scooter Walker
Knee walker Walking aid
Mobility aid Walking frame
Mobility device Walking stick
Mobility dis* Walking-aid
Mobility equipment Wheelchair
Mobility impair**Indicates truncated words/phrases
Bray et al. Health Economics Review (2020) 10:9 Page 5 of 38
Assessment of responsivenessResponsiveness refers to the extent to which a measurecan identify changes in health status [28]. A responsivemeasure should be able to detect clinically significantchanges in health outcomes over time [29]. Responsive-ness was determined by examining the relationship be-tween outcomes derived from PBMs and other relevantmeasures, before and after an intervention.
Evidence synthesisDescriptive synthesis of search results was conducted[27] and presented in tabulated form. Narrative synthesiswas undertaken to develop a structured narrative of re-sults; extracted results were grouped by type of mobilityimpairment and category of PBM performance.
ResultsSee Fig. 1 for search outcomes and the screening processflowchart. Searches were conducted from March to May2018. In total 1489 study articles were identified: 1332from the bibliographic searches and 157 articles fromother sources (i.e. CRD database, screening of referencelists and hand-searching), of which 410 duplicates wereremoved. After screening of titles and abstracts, 66
Table 2 Example of keyword search string
(“Assisted mobility” Or “Assistive mobility” Or “Brain damage*” Or “Braininjur*” Or Buggy Or Caliper Or Cane Or “Cerebral palsy” Or “Club foot”Or Clubfoot Or Crutch* Or Diplegi* Or Dysmelia Or Dystroph* Or“Electric chair” Or “Electric powered indoor outdoor chair” Or “Electricpowered indoor/outdoor chair” Or “Electric scooter” Or “Electricallypowered indoor outdoor chair” Or “Electrically powered indoor/outdoorchair” Or “Electronically powered indoor outdoor chair” Or “Electronicallypowered indoor/outdoor chair” Or encephal* Or EPIOC Or “Functionaldisab*” Or Handicap* Or Hemiplegi* Or Hydrocephalus Or “Kneescooter” Or “Knee walker” Or “Mobility aid” Or “Mobility device” Or“Mobility dis*” Or “Mobility equipment” Or “Mobility impair*” Or “Mobilityscooter” Or “Mobility technolog*” Or “Motor dis*” Or “Motorised scooter”Or Neurodisability Or “Neurological dis*” Or “Neuromotor dis*” Or“Neuromuscular dis*” Or Orthoti* Or “Osteogenesis imperfect” Or Paraly*Or Paraplegi* Or “Physical disab*” Or “Physical impair*” Or “Physicallydisab*” Or “Physically impaired” Or “Power chair” Or “Powered chair” OrPushchair Or Quadriplegi* Or Rollator Or Scooter Or “Spina bifida” Or“Spinal muscular atrophy” Or Talipes Or Tetraplegi* Or “Walk aid” Or“Walk-aid” Or Walker Or “Walking aid” Or “Walking frame” Or “Walkingstick” Or “Walking-aid” Or Wheelchair) AND (15D OR AQoL OR“Assessment of Quality of Life” OR “Child health utilities” OR “Childhealth utility” OR CHU9D OR “CHU-9D” OR EQ. 5D OR “EQ-5D” OREuroQoL OR “Health utilities” OR “Health-utilities” OR HUI OR HUI2 ORHUI3 OR “Preference-based” OR “Preference based” OR QALY OR “Qualityadjusted life year” OR “Quality of well-being scale” OR “Quality-adjustedlife year” OR “QWB-SA” OR “Short Form Six Dimension” OR “Short From 6Dimension” OR SF6D OR “SF-6D”)
Fig. 1 PRISMA flowchart for search outcomes and screening process
Bray et al. Health Economics Review (2020) 10:9 Page 6 of 38
studies were identified as potentially eligible. Followingfull review of full-texts, 35 studies were excluded for avariety of reasons (see Fig. 1). In total 31 studies wereidentified as relevant and eligible for inclusion in the sys-tematic review. Quality appraisal outcomes are presentedin Table 3.
Study and patient characteristicsStudy and participant characteristics are presented inTable 4. Most of the studies (22 of 31) were cross-sectional studies [15, 21, 22, 34, 36–38, 40–45, 47–52,54, 58, 59], six were prospective cohort studies [33, 35,39, 46, 53, 57], two were case-control studies [11, 56]and one was a randomised controlled trial [55]. The se-lected studies were conducted in a range of differentcountries; most were (n = 15) from Canada [21, 22, 33,35, 36, 38–43, 46, 49, 53, 57]. The sample sizes of rele-vant study groups ranged from 13 [15] to 770 [44].Fourteen studies included data relating to CP [22, 33,
35, 36, 39, 46, 48–51, 53, 55, 57, 58], six included datarelating to SB [21, 38, 51, 52, 54, 56] and five includeddata relating to childhood hydrocephalus [40–43, 45].Data for a range of other relevant conditions were alsofound in either single studies or from extractable sub-groups, these included muscular dystrophy [34, 44],spinal muscular atrophy [47], Morquio A syndrome [37],congenital clubfoot [59] and microcephaly [51]. Threestudies focused on multiple conditions where sub-groupdata could not be examined separately [11, 15, 51].The vast majority of studies (23 of 31) included only
child/adolescent participants [11, 15, 33, 35, 36, 38–44,46–54, 56, 58]. Of the remaining studies, six includedchildren/adolescents and adults [21, 22, 34, 37, 55, 57]and two included only adults [45, 59]. PBM reportingwas predominantly from proxies (13 of 23 studies);eleven studies included both self-reported and proxydata [15, 21, 22, 36, 43, 48–50, 53, 54, 57] and sevenstudies included only self-reported PBM data [11, 34, 37,45, 55, 58, 59].In terms of use of PBMs, 22 of the studies used a ver-
sion of the HUI [15, 21, 22, 33, 35, 36, 38–44, 46, 48, 49,51–54, 56, 57], eight used a version of the EQ-5D [11,15, 34, 37, 47, 50, 58, 59], three used a version of AQoLinstrument [21, 22, 57], one used the 15D [45] and oneused the SF-6D [55].
Narrative synthesisUtility outcomes are presented in Table 5 and PBM per-formance outcomes are presented in Table 6. The narra-tive synthesis is categorised by type of mobilityimpairment and PBM performance indicator. No studiesreported PBM reliability outcomes, therefore reliabilityresults have not been presented.
Cerebral palsyKnown-group analysesFive studies reported known-group analyses in CP [22,33, 51, 53, 58]. Petrou and Kupek [51] estimated that theadjusted HUI3 disutility of childhood CP from perfecthealth was − 0.72 (95% confidence interval [CI] -0.61 to− 0.85), and − 0.65 (95% CI − 0.54 to − 0.78) from child-hood norms. Two studies found that as CP severity (i.e.gross motor function) increased, average utility scores(measured using HUI3 or AQoL) decreased in adoles-cents and young adults with CP [22, 53]; Rosenbaumet al. [53] found statistically significant differences inmean utility scores between most Gross Motor FunctionClassification System (GMFCS) levels (p < 0.01). Onestudy demonstrated that the vision, pain and cognitiondimensions of the HUI3 steadily declined as GMFCSlevel increased, however statistical significance was notreported [33]. Vitale et al. [58] found that adolescentswith CP had significantly higher average EQ-5D utilityscores (0.92) compared to adolescents with scoliosis(with comorbidities) (0.73; p > 0.05), although selectionof these patient sub-groups was not explicitly justifiedand the version of the EQ-5D was not reported.
Convergent validity: comparing measuresTwo studies reported that GMFCS level was correlatedwith worsening utility scores [22, 53]. Young et al. [22]found that GMFCS level in childhood was responsiblefor between 45% (AQoL: β = − 0.148; p < 0.001) and 53%(HUI3: β = − 0.205; p < 0.001) of variance in utilityscores. Rosenbaum et al. [53] found a strong negativecorrelation between the HUI3 utility scores of adoles-cents with CP and their GMFCS level (r = − 0.81). Interms of individual PBM dimensions, results from fourstudies were varied [33, 35, 39, 49]; Kennes et al. [39]found that the dimension most associated with GMFCSlevel in children was ambulation (tau-b = 0.82; p < 0.01).Bartlett et al. [33] found that the HUI3 vision, pain andcognition dimensions generally worsened as GMFCSlevel increased in adolescents; however, there was no in-dication that these dimensions were determinants ofmotor capacity decline. Two studies [35, 49] reported asignificant negative association between the HUI3 paindimension and GMFCS level in children with CP, how-ever both of these studies only examined the HUI3 paindimension and not the full HUI3 system.Three studies reported various levels of correlation be-
tween PBMs and other outcome measures in CP [22, 46,53]. Young et al. [22] found moderate correlation be-tween utility score and SRH (r = 0.41; p < 0.001 for boththe HUI3 and AQoL). Two studies compared the Qual-ity of Life Instrument for People with DevelopmentalDisabilities (QOL Instrument) and the HUI3 [46, 53];Rosenbaum et al. [53] reported that adolescents’ HUI3
Bray et al. Health Economics Review (2020) 10:9 Page 7 of 38
utility scores explained between 3% (belonging) and 14%(being) of variance in QOL Instrument dimensionscores, while Livingston and Rosenbaum [46] reportedthat the HUI3 and QOL Instrument shared up to 23%variance, thus the relationship between the measureswas considered to be moderate at best.Two studies reported on the relationship between
different PBMs in CP [22, 57]. Although utility scoresderived from the HUI3 and AQoL were strongly cor-related (r = 0.87; p < 0.001) [22], HUI3 derived utilityscores tended to be lower than AQoL derived utilityscores [22, 57].
Convergent validity: comparing respondentsFour studies examined correlation between respondentsin CP [22, 48–50]. Morrow et al. [48] reported moderateagreement between parents and doctors of children withCP for the HUI2/3 dimensions of sensation (63.6%agreement; Kappa 0.41), cognition (70% agreement;Kappa 0.56), self-care (100% agreement; Kappa 1.00) andambulation (63.6% agreement; Kappa 0.46). Good correl-ation was also found between child self-reported HUI3pain scores and equivalent parent proxy scores (Good-man and Kruskal’s y statistic = 0.57; p < 0.001) [49].Perez Sousa et al. [50] reported a high level of dis-
agreement between parents and children on the EQ-5D
Table 3 Quality appraisal outcomes
Study reference (author, year) Tests of statisticalsignificanceconducted
Sub-groupanalysesconducted
Clinicalimplicationsdiscussed
Proportions ofmissing/incorrectdata reported
Response and/orcompletion ratesreported
Inclusion and/orexclusion criteriaexplicitly stated
Bartlett et al. (2010) [33] Y Y Y Y X Y
Bray et al. (2017) [15] Y Y Y Y Y Y
Burstrom et al. (2014) [11] Y Y X Y Y X
Cavazza et al. 2016 [34] X X X X X X
Christensen et al. (2016) [35] Y Y Y X Y X
Findlay et al. (2015) [36] Y Y Y Y Y X
Hendriksz etl al (2014) [37]. Y Y Y Y X Y
Karmur and Kulkarni (2018) [38] Y X Y Y X X
Kennes et al. (2002) [39] Y Y Y Y X Y
Kulkarni et al. (2004) [40] Y X Y X Y Y
Kulkarni (2006) [41] Y Y X Y X Y
Kulkarni et al. (2008) [42] Y X Y Y Y X
Kulkarni et al. (2008) [43] Y Y Y X Y Y
Landfeldt et al. (2016) [44] Y X Y X Y X
Lindquist et al. (2014) [45] Y Y Y X X Y
Livingston and Rosenbaum (2008) [46] Y X Y Y X X
Lopez-Bastida et al. (2017) [47] X X Y Y X X
Morrow et al. (2011) [48]. Y Y Y X Y Y
Penner et al. (2013) [49] Y Y Y X Y X
Perez Sousa et al. (2017) [50] Y Y Y Y Y Y
Petrou and Kupec (2009) [51] Y Y Y Y X Y
Rocque et al. (2015) [52] Y Y Y X X Y
Rosenbaum et al. (2007) [53] Y Y Y Y X Y
Sims-Williams et al. (2016) [54] Y Y Y Y X Y
Slaman et al. (2015) [55] Y X Y X X Y
Tilford et al. (2005) [56] Y X Y X Y X
Usuba et al. (2014) [57] Y Y Y Y Y Y
Vitale et al. (2001) [58] Y Y Y X X Y
Wallander et al. (2009) [59]. Y X Y X Y Y
Young et al. (2010) [22]. Y Y Y Y Y Y
Young et al. (2013) [21] Y Y Y X Y Y
Bray et al. Health Economics Review (2020) 10:9 Page 8 of 38
Table
4Stud
ycharacteristicsandparticipantde
mog
raph
icsforinclud
edstud
ies
Stud
yreference
Aim
sandob
jectives
Cou
ntry
Stud
ytype
Con
ditio
n(s)of
interest
Clinicalanddiagno
stic
inform
ation(%
ofsample)
Samplesize
Age
(mean±SD
;range
)and
gend
er(%
ofsample)
Bartlettet
al.
(2010)
[33]
Toexplorepo
ssiblereason
sfortheob
served
declinein
grossmotor
capacity
ofadolescentswith
cerebral
palsyin
GMFC
SlevelsIII,IV
andV
Canada
Prospe
ctive
coho
rtAdo
lescen
tswith
cerebral
palsy
GMFCSlevel
GMFC
SIII:38%
GMFC
SIV:35%
GMFC
SV:27%
Cerebralpalsy
sub-type
Diplegia:24%
Hem
iplegia:1%
Tetraplegia:71%
n=135
Mean14
yrs.(±2.4;rang
e11–17)
44%
female/56%
male
Bray
etal.
(2017)
[15]
Tocompare
how
children
with
mob
ility
impairm
ents
andtheirparents(byproxy)
repo
rtHRQ
oLusingstandard
outcom
emeasures
UK
Cross-sectio
nal
Childrenandadolescents
with
impairedmob
ility
(relevant
cond
ition
s:cerebral
palsy,he
miplegia,muscular
dystroph
y)
Mixed
diagno
ses
Cereb
ralp
alsy:85%
Hem
iplegia/stroke:8%
Musculardystroph
y:8%
n=13
Rang
e6-18
yrs.
39%
female/62%
male
Burstrom
etal.
(2014)
[11]
Totestthefeasibility
and
validity
oftheEQ
-5D-Y
ina
Swed
ishpatient
sampleof
childrenandadolescents
with
functio
nalm
otor,ortho
paed
icandmed
icaldisabil
ities
andto
compare
there
sults
with
age
neralp
opula
tionsample
Swed
enCase-control
Childrenandadolescents
with
functio
nalm
otor,
orthop
aedicandmed
ical
disabilities(re
levant
cond
ition
s:artrog
rypo
sis
multip
lecong
enital,
myelomen
ingo
cele,cereb
ral
palsy,orthop
aediclower
limb
deform
ities,achon
drop
lasia)
Mixed
diagno
ses
Artrogryposismultip
lecong
enital:14%
Myelomen
ingo
cele:17%
Cereb
ralp
alsy:20%
Ortho
paed
iclower
limb
deform
ities:7%
Achon
drop
lasia:6%
n=71
case
grou
pn=407control
grou
p
Case
group
Mean12
yrs.(±3.1;
rang
e7–17)
61%
female/39%
male
Controlgroup
Mean13
yrs.(±2.7;rang
e8–16)
49%
female/51%
male
Cavazza
etal.
(2016)
[34]
Tode
term
inetheecon
omic
burden
from
asocietal
perspe
ctiveandtheHRQ
oLof
patientswith
Duche
nne
musculardystroph
y,in
Europe
Multin
ational
(Bulgaria,France,
Germany,
Hun
gary,Italy,
Spain,Sw
eden
,UK)
Cross-sectio
nal
Ado
lescen
tsandadultswith
Duche
nnemuscular
dystroph
y
Not
stated
n=268
Meanagevariedby
coun
try,
from
11yrs.(±5.6)
inSw
eden
to23
yrs.(±15.8)inBu
lgaria.
70%
ofsamplewerechildren
(rang
e2–17)
7%female/93%
male
Christensen
etal.
(2016)
[35]
Toiden
tifyfactorsassociated
with
achange
inpain
over
timein
childrenwith
cerebral
palsy
Canada
Prospe
ctive
coho
rtChildrenandadolescents
with
cerebralpalsy
GMFCSlevel
GMFC
SI:21%
GMFC
SII:11%
GMFC
SIII:24%
GMFC
SIV:22%
GMFC
SV:22%
n=148
Mean8yrs.(rang
e3–19)
30%
female/70%
male
Find
layet
al.
(2015)
[36]
Toexplorewhe
ther
HRQ
oLcanbe
pred
ictedby
pain,
age,GMFC
Slevel,andsexin
childrenwith
cerebralpalsy
andwhe
ther
different
pain
aetio
logies
have
varying
effectson
HRQ
oL
Canada
Cross-sectio
nal
Childrenwith
cerebralpalsy
GMFCSlevel
GMFC
SI:26%
GMFC
SII:16%
GMFC
SIII:23%
GMFC
SIV:18%
GMFC
SV:17%
Cerebralpalsy
sub-type
Unilateral:19%
Bilateralspastic:75%
Dyskine
tic:4%
Other
(ataxicand
hypo
tonic):2%
n=248
Mean10
yrs.(±4.3)
37%
female/63%
male
Bray et al. Health Economics Review (2020) 10:9 Page 9 of 38
Table
4Stud
ycharacteristicsandparticipantde
mog
raph
icsforinclud
edstud
ies(Con
tinued)
Stud
yreference
Aim
sandob
jectives
Cou
ntry
Stud
ytype
Con
ditio
n(s)of
interest
Clinicalanddiagno
stic
inform
ation(%
ofsample)
Samplesize
Age
(mean±SD
;range
)and
gend
er(%
ofsample)
Hen
driksz
etal.
(2014)
[37]
Toassess
theglob
albu
rden
ofdiseaseam
ongpatients
with
Morqu
ioAsynd
rome,
includ
ingim
pact
onmob
ility/w
heelchairuse,
HRQ
oL,p
ainandfatig
ueand
theinteractionbe
tween
thesefactors
Multin
ational
(Brazil,Colom
bia,
Germany,Spain,
Turkey,U
K)
Cross-sectio
nal
Childrenandadultswith
Morqu
ioAsynd
rome
Comorbidities
Bone
deform
ity:75%
full
sample
Abn
ormalgait:96%
adult
grou
p/75%
child
grou
p
n=36
child
grou
pn=27
adult
grou
p
Child
group
Rang
e5-17
yrs.(47%
aged
10–14)
44%
female/56%
male
Adultgroup
Rang
e18-40yrs.(52%
aged
18–24)
44%
female/56%
male
Karm
urand
Kulkarni
(2018)
[38]
Toun
derstand
thequ
ality
oflifeof
patientswith
myelomen
ingo
celeand
shun
tedhydrocep
halus
Canada
Cross-sectio
nal
Childrenandadolescents
with
spinabifid
a(m
yelomen
ingo
cele)and
shun
tedhydrocep
halus
Not
stated
n=131
Mean12
yrs.(±3.7)
51%
female/49%
male
Kenn
eset
al.
(2002)
[39]
Tode
scrib
ethehe
alth
status
ofpre-adolescent
children
with
cerebralpalsy,andto
determ
inethestreng
thof
correlations
betw
eenthe
severityof
grossmotor
functio
nalimpairm
ent
andothe
raspe
ctsof
functio
nalh
ealth
status
(sen
sory,intellectual,
emotionaletc.)
Canada
Prospe
ctive
coho
rtChildrenwith
cerebralpalsy
GMFCSlevel
GMFC
SI:28%
GMFC
SII:12%
GMFC
SIII:20%
GMFC
SIV:20%
GMFC
SV:21%
n=408
Mean8yrs.(±1.9;rang
e5–15)
46%
female/54%
male
Kulkarni
etal.
(2004)
[40]
Tode
velopandtestthe
psycho
metric
prop
ertiesof
theHydroceph
alus
Outcome
Questionn
aire
(HOQ),as
ameasure
ofhe
alth
status
inclinicalresearch
projectsof
paed
iatrichydrocep
halus
Canada
Cross-sectio
nal
Childrenwith
hydrocep
halus
Hydroceph
alus
aetiology
Con
genital/aqu
eductal
sten
osis:36%
Myelomen
ingo
cele:13%
Other:51%
n=90
Mean10
yrs.(±3.5)
Gen
derdistrib
utionno
tstated
Kulkarni
etal.
(2006)
[41]
Tocompare
threeseparate
metho
dsforestablishing
interpretabilityfortheHOQ,
andto
calculatethe
conversion
ofnu
mericalHOQ
scores
into
utility
scores
obtained
from
HUI2
Canada
Cross-sectio
nal
Childrenwith
hydrocep
halus
Hydroceph
alus
aetiology
Con
genital/aqu
eductal
sten
osis:33%
Myelomen
ingo
cele:15%
Intraven
tricular
haem
orrhage:
13%
Other:40%
n=79
Mean10
yrs.(±3.5)
Gen
derdistrib
utionno
tstated
Kulkarni
etal.
(2008)
[42]
Tostud
ythefactors
associated
with
HRQ
oLin
Canadianchildrenwith
hydrocep
halus,usinga
compreh
ensive
mod
elof
determ
inantsof
child
health,
includ
ingsocioe
cono
mic
factors
Canada
Cross-sectio
nal
Childrenwith
hydrocep
halus
Hydroceph
alus
aetiology
Myelomen
ingo
cele:33%
Intraven
tricular
haem
orrhage
ofprem
aturity:9%
Aqu
eductalsteno
sis:10%
Post-in
fection:4%
Posteriorfossacyst:5%
Not
stated
:39%
n=340
Mean11
yrs.(±3.6)
Gen
derdistrib
utionno
tstated
Kulkarni
etal.
Toinvestigatethefeasibility
Canada
Cross-sectio
nal
Childrenwith
hydrocep
halus
Hydroceph
alus
aetiology
n=273
Mean14
yrs.(±2.6)
Bray et al. Health Economics Review (2020) 10:9 Page 10 of 38
Table
4Stud
ycharacteristicsandparticipantde
mog
raph
icsforinclud
edstud
ies(Con
tinued)
Stud
yreference
Aim
sandob
jectives
Cou
ntry
Stud
ytype
Con
ditio
n(s)of
interest
Clinicalanddiagno
stic
inform
ation(%
ofsample)
Samplesize
Age
(mean±SD
;range
)and
gend
er(%
ofsample)
(2008)
[43]
andscientificprop
ertiesof
achild-com
pleted
versionof
theHOQ(cHOQ)
Myelomen
ingo
cele:34%
Intraven
tricular
haem
orrhage
ofprem
aturity:11%
Aqu
eductalsteno
sis:10%
Post-in
fection:4%
Con
genitalcom
mun
icating:
3% Intracranialcyst:8%
Other:30%
47%
female/54%
male
Land
feldtet
al.
(2016)
[44]
Toestim
ateHRQ
oLin
patientswith
Duche
nne
musculardystroph
y
Multin
ational
(Germany,Italy,
UK,USA
)
Cross-sectio
nal
Childrenandadolescents
with
Duche
nnemuscular
dystroph
y
Ambulatorystatus
Early
ambu
latory
(5-7yrs):
20%
Late
ambu
latory
(8-11yrs):
33%
Early
non-am
bulatory
(12-15
yrs):20%
Late
non-am
bulatory
(≥16
yrs):27%
n=770
Mean14
yrs.(±7.0)
100%
male
Lind
quistet
al.
(2014)
[45]
Toanalysequ
ality
oflifein
avery
long
-term
follow-upof
now
adultindividu
als,treated
forhydrocep
halus(with
out
spinabifid
a)du
ringtheirfirst
year
oflife
Swed
enCross-sectio
nal
Adu
ltswho
expe
rienced
hydrocep
halusin
infancy
31%
ofstud
ygrou
pdiagno
sedwith
cerebralpalsy
and/or
epilepsy;
hydrocep
halusaetio
logies
notrepo
rted
n=29
stud
ygrou
pn=1613
controlg
roup
Studygroup
Mean34
yrs.(rang
e30–41)
38%
female/62%
male
Controlgroup
Matched
ageandge
nder
tocase
grou
p
Living
ston
and
Rosenb
aum
(2008)
[46]
Toassess
thestability
ofmeasuremen
tof
quality
oflifeandHRQ
oLover
the
course
of1year
amon
gadolescentswith
cerebral
palsy
Canada
Prospe
ctive
coho
rtAdo
lescen
tswith
cerebral
palsy
GMFCSlevel
GMFC
SI:30%
GMFC
SII:16%
GMFC
SIII:15%
GMFC
SIV:25%
GMFC
SV:15%
n=185
Mean16
yrs.(±1.75;range
13–20)
47%
female/54%
male
Lope
z-Bastida
etal.(2017)[47]
Tode
term
inetheecon
omic
burden
andhe
alth-related
quality
oflifeof
patientswith
spinalmuscularatroph
yand
theircaregiversin
Spain
Spain
Cross-sectio
nal
Childrenwith
spinalmuscular
atroph
ySpinal
muscularatroph
ytype
Type
I:10%
Type
II:74%
Type
III:16%
n=81
Mean7yrs.(±5.47)
58%
female/42%
male
Morrow
etal.
(2011)
[48]
Toevaluate
differences
betw
eenchildren’s,parents’
anddo
ctors’pe
rcep
tions
ofhe
alth
states
andHRQ
oLin
childrenwith
chronicillne
ssandexplorefactorswhich
explainthesedifferences
Australia
Cross-sectio
nal
Childrenwith
chronic
cond
ition
s(re
levant
cond
ition
:cereb
ralp
alsy)
Allparticipantsin
cerebral
palsysub-grou
pwerecate
gorised
asGMFC
SlevelV
Cerebralpalsy
sub-group
n=1child-
parent
pair
n=1child-
doctor
pair
n=11
parent-
doctor
pairs
Cerebralpalsy
sub-group
36%
aged
>12
yrs.
Gen
derdistrib
utionno
trepo
rted
Penn
eret
al.
(2013)
[49]
Tode
term
inetheim
pact
ofpain
onactivities
andto
iden
tifythecommon
Canada
Cross-sectio
nal
Childrenandadolescents
with
cerebralpalsy
GMFCSlevel
GMFC
SI:24%
GMFC
SII:13%
n=252
Mean9yrs.(±4.2;rang
e3–19)
36%
female/64%
male
Bray et al. Health Economics Review (2020) 10:9 Page 11 of 38
Table
4Stud
ycharacteristicsandparticipantde
mog
raph
icsforinclud
edstud
ies(Con
tinued)
Stud
yreference
Aim
sandob
jectives
Cou
ntry
Stud
ytype
Con
ditio
n(s)of
interest
Clinicalanddiagno
stic
inform
ation(%
ofsample)
Samplesize
Age
(mean±SD
;range
)and
gend
er(%
ofsample)
physician-iden
tifiedcauses
ofpain
inchildrenandyouth
aged
3to
19yearsacross
all
levelsof
severityof
cerebral
palsy
GMFC
SIII:21%
GMFC
SIV:19%
GMFC
SV:23%
PerezSousaet
al.
(2017)
[50]
Toanalysethelevelo
fagreem
entbe
tweenchildren
with
cerebralpalsyandtheir
parents,usingtheEQ
-5D-Y
questio
nnaire
andits
proxy
version
Spain
Cross-sectio
nal
Childrenandadolescents
with
cerebralpalsy
Functiona
lclassification
Grade
1(with
outactivity
limitatio
n):66%
Grade
2(m
ildor
mod
erate
activity
limitatio
n):34%
n=62
Mean10
yrs.(±2.3;rang
e6–17)
44%
female/56%
male
Petrou
andKu
pek
(2009)
[51]
Toaugm
entprevious
catalogu
esof
preferen
ce-
basedHRQ
oLweigh
tsby
estim
atingpreferen
ce-based
HUI3multiattrib
uteutility
scores
associated
with
awide
rang
eof
childho
odcond
ition
s
UK
Cross-sectio
nal
Childrenwith
childho
odcond
ition
s(re
levant
cond
ition
s:microceph
aly,
cerebralpalsy,spinal
muscularatroph
y,muscular
dystroph
y,spinabifid
a)
Not
stated
Relevant
sub-
groups
Microceph
aly
n=40
Cereb
ralp
alsy
n=178
Muscular
dystroph
yor
spinal
muscular
atroph
yn=45
Spinabifid
an=42
Relevant
sub-groups
Microceph
aly:mean11
yrs.
Cereb
ralp
alsy:m
ean11
yrs.
Musculardystroph
yor
spinal
muscularatroph
y:mean
12yrs.
Spinabifid
a:mean13
yrs.
Gen
derdistrib
utionpe
rsub-
grou
pno
trepo
rted
Rocque
etal.
(2015)
[52]
Tocharacterisethequ
ality
oflifeof
paed
iatricpatientswith
spinabifid
a,andto
analyse
factorsthat
influen
ceHRQ
oLandaidin
thede
term
ination
ofwhe
ther
acorrelation
existsbe
tweenvario
usdiseaseand/or
person
alcharacteristicsandHRQ
oLscores
USA
Cross-sectio
nal
Childrenandadolescents
with
spinabifid
aUnd
erlyingdiagno
sisMyelomen
ingo
cele:79%
Lipo
myelomen
ingo
cele:16%
Men
ingo
cele:3%
Filum
term
inale-related
patholog
y:2%
Sacralagen
esis:>
1%
n=159
Mean12
yrs.(rang
e5–20)
57%
female/43%
male
Rosenb
aum
etal.
(2007)
[53]
Torepo
rtself-
and
proxy-assessed
quality
oflife
alon
gwith
parentalaccoun
tsof
HRQ
oLof
acoho
rtof
adolescentswith
cerebral
palsyparticipatingin
along
itudinalstudy
chartin
gmob
ility
andself-care
throug
htheadolescent
years
Canada
Prospe
ctive
coho
rtAdo
lescen
tswith
cerebral
palsy
GMFCSlevel
GMFC
SI:30%
GMFC
SII:16%
GMFC
SIII:14%
GMFC
SIV:25%
GMFC
SV:16%
n=203
Mean16
yrs.(±1.75;range
13–20)
45%
female/55%
male
Sims-Williams
etal.(2017)[54]
Toascertainthequ
ality
oflife
ofsurvivingchildrenwith
spinabifid
aandto
determ
ine
whe
ther
thiswas
influen
ced
Ugand
aCross-sectio
nal
Childrenwith
spinabifid
a45%
ofsamplehad
comorbidhydrocep
halus
Walking
ability
Unableto
walk:47%
n=66
(63of
which
completed
HUI3)
Rang
e10-14yrs.
44%
female/56%
male
Bray et al. Health Economics Review (2020) 10:9 Page 12 of 38
Table
4Stud
ycharacteristicsandparticipantde
mog
raph
icsforinclud
edstud
ies(Con
tinued)
Stud
yreference
Aim
sandob
jectives
Cou
ntry
Stud
ytype
Con
ditio
n(s)of
interest
Clinicalanddiagno
stic
inform
ation(%
ofsample)
Samplesize
Age
(mean±SD
;range
)and
gend
er(%
ofsample)
bymob
ility,urin
ary
continen
ce,hydroceph
alus,
sex,family
size
andscho
olattend
ance
Walkwith
sticks/crutche
s:14%
Walkun
aide
d:39%
Slam
anet
al.
(2015)
[55]
Toevaluate
thecost-utility
ofalifestyleinterven
tions
amon
gadolescentsand
youn
gadultswith
cerebral
palsy
TheNethe
rland
sRand
omised
controlledtrial
(single-blind)
Ado
lescen
tsandyoun
gadultswith
cerebralpalsy
Interven
tiongrou
pGMFC
Slevel
GMFC
SI:61%
GMFC
SII:32%
GMFC
SIII:7%
GMFC
SIV:0%
Con
trol
grou
pGMFC
Slevel
GMFC
SI:55%
GMFC
SII:31%
GMFC
SIII:10%
GMFC
SIV:4%
n=20
interven
tion
grou
pn=20
control
grou
p
Interventiongroup
Mean20
yrs.(±3.0)
57%
female/43%
male
Controlgroup
Mean20
yrs.(±3.0)
48%
female/52%
male
Tilford
etal.
(2005)
[56]
Toprovideinform
ationon
thepreferen
cescores
ofchildrenwith
spinabifid
aaperta
andto
measure
the
impact
ofcarin
gforachild
with
spinabifid
aconsistent
with
econ
omicevaluatio
ns
USA
Case-control
Childrenwith
spinabifid
aCa
segrouplesio
nlevel
Sacral:42%
Lower
lumbar:34%
Thoracic:25%
n=80
case
grou
pn=30
gene
ral
popu
latio
ncontrolg
roup
Case
group
Mean9yrs.(±4.6)
61%
female/39%
male
Generalpopulationcontrol
group
Mean7yrs.(±4.0)
55%
female/45%
male
Usuba
etal.
(2014)
[57]
Toexplorethemagnitude
andtim
ingof
change
sin
grossmotor
functio
nand
HRQ
oLam
ongpe
rson
swith
cerebralpalsyover
an8-year
perio
d,with
specificinterest
incomparin
gthosewho
madethetransitio
nto
adult
services
Canada
Prospe
ctive
coho
rtAdo
lescen
tsandadultswith
cerebralpalsy
GMFCSlevel(fullsample)
GMFC
SI:22%
GMFC
SII:13%
GMFC
SIII:13%
GMFC
SIV:22%
GMFC
SV:30%
n=31
‘you
nger
adults’g
roup
n=23
‘older
adults’g
roup
‘Young
eradults’group
Mean15
yrs.(rang
e13–17)
‘Olderadults’group
Mean26
yrs.(rang
e23–32)
Fullsample
46%
female/54%
male
Vitaleet
al.
(2001)
[58]
Toexam
inewhe
ther
theSF-
36andEQ
-5Dwou
ldbe
use
fulfor
evaluatin
gqu
ality
oflifein
adolescentswith
ortho
paed
iccond
ition
s
USA
Cross-sectio
nal
Ado
lescen
tswith
orthop
aedic
prob
lems(re
levant
cond
ition
:cerebralpalsy)
Not
stated
n=14
cerebralpalsy
sub-grou
p
Cereb
ralp
alsy
sub-grou
page
data
notrepo
rted
(full
sample:mean14
yrs.
[rang
e10–18])
Gen
derdistrib
utionno
trepo
rted
Walland
eret
al.
(2009)
[59]
Toreview
agrou
pof
patients
over
60yearsof
agewho
had
been
treatedforcong
enital
talipes
equinu
svarus(CTEV)
ininfancy,usingge
neric
instrumen
tsforthe
assessmen
tof
quality
oflife
inge
neraland
aspecificfoot
andankleinstrumen
tfor
assessmen
tof
functio
n
Swed
enCross-sectio
nal
Adu
ltstreatedforCTEVin
infancy
Clubfoot
laterality
Unilateral:54%
Bilateral:46%
n=83
Mean64
yrs.(rang
e62–67)
24%
female/76%
male
Bray et al. Health Economics Review (2020) 10:9 Page 13 of 38
Table
4Stud
ycharacteristicsandparticipantde
mog
raph
icsforinclud
edstud
ies(Con
tinued)
Stud
yreference
Aim
sandob
jectives
Cou
ntry
Stud
ytype
Con
ditio
n(s)of
interest
Clinicalanddiagno
stic
inform
ation(%
ofsample)
Samplesize
Age
(mean±SD
;range
)and
gend
er(%
ofsample)
Youn
get
al.
(2010)
[22]
Tode
scrib
ethehe
alth
and
quality
oflifeou
tcom
esof
youthandyoun
gadultswith
cerebralpalsy,andto
explore
theim
pact
of3factors
(cereb
ralp
alsy
severity,age
andsex)on
quality
oflife
outcom
es
Canada
Cross-sectio
nal
Ado
lescen
tsandyoun
gadultswith
cerebralpalsy
‘Youth’group
GMFCSlevel
GMFC
SI:22%
GMFC
SII:12%
GMFC
SIII:18%
GMFC
SIV:25%
GMFC
SV:22%
‘Adult’groupGMFCSlevel
GMFC
SI:23%
GMFC
SII:14%
GMFC
SIII:19%
GMFC
SIV:25%
GMFC
SV:20%
n=129‘you
th’
grou
pn=70
‘adu
lt’grou
p
‘Youth’group
Mean15
yrs.(±1.36)
45%
female/55%
male
‘Adult’group
Mean26
yrs.(±2.63)
40%
female/60%
male
Youn
get
al.
(2013)
[21]
Tode
scrib
ethehe
alth
and
HRQ
oLou
tcom
esof
youths
andyoun
gadultswith
spina
bifid
a
Canada
Cross-sectio
nal
Ado
lescen
tsandyoun
gadultswith
spinabifid
a‘Youth’group
lesio
nlevelTho
racic:25%
High-lumbar:18%
Low-lu
mbar:30%
Sacral:28%
‘Adult’grouplesio
nlevel
Thoracic:15%
High-lumbar:23%
Low-lu
mbar:31%
Sacral:15%
Unkno
wn:15%
n=40
‘you
th’
grou
pn=13
‘adu
lt’grou
p
‘Youth’groupMean16
yrs.
(±1.3;rang
e13–17)65%
female/35%
male
‘Adult’group
Mean26
yrs.(±3.10;
rang
e23–32)
77%
female/23%
male
Bray et al. Health Economics Review (2020) 10:9 Page 14 of 38
Table
5Summaryof
utility
outcom
esforinclud
edstud
ies
Stud
yreference
Con
ditio
n(s)of
interest
Respon
dent
type
PBM
(s)
Other
relevant
outcom
emeasures
Overallutility
scores
(mean±SD
)Sub-grou
putility
scores
(mean±SD
)
Bartlettet
al.
(2010)
[33]
Ado
lescen
tswith
cerebral
palsy
Proxy(paren
t)HUI3
Gross
Motor
Functio
nMeasure
66Items(GMFM
-66)
SpinalAlignm
entandRang
eof
MotionMeasure
(SARO
MM)
NA
Overallutility
scores
notrepo
rted
,on
lyvision
,cog
nitio
nandpain
dimen
sion
sused
-am
bulatio
n,he
aring,
speech,d
exterityand
emotiondimen
sion
sallexclude
ddu
eto
concep
tualoverlapwith
othe
rmeasures
NA
Bray
etal.(2017)[15]
Childrenandadolescents
with
impairedmob
ility
(relevant
cond
ition
s:cerebral
palsy,he
miplegia,muscular
dystroph
y)
Self-repo
rted
and
proxy(paren
t);
matched
-pairs
HUI2,H
UI3and
EQ-5D-Y
Visualanalog
uescale(VAS)
Child
self-reported
HUI2:0.53(±0.07)
HUI3:0.22(±0.09)
EQ-5D-Y:0.24(±0.30)
Parent
proxy
HUI2:0.49(±0.09)
HUI3:0.16(±0.10)
EQ-5D-Y:0.01(±0.14)
NA
Burstrom
etal.
(2014)
[11]
Childrenandadolescents
with
functio
nalm
otor,
orthop
aedicandmed
ical
disabilities(re
levant
cond
ition
s:artrog
rypo
sis
multip
lecong
enital,
myelomen
ingo
cele,cereb
ral
palsy,orthop
aediclower
limb
deform
ities,achon
drop
lasia)
Self-repo
rted
EQ-5D-Y
VAS
KIDSC
REEN
-27
KIDSC
REEN
-10
Self-ratedhe
alth
(SRH
)Life
satisfactionladd
er(LSL)
NA
Overallutility
scores
notrepo
rted
,on
lydimen
sion
scores
repo
rted
NA
Cavazza
etal.
(2016)
[34]
Ado
lescen
tsandadultswith
Duche
nnemuscular
dystroph
y
Self-repo
rted
EQ-5D-Y
VAS
Barthe
lInd
exMean0.24
Variedby
coun
try,rang
ingfro
m−0.71
(±0.41)in
Swed
ento
0.66
(±0.08)in
Bulgaria
NA
Christensen
etal.
(2016)
[35]
Childrenandadolescents
with
cerebralpalsy
Proxy(careg
iver)
HUI3
Won
g-BakerFA
CES
Pain
Ratin
gScale
NA
Overallutility
scores
notrepo
rted
,on
lyHUI3pain
dimen
sion
measured
NA
Find
layet
al.
(2015)
[36]
Childrenwith
cerebralpalsy
Self-repo
rted
and
proxy(careg
iver);
prop
ortio
nof
proxy
data
notrepo
rted
HUI3
DISABKIDSChron
icGen
eric
mod
ule(DCGM-37)
DISABKIDSSm
ileymeasure
(DSM
)Won
g-BakerFA
CES
Pain
Ratin
gScale
NA
Overallutility
scores
notrepo
rted
,on
lyHUI3pain
dimen
sion
measured
NA
Hen
driksz
etal.
(2014)
[37]
Childrenandadultswith
Morqu
ioAsynd
rome
Self-repo
rted
EQ-5D-5L
BriefPain
InventoryShort
Form
(BPI-SF)
Ado
lescen
tPediatric
Pain
Tool
(APPT)
NA
Overallutility
scores
notrepo
rted
Child
group:utilityscoreby
wheelchairusefrequency
Nouse:0.534
Occasionalu
se:0.664
Full-tim
euse:−0.180
Adultgroup:utilityscoreby
wheelchairusefrequency
Nouse:0.846
Occasionalu
se:0.582
Full-tim
euse:0.057
Karm
urandKu
lkarni
(2018)
[38]
Childrenandadolescents
with
spinabifid
a(m
yelomen
ingo
cele)and
Proxy(careg
iver)
HUI2andHUI3
(version
used
tocalculateutility
Hydroceph
alus
Outcome
Questionn
aire
(HOQ)
0.51
(±0.28)
NA
Bray et al. Health Economics Review (2020) 10:9 Page 15 of 38
Table
5Summaryof
utility
outcom
esforinclud
edstud
ies(Con
tinued)
Stud
yreference
Con
ditio
n(s)of
interest
Respon
dent
type
PBM
(s)
Other
relevant
outcom
emeasures
Overallutility
scores
(mean±SD
)Sub-grou
putility
scores
(mean±SD
)
shun
tedhydrocep
halus
scores
notstated
)
Kenn
eset
al.
(2002)
[39]
Childrenwith
cerebralpalsy
Proxy(careg
iver)
HUI3
Functio
nalInd
epen
dence
Measure
forChildren
(WeeFIM)
Streng
thandDifficulties
Questionn
aire
(SDQ)
WideRang
eAchievemen
tTest(W
RAT)
NA
Overallutility
scores
notrepo
rted
,on
lydimen
sion
scores
repo
rted
NA
Kulkarni
etal.
(2004)
[40]
Childrenwith
hydrocep
halus
Proxy(paren
t)HUI2
HOQ
NA
Overallutility
scores
notrepo
rted
NA
Kulkarni
etal.
(2006)
[41]
Childrenwith
hydrocep
halus
Proxy(paren
t)HUI2
HOQ
NA
Overallutility
scores
notrepo
rted
NA
Kulkarni
etal.
(2008)
[42]
Childrenwith
hydrocep
halus
Proxy(careg
iver)
HUI3
HOQ
0.58
(±0.32)
NA
Kulkarni
etal.
(2008)
[43]
Childrenwith
hydrocep
halus
Self-repo
rted
and
proxy(careg
iver);
proxy-repo
rted
HUI3
andHOQ,self-
repo
rted
cHOQ
HUI3
Child-com
pleted
versionof
theHOQ(cHOQ)
0.71
(±0.27)
NA
Land
feldtet
al.
(2016)
[44]
Childrenandadolescents
with
Duche
nnemuscular
dystroph
y
Proxy(careg
iver);
althou
ghpatients
includ
edin
data
collection,
only
proxyutility
data
repo
rted
HUI(versionno
tstated
)Pediatric
Qualityof
Life
Inventory(Ped
sQL)
neurom
uscularmod
ule3.0
NA
Overallutility
scores
notrepo
rted
Utility
scoreby
ambulatorystatus
Early
ambu
latory:0.75
Late
non-am
bulatory:0.15
Lind
quistet
al.
(2014)
[45]
Adu
ltswho
expe
rienced
hydrocep
halusin
infancy
Self-repo
rted
15D
NA
Stud
ygrou
p:0.92
Con
trol
grou
p:0.95
Living
ston
and
Rosenb
aum
(2008)
[46]
Ado
lescen
tswith
cerebral
palsy
Proxy(paren
t)HUI3
Qualityof
Life
Instrumen
tfor
Peop
lewith
Develop
men
tal
Disabilities(QOLInstrumen
t)
NA
Overallutility
scores
notrepo
rted
NA
Lope
z-Bastidaet
al.
(2017)
[47]
Childrenwith
spinalmuscular
atroph
yProxy(careg
iver)
EQ-5D-3L
VAS
Barthe
lInd
ex0.16
(±0.44)
Utility
scoreby
spinal
muscular
atroph
ytype
Type
IIspinalmuscularatroph
ysub-grou
p:−0.01
(±0.35)
Scores
forType
Iand
IIIno
trepo
rted
Morrow
etal.
(2011)
[48]
Childrenwith
chronic
cond
ition
s(re
levant
cond
ition
:cereb
ralp
alsy)
Self-repo
rted
and
proxy(paren
tand
doctor);matched
grou
ps
HUI2andHUI3
NA
NA
Overallutility
scores
notrepo
rted
forrelevant
sub-grou
p
NA
Penn
eret
al.
(2013)
[49]
Childrenandadolescents
with
cerebralpalsy
Self-repo
rted
and
proxy(paren
tand
physician);p
roxy-
repo
rted
HUI3pain
score,self-repo
rted
Won
g-BakerFA
CES
Pain
Scale
HUI3
Won
g-BakerFA
CES
Pain
Ratin
gScale
NA
Overallutility
scores
notrepo
rted
,on
lyHUI3pain
dimen
sion
measured
NA
PerezSousaet
al.
Childrenandadolescents
Self-repo
rted
and
EQ-5D-Y
VAS
NA
NA
Bray et al. Health Economics Review (2020) 10:9 Page 16 of 38
Table
5Summaryof
utility
outcom
esforinclud
edstud
ies(Con
tinued)
Stud
yreference
Con
ditio
n(s)of
interest
Respon
dent
type
PBM
(s)
Other
relevant
outcom
emeasures
Overallutility
scores
(mean±SD
)Sub-grou
putility
scores
(mean±SD
)
(2017)
[50]
with
cerebralpalsy
proxy(m
othe
rand
father);matched
grou
ps
Overallutility
scores
notrepo
rted
,on
lydimen
sion
scores
repo
rted
Petrou
andKu
pek
(2009)
[51]
Childrenwith
childho
odcond
ition
s(re
levant
cond
ition
s:microceph
aly,
cerebralpalsy,spinal
muscularatroph
y,muscular
dystroph
y,spinabifid
a)
Proxy(careg
iver)
HUI3
NA
Meanutilityby
cond
ition
Microceph
aly:0.141
Cereb
ralp
alsy:0.276
Musculardystroph
yor
spinal
muscularatroph
y:0.386
Spinabifid
a:0.452
NA
Rocque
etal.
(2015)
[52]
Childrenandadolescents
with
spinabifid
aSelf-repo
rted
and
proxy(careg
iver);
pred
ominantly
proxy
(11%
self-repo
rted
)
HUI3
NA
NA
Overallutility
scores
notrepo
rted
Utility
scoreby
type
ofspinabifida
Myelomen
ingo
cele:0.51
Closedspinaldysraphism
:0.77
Utility
scoreby
treatm
enthistory
Nohistoryof
shun
tor
CM-II
decompression
:0.74
Shun
tbu
tno
CM-II
decompression
:0.49
Shun
tandCM-II
decompression
:0.29
Rosenb
aum
etal.
(2007)
[53]
Ado
lescen
tswith
cerebral
palsy
Self-repo
rted
and
proxy(paren
t);
proxy-repo
rted
HUI3,
34%
self-repo
rted
QOLInstrumen
t
HUI3
QOLInstrumen
t0.42
(±0.41)
Utility
scoreby
GMFCSlevel
GMFC
SI:0.84
(±0.20)
GMFC
SII:0.50
(±0.31)
GMFC
SIII:0.39(±0.31)
GMFC
SIV:0.16(±0.26)
GMFC
SV:−0.08
(±0.23)
Sims-Williamset
al.
(2017)
[54]
Childrenwith
spinabifid
aSelf-repo
rted
and
proxy(careg
iver);
matched
pairs
HUI3
VAS
Child
self-reported
0.58
(95%
CI0.49–0.66)
Caregiverproxy
0.55
(95%
CI0.47–0.63)
NA
Slam
anet
al.
(2015)
[55]
Ado
lescen
tsandyoun
gadultswith
cerebralpalsy
Self-repo
rted
SF-6D
36-Item
ShortForm
Survey
(SF-36)-SF-6Dutility
out
comes
derived
from
SF-36
Baseline
Con
trol:0.74(±0.12)
Interven
tion:
0.75
(±0.10)
6mon
thspostintervention
Con
trol:0.77(±0.12)
Interven
tion:
0.80
(±0.03)
NA
Tilford
etal.
(2005)
[56]
Childrenwith
spinabifid
aProxy(careg
iver)
HUI2
Qualityof
Well-Being
scale
(QWB)
Casegrou
p:0.55
(±0.24)
Utility
bycase
grouplesio
nlevel
Sacrallesion
:0.61(±0.26)
Lower
lumbarlesion
:0.54(±0.19)
Thoraciclesion
:0.45(±0.25)
Gen
eralpo
pulatio
ncontrol
grou
p:0.93
(±0.11)
Usuba
etal.
(2014)
[57]
Ado
lescen
tsandadultswith
cerebralpalsy
Self-repo
rted
and
proxy(und
efined
);pred
ominantly
proxy
(40%
self-repo
rted
)
HUI3andAQoL
SRH
Baseline(bothgroups)
HUI3:0.29(±0.39)
AQoL:0.35(±0.33)
8-year
follow-up(bothgroups)
HUI3:0.29(±0.38)
AQoL:0.35(±0.32)
NA
Vitaleet
al.
(2001)
[58]
Ado
lescen
tswith
orthop
aedicprob
lems
(relevant
cond
ition
:cereb
ral
palsy)
Self-repo
rted
EQ-5D(version
not
stated
)SF-36
Cereb
ralp
alsy
sub-grou
p:0.922
NA
Bray et al. Health Economics Review (2020) 10:9 Page 17 of 38
Table
5Summaryof
utility
outcom
esforinclud
edstud
ies(Con
tinued)
Stud
yreference
Con
ditio
n(s)of
interest
Respon
dent
type
PBM
(s)
Other
relevant
outcom
emeasures
Overallutility
scores
(mean±SD
)Sub-grou
putility
scores
(mean±SD
)
Walland
eret
al.
(2009)
[59]
Adu
ltstreatedforCTEVin
infancy
Self-repo
rted
EQ-5D(version
not
stated
)SF-36
VAS
American
Acade
myof
Ortho
paed
icSurgeo
nsFoot
andAnkleQuestionn
aire
NA
Overallutility
scores
notrepo
rted
NA
Youn
get
al.
(2010)
[22]
Ado
lescen
tsandyoun
gadultswith
cerebralpalsy
Self-repo
rted
and
proxy(careg
iver);
pred
ominantly
self-
repo
rted
(45%
proxy)
HUI3andAQoL
SRH
Health
Assessm
ent
Questionn
aire
(HAQ)
Com
bine
dagegrou
psHUI3:0.30(±0.42)
AQoL:0.28(±0.33)
Utility
scoreby
agegroup
(HUI3/AQ
oL)
‘You
th’g
roup
:0.30(±0.43)/
0.28
(±0.34)
‘Adu
lt’grou
p:0.31
(±0.40)/
0.28
(±0.314)
Youthgroup:utilityscoreby
GMFCSlevel(HUI3/AQ
oL)
GMFC
SI:0.67
(±0.32)/0.58
(±0.31)
GMFC
SII:0.59
(±0.35)/0.53
(±0.34)
GMFC
SIII:0.43(±0.39)/0.31
(±0.32)
GMFC
SIV:0.08(±0.25)/0.06
(±0.12)
GMFC
SV:−0.13
(±0.19)/0.01
(±0.07)
Adultgroup:utilityscoreby
GMFCS
level(HUI3/AQ
oL)GMFC
SI:0.64
(±0.30)/0.52
(±0.32)
GMFC
SII:0.50
(±0.39)/0.33
(±0.24)
GMFC
SIII:0.53(±0.27)/0.39
(±0.27)
GMFC
SIV:0.06(±0.21)/0.10
(±0.20)
GMFC
SV:−0.14
(±0.20)/0.02
(±0.06)
Youn
get
al.
(2013)
[21]
Ado
lescen
tsandyoun
gadultswith
spinabifid
aSelf-repo
rted
and
proxy(careg
iver);
pred
ominantly
self-
repo
rted
(15%
proxy)
HUI3andAQoL
SRHHAQ
Com
bine
dagegrou
psHUI3:0.52(±0.28)
AQoL:0.34(±0.24)
Utility
scoreby
agegroup(HUI3/
AQoL)
‘You
th’g
roup
:0.58(±0.27)/0.37
(±0.26)
‘Adu
lt’grou
p:0.36
(±0.27)/0.25
(±0.17)
Utility
scoreby
lesio
nlevel(HUI3/
AQoL)
Thoracic:0.29(±0.14)/0.22
(±0.14)
High-lumbar:0.44
(±0.31)/0.28
(±0.23)
Low-lu
mbar:0.63
(±0.23)/0.39
(±0.21)
Sacral:0.76(±0.20)/0.51
(±0.33)
Unkno
wn:
0.22
(±0.02)/0.09
(±0.04)
Bray et al. Health Economics Review (2020) 10:9 Page 18 of 38
Table
6Summaryof
PBM
perfo
rmance
forinclud
edstud
ies
Stud
yreference
Con
ditio
n(s)of
interest
Know
n-grou
panalyses
Con
struct
validity:
comparin
gou
tcom
esCon
struct
validity:
comparin
gPBMs
Con
struct
validity:
comparin
grespon
dents
Respon
sivene
ss
Bartlettet
al.
(2010)
[33]
Ado
lescen
tswith
cerebral
palsy
HUI3vision
,cog
nitio
nandpain
dimen
sion
ssteadilyde
clined
asGMFC
Slevel
increased,
statistical
sign
ificanceno
trepo
rted
.
Exam
iningcorrelation
coefficients,therewas
noindicatio
nthat
theHUI3
vision
(r=0.01;p
=0.92),
cogn
ition
(r=−0.05;p
=0.59)or
pain
dimen
sion
s(r=0.16;p
=0.07)w
ere
determ
inantsof
motor
capacity,asmeasured
usingtheGMFM
-66.
NA
NA
Individu
alswith
aGMFC
Slevelo
fVexhibitedthe
largestmeande
creasesin
HUI3dimen
sion
levels
over
time,rang
ingfro
m−0.2(±1.2)
fortheHUI3
vision
dimen
sion
to−0.3
(±1.3)
fortheHUI3
cogn
ition
andpain
dimen
sion
s.
Bray
etal.
(2017)
[15]
Childrenandadolescents
with
impairedmob
ility
(relevant
cond
ition
s:cerebralpalsy,
hemiplegia,muscular
dystroph
y)
NA
NA
Largevariancebe
tween
meanutility
scores
derived
from
different
PBMs,rang
ingfro
m0.24
(EQ-5D-Y)to0.53
(HUI2)
forchild
self-repo
rted
utility,and
from
0.01
(EQ-
5D-Y)to
0.49
(HUI2)for
parent-rep
ortedproxy
utility.
Asign
ificant
respon
dent
type
effect
was
foun
d,with
meanchild
self-
repo
rted
utility
scores
sign
ificantly(p
≤0.021)
high
erthan
equivalent
proxieson
allP
BMs.
Sign
ificant
strong
correlations
werefoun
dbe
tweenutility
scores
for
children/parent
proxies
usingallm
easures:EQ
-5D
-Y(r=0.67;p
=0.026),
HUI2(r=0.73;p
=0.005)
andHUI3(r=0.84;p
<0.001).
Using
Bland-Altm
anplots,
sufficien
tagreem
ent
betw
eenutility
scores
for
children/parent
proxies
was
foun
dfortheHUI2
(CL=0.22)andHUI3
(CL=0.22).EQ
-5D-Y
exhibitedclinically
impo
rtantdiscrepancies
betw
eenchild
andparent
proxyrespon
ses
(CL=1.04).
NA
Burstrom
etal.
(2014)
[11]
Childrenandadolescents
with
functio
nalm
otor,
orthop
aedicandmed
ical
disabilities(re
levant
cond
ition
s:artrog
rypo
sis
multip
lecong
enital,
myelomen
ingo
cele,
cerebralpalsy,
orthop
aediclower
limb
deform
ities,
Statisticallysign
ificant
(p≤0.001)
differences
betw
eencase
and
controlg
roup
son
all
dimen
sion
s.Mean
dimen
sion
scores
not
repo
rted
,propo
rtions
ofpatientschoo
sing
each
dimen
sion
level
indicate
case
grou
p
Strong
sign
ificant
correlationwas
foun
dbe
tweentheEQ
-5D-Y
anxiety/de
pression
dimen
sion
andthe
KIDSC
REEN
-27psycho
logicalw
ell-b
eing
dimen
sion
(r=−0.51;p
=0.001);
KIDSC
REEN
-27ph
ysical
well-b
eing
dimen
sion
NA
NA
NA
Bray et al. Health Economics Review (2020) 10:9 Page 19 of 38
Table
6Summaryof
PBM
perfo
rmance
forinclud
edstud
ies(Con
tinued)
Stud
yreference
Con
ditio
n(s)of
interest
Know
n-grou
panalyses
Con
struct
validity:
comparin
gou
tcom
esCon
struct
validity:
comparin
gPBMs
Con
struct
validity:
comparin
grespon
dents
Respon
sivene
ss
acho
ndroplasia)
repo
rted
more
prob
lemson
all
dimen
sion
sthan
the
controlg
roup
:83%
ofcase
grou
prepo
rted
some/alotof
prob
lemson
any
dimen
sion
,com
pared
to37%
ofcontrol
grou
p;likew
ise21%
ofcase
grou
prepo
rted
extrem
eprob
lemson
any
dimen
sion
,com
pared
to2%
ofcontrol
grou
p.
(r=−0.53;p
=0.001);and
lifesatisfactionladd
er(LSL)(r=
−0.54;p
<0.001).M
oderatecorrel
ationalso
foun
dbe
tween
thisdimen
sion
andthe
KIDSC
REEN
-10HRQ
oLinde
xscore(r=−0.50;
p=0.001)
andself-
repo
rted
health
(SRH
)(0.42;p=0.007).
Theself-care
EQ-5D-Y
dimen
sion
exhibitedmod
eratecorrelationwith
the
KIDSC
REEN
-27ph
ysical
wellbeing
dimen
sion
(r=
−0.37;p
=0.028)
andthe
usualactivities
EQ-5D-Y
dimen
sion
was
mod
eratelycorrelated
with
theKIDSC
REEN
-27
psycho
logicalw
ellbeing
dimen
sion
(r=−0.35;p
=0.027).A
llothe
rcorrelations
were
weakor
absent.
Cavazza
etal.
(2016)
[34]
Ado
lescen
tsandadults
with
Duche
nnemuscular
dystroph
y
NA
NA
NA
NA
NA
Christensen
etal.
(2016)
[35]
Childrenandadolescents
with
cerebralpalsy
NA
Using
multivariate
linear
regression
,asign
ificant
associationwas
foun
dbe
tweentheHUI3pain
dimen
sion
scoreat
baselineandGMFC
Slevel
(b=−0.11,β
=−0.15;
p<0.036;95%
CI−
0.21
to−0.01):high
erpain
scoreat
baselinewas
associated
with
greater
improvem
entin
pain
status
inGMFC
SlevelI
comparedto
levelV
.
NA
NA
Using
one-way
ANOVA
analysis,a
sign
ificant
associationwas
foun
dbe
tweenph
ysician
prim
arypain
aetio
logy
andchange
inHUI3pain
status
(p=0.001):children
with
musculoskeletalpain
atbaselineshow
edsig
nificantim
provem
ents
(meanchange
0.55;95%
CI0.053
to1.05)com
paredto
childrenwith
out
pain
atbaseline(m
ean
change
−0.39;95%
CI−
0.62
to−0.15)(p=0.006).
Noothe
rassociations
wererepo
rted
.HUI3pain
dimen
sion
scores
didno
t
Bray et al. Health Economics Review (2020) 10:9 Page 20 of 38
Table
6Summaryof
PBM
perfo
rmance
forinclud
edstud
ies(Con
tinued)
Stud
yreference
Con
ditio
n(s)of
interest
Know
n-grou
panalyses
Con
struct
validity:
comparin
gou
tcom
esCon
struct
validity:
comparin
gPBMs
Con
struct
validity:
comparin
grespon
dents
Respon
sivene
ss
change
sign
ificantlyover
time:med
ianscoreof
2(out
of5)
atbo
thvisits.
How
ever,55%
ofchildren
change
dpain
status
over
time:34%
worsening
,21%
improving.
Find
layet
al.
(2015)
[36]
Childrenwith
cerebral
palsy
NA
NA
NA
NA
NA
Hen
driksz
etal.
(2014)
[37]
Childrenandadultswith
Morqu
ioAsynd
rome
Asign
ificant
effect
ofwhe
elchairuseon
utility
outcom
eswas
repo
rted
;significant
differences
repo
rted
betw
een:adultno
n-whe
elchairusersand
occasion
alwhe
elchair
users(p
=0.0115);
adultoccasion
alwhe
elchairusersand
full-tim
ewhe
elchair
users(p
=0.0007);
child
occasion
alwhe
elchairusersand
full-tim
ewhe
elchair
users(p=0.0018);and
child
non-whe
elchair
usersandfull-tim
ewhe
elchairusers(p
=0.0018).Childrenwho
occasion
allyused
awhe
elchairhada
high
ermeanutility
scoreon
averagethan
non-whe
elchairusers,
althou
ghbo
thgrou
pshadhigh
eraverage
utility
scores
than
full-
timewhe
elchairusers.
NA
NA
NA
NA
Karm
urand
Kulkarni
(2018)
[38]
Childrenandadolescents
with
spinabifid
a(m
yelomen
ingo
cele)and
shun
tedhydrocep
halus
Using
multivariate
regression
analysis,
anatom
icallevelo
fmyelomen
ingo
cele
hadasign
ificant
effect
onutility
score
(p=0.01):lower
anatom
icallevelo
fmyelomen
ingo
cele
NA
NA
NA
NA
Bray et al. Health Economics Review (2020) 10:9 Page 21 of 38
Table
6Summaryof
PBM
perfo
rmance
forinclud
edstud
ies(Con
tinued)
Stud
yreference
Con
ditio
n(s)of
interest
Know
n-grou
panalyses
Con
struct
validity:
comparin
gou
tcom
esCon
struct
validity:
comparin
gPBMs
Con
struct
validity:
comparin
grespon
dents
Respon
sivene
ss
was
associated
with
ahigh
erutility
score.
Kenn
eset
al.
(2002)
[39]
Childrenwith
cerebral
palsy
NA
Using
Kend
all’s
tau-btest
ofassociation,theHUI3
dimen
sion
mostassoci
ated
with
GMFC
Slevel
was
ambu
latio
n(tau-b
=0.82;p
<0.01):high
erGMFC
Slevelw
asassociated
with
increased
mob
ility
impairm
ent.
Mod
eratecorrelations
(rang
ingfro
mtau-b=
0.36
to0.58;p
<0.01)
werefoun
dbe
tween
GMFC
Sleveland
the
vision
,spe
echand
dexterity
dimen
sion
s.Overallpatterns
were
similarto
theam
bulatio
ndimen
sion
,but
correlations
wereweaker.
Hearin
g(tau-b
=0.16;p
=0.04)andcogn
ition
(tau-
b=0.27;p
<0.01)
dimen
sion
shad
statisticallysign
ificant
but
low
associationwith
GMFC
Slevel.The
emotion(tau-b
=0.03;
p=0.24)andpain
(tau-b
=0.07;p
=0.37)
dimen
sion
swereno
tsign
ificantlyassociated
with
GMFC
Slevel.
NA
NA
NA
Kulkarni
etal.
(2004)
[40]
Childrenwith
hydrocep
halus
NA
Strong
correlationwas
foun
dbe
tweenutility
scoreandtheHOQ
overallh
ealth
(r=0.81),
physicalhe
alth
(r=0.88),
social-emotional(r=
0.56)
andcogn
itive
(r=0.57)
scores.
NA
NA
NA
Kulkarni
etal.
(2006)
[41]
Childrenwith
hydrocep
halus
NA
Correlatio
nbe
tween
utility
scoreandHOQ
overallh
ealth
scorewas
high
(r=0.81),a
scatterplotde
mon
strated
NA
NA
NA
Bray et al. Health Economics Review (2020) 10:9 Page 22 of 38
Table
6Summaryof
PBM
perfo
rmance
forinclud
edstud
ies(Con
tinued)
Stud
yreference
Con
ditio
n(s)of
interest
Know
n-grou
panalyses
Con
struct
validity:
comparin
gou
tcom
esCon
struct
validity:
comparin
gPBMs
Con
struct
validity:
comparin
grespon
dents
Respon
sivene
ss
astrong
linear
relatio
nship.
Simpleand
complex
linearregression
mod
elsbo
thaccoun
ted
foralargeprop
ortio
nof
HUI2variability(adjusted
R2=0.66
and0.80
respectively).
Kulkarni
etal.
(2008)
[42]
Childrenwith
hydrocep
halus
NA
Meanutility
score(0.58±
0.63)closeto
cHOQ
meanscores
foroverall
health
(0.65±0.20),
Physicalhe
alth
(0.66±
0.25),Cog
nitivehe
alth
(0.55±0.28)andSocial-
emotionalh
ealth
(0.71±
0.19).Statistical
sign
ificanceno
trepo
rted
.
NA
NA
NA
Kulkarni
etal.
(2008)
[43]
Childrenwith
hydrocep
halus
NA
Sign
ificant
correlationwas
foun
dbe
tweenthecH
OQ
overallh
ealth
score(self-
repo
rted
)and
utility
score
(proxy-rep
orted)
(r=0.60;
p<0.001).
NA
NA
NA
Land
feldtet
al.
(2016)
[44]
Childrenandadolescents
with
Duche
nnemuscular
dystroph
y
Ambu
latory
classwas
sign
ificantly
associated
with
proxy-
repo
rted
utility
scores
(p<0.001),d
ecreasing
from
early
ambu
latory
class(m
ean0.75)to
late
non-am
bulatory
class(m
ean0.15).
Careg
iver
assessmen
tsof
health
andmen
tal
status
weresign
ificantlyassociated
with
utility
score
(p<0.001).
NA
NA
NA
NA
Lind
quistet
al.
(2014)
[45]
Adu
ltswho
expe
rienced
hydrocep
halusin
infancy
Thestud
ygrou
phad
sign
ificantly(p
≤0.004)
lower
dimen
sion
scores
comparedto
the
controlg
roup
invision
,eating,
usual
activities,m
ental
NA
NA
NA
NA
Bray et al. Health Economics Review (2020) 10:9 Page 23 of 38
Table
6Summaryof
PBM
perfo
rmance
forinclud
edstud
ies(Con
tinued)
Stud
yreference
Con
ditio
n(s)of
interest
Know
n-grou
panalyses
Con
struct
validity:
comparin
gou
tcom
esCon
struct
validity:
comparin
gPBMs
Con
struct
validity:
comparin
grespon
dents
Respon
sivene
ss
functio
ndimen
sion
s.Themostrepo
rted
prob
lemswere
associated
with
the
neuroimpairm
ent
subg
roup
(mean
utility
score0.87
comparedto
0.94
intheno
neuroimpairm
ent
subg
roup
).The
subg
roup
with
out
neuroimpairm
ent
wereno
tsign
ificantly
different
tothe
controlinterm
sof
meanutility
score
(0.94and0.95
respectively,pvalue
notrepo
rted
).
Living
ston
and
Rosenb
aum
(2008)
[46]
Ado
lescen
tswith
cerebralpalsy
NA
Disattenu
ated
correlation
coefficientsbe
tween
utility
scores
andQOL
Instrumen
tscores
demon
stratedweakto
mod
eratecorrelationfor
thebe
ing(r=0.48);
belong
ing(r=0.35);
becoming(r=0.29)and
overallq
ualityof
life
(r=0.35)scales.The
two
measure
shared
upto
23%
variance.
NA
NA
Gen
eralizability
coefficientswere
calculated
toassess
variabilityof
scores
over
time.Dim
ension
swith
greaterstability
(i.e.larger
Gscores)hadless
variabilitybe
tween
individu
alsover
time.The
ambu
latio
ndimen
sion
(G=0.94)andoverall
utility
(G=0.91)were
foun
dto
behigh
lystable;
whilethespeech
(G=
0.87),vision
(G=0.87);
dexterity
(G=0.82),
cogn
ition
(G=0.81),and
hearing(G
=0.72)
dimen
sion
swere
mod
eratelystable.The
pain
(G=0.48)and
emotion(G
=0.24)
dimen
sion
swerefoun
dto
have
low
stability.
Lope
z-Bastida
etal.(2017)[47]
Childrenwith
spinal
muscularatroph
yChildrenwith
Type
IIspinalmuscular
atroph
ywerefoun
dto
have
alower
NA
NA
NA
NA
Bray et al. Health Economics Review (2020) 10:9 Page 24 of 38
Table
6Summaryof
PBM
perfo
rmance
forinclud
edstud
ies(Con
tinued)
Stud
yreference
Con
ditio
n(s)of
interest
Know
n-grou
panalyses
Con
struct
validity:
comparin
gou
tcom
esCon
struct
validity:
comparin
gPBMs
Con
struct
validity:
comparin
grespon
dents
Respon
sivene
ss
averageutility
score
(−0.012)
than
the
combine
daverageof
childrenwith
alltypes
ofspinalmuscular
atroph
y(0.158).
Statisticalsign
ificance
was
notrepo
rted
.
Morrow
etal.
(2011)
[48]
Childrenwith
chronic
cond
ition
s(re
levant
cond
ition
:cereb
ralp
alsy)
NA
NA
NA
Agreemen
tbe
tween
respon
dentswas
assessed
usingCoh
en’skapp
acoefficient.M
oderate
agreem
entwas
foun
dbe
tweenparentsof
childrenwith
cerebral
palsyanddo
ctorsforthe
HUI2dimen
sion
sof
sensation(63.6%
agreem
ent;Kapp
a0.41),
cogn
ition
(70%
;Kappa
0.56)andself-care
(100%;
Kapp
a1).O
nlytheHUI3
ambu
latio
ndimen
sion
(63.6%
;Kappa
0.46)de
mon
stratedmod
erate
agreem
ent.Allothe
rdi
men
sion
sexhibitedslight
orfairagreem
ent
NA
Penn
eret
al.
(2013)
[49]
Childrenandadolescents
with
cerebralpalsy
Asign
ificant
negative
correlationwas
foun
dbe
tweentheHUI3pain
dimen
sion
andGMFC
Slevel(r=
0.36;p
<0.001).
NA
Goo
dcorrelationwas
foun
dbe
tweenchild
self-
repo
rted
pain
(Won
g-Bakerscale)
andproxyre
ported
pain
(HUI3pain
dimen
sion
)(Goo
dman
andKruskall’sy=0.57;
p<0.001).U
sing
the
HUI3pain
dimen
sion
,ph
ysicians
iden
tifiedpain
in4.4%
ofcaseswhe
reparent
proxiesdidno
tiden
tifypain.In17.4%
ofcases,parent
proxies
iden
tifiedpain
whe
nph
ysicians
didno
t.Statisticalsign
ificance
notrepo
rted
.
NA
PerezSousaet
al.
(2017)
[50]
Childrenandadolescents
with
cerebralpalsy
NA
NA
NA
Child/fathe
ragreem
ent
was
poor
forallEQ-5D-Y
NA
Bray et al. Health Economics Review (2020) 10:9 Page 25 of 38
Table
6Summaryof
PBM
perfo
rmance
forinclud
edstud
ies(Con
tinued)
Stud
yreference
Con
ditio
n(s)of
interest
Know
n-grou
panalyses
Con
struct
validity:
comparin
gou
tcom
esCon
struct
validity:
comparin
gPBMs
Con
struct
validity:
comparin
grespon
dents
Respon
sivene
ss
dimen
sion
s(Kappa
rang
e0.016–0.067;no
n-sign
ificant).Child/m
othe
ragreem
entbe
tweendi
men
sion
swas
mostly
poor
(Kappa
rang
e0.057–
0.389;no
n-sign
ificant),
however
forthemob
ility
dimen
sion
sagreem
ent
was
good
(Kappa
0.713;
p=0.000)
andforthe
usualactivities
dimen
sion
agreem
entwas
mod
erate
(Kappa
0.436;p=0.000).
Mothe
rsandfather
both
tend
edto
repo
rtfewer
prob
lems(byproxy)than
thechild.
Petrou
andKu
pek
(2009)
[51]
Childrenwith
childho
odcond
ition
s(re
levant
cond
ition
s:microceph
aly,
cerebralpalsy,spinal
muscularatroph
y,musculardystroph
y,spina
bifid
a)
Microceph
aly:
Adjusteddisutility
from
perfe
cthe
alth
was
estim
ated
tobe
−0.820(95%
CI−
0.670to
−0.970).
Adjusteddisutility
from
childho
odno
rmswas
estim
ated
tobe
−0.745(95%
CI
−0.598to
−0.899).
Cereb
ralp
alsy:
Adjusteddisutility
from
perfe
cthe
alth
was
estim
ated
tobe
−0.726(95%
CI−
0.607to
−0.846).
Adjusteddisutility
from
childho
odno
rmswas
estim
ated
tobe
−0.652(95%
CI
−0.536to
−0.775).
Musculardystroph
yandspinalmuscular
atroph
y:Adjusted
disutility
from
perfe
cthe
alth
was
estim
ated
tobe
−0.616(95%
CI
−0.471to
−0.761).
Adjusteddisutility
from
childho
od
NA
NA
NA
NA
Bray et al. Health Economics Review (2020) 10:9 Page 26 of 38
Table
6Summaryof
PBM
perfo
rmance
forinclud
edstud
ies(Con
tinued)
Stud
yreference
Con
ditio
n(s)of
interest
Know
n-grou
panalyses
Con
struct
validity:
comparin
gou
tcom
esCon
struct
validity:
comparin
gPBMs
Con
struct
validity:
comparin
grespon
dents
Respon
sivene
ss
norm
swas
estim
ated
tobe
−0.541(95%
CI
−0.400to
−0.690).
Spinabifid
a:Adjusted
disutility
from
perfe
cthe
alth
was
estim
ated
tobe
−0.552(95%
CI
−0.404to
−0.701).
Adjusteddisutility
from
childho
odno
rmswas
estim
ated
tobe
−0.478(95%
CI
−0.333to
−0.630).
Rocque
etal.
(2015)
[52]
Childrenandadolescents
with
spinabifid
aDiagn
osisandtype
ofspinabifid
ahada
sign
ificant
effect
onoverallu
tility
and
certaindimen
sion
scores.O
verallutility
was
foun
dto
besign
ificantlylower
(p<0.001)
inchildren
andadolescentswith
myelomen
ingo
cele
comparedto
individu
alswith
closed
dysraphism
.TheHUI3am
bulatio
ndimen
sion
was
sign
ificantlylower
(p<0.001)
inindividu
alswith
open
myelomen
ingo
cele
comparedto
individu
alswith
closed
neuraltube
defects.TheHUI3
cogn
ition
dimen
sion
was
sign
ificantly
lower
(p=0.039)
inindividu
alswith
open
myelomen
ingo
cele
comparedto
individu
alswith
closed
neural
tube
defects
Utility
andam
bulatio
nscores
were
sign
ificantly
NA
NA
NA
NA
Bray et al. Health Economics Review (2020) 10:9 Page 27 of 38
Table
6Summaryof
PBM
perfo
rmance
forinclud
edstud
ies(Con
tinued)
Stud
yreference
Con
ditio
n(s)of
interest
Know
n-grou
panalyses
Con
struct
validity:
comparin
gou
tcom
esCon
struct
validity:
comparin
gPBMs
Con
struct
validity:
comparin
grespon
dents
Respon
sivene
ss
associated
with
bowel/bladd
ercontinen
ce(p
<0.05).
Thosewith
bowel
continen
cehad
high
eraverageutility
scores.
Utility
scores
were
sign
ificantlylower
for
patientswith
ahistory
ofreceivingshun
tand/or
CM-II
decom
pression
interven
tions
(p<0.005).The
dimen
sion
sof
cogn
ition
(p=0.17)
andam
bulatio
n(p=
0.002)
werefoun
dto
besign
ificantlylower
forindividu
alswith
ahistoryof
shun
ting,
whilethedimen
sion
sof
speech
(p=0.01)
andcogn
ition
(p=
0.005)
weresign
ificantlylower
for
individu
alswith
ahistoryof
CM-II
decompression
.The
results
remaine
dsign
ificant
whe
ncontrolledforage;
shun
tstatus
accoun
tedfor10.5%
ofvariabilityin
overall
utility
scores.Shu
ntrevision
status
weakly
correlated
with
utility
score(r=−0.197;p=
0.048)
anddimen
sion
scores
forvision
,speech,ambu
latio
n(coe
fficien
tsno
tgiven;pvalues
rang
edfro
m0.01
to0.05).
Rosenb
aum
etal.
(2007)
[53]
Ado
lescen
tswith
cerebral
palsy
GMFC
Slevelw
asfoun
dto
have
asign
ificant
impact
on
Astrong
negative
correlationwas
foun
dbe
tweenutility
scoreand
NA
NA
NA
Bray et al. Health Economics Review (2020) 10:9 Page 28 of 38
Table
6Summaryof
PBM
perfo
rmance
forinclud
edstud
ies(Con
tinued)
Stud
yreference
Con
ditio
n(s)of
interest
Know
n-grou
panalyses
Con
struct
validity:
comparin
gou
tcom
esCon
struct
validity:
comparin
gPBMs
Con
struct
validity:
comparin
grespon
dents
Respon
sivene
ss
utility
score(p
<0.01);
averageutility
scores
decreasedsteadilyas
GMFC
Slevel
increased.
Post-hoc
Bonferon
icorrection
confirm
edsign
ificant
differences
inmean
overallu
tility
scores
betw
eenallG
MFC
Slevels(p
values
not
repo
rted
),except
betw
eenlevelsIIandIII
(p=0.82).HUI3di
men
sion
levelswere
also
sign
ificantly
associated
with
GMFC
Slevelfor
all
dimen
sion
s(p
<0.05).
GMFC
Slevel(r=
−0.81;p
valueno
trepo
rted
)Ado
lescen
ts’abilityto
self-repo
rtusingtheQOL
Instrumen
twas
sign
ificantlyassociated
(p<0.01)with
parent-rep
ortedHUI3
ratin
gsof
speech
(tau-b
=0.52);cogn
ition
(tau-b
=0.50);de
xterity
(tau-b
=0.35);am
bulatio
n(tau-b
=0.31);vision
(tau-b
=0.22)and
hearing(tau-b
=0.19).
Utility
scorewas
sign
ificantly(p
valueno
trepo
rted
)but
weakly
correlated
with
scores
ontheQOLInstrumen
tfor
Being(r=0.37),
Belong
ing(r=0.17),
Becoming(r=0.20),and
Overallqu
ality
oflife
(r=0.28).Utility
scores
explaine
dbe
tween2.9%
(Belon
ging
)and14%
(Being
)of
variancein
QOLInstrumen
tscores.
Sims-Williams
etal.(2016)[54]
Childrenwith
spinabifid
aMod
eratecorrelationwas
foun
dbe
tweenchild-
repo
rted
(r=0.49)and
proxy-repo
rted
(r=0.38)
utility
scores
andchild-
repo
rted
VAS-theau
thorsde
scrib
ethisas
a‘poo
r’correlation,p
values
notrepo
rted
.
NA
Child
self-repo
rted
and
caregiverproxyutility
scores
werehigh
lycorrelated
(r=0.85;p
valueno
trepo
rted
).
NA
Slam
anet
al.
(2015)
[55]
Ado
lescen
tsandyoun
gadultswith
cerebralpalsy
NA
NA
NA
NA
Nosign
ificant
difference
betw
eenthecontroland
interven
tiongrou
psat
theen
dof
thetrial(p=
0.42).QALYsgained
equatedto
0.78
forthe
controlg
roup
,and
0.79
fortheinterven
tion
grou
p;theincrem
ental
differenceof
0.013was
Bray et al. Health Economics Review (2020) 10:9 Page 29 of 38
Table
6Summaryof
PBM
perfo
rmance
forinclud
edstud
ies(Con
tinued)
Stud
yreference
Con
ditio
n(s)of
interest
Know
n-grou
panalyses
Con
struct
validity:
comparin
gou
tcom
esCon
struct
validity:
comparin
gPBMs
Con
struct
validity:
comparin
grespon
dents
Respon
sivene
ss
notsign
ificant
(p=0.76)
Tilford
etal.
(2005)
[56]
Childrenwith
spinabifid
aATren
dtestacross
thecase
grou
prevealed
that
lesion
levelh
adasign
ificant
effect
(atp=0.01
level)on
average
utility
score,with
individu
alswith
thoraciclesion
sscoringthelowest
utility
scores
onaverage.
NA
NA
NA
NA
Usuba
etal.
(2014)
[57]
Ado
lescen
tsandadults
with
cerebralpalsy
NA
NA
Average
utility
scores
variedde
pend
ingon
PBM
used
;HUI3de
rived
utility
scores
werelower
atbaseline(0.29)
and8-
year
follow-up(0.29)
than
equivalent
AQoL
derived
utility
scores
(0.35and
0.32
respectively);statis
ticalsign
ificanceno
trepo
rted
.
NA
The‘older
adult’grou
pweremorelikelyto
repo
rtutility
deterio
ratio
nthan
the‘you
nger
adult’grou
p(HUI3:Relativerisk[RR]=
1.19;95%
CI0.66–2.15
/AQoL:RR=3.17;95%
CI
1.12–9.00).
Asign
ificant
interaction
was
foun
dbe
tweenage
grou
pandtim
eof
survey
usingtheAQoL
(p=
0.002);A
QoL
derived
utility
improved
over
time
inthe‘you
nger
adult’
grou
pbu
ttheop
posite
occurred
inthe‘older
adult’grou
p.Thedistrib
utionof
PBM
dimen
sion
scores
was
stableacross
the8-year
follow-up,
except
forthe
AQoL
socialrelatio
nships,
AQoL
inde
pend
entliving
andHUI3am
bulatio
ndi
men
sion
s,which
allim
proved
inthe‘you
nger
adult’grou
pbu
tde
terio
ratedin
‘older
adult’grou
p.
Vitaleet
al.
(2001)
[58]
Ado
lescen
tswith
orthop
aedicprob
lems
(relevant
cond
ition
:cerebralpalsy)
Therewas
foun
dto
beasign
ificant
difference(p
>0.05)
betw
eenthemean
utility
scores
of
NA
NA
NA
NA
Bray et al. Health Economics Review (2020) 10:9 Page 30 of 38
Table
6Summaryof
PBM
perfo
rmance
forinclud
edstud
ies(Con
tinued)
Stud
yreference
Con
ditio
n(s)of
interest
Know
n-grou
panalyses
Con
struct
validity:
comparin
gou
tcom
esCon
struct
validity:
comparin
gPBMs
Con
struct
validity:
comparin
grespon
dents
Respon
sivene
ss
childrenwith
cerebral
palsy(m
ean0.92)and
childrenwith
scoliosis
with
comorbidities
(mean0.725),b
utno
tchildrenwith
idiopathicscoliosis
(mean0.889).
Justificatio
nforthese
grou
ping
swas
based
onsamplesize,thu
sthereisno
explicit
explanationas
towhy
utility
differences
betw
eenthese
grou
pswou
ldbe
expe
cted
.
Walland
eret
al.
(2009)
[59]
Adu
ltstreatedforCTEVin
infancy
Maleparticipantshad
sign
ificantlyhigh
eraverageutility
score
than
thecomparable
norm
grou
p(p
=0.027).M
ale
participantsrepo
rted
sign
ificantlyless
mod
erate/extrem
eprob
lemswith
anxiety/de
pression
than
thecomparable
norm
grou
p(6.3%
vs21.2%;p
=0.007).
Femaleparticipants
hadworse
utility
onaveragethan
acomparableno
rmgrou
p,bu
tno
tsign
ificantly.Fem
ale
participantsrepo
rted
sign
ificantlymore
mod
erate/extrem
eprob
lemswith
mob
ility
(45%
vs12.2%;p
<0.001)
and
usualactivities
(35%
vs12.5%;p
=0.006)
than
acomparable
norm
grou
p.
NA
NA
NA
NA
Youn
get
al.
(2010)
[22]
Ado
lescen
tsandyoun
gadultswith
cerebralpalsy
Inbo
ththe‘you
th’
and‘adu
lt’grou
ps,
Line
arregression
analysis
revealed
that
GMFC
SUtility
scores
derived
from
HUI3wereon
Proxyutility
scores
were
gene
rally
lower
(by0.16)
NA
Bray et al. Health Economics Review (2020) 10:9 Page 31 of 38
Table
6Summaryof
PBM
perfo
rmance
forinclud
edstud
ies(Con
tinued)
Stud
yreference
Con
ditio
n(s)of
interest
Know
n-grou
panalyses
Con
struct
validity:
comparin
gou
tcom
esCon
struct
validity:
comparin
gPBMs
Con
struct
validity:
comparin
grespon
dents
Respon
sivene
ss
utility
scores
deterio
ratedsteadily
asGMFC
Slevel
increased.
Statistical
sign
ificanceno
trepo
rted
levelinchildho
odwas
themostim
portant
influen
ceon
utility
scores;
respon
siblefor53.2%
ofvariancein
HUI3utility
outcom
es(β=−0.205;
p<0.001)
and45.2%
ofvariancein
AQoL
utility
outcom
es(β=−0.148;
p<0.001).
TheSRHwas
foun
dto
bemod
eratelycorrelated
with
utility
scores
derived
from
theHUI3(r=0.41;
p<0.001)
andAQoL
(r=0.41;p
<0.001).
averagehigh
erthan
thosede
rived
from
AQoL
across
allG
MFC
Slevels;
however
strong
correlationwas
foun
dbe
tweentheHUI3and
AQoL
(r=0.87;p
<0.001).
whe
nadjusted
for
cogn
ition
,gen
eralhe
alth
andCPseverity.Statistical
sign
ificanceno
trepo
rted
.
Youn
get
al.
(2013)
[21]
Ado
lescen
tsandyoun
gadultswith
spinabifid
aUtility
scores
derived
from
both
HUI3and
AQoL
variedin
the
sameway
according
tolesion
level,with
thoraciclesion
sassociated
with
lowestutility.Linear
regression
analysis
revealed
that
the
mostim
portantsing
lefactor
contrib
utingto
utility
outcom
eswas
surgicallesion
level;
respon
siblefor40%
ofvariancein
HUI3
utility
scores
and18%
inAQoL
utility
scores.
Both
lesion
leveland
agewereim
portant
whe
ncombine
d,accoun
tingfor48%
variancein
HUI3utility
scores
and22%
variancein
AQoL
utility
scores.
TheSRHwas
foun
dto
bemod
erately/strong
lycorrelated
with
utility
scores
derived
from
the
HUI3(r=−0.45;p
<0.001)
andAQoL
(r=−
0.58;p
<0.001).The
HAQ
was
also
strong
lycorrelated
with
utility
scores
derived
from
the
HUI3(r=0.79;p
<0.001)
andAQoL
(r=0.70;p
<0.001).
Utility
scores
derived
from
theAQoL
were
high
erthan
from
the
HUI3across
alllesion
levelsub
-group
s;ho
wever
theAQoL
andHUI3
exhibitedstrong
correl
ation(r=0.73;p
<0.001).
Meanutility
scores
were
slightlyhigh
erin
theself-
repo
rted
grou
p(HUI3
mean+0.04;A
QoL
mean+0.03)ho
wever
thiswas
basedon
the
comparison
ofyouthand
adultdata
andtheregres
sion
mod
elsremaine
dun
change
d.
NA
Bray et al. Health Economics Review (2020) 10:9 Page 32 of 38
youth version (EQ-5D-Y); parents reported a lower fre-quency of problems on all EQ-5D-Y proxy dimensions,particularly fathers.
ResponsivenessFive studies allowed analysis of PBM responsiveness inCP [33, 35, 46, 55, 57]. Adolescents with a GMFCS levelof V exhibited the largest decreases in HUI3 dimensionlevels over time, compared to adolescents with GMFCSlevels of III and IV [33]. Christensen et al. [35] reporteda significant association between physicians’ primarypain aetiology and change in HUI3 pain status (p =0.001). Conversely, in two studies utility outcomes didnot change significantly over time; HUI3 utility out-comes were found to be stable over a 1 year period foradolescents with CP (G = 0.91) [46], likewise utility out-comes (derived from HUI3 and AQoL) did not signifi-cantly change over an 8 year follow-up period [57].Slaman et al. [55] utilised the SF-6D in a randomisedcontrolled trial, but did not find a significant differencebetween the control and intervention groups at the endof the trial (p = 0.42).
Spina bifidaKnown-group analysesThree studies reported known-group analyses in SB [51,52, 56], two of which found that clinical factors had asignificant impact on utility scores. Petrou and Kupek[51] estimated that the adjusted HUI3 disutility of child-hood SB from perfect health was − 0.55 (95% CI − 0.40to − 0.70), and − 0.48 (95% CI − 0.33 to − 0.63) fromchildhood norms. A statistically significant effect of SBdiagnosis on HUI3 utility score (p < 0.001) was reported[52]; children diagnosed with myelomeningocele (meanutility score = 0.51) tended to have lower utility scorescompared to children with closed dysraphism (meanutility score = 0.77). Tilford et al. [56] reported that le-sion location in childhood SB had a significant impacton overall utility (p < 0.01); individuals with sacral le-sions had the highest overall mean utility (0.61; ±0.26).Two studies reported correlation between clinical fac-
tors and utility scores in SB [21, 38]. Anatomical myelo-meningocele level was found to have a significant effecton the HUI utility scores of children with myelomenin-gocele and shunted hydrocephalus (mean HUI score =0.03; 95% CI 0.01 to 0.05; p = 0.01) [38], with lower mye-lomeningocele level showing association with higherutility scores. A similar trend was reported by Younget al. [21], who found that the most important single fac-tor contributing to utility outcomes in SB was surgicallesion level, which was responsible for between 18%(AQoL) and 40% (HUI3) of variance in utility scores.
Convergent validity: comparing measuresTwo studies reported various levels of correlation betweenPBMs and other outcome measures in SB [21, 54]. Childself-reported HUI3 utility scores were not found to behighly correlated with VAS scores (r = 0.488) [54]. Like-wise, the SRH was only moderately correlated with utilityscores in SB (HUI3: r = − 0.45; p < 0.001 / AQoL: r = −0.58; p < 0.001) [21]. Only the HAQ was found to bestrongly correlated with utility score (HUI3: r = 0.79; p <0.001 / AQoL: r = 0.70; p < 0.001) [21].Young et al. [21] compared results from different
PBMs in SB, and found that mean utility scores on theAQoL were lower than mean utility scores on the HUI3for all sub-groups, despite strong correlation betweenthese measures (r = 0.73; p < 0.001).
Convergent validity: comparing respondentsSims-Williams et al. [54] reported that proxy and self-reported HUI3 utility scores were highly correlated forchildren with SB (r = 0.85; significance not reported).Young et al. [21] found that mean self-reported utilityscores were slightly higher than equivalent proxy scores(HUI3 mean + 0.04; AQoL mean + 0.03) however re-spondent type was not influential in their regressionanalysis.PBM responsiveness outcomes in SB were not found
in the literature.
Mixed patient groups and mobility impairmentsThis section includes results from studies which did notfocus on specific conditions, and where sub-group datacould not be examined separately. Relevant conditions/mobility impairments included muscular dystrophy,spinal muscular atrophy, CP, SB, orthopaedic lower limbdeformities, artrogryposis multiple congenital, achondro-plasia and hemiplegia.
Known-group analysesTwo studies reported known-group analyses in studies ofmixed patient groups [11, 51]. Petrou and Kupek [51]found that children with muscular dystrophy or spinalmuscular atrophy had a mean utility score of 0.39, equat-ing to an adjusted disutility from perfect health of − 0.62(95% CI − 0.471 to − 0.761), and an adjusted disutilityfrom childhood norms of − 0.54 (95% CI − 0.400 to −0.690). Burstrom et al. [11] reported that children with afunctional disability (64% congenital; see Table 4 for pa-tient characteristics) had significantly lower EQ-5D-Y di-mension scores than the general population (p < 0.001).
Convergent validity: comparing outcomesModerate correlation was found between the individualdimensions of the EQ-5D-Y and the dimensions of othermeasures (KIDSCREEN-27 and KIDSCREEN-10) and
Bray et al. Health Economics Review (2020) 10:9 Page 33 of 38
the life satisfaction ladder (LSL), when completed bychildren with functional disabilities [11] . The EQ-5D-Yanxiety/depression dimension exhibited significant cor-relation with the KIDSCREEN-27 psychological well-being dimension (r = − 0.51; p = 0.001), theKIDSCREEN-27 physical well-being dimension (r = −0.53; p = 0.001), and the LSL (r = − 0.54; p < 0.001) [11].Large variance was observed between utility scores de-
rived from different PBMs for young wheelchair users;mean utility scores ranged from 0.24 (EQ-5D-Y) to 0.53(HUI2) for self-reporting children, and from 0.01 (EQ-5D-Y) to 0.49 (HUI2) for parent proxies [15].
Convergent validity: comparing respondentsA significant strong correlation was observed betweendyads of young wheelchair users (84.6% children withCP, see Table 4 for patient characteristics) and parentproxies for a number of utility measures: EQ-5D-Y (r =0.67; p = 0.026), HUI2 (r = 0.73; p = 0.005) and HUI3(r = 0.84; p < 0.001) [15]. Using Bland-Altman plots [60],sufficient agreement was observed between dyads for theHUI2 (CL = 0.22) and HUI3 (CL = 0.22), but not the EQ-5D-Y (CL = 1.04). A significant respondent type effectwas found for all measures, with child self-reported util-ity scores significantly higher for the EQ-5D-Y (p =0.012), HUI2 (p = 0.021) and HUI3 (p = 0.009) thanequivalent proxy measures [15].PBM responsiveness outcomes were not found in the
literature for this patient group.
Childhood hydrocephalus (mixed aetiology)Known-group analysesLindquist et al. [45] found that adults with a history ofhydrocephalus (with or without neuro-impairment) hadsignificantly lower 15D dimension scores, compared to acontrol group, in the dimensions of vision (p = 0.001),eating (p = 0.000), usual activities (p = 0.004) and mentalfunction (p = 0.000).
Convergent validity: comparing measuresFour studies examined the relationship between theHydrocephalus Outcome Questionnaire (HOQ) andPBMs [40–43], however only three of these studies per-formed relevant statistical analyses [40, 41, 43]. Kulkarni[41] reported strong correlation (0.81) and a strong lin-ear relationship between HUI2 utility score and HOQoutcomes for children with hydrocephalus. Simple andcomplex linear regression models both accounted for alarge proportion of HUI2 variability (adjusted R2 = 0.66and 0.80 respectively). Similarly, a strong correlation wasfound between HUI2 utility score and the HOQ scoresfor overall health (r = 0.81), physical health (r = 0.88),social-emotional (r = 0.56) and cognitive (r = 0.57) [40].Furthermore, a significant positive correlation was
exhibited between self-reported scores on the child-completed version of the HOQ (cHOQ) and proxy-reported utility scores on the HUI3 (r = 0.60; p < 0.001)[43].PBM responsiveness outcomes in childhood hydro-
cephalus were not found in the literature.
Other conditions and mobility impairmentsOnly known-group analyses were reported for the fol-lowing conditions associated with mobility impairment:
Muscular dystrophyLandfeldt et al. [44] reported that ambulatory status andage were significantly associated with HUI utility scoresin muscular dystrophy (HUI version not stated; p <0.001); young ambulators (5–7 years old) had the highestutility scores on average (0.75), whilst older non-ambulators (≥16 years old) had the lowest utility scoreson average (0.15).
Spinal muscular atrophyLopez-Bastida et al. [47] reported that children with TypeII spinal muscular atrophy tended to have lower meanproxy utility scores (− 0.01; ±0.35) than the combinedaverage for all forms of spinal muscular atrophy (0.16; ±0.44), however statistical analysis was not undertaken.
Morquio A syndromeOne study found that for both adults and children withMorquio A syndrome, wheelchair use was significantlyassociated with lower utility scores [37]. Significant dif-ferences were reported in the adult group between non-wheelchair users and occasional wheelchair users (p =0.0115) and between occasional wheelchair users andfull-time wheelchair users (p = 0.0007). In the childgroup significant differences were reported betweennon-wheelchair users and full-time wheelchair users(p = 0.0018) and occasional wheelchair users and full-time wheelchair users (p = 0.0007).
Congenital clubfootWallander et al. [59] found that male adult patients withcongenital clubfoot (CCF) had significantly better overallutility scores than the male norm group (p = 0.027). Thefemale CCF group had a lower average utility score thanthe norm group, but not significantly (significance levelnot reported).
MicrocephalyPetrou and Kupek [51] found that children with micro-cephaly had a mean utility score of 0.14, equating to anadjusted disutility from perfect health of − 0.82 (95% CI− 0.67 to − 0.97), and an adjusted disutility from child-hood norms of − 0.75 (95% CI − 0.60 to − 0.90).
Bray et al. Health Economics Review (2020) 10:9 Page 34 of 38
DiscussionThe results from this systematic review demonstrate thatPBMs have been used in a relatively small number ofstudies relating to congenital mobility impairments. Inconditions such as CP and SB, increased clinical severityappears to be associated with decreased utility. This isparticularly evident using the HUI3, which was also themost commonly used PBM found in this review. In par-ticular, there appears to be a relationship between utilityoutcomes and GMFCS level in CP, and clinical factorssuch as lesion level in SB. In case-control studies, utilityoutcomes tended to be significantly lower in the casegroups, although it is worth noting that in one study themale case-group had significantly higher utility out-comes compared to the control [59].In order to demonstrate sufficient applicability and
sensitivity in a specific disease or disability, associationbetween PBMs and validated clinical/condition-specificoutcomes is of key importance. In this respect existingPBMs show weakness in various conditions associatedwith congenital mobility impairments; exhibiting gener-ally limited correlation with measures such as the QOLInstrument, VAS, SRH, KIDSCREEN-27, KIDSCREEN-10 and LSL. Only the GMFCS and HOQ/cHOQ ap-peared to be well correlated with PBMs across a numberof studies, although it is important to note that GMFCSis a classification system of gross motor function andnot an outcome measure.The results from this systematic review highlight im-
portant considerations for the use of PBMs in healthstates associated with congenital mobility impairment. Itis first of note that the use of PBMs has been dominatedby studies in CP, followed by SB. This is somewhat un-surprising given that these are two of the most prevalentcongenital disabilities which affect mobility. Secondly, itis important to acknowledge that only two studies fo-cussed on adults alone, and that both of these studieswere measuring utility outcomes in adults following con-ditions experienced in childhood [45, 59]. This focus onchildren and adolescents is again unsurprising, as manycongenital conditions can be life-limiting, thus examin-ing health outcomes at a young age becomes particularlyimportant. However, this also shows a lack of focus onthe health outcomes of adults who have life-long experi-ence of mobility impairment, and the potential ways inwhich their health outcomes could change over time orbe improved.There is some evidence to demonstrate that different
PBMs vary in their estimation of utility outcomes instates of impaired mobility. For instance, Bray et al. [15]found large variation between the EQ-5D-Y and HUI2/3utility scores for wheelchair users. Likewise, Usuba et al.[57] and Young et al. [21, 22] found that utility out-comes were generally higher when derived from the
AQoL than the HUI3 for individuals with CP, but viceversa for individuals with SB. This is despite the strongcorrelation between the AQoL and HUI3 in these popu-lations [21, 22]. Unfortunately, statistical differences be-tween these measures were not reported, but it is stillimportant to consider the implications that these differ-ences could have on subsequent QALY outcomes. Forinstance, if one PBM were to produce significantlyhigher utility scores than another, this would mean sig-nificantly different estimates of cost per QALYs and thusestimates of cost-effectiveness. At present there is lim-ited evidence to enable health economists and re-searchers to choose between these different PBMs foruse in congenital mobility impairments.Despite documented limitations [61], QALYs have be-
come increasingly influential in health policy as a meansto determine the cost-effectiveness of new treatments andservices, and therefore guide healthcare funding and pri-oritisation decisions. It is therefore imperative that thePBMs used to develop QALYs are accurate. Otherwise,the cost-effectiveness of certain interventions in certainpatient groups could be underestimated. This in turncould impact funding and prioritisation decisions. In thecontext of congenital mobility impairment, this could im-pact the provision of AMT and other mobility-enhancinginterventions, and subsequently impact patient outcomes.This is particularly important considering the ongoing is-sues of unmet need in assistive technology provision. TheWHO Priority Assistive Products List (APL) was launchedin 2016 to help tackle unmet need [62]. The APL contains50 priority assistive technology products, and was pro-duced through consultation with users, experts and otherkey stakeholders. Appropriate PBM data, and subsequentestimates of cost-effectiveness, could help to inform theAPL and ensure that the most effective and cost-effectiveAMTs are prioritised.Considering that the majority of evidence found in this
review related to children, the relationship between self-reported and proxy utility outcomes is particularly sig-nificant. The results from various studies in this reviewdemonstrate that proxy-reporting of utility is consist-ently different to that of self-reporters, including signifi-cant respondent-type effects and limited agreementbetween respondents. Interestingly, proxy respondentswere found to both underestimate and overestimate util-ity outcomes compared to self-reporters. It is thereforeimportant to prioritise outcome measurement from thepatient, although this can be challenging in populationswho may lack capacity, such as young children and indi-viduals with cognitive impairments.
Methodological implicationsDue to the underlying trade-off between quantity andquality of life in the calculation of utility values and
Bray et al. Health Economics Review (2020) 10:9 Page 35 of 38
subsequent QALYs, there is a tendency for lower valueto be placed on extending the length of life of peoplewith long-term disabilities [10, 63], as their quality of lifeis routinely considered to be worse than that of an able-bodied person. Thus, when using the QALY frameworkto assess the outcomes of individuals with disabilities, itis difficult to achieve substantially higher quality of lifewhen compared to individuals without disabilities, rais-ing concerns about bias [10]. To some extent these is-sues could be a result of using generic PBMs to valuedisabled health states.One of the underlying issues of using PBMs in disabil-
ity is that the definition of HRQoL differs profoundly be-tween people with disabilities and the general public[64]. When asked to define HRQoL, young wheelchairusers focus on a number of concepts not explicitly mea-sured using generic PBMs, such as ability to adapt,achievement and independence [17]. The experience ofdisability also affects HRQoL perceptions. Mechanismsof adaptation, coping and adjustment can help individ-uals with disabilities to experience diminishing effects totheir HRQoL over time. These processes are also influ-enced by the onset of disability, as individuals with con-genital disabilities demonstrate better adaptation thanindividuals with acquired disabilities [2]. The evaluationof states of disability by non-disabled individuals maytherefore cause such states to have an exaggerated per-ceived impact on HRQoL and health status [65], particu-larly with regards to congenital disability.When assessing the desirability of hypothetical health
states, individuals focus on the transition from their ownhealth state to the hypothetical health state, thus generalpublic beliefs about the impact of disability do not al-ways reflect the lived experience [66, 67]. Focus on per-sonal transition means that processes such as adaptationare not accounted for, causing a discrepancy in howstates of disability impact HRQoL [9].An alternative approach is to use more sensitive
condition-specific measures to model utility values ongeneric PBMs. Although this approach is advocated byNICE [68] there are serious concerns about the validityof modelled utility values [69], as it cannot be assumedthat modelled utility values are representative of directlymeasured utility values [70]. Using modelled utility datato guide funding and resource allocation decisions istherefore controversial. For instance, Sidovar et al. [71]mapped the 12-Item Multiple Sclerosis Walking Scale(MSWS-12) onto the EQ-5D-3 L. While prediction esti-mates were relatively precise for patients with moder-ately impaired mobility, they were significantly lessaccurate for individuals with severe impairments.Neilson et al. [72] suggest supplementing PBMs with
additional questions relating to functional activitieswhich have a large impact on overall quality of life, such
as ‘sitting’ for AMT users. For instance, Persson et al.[73] added complimentary mobility and social relation-ship items to the EQ-5D-3 L to increase sensitivity indisabled populations. However, this approach assumesthat supplemental questions can be mapped on to thehealth state preference values of existing measures with-out impacting accuracy.
Limitations and challengesDefining the term congenital mobility impairment wasone of the key challenges of designing this systematic re-view. An NHS posture and mobility service manager wasconsulted to help construct the search terms, and anumber of preliminary searches were carried out to testsearch terms for both sensitivity and scope. A multitudeof conditions and disabilities can affect mobility frombirth or early infancy, thus we attempted to cover a widevariety of these in our search terms, but accept thatthere are likely to be conditions and disabilities whichwere missed. Given the search results, we are confidentthat we have captured the vast majority of relevant stud-ies relating to at least the most common forms of con-genital mobility impairment. One key issue wasconsidering whether to include studies relating to hydro-cephalus in this review. Although hydrocephalus is notalways associated with mobility impairment, it can causemovement issues and is commonly related to relevantconditions such as CP and SB. We therefore chose to in-clude studies related to childhood hydrocephalus.We chose specifically to focus on congenital mobility
impairment due to the significant differences in life sat-isfaction, self-identity and self-efficacy experienced bypeople with congenital disabilities compared to peoplewith acquired disabilities [2]. Further research couldcompare PBM performance in congenital and acquiredmobility impairments. Previous reviews have reportedmixed results regarding the validity and responsivenessof existing PBMs in the assessment of health states asso-ciated with acquired mobility impairments, such asrheumatoid arthritis [24] and multiple sclerosis [26].
ConclusionTo our knowledge, this is the first systematic review ofthe use of PBMs in congenital mobility impairment. Evi-dence suggests that existing generic PBMs exhibit im-portant issues relating to validity and responsiveness,and thus care must be taken when selecting a PBM asan outcome measure in this context. Condition or dis-ability specific approaches to utility measurement, suchas the mobility and quality of life (MobQoL) outcomemeasure [74], could improve the sensitivity and applic-ability of utility measurement in this context.
Bray et al. Health Economics Review (2020) 10:9 Page 36 of 38
AbbreviationsAMT: Assistive mobility technologies; APL: WHO Priority Assistive ProductsList; APPT: Adolescent Pediatric Pain Tool; AQoL: Assessment of Quality ofLife; BPI-SF: Brief Pain Inventory Short Form; CCF: Congenital clubfoot;cHOQ: Child-completed version of the Hydrocephalus OutcomeQuestionnaire; CI: Confidence interval; CP: Cerebral palsy; CRD: Centre forReviews and Dissemination; CTEV: Congenital talipes equinus varus;DARE: Database of Abstracts of Reviews of Effects; DCGM-37: DISABKIDSChronic Generic Module; DSM: DISABKIDS Smiley Measure; EQ-5D: EuroQoLFive-Dimension; EQ-5D-Y: EuroQoL Five-Dimension - Youth version; EQ-5D-3L: EuroQoL Five-Dimension - Three Level version; EQ-5D-5 L: EuroQoL Five-Dimension - Five Level version; GMFCS: Gross Motor Function ClassificationSystem; GMFM-66: Gross Motor Function Measure 66 Items; HAQ: HealthAssessment Questionnaire; HOQ: Hydrocephalus Outcome Questionnaire;HUI: Health Utilities Index; HRQoL: Health-related quality of life; HTA: HealthTechnology Assessment; LSL: Life Satisfaction Ladder; MeSH: Medical SubjectHeadings; MSWS-12: 12-Item Multiple Sclerosis Walking Scale; NHS: NationalHealth Service; NHS EED: NHS Economic Evaluation Database; NICE: NationalInstitute for Health and Care Excellence; PBM: Preference-based measure;PedsQL: Pediatric Quality of Life Inventory; QALY: Quality-adjusted life year;QOL Instrument: Quality of Life Instrument for People with DevelopmentalDisabilities; QWB: Quality of Well-Being scale; RR: Relative risk;SAROMM: Spinal Alignment and Range of Motion Measure; SB: Spina bifida;SD: Standard deviation; SDQ: Strength and Difficulties Questionnaire; SF-36: 36-Item Short Form Health Survey; SF-6D: Short-Form Six-Dimension;SRH: Self-reported health; UN: United Nations; VAS: Visual analogue scale;WeeFIM: Functional Independence Measure for Children; WHO: World HealthOrganization; WRAT: Wide Range Achievement Test
AcknowledgementsThe author’s would like to acknowledge the support of Dr. Hareth Al-Janabi,Professor Joanna Coast, Professor Deborah Fitszimmons, Ms. Krys Jarvis andProfessor Katherine Payne for acting as advisors to the project.
Authors’ contributionsAll authors were responsible for designing the systematic review, developingthe original review protocol and editing the manuscript. NB devised thesystematic review, prepared the manuscript and synthesised extracted data.LHS conducted the review searches and led the study appraisal process. NBand LHS screened papers for eligibility and quality, and extracted evidencewith support from RTE. All author(s) read and approved the final manuscript.
FundingThis systematic review was funded by the Welsh Government throughHealth and Care Research Wales. The funding body had no role in thedesign, conduct or reporting of this systematic review.
Availability of data and materialsAs this is a systematic review, there was no primary data generated as partof this research. All of the data utilised in this review is available from thecited literature, and has been summarised in Tables 4, 5 and 6.
Ethics approval and consent to participateNot applicable, this is a systematic review.
Consent for publicationNot applicable, this is a systematic review.
Competing interestsThe authors declare that they have no competing interests.
Received: 4 December 2019 Accepted: 8 April 2020
References1. Department for Work and Pensions. Family Resources Survey 2016/2017.
London: Department for Work and Pensions; 2018.2. Bogart KR. The role of disability self-concept in adaptation to congenital or
acquired disability. Rehabil Psychol. 2014;59(1):107–15.3. Medicines and Healthcare Products Regulatory Agency. Assistive
technology: definition and safe use. London: Department of Health; 2017.
4. Tebbutt E, Brodmann R, Borg J, et al. Assistive products and the sustainabledevelopment goals (SDGs). Glob Health. 2016;12(1):79.
5. UN General Assembly. Convention on the rights of persons with disabilities.2006. Available at: https://www.un.org/development/desa/disabilities/convention-on-the-rights-of-persons-with-disabilities.html [Accessed 30March 2020].
6. World Health Organization. Improving access to assistive technology: reportby the secretariat. Geneva: WHO; 2016.
7. NHS Improving Quality. Right chair, right time, right now. London: NHSImproving Quality; 2014.
8. Ogden J. QALYs and their role in the NICE decision-making process.Prescriber. 2017;28(4):41–3.
9. Versteegh MM, Brouwer WB. Patient and general public preferences forhealth states: a call to reconsider current guidelines. Soc Sci Med. 2016;16:16–74.
10. Pettitt DA, Raza S, Naughton B, Roscoe A, Ramakrishnan A, et al. Thelimitations of QALY: a literature review. J Stem Cell Res Ther. 2016;6:334.
11. Burström K, Bartonek BEW, Sun S, Egmar AC. EQ-5D-Y as a health-relatedquality of life measure in children and adolescents with functionaldisability in Sweden: testing feasibility and validity. Acta Paediatr. 2014;103(4):426–35.
12. Hagberg L, Hermansson L, Fredriksson C, Pettersson I. Cost-effectiveness ofpowered mobility devices for elderly people with disability. Disabil RehabilAssist Technol. 2015;30:1–6.
13. Morris C, Janssens A, Allard A, Coon JT, Shilling V, Tomlinson R, et al.Informing the NHS Outcomes Framework: evaluating meaningful healthoutcomes for children with neurodisability using multiple methodsincluding systematic review, qualitative research, Delphi survey andconsensus meeting. Heal Serv Deliv Res. 2014;2:15.
14. Persson J, Husberg M. Can we rely on QALYs for assistive technologies?Technol Disabil. 2012;24(1):93–100.
15. Bray N, Noyes J, Harris N, Edwards RT. Measuring the health-related qualityof life of children with impaired mobility: examining correlation andagreement between children and parent proxies. BMC Res Notes. 2017;10(1):1–8.
16. Bartonek A, Saraste H, Danielsson D. Health-related quality of life andambulation in children with myelomeningocele in a Swedish population.Acta Paediatr. 2012;101:953–6.
17. Bray N, Noyes J, Harris N, Edwards R. Defining health-related quality of lifefor young wheelchair users: a qualitative health economics study. PLoS One.2017;12(6):1–20.
18. Scott DL, Khoshaba B, Choy EH, Kingsley GH. Limited correlation betweenthe health assessment questionnaire (HAQ) and EuroQol in rheumatoidarthritis: questionable validity of deriving quality adjusted life years fromHAQ. Ann Rheum Dis. 2007;66(11):1534–7.
19. Wolfe F, Hawley DJ. Measurement of the quality of life in rheumaticdisorders using the Euroqol. Br J Rheumatol. 1997;36:786–93.
20. Hulme C, Long AF, Kneafsey R, Reid G. Using the EQ-5D to assess health-related quality of life in older people. Age Ageing. 2004;33(5):504–7.
21. Young NL, Sheridan K, Burke TA, Mukherjee S, McCormick A. Healthoutcomes among youths and adults with Spina bifida. J Pediatr. 2013;162(5):993–8.
22. Young N, Rochon T, McCormick A, Law M, Wedge J, Fehlings D. The healthand quality of life outcomes among youth and Young adults with cerebralpalsy. Arch Phys Med Rehabil. 2010;91(1):143–8.
23. Andresen EM, Fouts BS, Romeis JC, Brownson CA. Performance of health-related quality-of-life instruments in a spinal cord injured population. ArchPhys Med Rehabil. 1999;80(8):877–84.
24. Harrison MJ, Davies LM, Bansback NJ, Ingram M, Anis AH, Symmons DP. Thevalidity and responsiveness of generic utility measures in rheumatoidarthritis: a review. J Rheumatol. 2008;35(4):592.
25. Mpundu-Kaambwa C, Chen G, Huynh E, Russo R, Ratcliffe J. A review ofpreference-based measures for the assessment of quality of life in childrenand adolescents with cerebral palsy. Qual Life Res. 2018;27:1781.
26. Kuspinar A, Mayo N. A review of the psychometric properties of genericutility measures in multiple sclerosis. Pharmacoeconomics. 2014;32:759–73.
27. Centre for Reviews and Dissemination. Systematic Reviews: CRD’s Guidancefor Undertaking Reviews in Health Care. York: University of York; 2009.
28. Tosh J, Brazier J, Evans P, Longworth L. A review of generic preference-based measures of health-related quality of life in visual disorders. ValueHealth. 2012;15(1):118–27.
Bray et al. Health Economics Review (2020) 10:9 Page 37 of 38
29. Papaioannou D, Brazier J, Parry G. How valid and responsive are generichealth status measures, such as EQ-5D and SF-36, in schizophrenia? Asystematic review. Value Health. 2011;14(6):907–20.
30. Fitzsimmons D, Gilbert J, Howse F, Young T, Arrarras JI, Brédart A, et al. Asystematic review of the use and validation of health-related quality of lifeinstruments in older cancer patients. Eur J Cancer. 2009;45(1):19–32.
31. Mokkink LB, Terwee CB, Patrick DL, Alonso J, Stratford PW, Knol DL, et al.The COSMIN study reached international consensus on taxonomy,terminology, and definitions of measurement properties for health-relatedpatient-reported outcomes. J Clin Epidemiol. 2010;63(7):737–45.
32. Juniper EF, Guyatt GH, Jaeshke R. How to develop and validate a newhealth-related quality of life instrument. In: quality of life andPharmacoeconomics in clinical trials. Philadelphia: Lippincott-RavenPublishers; 1996.
33. Bartlett DJ, Hanna SE, Avery L, Stevenson RD, Galuppi B. Correlates ofdecline in gross motor capacity in adolescents with cerebral palsy in grossmotor function classification system levels III to V: an exploratory study. DevMed Child Neurol. 2010;52(7):155–60.
34. Cavazza M, Kodra Y, Armeni P, De Santis M, López-Bastida J, Linertová R,et al. Social/economic costs and health-related quality of life in patientswith Duchenne muscular dystrophy in Europe. Eur J Health Econ. 2016;17(S1):19–29.
35. Christensen R, MacIntosh A, Switzer L, Fehlings D. Change in pain status inchildren with cerebral palsy. Dev Med Child Neurol. 2017;59(4):374–9.
36. Findlay B, Switzer L, Narayanan U, Chen S, Fehlings D. Investigating theimpact of pain, age, gross motor function classification system, and sex onhealth-related quality of life in children with cerebral palsy. Dev Med ChildNeurol. 2016;58(3):292–7.
37. Hendriksz CJ, Lavery C, Coker M, Ucar S, Jain M, Bell L, et al. Burden ofdisease in patients with Morquio a syndrome: results from an internationalpatient-reported outcomes survey. Orphanet J Rare Dis. 2014;9(1):32.
38. Karmur BS, Kulkarni AV. Medical and socioeconomic predictors of quality oflife in myelomeningocele patients with shunted hydrocephalus. Childs NervSyst. 2018;34(4):741–7.
39. Kennes J, Raina P, Care H. Health status of with cerebral palsy :information from a population-based sample. Dev Med Child Neurol.2002;44(2):240–7.
40. Kulkarni AV, Rabin D, Drake JM. An instrument to measure the health statusin children with hydrocephalus: the hydrocephalus outcome questionnaire.J Neurosurg Pediatr. 2004;101(2):134–40.
41. Kulkarni AV. Distribution-based and anchor-based approaches provideddifferent interpretability estimates for the hydrocephalus outcomequestionnaire. J Clin Epidemiol. 2006;59(2):176–84.
42. Kulkarni AV, Cochrane DD, McNeely PD, Shams I. Medical, social, andeconomic factors associated with health-related quality of life in Canadianchildren with hydrocephalus. J Pediatr. 2008;153(5):689–95.
43. Kulkarni AV, Cochrane DD, Mcneely PD, Shams I. Comparing children’s andparents’ perspectives of health outcome in paediatric hydrocephalus. DevMed Child Neurol. 2008;50(8):587–92.
44. Landfeldt E, Lindgren P, Bell CF, Guglieri M, Straub V, Lochmüller H, et al.Health-related quality of life in patients with Duchenne muscular dystrophy:a multinational, cross-sectional study. Dev Med Child Neurol. 2016;58(5):508–15.
45. Lindquist B, Fernell E, Persson E-K, Uvebrant P. Quality of life in adultstreated in infancy for hydrocephalus. Childs Nerv Syst. 2014;30(8):1413–8.
46. Livingston MH, Rosenbaum PL. Adolescents with cerebral palsy: stability inmeasurement of quality of life and health-related quality of life over 1 year.Dev Med Child Neurol. 2008;50(9):696–701.
47. López-Bastida J, Peña-Longobardo LM, Aranda-Reneo I, Tizzano E, Sefton M,Oliva-Moreno J. Social/economic costs and health-related quality of life inpatients with spinal muscular atrophy (SMA) in Spain. Orphanet J Rare Dis.2017;12(1):1–7.
48. Morrow AM, Hayen A, Quine S, Scheinberg A, Craig JC. A comparison ofdoctors’, parents’ and children’s reports of health states and health-relatedquality of life in children with chronic conditions. Child Care Health Dev.2012;38(2):186–95.
49. Penner M, Xie WY, Binepal N, Switzer L, Fehlings D. Characteristics of pain inchildren and youth with cerebral palsy. Pediatrics. 2013;132(2):e407–13.
50. Perez-Sousa MA, Sanchez-Toledo O, Fuerte NG. Parent-child discrepancy inthe assessment of health- related quality of life using the EQ-5D-Yquestionnaire. Arch Argent Pediatr. 2017;115:6.
51. Petrou S, Kupek E. Estimating preference-based health utilities index mark 3utility scores for childhood conditions in England and Scotland. Med DecisMak. 2009;29(3):291–303.
52. Rocque BG, Bishop ER, Scogin MA, Hopson BD, Arynchyna AA, Boddiford CJ,et al. Assessing health-related quality of life in children with spina bifida. JNeurosurg Pediatr. 2015;15:144–9.
53. Rosenbaum PL, Livingston MH, Palisano RJ, Galuppi BE, Russell DJ. Quality oflife and health-related quality of life of adolescents with cerebral palsy. DevMed Child Neurol. 2007;49(7):516–21.
54. Sims-Williams HJ, Sims-Williams HP, Mbabazi Kabachelor E, Warf BC. Qualityof life among children with spina bifida in Uganda. Arch Dis Child. 2017;102(11):1057–61.
55. Slaman J, Van Den Berg-Emons R, Tan SS, Russchen H, Van Meeteren J,Stam H, et al. Cost-utility of a lifestyle intervention in adolescents andyoung adults with spastic cerebral palsy. J Rehabil Med. 2015;47(4):338–45.
56. Tilford JM, Grosse SD, Robbins JM, Pyne JM, Cleves MA, Hobbs CA. Healthstate preference scores of children with spina bifida and their caregivers.Qual Life Res. 2005;14(4):1087–98.
57. Usuba K, Oddson B, Gauthier A, Young NL. Changes in gross motor functionand health-related quality of life in adults with cerebral palsy: an 8-yearfollow-up study. Arch Phys Med Rehabil. 2014;95(11):2071–7.
58. Vitale MG, Levy DE, Johnson MG, Gelijns AC, Moskowitz AJ, Roye BP, et al.Assessment of quality of life in adolescent patients with orthopaedic problems:are adult measures appropriate? J Pediatr Orthop. 2001;21(5):622–8.
59. Wallander H, Larsson S, Bjönness T, Hansson G. Patient-reported outcome at62 to 67 years of age in 83 patients treated for congenital clubfoot. J BoneJoint Surg (Br). 2009;91–B(10):1316–21.
60. Bland MJ, Altman D. Statistical methods for assessing agreement betweentwo methods of clinical measurement. Lancet. 1986;327(8476):307–10.
61. Whitehead SJ, Ali S. Health outcomes in economic evaluation: the QALYand utilities. Br Med Bull. 2010;96(1):5–21.
62. World Health Organization. Priority assistive products list. Geneva: WHO; 2016.63. Singer P, McKie J, Kuhse H, Richardson J. Double jeopardy and the use of
QALYs in health care allocation. J Med Ethics. 1995;21(3):144–50.64. Andresen EM, Meyers AR. Health-related quality of life outcomes measures.
Arch Phys Med Rehabil. 2000;81(12 SUPPL. 2):30–45.65. Edejer T, Tessa R, Blatussen R, Adam T, Hutubessy R, Acharya A, et al.
Methods for generalized cost- effectiveness analysis, in making choices inhealth: WHO guide to cost-effectiveness analysis. Geneva: World HealthOrganization; 2004.
66. Dolan P, Kahneman D. Interpretations of utility and their implications forthe valuation of health. Econ J. 2008;118:215–34.
67. Ubel PA, Loewenstein G, Jepson C. Whose quality of life? A commentaryexploring discrepancies between health state evaluations of patients andthe general public. Qual Life Res. 2003;12(6):599–607.
68. National Institute for Health and Clinical Excellence. Guide to the methodsof technology appraisal. London; 2013.
69. Round J, Hawton A. Statistical alchemy: conceptual validity and mapping togenerate health state utility values. Pharmacoeconomics. 2017;1(4):1–7.
70. McCabe C, Edlin R, Meads D, Brown C, Kharroubi S. Constructing indirectutility models: some observations on the principles and practice of mappingto obtain health state utilities. Pharmacoeconomics. 2013;31(8):635–41.
71. Sidovar MF, Limone BL, Coleman CI. Mapping of multiple sclerosis walkingscale (MSWS-12) to five-dimension EuroQol (EQ-5D) health outcomes: anindependent validation in a randomized control cohort. Patient RelatOutcome Meas. 2016;3(7):13–8.
72. Neilson A, Hogg J, Malek M, Rowley D. Impact of surgical and orthoticintervention on the quality of life of people with profound intellectual andmultiple disabilities and their carers. J Appl Res Intellect Disabil. 2000;13:216–38.
73. Persson J, Andrich R, Van Beekum T, Brodin H, Lorentsen O, Wessels R, et al.Preference based assessment of the quality of life of disabled persons.Technol Disabil. 2002;14(3):119–24.
74. Bray N, Spencer LH, Tuersley L, Edwards RT. Development of the MobQoLpatient reported outcome measure for mobility-related quality of life.Disability and Rehabilitation. 2020; https://doi.org/10.1080/09638288.2020.1741701.
Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.
Bray et al. Health Economics Review (2020) 10:9 Page 38 of 38
Top Related