Exploring the Factors that Drive Clinical Negligence Claims: Stated Preferences of Those Who Have Experienced Unintended Harm.

BACKGROUND: Better understanding of the factors that influence patients to make a financial claim for compensation is required to inform policy decisions. This study aimed to assess the relative importance of factors that influence those who have experienced a patient safety incident (PSI) to make a claim for compensation. METHOD: Participants completed an online discrete choice experiment (DCE) involving 10 single profile tasks where they chose whether or not to file a claim. DCE data were modelled using logistic, mixed logit and latent class regressions; scenario analyses, external validity, and willingness to accept were also conducted. RESULTS: A total of 1029 participants in the United Kingdom responded to the survey. An appropriate apology and a satisfactory investigation reduced the likelihood of claiming. Respondents were more likely to claim if they could hold those responsible accountable, if the process was simple and straightforward, if the compensation amount was higher, if the likelihood of compensation was high or uncertain, if the time to receive a decision was quicker, and if they used the government compensation scheme. Men are more likely to claim for low impact PSIs. DISCUSSION AND CONCLUSIONS: The actions taken by the health service after a PSI, and people's perceptions about the probability of success and the size of potential reward, can influence whether a claim is made. Results show the importance of giving an appropriate apology and conducting a satisfactory investigation. This stresses the importance around how patients are treated after a PSI in influencing the clinical negligence claims that are made.

Does EQ-5D Tell the Whole Story? Statistical Methods for Comparing the Thematic Coverage of Clinical and Generic Outcome Measures, With Application to Breast Cancer.

OBJECTIVES: This study aimed to develop the following: (1) methods for assessing claims in any specific application that a generic outcome measure, such as EQ-5D is deficient in its coverage of 1 or more specified domains, and (2) a simple method of judging whether any such deficiency is likely to be quantitatively important enough to call into question evaluations based on the generic instrument. Also to demonstrate the applicability of the methods in the important area of breast cancer. METHODS: The methodology requires a data set with observations from a generic instrument (eg, EQ-5D) and also a more comprehensive clinical instrument (eg, FACT-B [Functional Assessment of Cancer Therapy - Breast]). A standardized 3-component statistical analysis is proposed for investigating the claim that the generic measure inadequately captures some specified dimension covered by the latter instrument. A theoretically based upper bound on the bias induced by deficient coverage is derived based on the assumption that the designers of the (k-dimensional) generic instrument did succeed in identifying the k most important domains. RESULTS: Data from the MARIANNE breast cancer trial were analyzed and results suggested that impacts on personal appearance and relationships may be inadequately represented by EQ-5D. Nevertheless, the indications are that the bias in quality-adjusted life-year differences from deficient coverage by EQ-5D is likely to be modest. CONCLUSIONS: The methodology offers a systematic approach to determining whether there is clear evidence consistent with any claim that a generic outcome measure such as EQ-5D misses an important specific domain. The approach is readily implementable using data sets that are available in many randomized controlled trials.

Estimating the Relationship Between EQ-5D-5L and EQ-5D-3L: Results from a UK Population Study.

OBJECTIVES: The aim of this study was to estimate the relationship between EQ-5D-3L and EQ-5D-5L, in both directions, using a single model. METHODS: An online survey containing both variants of EQ-5D, with randomised ordering, was administered to a large UK sample in 2020. A joint statistical model of the ten EQ-5D responses (five at 5L, five at 3L), using a multi-equation ordinal regression framework was estimated. The joint model ensures mappings in either direction are fully consistent with the information in the sample and satisfy Bayes' rule. Three extensions enhance model flexibility: a copula specification allows differing degrees of correlation between the 3L and 5L responses at the upper and lower extremes of health; a normal mixture residual distribution gives flexibility in the distributional form of responses; and a common factor captures correlations in responses across the five dimensions. RESULTS: Almost 50,000 responses were received. Thirty-five percent of respondents reported an existing medical condition. Ninety percent of possible 3L and 43% of possible 5L health states were observed. The preferred model specification includes age, sex and the responses to the EQ-5D instrument. Close alignment to the observed data was observed both in within-sample and out-of-sample comparisons. CONCLUSION: The results from this study provide a means of translating evidence to or from EQ-5D-3L to or from 5L based on a large-scale UK population survey with randomised ordering. Mapping can be performed either using descriptive system responses, individual utility scores or summary statistics.

Convergent validity of EQ-5D with core outcomes in dementia: a systematic review.

OBJECTIVES: To explore through a systematic review, the convergent validity of EQ-5D (EQ-5D-3L and EQ-5D-5L (total score and dimensions)) with core outcomes in dementia and investigate how this may be impacted by rater-type; with the aim of informing researchers when choosing measures to use in dementia trials. METHODS: To identify articles relevant to the convergent validity of EQ-5D with core dementia outcomes, three databases were electronically searched to September 2022. Studies were considered eligible for inclusion within the review if they included individual level data from people with dementia of any type, collected self and/or proxy reported EQ-5D and collected at least one core dementia outcome measure. Relevant data such as study sample size, stage of dementia and administration of EQ-5D was extracted, and a narrative synthesis was adopted. RESULTS: The search strategy retrieved 271 unique records, of which 30 met the inclusion criteria for the review. Twelve different core outcome measures were used to capture dementia outcomes: cognition, function, and behaviour/mood across the studies. Most studies used EQ-5D-3L (n = 27). Evidence related to the relationship between EQ-5D and measures of function and behaviour/mood was the most robust, with unanimous directions of associations, and more statistically significant findings. EQ-5D dimensions exhibited associations with corresponding clinical outcomes, whereby relationships were stronger with proxy-EQ-5D (than self-report). CONCLUSION: Measuring health-rated quality of life in dementia populations is a complex issue, particularly when considering balancing the challenges associated with both self and proxy report. Published evidence indicates that EQ-5D shows evidence of convergent validity with the key dementia outcomes, therefore capturing these relevant dementia outcomes. The degree of associations with clinical measures was stronger when considering proxy-reported EQ-5D and differed by EQ-5D dimension type. This review has revealed that, despite the limited targeted psychometric evidence pool and reliance on clinical and observational studies, EQ-5D exhibits convergent validity with other dementia outcome measures.

Assessing the psychometric performance of EQ-5D-5L in dementia: a systematic review.

BACKGROUND: EQ-5D is widely used for valuing changes in quality of life for economic evaluation of interventions for people with dementia. There are concerns about EQ-5D-3L in terms of content validity, poor inter-rater agreement and reliability in the presence of cognitive impairment, but there is also evidence to support its use with this population. An evidence gap remains regarding the psychometric properties of EQ-5D-5L. OBJECTIVES: To report psychometric evidence around EQ-5D-5L in people with dementia. METHODS: A systematic review identified primary studies reporting psychometric properties of EQ-5D-5L in people with dementia. Searches were completed up to November 2020. Study selection, data extraction and quality assessment were undertaken independently by at least 2 researchers. RESULTS: Evidence was extracted from 20 articles from 14 unique studies covering a range of dementia severity. Evidence of known group validity from 5 of 7 studies indicated that EQ-5D-5L distinguishes severity of disease measured by cognitive impairment, depression, level of dependence and pain. Convergent validity (9 studies) showed statistically significant correlations of weak and moderate strengths, between EQ-5D-5L scores and scores on other key measures. Statistically significant change was observed in only one of 6 papers that allowed this property to be examined. All seven studies showed a lack of inter-rater reliability between self and proxy reports with the former reporting higher EQ-5D-5L scores than those provided by proxies. Five of ten studies found EQ-5D-5L to be acceptable, assessed by whether the measure could be completed by the PwD and/or by the amount of missing data. As dementia severity increased, the feasibility of self-completing EQ-5D-5L decreased. Three papers reported on ceiling effects, two found some evidence in support of ceiling effects, and one did not. CONCLUSIONS: EQ-5D-5L seems to capture the health of people with dementia on the basis of known-group validity and convergent validity, but evidence is inconclusive regarding the responsiveness of EQ-5D-5L. As disease progresses, the ability to self-complete EQ-5D-5L is diminished.

Should We Consider Including a Value for "Hope" as an Additional Benefit Within Health Technology Assessment?

Health technology assessment (HTA) typically uses average health-related quality of life gain as its main measure of benefit used in economic evaluation. Nevertheless, there have been calls to consider novel aspects of benefit including the "value of hope," defined as a patients' potential preferences for a wider distribution of treatment benefit with a positive skew, in the hope that they will be one of the lucky ones. The value of hope may also derive from feeling hopeful as a positive mental state, which may be missing from current measures of health-related quality of life. The value attributed to feeling hopeful could be related to, or additional to, the value derived from possible risk-seeking preferences. Here, we reflect upon the strength of the case for the inclusion of the "value of hope" taking a critical look at the commonly referenced evidence for including the "value of hope" as risk-seeking preferences. We also draw attention to other conceptions of hope-as an emotion, a cognitive process, or a combination of both-and reflect upon the potential of including these broader notions of hope into HTA. The case for the inclusion of the "value of hope" based on risk-seeking preferences is weak. We suggest research questions that could give further evidence on whether hope is an important missing value from HTA.

A Discrete Choice Experiment to Elicit General Population Preferences Around the Factors Influencing the Choice to Make Clinical Negligence Claims.

OBJECTIVES: This article determines public stated preferences around different factors that influence the choice to make clinical negligence claims against a national healthcare system. METHODS: A large online survey was conducted using a discrete choice experiment (DCE) with the UK general population (N = 1013). DCE tasks involved a single profile and participants chose whether to make a claim for compensation (yes/no) after one of 3 randomly allocated patient safety incident (PSI) "scenarios" of different severities (mild, moderate, severe). DCE attributes described the actions of the healthcare system after a PSI and characteristics of the clinical negligence claims process. The data were modeled separately for each scenario (mild, moderate, severe) using logistic regression. Marginal effects and the probability of making a claim in a baseline case were estimated. RESULTS: Probability of choosing to claim was reduced by receipt of an apology, investigation and prevention of recurrence of the PSI, and longer time until claim decision and increased by an easy and straightforward claims process and high chance of compensation and for the mild scenario higher compensation amounts. Marginal effects and baseline case probabilities differed by scenario severity. CONCLUSIONS: The results suggest the actions of the healthcare system after a PSI and characteristics of the claims process have a larger impact on the probability of making a claim for milder PSIs. For more severe PSIs, a larger probability of making a claim was observed, and the choice was less influenced by the actions of the healthcare system after the PSI and characteristics of the claims process.

Target Trial Emulation for Transparent and Robust Estimation of Treatment Effects for Health Technology Assessment Using Real-World Data: Opportunities and Challenges.

Evidence about the relative effects of new treatments is typically collected in randomised controlled trials (RCTs). In many instances, evidence from RCTs falls short of the needs of health technology assessment (HTA). For example, RCTs may not be able to capture longer-term treatment effects, or include all relevant comparators and outcomes required for HTA purposes. Information routinely collected about patients and the care they receive have been increasingly used to complement RCT evidence on treatment effects. However, such routine (or real-world) data are not collected for research purposes, so investigators have little control over the way patients are selected into the study or allocated to the different treatment groups, introducing biases for example due to selection or confounding. A promising approach to minimise common biases in non-randomised studies that use real-world data (RWD) is to apply design principles from RCTs. This approach, known as 'target trial emulation' (TTE), involves (1) developing the protocol with respect to core study design and analysis components of the hypothetical RCT that would answer the question of interest, and (2) applying this protocol to the RWD so that it mimics the data that would have been gathered for the RCT. By making the 'target trial' explicit, TTE helps avoid common design flaws and methodological pitfalls in the analysis of non-randomised studies, keeping each step transparent and accessible. It provides a coherent framework that embeds existing analytical methods to minimise confounding and helps identify potential limitations of RWD and the extent to which these affect the HTA decision. This paper provides a broad overview of TTE and discusses the opportunities and challenges of using this approach in HTA. We describe the basic principles of trial emulation, outline some areas where TTE using RWD can help complement RCT evidence in HTA, identify potential barriers to its adoption in the HTA setting and highlight some priorities for future work.

Exploring the Issues of Valuing Child and Adolescent Health States Using a Mixed Sample of Adolescents and Adults.

Preferences for child and adolescent health states used to generate health state utility values can be elicited from adults, young adults, adolescents, or combinations of these. This commentary paper provides a critical overview of issues and implications arising from valuing child and adolescent health states using a novel approach of a mixed sample of adolescents and adults. The commentary is informed by critical analysis of normative, ethical, practical and theoretical arguments in the health state valuation literature. Discussion focusses upon adolescent empowerment, understanding and psychosocial maturity; ethical concerns; elicitation tasks; perspective; and selection of sample proportions across adolescents and adults. It is argued that valuation of child and adolescent health states by both adolescents and adults could involve all participants completing the same preference elicitation task using the same perspective (e.g. time trade-off imagining they are living in the health state), and all preferences being modelled to generate a combined value set that reflects both adolescent and adult preferences. It is concluded that the valuation of child and adolescent health states by a mixed adolescent and adult sample appears feasible and has the advantage that it includes some of the population who can potentially experience the health states, thus enabling adolescents to express their views around matters that may affect them, and the population that are taxpayers and voters. However, both the relative proportion of adults and adolescents to include in a valuation sample and the elicitation technique require careful consideration.

Measuring quality of life of patients with axial spondyloarthritis for economic evaluation.

OBJECTIVES: To estimate the relationship between EQ5D (three levels, UK version) and the Ankylosing Spondylitis Disease Activity Score (ASDAS) for use in the economic evaluation of health technologies for people with axial spondyloarthritis (axSpA). To compare against the relationship with the Bath Ankylosing Spondylitis Disease Activity Index (BASDAI). METHODS: An electronic, prospective, Portuguese, nationwide, rheumatic disease register (Reuma.pt) provided data on 1140 patients (5483 observations) with a confirmed diagnosis of axSpA. We estimated models of EQ5D as a function of ASDAS, alone or in combination with measures of functional impairment, using bespoke mixture models which reflect the complex distributional features of EQ5D. The SPondyloArthritis Caught Early cohort provided data from 344 patients (1405 observations) in four European countries and was used for validation. A previously published model of BASDAI/Bath Ankylosing Spondylitis Functional Index (BASFI) was also used to generate predicted EQ5D scores and model performance compared. RESULTS: A non-linear relationship exists between EQ5D from ASDAS. The final model included ASDAS, ASDAS squared, age and age squared and demonstrated close fit in both datasets except where data were sparse for patients with very high levels of disease activity (ASDAS >4). This finding held in the validation dataset. Models that included BASFI improved model fit. The ASDAS based models fit the data marginally less well than models using BASDAI. CONCLUSIONS: Mapping models linking ASDAS to EQ5D allow results from clinical studies to be used in economic evaluation of health technologies with confidence. There is some loss of information compared with BASDAI but this has only a marginal impact.

A Cost-Effectiveness Analysis of Remdesivir for the Treatment of Hospitalized Patients With COVID-19 in England and Wales.

OBJECTIVES: COVID-19 is associated with significant morbidity and mortality. This study aims to synthesize evidence to assess the cost-effectiveness of remdesivir (RDV) for the treatment of hospitalized patients with COVID-19 in England and Wales. METHODS: A probabilistic cost-effectiveness analysis was conducted informed by 2 large trials and uses a partitioned survival approach to assess short- and long-term clinical consequences and costs associated with COVID-19 in a hypothetical cohort of hospitalized patients requiring supplemental oxygen at the start of treatment. Given that it is uncertain whether RDV reduces death, 2 analyses are presented, assuming RDV either reduces death or does not. Published sources were used for long-term clinical, quality of life, and cost parameters. RESULTS: Under the assumption that RDV reduces death, the incremental cost-effectiveness ratio for RDV is estimated at £11 881 per quality-adjusted life-year gained compared with standard of care (SoC) (probabilistic incremental cost-effectiveness ratio £12 400). The probability for RDV to be cost-effective is 74% at a willingness-to-pay threshold of £20 000 per quality-adjusted life-year gained. RDV was no longer cost-effective when the hazard ratio for overall survival compared with SoC was >0·915. CONCLUSIONS: Results from this study suggest that using RDV for the treatment of hospitalized patients with COVID-19 is likely to represent a cost-effective use of National Health Service resources at current willingness-to-pay threshold in England and Wales, only if it prevents death. Results needs to be interpreted caution as vaccination was introduced and the SoC and evidence available have also evolved considerably since the analysis is conducted.

Mapping the EORTC QLQ-C30 to EQ-5D-3L in patients with breast cancer.

BACKGROUND: The types of outcomes measured collected in clinical studies and those required for cost-effectiveness analysis often differ. Decision makers routinely use quality adjusted life years (QALYs) to compare the benefits and costs of treatments across different diseases and treatments using a common metric. QALYs can be calculated using preference-based measures (PBMs) such as EQ-5D-3L, but clinical studies often focus on objective clinician or laboratory measured outcomes and non-preference-based patient outcomes, such as QLQ-C30. We model the relationship between the generic, preference-based EQ-5D-3L and the cancer specific quality of life questionnaire, QLQ-C30 in patients with breast cancer. This will result in a mapping that allows users to convert QLQ-C30 scores into EQ-5D-3L scores for the purposes of cost-effectiveness analysis or economic evaluation. METHODS: We use data from a randomized trial of 602 patients with HER2-positive advanced breast cancer provided 3766 EQ-5D-3L observations. Direct mapping using adjusted, limited dependent variable mixture models (ALDVMM) is compared to a random effects linear regression and indirect mapping using seemingly unrelated ordered probit models. EQ-5D-3L was estimated as a function of the summary scales of the QLQ-C30 and other patient characteristics. RESULTS: A four component mixture model outperformed other models in terms of summary fit statistics. A close fit to the observed data was observed across the range of disease severity. Simulated data from the model closely aligned to the original data and showed that mapping did not significantly underestimate uncertainty. In the simulated data, 22.15% were equal to 1 compared to 21.93% in the original data. Variance was 0.0628 in the simulated data versus 0.0693 in the original data. The preferred mapping is provided in Excel and Stata files for the ease of users. CONCLUSION: A four component adjusted mixture model provides reliable, non-biased estimates of EQ-5D-3L from the QLQ-C30, to link clinical studies to economic evaluation of health technologies for breast cancer. This work adds to a growing body of literature demonstrating the appropriateness of mixture model based approaches in mapping.

An International Comparison of EQ-5D-5L and EQ-5D-3L for Use in Cost-Effectiveness Analysis.

OBJECTIVES: To estimate the impact of using EQ5D-5L (5L) compared with EQ5D-3L (3L) in cost-effectiveness analyses in 6 countries with 3L and 5L values: Germany, Japan, Korea, The Netherlands, China, and Spain. METHODS: Eight cost-effectiveness analyses based on clinical studies with 3L provided 11 pairwise comparisons. We estimated cost-effectiveness by applying the appropriate country values for 3L to observed responses. We re-estimated cost-effectiveness for each country by predicting the 5L tariff score for each respondent, for each country, using a previously published mapping method. We compared results in terms of impact on estimated incremental quality-adjusted life-year (QALY) gain and cost-effectiveness ratios. RESULTS: For most countries the impact of moving from 3L to 5L is to lower the incremental QALY gain in the majority of comparisons. The only exception to this was Japan, where 4 out of 11 cases (37%) saw lower QALYs gained when using 5L. The mean and median reductions in health gain, in those case studies where 5L does lead to lower health gain, are largest in The Netherlands (84% mean reduction, 41% median reduction), Germany (68% and 27%), and Spain (30% and 31%). For most countries, those studies where 5L leads to lower health gain see larger reductions than the gains in studies showing the opposite tendency. CONCLUSIONS: Overall, 3L and 5L are not interchangeable in these countries. Differences between results are large, but the direction of change can be unpredictable. These findings should prompt further investigation into the reasons for differences.

A Review of the Psychometric Performance of Selected Child and Adolescent Preference-Based Measures Used to Produce Utilities for Child and Adolescent Health.

OBJECTIVE: This review examined the psychometric performance of 4 generic child- and adolescent-specific preference-based measures that can be used to produce utilities for child and adolescent health. METHODS: A systematic search was undertaken to identify studies reporting the psychometric performance of the Child Health Utility (CHU9D), EQ-5D-Y (3L or 5L), and Health Utilities Index Mark 2 (HUI2) or Mark 3 (HUI3) in children and/or adolescents. Data were extracted to assess known-group validity, convergent validity, responsiveness, reliability, acceptability, and feasibility. Data were extracted separately for the dimensions and utility index where this was reported. RESULTS: The review included 76 studies (CHU9D n = 12, EQ-5D-Y-3L n = 20, HUI2 n = 26,HUI3 n = 43), which varied considerably across conditions and sample size. EQ-5D-Y-3L had the largest amount of evidence of good psychometric performance in proportion to the number of studies examining performance. The majority of the evidence related to EQ-5D-Y-3L was based on dimensions. CHU9D was assessed in fewer studies, but the majority of studies found evidence of good psychometric performance. Evidence for HUI2 and HUI3 was more mixed, but the studies were more limited in sample size and statistical power, which was likely to have affected performance. CONCLUSIONS: The heterogeneity of published studies means that the evidence is based on studies across a range of countries, populations and conditions, using different study designs, different languages, different value sets and different statistical techniques. Evidence for CHU9D in particular is based on a limited number of studies. The findings raise concerns about the comparability of self-report and proxy-report responses to generate utility values for children and adolescents.

The clinical effectiveness of intensive management in moderate established rheumatoid arthritis: The titrate trial.

OBJECTIVES: Many trials have shown that intensive management is effective in patients with early active rheumatoid arthritis (RA). But its benefits are unproven for the large number of RA patients seen in routine care who have established, moderately active RA and are already taking conventional synthetic disease modifying anti-rheumatic drugs (csDMARDs). The TITRATE trial studied whether these patients also benefit from intensive management and, in particular, achieve more remissions. METHODS: A 12-month multicentre individually randomised trial compared standard care with monthly intensive management appointments which was delivered by specially trained healthcare professionals and incorporated monthly clinical assessments, medication titration and psychosocial support. The primary outcome was 12-month remission assessed using the Disease Activity Score for 28 joints using ESR (DAS28-ESR). Secondary outcomes included fatigue, disability, harms and healthcare costs. Intention-to-treat multivariable logistic- and linear regression analyses compared treatment arms with multiple imputation used for missing data. RESULTS: 459 patients were screened and 335 were randomised (168 intensive management; 167 standard care); 303 (90%) patients provided 12-month outcomes. Intensive management increased DAS28-ESR 12-month remissions compared to standard care (32% vs 18%, p = 0.004). Intensive management also significantly increased remissions using a range of alternative remission criteria and increased patients with DAS28-ESR low disease activity scores. (48% vs 32%, p = 0.005). In addition it substantially reduced fatigue (mean difference -18; 95% CI: -24, -11, p<0.001). There was no evidence that serious adverse events (intensive management =15 vs standard care =11) or other adverse events (114 vs 151) significantly increase with intensive management. INTERPRETATION: The trial shows that intensive management incorporating psychosocial support delivered by specially trained healthcare professions is effective in moderately active established RA. More patients achieve remissions, there were greater improvements in fatigue, and there were no more harms.

A Review of the Methods Used to Generate Utility Values in NICE Technology Assessments for Children and Adolescents.

OBJECTIVE: This review summarizes and critically examines methods used to generate utilities for child and adolescent health states in previous National Institute for Health and Care Excellence (NICE) technology assessments (TA) and highly specialized technology (HST) evaluations. METHODS: We identified all NICE TA and HST evaluations in which the licensed indication for the technology included people younger than 18 and included in the review all evaluations using a cost-utility analysis. RESULTS: The review includes 40 TA and HST evaluations. Most assessments generated utility values with the EQ-5D scored using the adult version of the EQ-5D either exclusively (n = 16) or alongside other utility measures and direct elicitation methods of patient own utility (n = 17), although 7 did not use the EQ-5D. Eight assessments used both the EQ-5D child- and adolescent-specific preference-based measures: Health Utilities Index Mark 2 (n = 6), child- and adolescent-specific preference-based measure for atopic dermatitis (n = 1), and youth version of the EQ-5D (EQ-5D-Y) valued using the adult EQ-5D value set (n = 1) or generated using mapping and valued using the adult EQ-5D value set (n = 2). Some cost-utility analyses used age adjustment (utility subtractions, weights, and published mapping formulae) from the adult EQ-5D UK population norms to reflect the general population or disease-free health for children and adolescents (n = 9), and 1 assessment assumed full health (utility value of 1). CONCLUSION: The review found limited use of child and adolescent population-specific measures to generate health state utility values for children and adolescents in NICE technology assessments. Often assessments involve the use of an adult-specific measure to reflect the health of children.

Mapping clinical outcomes to generic preference-based outcome measures: development and comparison of methods.

BACKGROUND: Cost-effectiveness analysis using quality-adjusted life-years as the measure of health benefit is commonly used to aid decision-makers. Clinical studies often do not include preference-based measures that allow the calculation of quality-adjusted life-years, or the data are insufficient. 'Mapping' can bridge this evidence gap; it entails estimating the relationship between outcomes measured in clinical studies and the required preference-based measures using a different data set. However, many methods for mapping yield biased results, distorting cost-effectiveness estimates. OBJECTIVES: Develop existing and new methods for mapping; test their performance in case studies spanning different preference-based measures; and develop methods for mapping between preference-based measures. DATA SOURCES: Fifteen data sets for mapping from non-preference-based measures to preference-based measures for patients with head injury, breast cancer, asthma, heart disease, knee surgery and varicose veins were used. Four preference-based measures were covered: the EuroQoL-5 Dimensions, three-level version (n = 11), EuroQoL-5 Dimensions, five-level version (n = 2), Short Form questionnaire-6 Dimensions (n = 1) and Health Utility Index Mark 3 (n = 1). Sample sizes ranged from 852 to 136,327. For mapping between generic preference-based measures, data from FORWARD, the National Databank for Rheumatic Diseases (which includes the EuroQoL-5 Dimensions, three-level version, and EuroQoL-5 Dimensions, five-level version, in its 2011 wave), were used. MAIN METHODS DEVELOPED: Mixture-model-based approaches for direct mapping, in which the dependent variable is the health utility value, including adaptations of methods developed to model the EuroQoL-5 Dimensions, three-level version, and beta regression mixtures, were developed, as were indirect methods, in which responses to the descriptive systems are modelled, for consistent multidirectional mapping between preference-based measures. A highly flexible approach was designed, using copulas to specify the bivariate distribution of each pair of EuroQoL-5 Dimensions, three-level version, and EuroQoL-5 Dimensions, five-level version, responses. RESULTS: A range of criteria for assessing model performance is proposed. Theoretically, linear regression is inappropriate for mapping. Case studies confirm this. Flexible, direct mapping methods, based on different variants of mixture models with appropriate underlying distributions, perform very well for all preference-based measures. The precise form is important. Case studies show that a minimum of three components are required. Covariates representing disease severity are required as predictors of component membership. Beta-based mixtures perform similarly to the bespoke mixture approaches but necessitate detailed consideration of the number and location of probability masses. The flexible, bi-directional indirect approach performs well for testing differences between preference-based measures. LIMITATIONS: Case studies drew heavily on EuroQoL-5 Dimensions. Indirect methods could not be undertaken for several case studies because of a lack of coverage. These methods will often be unfeasible for preference-based measures with complex descriptive systems. CONCLUSIONS: Mapping requires appropriate methods to yield reliable results. Evidence shows that widely used methods such as linear regression are inappropriate. More flexible methods developed specifically for mapping show that close-fitting results can be achieved. Approaches based on mixture models are appropriate for all preference-based measures. Some features are universally required (such as the minimum number of components) but others must be assessed on a case-by-case basis (such as the location and number of probability mass points). FUTURE RESEARCH PRIORITIES: Further research is recommended on (1) the use of the monotonicity concept, (2) the mismatch of trial and mapping distributions and measurement error and (3) the development of indirect methods drawing on methods developed for mapping between preference-based measures. FUNDING: This project was funded by the National Institute for Health Research (NIHR) Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 24, No. 34. See the NIHR Journals Library website for further project information. This project was also funded by a Medical Research Council grant (MR/L022575/1).

Coherent decisions about which health services and treatments to provide rely on economic analysis to weigh potential health benefits against costs. For decisions to be consistent across the whole health service, benefits need to be counted in the same way for patients with different health problems. This is accomplished by using a unit of measurement for treatment outcomes called the quality-adjusted life-year. The best way to calculate quality-adjusted life-years is to ask patients taking part in clinical studies to fill in specially designed questionnaires to describe their health in a simple, standardised way. However, clinical trials often record patient outcomes in different ways, leaving economic analysts without the necessary information to calculate quality-adjusted life-years. A way to overcome this problem (known as 'statistical mapping') is to use the available clinical data to predict the responses that would have been made by trial participants to the standard questionnaire. This requires analysis of data from an additional study in which patients have provided both types of outcome data to construct a statistical 'mapping model'. Mapping is widely used in practice, but it is often based on simple mapping models that in some circumstances systematically mispredict and may consequently give a false picture of the real health benefits of treatments. This is important because it influences decisions about which treatments are available in the NHS; it has real effects on patients, clinicians, industry and the general public. Our objectives are to develop promising new statistical mapping models specifically designed for different clinical contexts and to compare them using patient data in different disease areas. We have also developed an approach for judging the outcome of a mapping study. We find that the new methods work better than existing methods in terms of their ability to fit the data and avoid systematic bias.

The EQ-5D-5L Value Set for England: Findings of a Quality Assurance Program.

OBJECTIVES: Five-level EuroQol 5-dimensional questionnaire (EQ-5D-5L) values for several countries now exist. Decision makers require confidence in the underlying data and statistical analyses before advocating their use. Independent quality assurance of the published English value set is reported here. METHODS: Data from 996 participants, and code to run published statistical models, were provided for inspection. The main elements of the study were 10 lead-time trade-off (TTO) experiments and 7 discrete choice experiments (DCEs). Data quality was examined and tested with respect to subsequent assumptions made in the statistical analysis. We examined the statistical analysis including model specification and estimation methods. RESULTS: The TTO experiments covered less than 3% of the possible 3125 5-level health states. There is strong evidence, both direct (self-reported) and indirect (poor data quality), that many participants found tasks difficult or did not engage effectively. Forty-seven percent of respondents valued more than 20% of states inconsistently, which is double the 3-level rate. The DCEs covered 12.5% of possible states and 0.01% of possible 2-state comparisons. The design precludes examination of inconsistent responses. Several aspects of the statistical model conflict with the data and underlying experimental design. The model is unidentified. The Bayesian approach relies on unjustified, informative priors. There is a clear failure to achieve convergence. CONCLUSION: Significant limitations are identified with the quality of the valuation data and the subsequent statistical analysis that underpin the English EQ-5D-5L value set. A new program of further development, including a new data collection initiative, should be considered to put the EQ-5D-5L on a sufficiently firm evidential basis for healthcare decision making.

MRI in the diagnosis of fetal developmental brain abnormalities: the MERIDIAN diagnostic accuracy study.

BACKGROUND: Ultrasonography has been the mainstay of antenatal screening programmes in the UK for many years. Technical factors and physical limitations may result in suboptimal images that can lead to incorrect diagnoses and inaccurate counselling and prognostic information being given to parents. Previous studies suggest that the addition of in utero magnetic resonance imaging (iuMRI) may improve diagnostic accuracy for fetal brain abnormalities. These studies have limitations, including a lack of an outcome reference diagnosis (ORD), which means that improvements could not be assessed accurately. OBJECTIVES: To assess the diagnostic impact, acceptability and cost consequence of iuMRI among fetuses with a suspected fetal brain abnormality. DESIGN: A pragmatic, prospective, multicentre, cohort study with a health economics analysis and a sociological substudy. SETTING: Sixteen UK fetal medicine centres. PARTICIPANTS: Pregnant women aged ≥ 16 years carrying a fetus (at least 18 weeks' gestation) with a suspected brain abnormality detected on ultrasonography. INTERVENTIONS: Participants underwent iuMRI and the findings were reported to their referring fetal medicine clinician. MAIN OUTCOME MEASURES: Pregnancy outcome was followed up and an ORD from postnatal imaging or postmortem autopsy/imaging collected when available. Developmental data from the Bayley Scales of Infant Development and questionnaires were collected from the surviving infants aged 2-3 years. Data on the management of the pregnancy before and after the iuMRI were collected to inform the economic evaluation. Two surveys collected data on patient acceptability of iuMRI and qualitative interviews with participants and health professionals were undertaken. RESULTS: The primary analysis consisted of 570 fetuses. The absolute diagnostic accuracies of ultrasonography and iuMRI were 68% and 93%, respectively [a difference of 25%, 95% confidence interval (CI) 21% to 29%]. The difference between ultrasonography and iuMRI increased with gestational age. In the 18-23 weeks group, the figures were 70% for ultrasonography and 92% for iuMRI (difference of 23%, 95% CI 18% to 27%); in the ≥ 24 weeks group, the figures were 65% for ultrasonography and 94% for iuMRI (difference of 29%, 95% CI 23% to 36%). Patient acceptability was high, with at least 95% of respondents stating that they would have iuMRI again in a similar situation. Health professional interviews suggested that iuMRI was acceptable to clinicians and that iuMRI was useful as an adjunct to ultrasonography, but not as a replacement. Across a range of scenarios, iuMRI resulted in additional costs compared with ultrasonography alone. The additional cost was consistently < £600 per patient and the cost per management decision appropriately changed was always < £3000. There is potential for reporting bias from the referring clinicians on the diagnostic and prognostic outcomes. Lower than anticipated follow-up rates at 3 years of age were observed. CONCLUSIONS: iuMRI as an adjunct to ultrasonography significantly improves the diagnostic accuracy and confidence for the detection of fetal brain abnormalities. An evaluation of the use of iuMRI for cases of isolated microcephaly and the diagnosis of fetal spine abnormalities is recommended. Longer-term follow-up studies of children diagnosed with fetal brain abnormalities are required to fully assess the functional significance of the diagnoses. TRIAL REGISTRATION: Current Controlled Trials ISRCTN27626961. FUNDING: This project was funded by the National Institute for Health Research (NIHR) Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 23, No. 49. See the NIHR Journals Library website for further project information.

Ultrasonography is routine in pregnancy to check that the baby's brain is developing as expected. However, no medical test is perfect and ultrasonography may miss some brain abnormalities, may get some brain abnormalities wrong or may diagnose an abnormality that is not really present. Magnetic resonance imaging (MRI) may help clarify difficult cases during pregnancy. We wanted to find out if MRI was better than ultrasonography alone in making an accurate diagnosis. We recruited pregnant women whose ultrasound scan, performed by an expert, suggested that their baby had a brain abnormality, and referred them for a MRI scan. The results of the two tests were compared with each other and to the final outcome of the pregnancy. Our results showed that using MRI in addition to ultrasonography improved the accuracy of the diagnosis in about one in four pregnancies. It changed the prediction of how the baby would develop in at least one in five cases. In many cases, the pregnancy was managed differently because of the MRI result. The MRI was acceptable to women, with 95% saying that they would have MRI again in a similar situation. Neither MRI nor ultrasonography accurately identified children who went on to have delayed development at the age of 2­3 years, but MRI was better than ultrasonography at ruling out developmental problems at this age. The MRI cost more than ultrasonography alone; therefore, whether or not it is worthwhile depends on the value placed on the decisions that changed as a result of its use.