Search

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.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.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.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.

Background Adherence to preventative inhaled therapies in people with cystic fibrosis (CF) is low, resulting in potentially avoidable health losses and the need for costly rescue therapies. Objectives To estimate the cost-effectiveness of the CFHealthHub (CFHH) intervention to support adherence to inhaled medications. Methods A state transition model was developed to assess the cost-effectiveness of the CFHH intervention versus usual care from the perspective of the UK National Health Service and Personal Social Services over a lifetime horizon. Costs and health outcomes were discounted at a rate of 3.5 percent per annum. Costs were valued at 2021/22 prices. The model structure includes health states defined by survival status, level of lung function, and transplant history. Treatment effects were modeled by changing the probabilities of transitioning between lung function states and reducing exacerbation rates. Model parameters were informed by the CFHH trial, CF Registry data, routine cost databases, literature, and expert opinion. Deterministic and probabilistic sensitivity analyses were undertaken to assess uncertainty. Results The CFHH intervention is expected to generate additional health gains and cost savings compared with usual care. Assuming that it is delivered for 10 years, the CFHH intervention is expected to generate 0.17 additional quality-adjusted life years and cost savings of GBP 1,600 (EUR 1,662) per patient. Conclusions The CFHH intervention is expected to dominate usual care, irrespective of the duration over which the intervention is delivered. The modeled benefits and cost savings are smaller than initially expected and are sensitive to relative treatment effects on lung function.

Background and aims\ud \ud Smoking prevalence has been falling in England for over 50 years but remains a prevalent and major public health problem. This study used an Age-Period-Cohort (APC) approach to measure lifecycle, historical and generational patterns of individual smoking behaviour.\ud \ud \ud \ud Design\ud \ud APC analysis of repeated cross-sectional smoking prevalence data obtained from three nationally representative surveys.\ud \ud \ud \ud Setting\ud \ud England (1972 – 2019).\ud \ud \ud \ud Participants\ud \ud Individuals aged 18 to 90 years old.\ud \ud \ud \ud Measurements\ud \ud We studied relative odds of current smoking in relation to age in single years from 18 to 90, 24 groups of 2-year survey periods (1972-1973 to 2018-2019), and 20 groups of 5-year birth cohorts (1907-1911 to 1997-2001). Age and period rates were studied for two groups of birth cohorts: those aged 18 to 25 years and those aged over 25 years.\ud \ud \ud \ud Findings\ud \ud Relative to age 18, the odds of current smoking increased with age until around age 25 (odds ratio (OR): 1.48; 95% confidence interval (CI):1.41 to 1.56) and then decreased progressively to age 90 (OR: 0.06; 95% CI: 0.04 to 0.08). They also decreased almost linearly with period relative to 1972-1973 (for 2018-2019 OR: 0.30; 95% CI: 0.26 to 0.34) and with birth cohort relative to 1902-1906, with the largest decreased observed for birth cohort 1992-1996 (OR: 0.44; 95% CI: 0.35 to 0.46) and 1997-2001 (OR: 0.35; 95% CI: 0.74 to 0.88). Smoking declined in the 18-25 age group by an average of 7% over successive 2-year periods, and by an average of 5% in those aged over 25.\ud \ud \ud \ud Conclusions\ud \ud Smoking in England appears to have declined over recent decades mainly as a result of reduced smoking uptake before age 25, and to a lesser extent to smoking cessation after age 25.

Background\ud \ud Relatively little is known about how risky drinkers attempt to moderate their drinking in the absence of specialist support. The broader literature has identified multiple potential strategies that people use to cope with temptation when trying to control health-risk behaviours. This study aims to identify types of alcohol moderation strategies used by British adults, and to explore how concurrent alcohol consumption differs across moderation strategies, focusing on the important role of usual drinking frequency.\ud \ud \ud \ud Methods\ud \ud We use a continuous repeat cross-sectional survey and one-week drinking diary collected by the market research company Kantar; these provide detailed information on alcohol consumption during a diary week and on how individuals try to moderate alcohol use for 49,204 British adults trying to reduce their drinking from 2013 to 2019. We use Latent Class Analysis (LCA) to identify predominant types of moderation strategies. With a three-step method, we also analyse the associations between adopting different moderation strategies, measures of frequency and intensity of drinking events, and usual drinking frequency.\ud \ud \ud \ud Results\ud \ud We found evidence of four alcohol moderation strategies: 29% of individuals use a pre-commitment-focused strategy (having fewer drinking occasions), two set of individuals adopt self-control strategies within drinking occasions (specifically 28% select smaller drinks and 5% have fewer drinks), while 38% adopt a mixed strategy that involves all three. Those using commitment tend to have a higher average consumption per drinking occasion but lower overall weekly consumption compared to those using self-control. Weekly alcohol consumption is particularly high among individuals who are usual everyday drinkers and use self-control to cut down drinking.\ud \ud \ud \ud Conclusion\ud \ud This analysis provides a useful platform for further work, using prospective or intervention designs, to test the relative effectiveness of different moderation strategies for alcohol consumers who want to reduce their alcohol consumption.

Objectives\ud 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.\ud \ud Methods\ud 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.\ud \ud Results\ud 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.\ud \ud Conclusion\ud 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.

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.

Background People with cystic fibrosis frequently have low levels of adherence to inhaled medications. Objectives The objectives were to develop and evaluate an intervention for adults with cystic fibrosis to improve adherence to their inhaled medication. Design We used agile software methods to develop an online platform. We used mixed methods to develop a behaviour change intervention for delivery by an interventionist. These were integrated to become the CFHealthHub intervention. We undertook a feasibility study consisting of a pilot randomised controlled trial and process evaluation in two cystic fibrosis centres. We evaluated the intervention using an open-label, parallel-group randomised controlled trial with usual care as the control. Participants were randomised in a 1 : 1 ratio to intervention or usual care. Usual care consisted of clinic visits every 3 months. We undertook a process evaluation alongside the randomised controlled trial, including a fidelity study, a qualitative interview study and a mediation analysis. We undertook a health economic analysis using both a within-trial and model-based analysis. Setting The randomised controlled trial took place in 19 UK cystic fibrosis centres. Participants Participants were people aged ≥ 16 years with cystic fibrosis, on the cystic fibrosis registry, not post lung transplant or on the active transplant list, who were able to consent and not using dry-powder inhalers. Intervention People with cystic fibrosis used a nebuliser with electronic monitoring capabilities. This transferred data automatically to a digital platform. People with cystic fibrosis and clinicians could monitor adherence using these data, including through a mobile application (app). CFHealthHub displayed graphs of adherence data as well as educational and problem-solving information. A trained interventionist helped people with cystic fibrosis to address their adherence. Main outcome measures Randomised controlled trial – adjusted incidence rate ratio of pulmonary exacerbations meeting the modified Fuchs criteria over a 12-month follow-up period (primary outcome); change in percentage adherence; and per cent predicted forced expiratory volume in 1 second (key secondary outcomes). Process evaluation – percentage fidelity to intervention delivery, and participant and interventionist perceptions of the intervention. Economic modelling – incremental cost per quality-adjusted life-year gained. Results Randomised controlled trial – 608 participants were randomised to the intervention (n = 305) or usual care (n = 303). To our knowledge, this was the largest randomised controlled trial in cystic fibrosis undertaken in the UK. The adjusted rate of exacerbations per year (primary outcome) was 1.63 in the intervention and 1.77 in the usual-care arm (incidence rate ratio 0.96, 95% confidence interval 0.83 to 1.12; p = 0.638) after adjustment for covariates. The adjusted difference in mean weekly normative adherence was 9.5% (95% confidence interval 8.6% to 10.4%) across 1 year, favouring the intervention. Adjusted mean difference in forced expiratory volume in 1 second (per cent) predicted at 12 months was 1.4% (95% confidence interval –0.2% to 3.0%). No adverse events were related to the intervention. Process evaluation – fidelity of intervention delivery was high, the intervention was acceptable to people with cystic fibrosis, participants engaged with the intervention [287/305 (94%) attended the first intervention visit], expected mechanisms of action were identified and contextual factors varied between randomised controlled trial sites. Qualitative interviews with 22 people with cystic fibrosis and 26 interventionists identified that people with cystic fibrosis welcomed the objective adherence data as proof of actions to self and others, and valued the relationship that they built with the interventionists. Economic modelling – the within-trial analysis suggests that the intervention generated 0.01 additional quality-adjusted life-years at an additional cost of £865.91 per patient, leading to an incremental cost-effectiveness ratio of £71,136 per quality-adjusted life-year gained. This should be interpreted with caution owing to the short time horizon. The health economic model suggests that the intervention is expected to generate 0.17 additional quality-adjusted life-years and cost savings of £1790 over a lifetime (70-year) horizon; hence, the intervention is expected to dominate usual care. Assuming a willingness-to-pay threshold of £20,000 per quality-adjusted life-year gained, the probability that the intervention generates more net benefit than usual care is 0.89. The model results are dependent on assumptions regarding the duration over which costs and effects of the intervention apply, the impact of the intervention on forced expiratory volume in 1 second (per cent) predicted and the relationship between increased adherence and drug-prescribing levels. Limitations Number of exacerbations is a sensitive and valid measure of clinical change used in many trials. However, data collection of this outcome in this context was challenging and could have been subject to bias. It was not possible to measure baseline adherence accurately. It was not possible to quantify the impact of the intervention on the number of packs of medicines prescribed. Conclusions We developed a feasible and acceptable intervention that was delivered to fidelity in the randomised controlled trial. We observed no statistically significant difference in the primary outcome of exacerbation rates over 12 months. We observed an increase in normative adherence levels in a disease where adherence levels are low. The magnitude of the increase in adherence may not have been large enough to affect exacerbations. Future work Given the non-significant difference in the primary outcome, further research is required to explore why an increase in objective normative adherence did not reduce exacerbations and to develop interventions that reduce exacerbations. Trial registration Work package 3.1: Current Controlled Trials ISRCTN13076797. Work packages 3.2 and 3.3: Current Controlled Trials ISRCTN55504164. Funding This project was funded by the National Institute for Health Research (NIHR) Programme Grants for Applied Research programme and will be published in full in Programme Grants for Applied Research; Vol. 9, No. 11. See the NIHR Journals Library website for further project information.

Background Preference-based measures of health, such as the three-level EuroQol five-dimensional questionnaire (EQ-5D-3L), are required to calculate quality-adjusted life-years for use in cost-effectiveness analysis, but are often not recorded in clinical studies. In these cases, mapping can be used to estimate preference-based measures. Objectives To model the relationship between the EQ-5D-3L and the Functional Assessment of Cancer Therapy—Breast Cancer (FACT-B) instrument, comparing indirect and direct mapping methods, and the use of FACT-B summary score versus FACT-B subscale scores. Methods We used data from three clinical studies for advanced breast cancer providing 11,958 observations with full information on FACT-B and the EQ-5D-3L. We compared direct mapping using adjusted limited dependent variable mixture models (ALDVMMs) with indirect mapping using seemingly unrelated ordered probit models. The EQ-5D-3L was estimated as a function of FACT-B and other patient-related covariates. Results The use of FACT-B subscale scores was better than using the total FACT-B score. A good fit to the observed data was observed across the entire range of disease severity in all models. ALDVMMs outperformed the indirect mapping. The breast cancer–specific scale had a strong influence in predicting the pain/discomfort and self-care dimensions of the EQ-5D-3L. Conclusions This article adds to the growing literature that demonstrates the performance of the ALDVMM method for mapping. Regardless of which model is used, the subscales of FACT-B should be included as independent variables wherever possible. The breast cancer–specific subscale of FACT-B is important in predicting the EQ-5D-3L. This suggests that generic cancer measures should not be used for utility mapping in patients with breast cancer.

Objectives\ud \ud 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.\ud \ud \ud \ud Methods\ud \ud 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.\ud \ud \ud \ud Results\ud \ud 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.\ud \ud \ud \ud Conclusions\ud \ud 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.

Objectives Mappings to convert clinical measures to preference-based measures of health such as the EQ-5D-3L are sometimes required in cost-utility analyses. We developed mappings to convert best-corrected visual acuity (BCVA) to the EQ-5D-3L, the EQ-5D-3L with a vision bolt-on (EQ-5D V), and the Visual Functioning Questionnaire-Utility Index (VFQ-UI) in patients with macular edema caused by central retinal vein occlusion. Methods We used data from Lucentis, Eylea, Avastin in vein occlusion (LEAVO), which is a phase-3 randomized controlled trial comparing ranibizumab, aflibercept, and bevacizumab in 463 patients with observations at 6 time points. We estimated adjusted limited dependent variable mixture models consisting of 1 to 4 distributions (components) using BCVA in each eye, age, and sex to predict utility within the components and BCVA as a determinant of component membership. We compared model fit using mean error, mean absolute error, root mean square error, Akaike information criteria, Bayesian information criteria, and visual inspection of mean predicted and observed utilities and cumulative distribution functions. Results Mean utility scores were 0.82 for the EQ-5D-3L, 0.79 for the EQ-5D V, and 0.88 for the VFQ-UI. The best-fitting models for the EQ-5D and EQ-5D V had 2 components (with means of approximately 0.44 and 0.85), and the best-fitting model for VFQ-UI had 3 components (with means of approximately 0.95, 0.74, and 0.90). Conclusions Models with multiple components better predict utility than those with single components. This article provides a valuable addition to the literature, in which previous mappings in visual acuity have been limited to linear regressions, resulting in unfounded assumptions about the distribution of the dependent variable., Highlights • Adjusted, limited dependent variable mixture models with multiple components better predict utility using the EQ-5D-3L with and without vision bolt-on and Visual Functioning Questionnaire-Utility Index (VFQ-UI) than mappings with only 1 component. Previous mappings in visual acuity have considered only linear regression using ordinary least squares (1 component). • Our mappings can be used to predict 3 different measures of utility as a function of age, sex, and best-corrected visual acuity in both eyes. These variables are routinely included in economic evaluations of visual disorders.

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).

• Clinical trials often do not contain preference based outcome measure • Cost per QALY evaluation requires such outcomes • Linking clinical to preference based outcomes via statistical model is commonplace in economic evaluation • 22% of NICE appraisals include “mapping”, 100% in RA • Typically done using simple regression techniques • EQ-5D has an odd distribution • Standard methods perform badly • Develop and refine a new method here

Purpose To develop a mapping model to estimate EQ-5D-3L from the Knee Injury and Osteoarthritis Outcome Score (KOOS). Methods The responses to EQ-5D-3L and KOOS questionnaires (n = 40,459 observations) were obtained from the Swedish National anterior cruciate ligament (ACL) Register for patients ≥ 18 years with the knee ACL injury. We used linear regression (LR) and beta-mixture (BM) for direct mapping and the generalized ordered probit model for response mapping (RM). We compared the distribution of the original data to the distributions of the data generated using the estimated models. Results Models with individual KOOS subscales performed better than those with the average of KOOS subscale scores (KOOS5, KOOS4). LR had the poorest performance overall and across the range of disease severity particularly at the extremes of the distribution of severity. Compared with the RM, the BM performed better across the entire range of disease severity except the most severe range (KOOS5 Conclusions We developed mapping models to estimate EQ-5D-3L from KOOS facilitating application of KOOS in cost-utility analyses. The BM showed superior performance for estimating EQ-5D-3L from KOOS. Further validation of the estimated models in different independent samples is warranted.

• Clinical trials often do not contain preference based outcome measure • Cost per QALY evaluation requires such outcomes • Linking clinical to preference based outcomes via statistical model is commonplace in economic evaluation • 22% of NICE appraisals include “mapping”, 100% in RA • Typically done using simple regression techniques • EQ-5D has an odd distribution • Standard methods perform badly • Develop and refine a new method here

Background: The EuroQol-5 Dimension (EQ-5D) instrument is the National Institute for Health and Care Excellence (NICE)’s preferred measure of health-related quality of life (QoL) in adults. The three-level (3L) value set is currently recommended for use, but the five-level (5L) set is increasingly being used in practice. Objective: We aimed to explore the impact of moving from 3L to 5L in NICE appraisals. Methods: We adapted our existing mapping for use with health state utility values derived from a population where the original distribution of utilities was unknown. We used this mapping to estimate 5L utilities for 21 comparisons of interventions from models used in NICE technology appraisal decision making, covering a range of disease areas. Results: All utilities increased using 5L, and the differences between highest and lowest utilities decreased. In ten oncology comparisons, using 5L generally increased the incremental quality-adjusted life-years (QALYs) as the benefit from improving survival increased. In four non-oncology comparisons where the intervention improved QoL only, the incremental QALYs decreased as the benefit of improving QoL was reduced. In seven non-oncology comparisons where interventions improved survival and QoL, there was a trade-off between increasing the benefit from survival and decreasing the benefit from improving QoL. Conclusion: 3L and 5L lead to substantially different estimates of incremental QALYs and cost effectiveness. The direction and magnitude of the change is not consistent across case studies. Using 5L instead of 3L may lead to different reimbursement decisions. NICE will face inconsistencies in decision making if it uses 3L and 5L concurrently.

In this article, we describe a new command, eq5dmap, for conditional prediction of the utility values of EQ-5D-5L (EQ-5D-3L) from observed or specified values of EQ-5D-3L (EQ-5D-5L) conditional on age and gender. Predictions can be made either from the five-item health descriptions or from the (exact or approximate) utility score. The prediction process is based on a joint statistical model of the two variants of EQ-5D that have been fit to alternative reference datasets (the National Data Bank for Rheumatic Diseases and a EuroQol Group coordinated data-collection study). The underlying model is a system of ordinal regressions with a flexible residual distribution specified as Gaussian or as a copula mixture. Use of the command is illustrated with an application that includes an investigation of the sensitivity of the mapping outcomes to the choice of reference dataset.

In this article, we describe the betamix command, which fits mixture regression models for dependent variables bounded in an interval. The model is a generalization of the truncated inflated beta regression model introduced in Pereira, Botter, and Sandoval (2012, Communications in Statistics—Theory and Methods 41: 907–919) and the mixture beta regression model in Verkuilen and Smithson (2012, Journal of Educational and Behavioral Statistics 37: 82–113) for variables with truncated supports at either the top or the bottom of the distribution. betamix accepts dependent variables defined in any range that are then transformed to the interval (0, 1) before estimation.

OBJECTIVE: This paper estimates productivity loss using the health of the patient in order to allow indirect estimation of these costs for inclusion in economic evaluation. METHODS: Data from two surveys of inpatients [Health outcomes data repository (HODaR) sample (n = 42,442) and health improvement and patient outcomes (HIPO) sample (n = 6046)] were used. The number of days off paid employment or normal activities (excluding paid employment) was modelled using the health of the patients measured by the EQ-5D, international classification of diseases (ICD) chapters, and other health and sociodemographic data. Two-part models (TPMs) and zero-inflated negative binomial (ZINB) models were identified as the most appropriate specifications, given large spikes at the minimum and maximum days for the dependent variable. Analysis was undertaken separately for the two datasets to account for differences in recall period and identification of those who were employed. RESULTS: Models were able to reflect the large spike at the minimum (zero days) but not the maximum, with TPMs doing slightly better than the ZINB model. The EQ-5D was negatively associated with days off employment and normal activities in both datasets, but ICD chapters only had statistically significant coefficients for some chapters in the HODaR. CONCLUSIONS: TPMs can be used to predict productivity loss associated with the health of the patient to inform economic evaluation. Limitations include recall and response bias and identification of who is employed in the HODaR, while the HIPO suffers from a small sample size. Both samples exclude some patient groups.

In this study, we use the UK Millennium Cohort Study to estimate a dynamic factor model of child development. Our model follows the children from birth until 7 years of age and allows for both cognitive and noncognitive abilities in children. We find a significant self-productivity effect in both cognitive and noncognitive development, as well as some evidence of dynamic dependence across different abilities. The activities that parents carry out with children at home (parental investment) have a significant effect on children’s development; we find substantial evidence of two distinct latent parental investment variables with differential effects across the two abilities. Electronic supplementary material The online version of this article (doi:10.1007/s13524-017-0554-6) contains supplementary material, which is available to authorized users.

The assessment of health-related quality of life is crucially important in the evaluation of healthcare technologies and services. In many countries, economic evaluation plays a prominent role in informing decision making often requiring preference-based measures (PBMs) to assess quality of life. These measures comprise two aspects: a descriptive system where patients can indicate the impact of ill health, and a value set based on the preferences of individuals for each of the health states that can be described. These values are required for the calculation of quality adjusted life years (QALYs), the measure for health benefit used in the vast majority of economic evaluations. The National Institute for Health and Care Excellence (NICE) has used cost per QALY as its preferred framework for economic evaluation of healthcare technologies since its inception in 1999. However, there is often an evidence gap between the clinical measures that are available from clinical studies on the effect of a specific health technology and the PBMs needed to construct QALY measures. Instruments such as the EQ-5D have preference-based scoring systems and are favored by organizations such as NICE but are frequently absent from clinical studies of treatment effect. Even where a PBM is included this may still be insufficient for the needs of the economic evaluation. Trials may have insufficient follow-up, be underpowered to detect relevant events, or include the wrong PBM for the decision- making body. Often this gap is bridged by “mapping”—estimating a relationship between observed clinical outcomes and PBMs, using data from a reference dataset containing both types of information. The estimated statistical model can then be used to predict what the PBM would have been in the clinical study given the available information. There are two approaches to mapping linked to the structure of a PBM. The indirect approach (or response mapping) models the responses to the descriptive system using discrete data models. The expected health utility is calculated as a subsequent step using the estimated probability distribution of health states. The second approach (the direct approach) models the health state utility values directly. Statistical models routinely used in the past for mapping are unable to consider the idiosyncrasies of health utility data. Often they do not work well in practice and can give seriously biased estimates of the value of treatments. Although the bias could, in principle, go in any direction, in practice it tends to result in underestimation of cost effectiveness and consequently distorted funding decisions. This has real effects on patients, clinicians, industry, and the general public. These problems have led some analysts to mistakenly conclude that mapping always induces biases and should be avoided. However, the development and use of more appropriate models has refuted this claim. The need to improve the quality of mapping studies led to the formation of the International Society for Pharmacoeconomics and Outcomes Research (ISPOR) Mapping to Estimate Health State Utility values from Non-Preference-Based Outcome Measures Task Force to develop good practice guidance in mapping.

ObjectivesRheumatoid arthritis (RA) is a chronic inflammatory disease associated with increasing disability, reduced quality of life and substantial costs (as a result of both intervention acquisition and hospitalisation). The objective was to assess the clinical effectiveness and cost-effectiveness of seven biologic disease-modifying antirheumatic drugs (bDMARDs) compared with each other and conventional disease-modifying antirheumatic drugs (cDMARDs). The decision problem was divided into those patients who were cDMARD naive and those who were cDMARD experienced; whether a patient had severe or moderate to severe disease; and whether or not an individual could tolerate methotrexate (MTX).Data sourcesThe following databases were searched: MEDLINE from 1948 to July 2013; EMBASE from 1980 to July 2013; Cochrane Database of Systematic Reviews from 1996 to May 2013; Cochrane Central Register of Controlled Trials from 1898 to May 2013; Health Technology Assessment Database from 1995 to May 2013; Database of Abstracts of Reviews of Effects from 1995 to May 2013; Cumulative Index to Nursing and Allied Health Literature from 1982 to April 2013; and TOXLINE from 1840 to July 2013. Studies were eligible for inclusion if they evaluated the impact of a bDMARD used within licensed indications on an outcome of interest compared against an appropriate comparator in one of the stated population subgroups within a randomised controlled trial (RCT). Outcomes of interest included American College of Rheumatology (ACR) scores and European League Against Rheumatism (EULAR) response. Interrogation of Early Rheumatoid Arthritis Study (ERAS) data was undertaken to assess the Health Assessment Questionnaire (HAQ) progression while on cDMARDs.MethodsNetwork meta-analyses (NMAs) were undertaken for patients who were cDMARD naive and for those who were cDMARD experienced. These were undertaken separately for EULAR and ACR data. Sensitivity analyses were undertaken to explore the impact of including RCTs with a small proportion of bDMARD experienced patients and where MTX exposure was deemed insufficient. A mathematical model was constructed to simulate the experiences of hypothetical patients. The model was based on EULAR response as this is commonly used in clinical practice in England. Observational databases, published literature and NMA results were used to populate the model. The outcome measure was cost per quality-adjusted life-year (QALY) gained.ResultsSixty RCTs met the review inclusion criteria for clinical effectiveness, 38 of these trials provided ACR and/or EULAR response data for the NMA. Fourteen additional trials contributed data to sensitivity analyses. There was uncertainty in the relative effectiveness of the interventions. It was not clear whether or not formal ranking of interventions would result in clinically meaningful differences. Results from the analysis of ERAS data indicated that historical assumptions regarding HAQ progression had been pessimistic. The typical incremental cost per QALY of bDMARDs compared with cDMARDs alone for those with severe RA is > £40,000. This increases for those who cannot tolerate MTX (£50,000) and is > £60,000 per QALY when bDMARDs were used prior to cDMARDs. Values for individuals with moderate to severe RA were higher than those with severe RA. Results produced using EULAR and ACR data were similar. The key parameter that affected the results is the assumed HAQ progression while on cDMARDs. When historic assumptions were used typical incremental cost per QALY values fell to £38,000 for those with severe disease who could tolerate MTX.ConclusionsbDMARDs appear to have cost per QALY values greater than the thresholds stated by the National Institute for Health and Care Excellence for interventions to be cost-effective. Future research priorities include: the evaluation of the long-term HAQ trajectory while on cDMARDs; the relationship between HAQ direct medical costs; and whether or not bDMARDs could be stopped once a patient has achieved a stated target (e.g. remission).Study registrationThis study is registered as PROSPERO CRD42012003386.FundingThe National Institute for Health Research Health Technology Assessment programme.

In this article, we describe a new Stata command, bicop, for fitting a model consisting of a pair of ordinal regressions with a flexible residual distribution, with each marginal distribution specified as a two-part normal mixture, and stochastic dependence governed by a choice of copula functions. The bicop command generalizes the existing biprobit and bioprobit commands, which assume a bivariate normal residual distribution. We present and explain the bicop estimation command and the available postestimation commands using data on financial well-being from the UK Understanding Society Panel Survey.

Background\ud The prevalence of childhood obesity has been increasing but the causes are not fully understood. Recent public\ud health interventions and guidance aiming to reduce childhood obesity have focused on the whole family, as\ud opposed to just the child but there remains a lack of empirical evidence examining this relationship.\ud \ud Methods\ud Using data from the longitudinal Millennium Cohort Study (MCS), we investigate the dynamic relationship\ud between underlying family lifestyle and childhood obesity during early childhood. The MCS interviewed\ud parents shortly after the birth of their child and follow up interviews were carried out when the child was 3, 5\ud and 7 years. We use a dynamic latent factor model, an approach that allows us to identify family lifestyle, its\ud evolution over time (in this case between birth and 7 years) and its influence on childhood obesity and other\ud observable outcomes.\ud \ud Results\ud We find that family lifestyle is persistent, 87.43% of families which were above the 95th percentile on the\ud lifestyle distribution, remained above the 95th percentile when the child was 7 years old. Family lifestyle has a\ud significant influence on all outcomes in the study, including diet, exercise and parental weight status; family\ud lifestyle accounts for 11.3% of the variation in child weight by age 7 years.\ud \ud Conclusion\ud The analysis suggests that interventions should therefore be prolonged and persuasive and target the underlying\ud lifestyle of a family as early as possible during childhood in order to have the greatest cumulative influence.\ud Our results suggest that children from advantaged backgrounds are more likely to be exposed to healthier\ud lifestyles and that this leads to inequalities in the prevalence of obesity. To reduce inequalities in childhood\ud obesity, policy makers should target disadvantaged families and design interventions specifically for these\ud families.

OBJECTIVES: To model the relationship between the three-level (3L) and the five-level (5L) EuroQol five-dimensional questionnaire and examine how differences have an impact on cost effectiveness in case studies. METHODS: We used two data sets that included the 3L and 5L versions from the same respondents. The EuroQol Group data set (n = 3551) included patients with different diseases and a healthy cohort. The National Data Bank data set included patients with rheumatoid disease (n = 5205). We estimated a system of ordinal regressions in each data set using copula models to link responses of the 3L instrument to those of the 5L instrument and its UK tariff, and vice versa. Results were applied to nine cost-effectiveness studies. RESULTS: Best-fitting models differed between the EuroQol Group and the National Data Bank data sets in terms of the explanatory variables, copulas, and coefficients. In both cases, the coefficients of the covariates and latent factors between the 3L and the 5L instruments were significantly different, indicating that moving between instruments is not simply a uniform re-alignment of the response levels for most dimensions. In the case studies, moving from the 3L to the 5L caused a decrease of up to 87% in incremental quality-adjusted life-years gained from effective technologies in almost all cases. Incremental cost-effectiveness ratios increased, often substantially. Conversely, one technology with a significant mortality gain saw increased incremental quality-adjusted life-years. CONCLUSIONS: The 5L shifts mean utility scores up the utility scale toward full health and compresses them into a smaller range, compared with the 3L. Improvements in quality of life are valued less using the 5L than using the 3L. The 3L and the 5L can produce substantially different estimates of cost effectiveness. There is no simple proportional adjustment that can be made to reconcile these differences. Copyright (c) 2018 International Society for Pharmacoeconomics and Outcomes Research (ISPOR). Published by Elsevier Inc. All rights reserved.

Objective.To ascertain whether strategies of treatment with a biological disease-modifying antirheumatic drug (bDMARD) are cost-effective in an English setting. Results are presented for those patients with moderate to severe rheumatoid arthritis (RA) and those with severe RA.Methods.An economic model to assess the cost-effectiveness of 7 bDMARD was developed. A systematic literature review and network metaanalysis was undertaken to establish relative clinical effectiveness. The results were used to populate the model, together with estimates of Health Assessment Questionnaire (HAQ) score following European League Against Rheumatism response; annual costs, and utility, per HAQ band; trajectory of HAQ for patients taking bDMARD; and trajectory of HAQ for patients using nonbiologic therapy (NBT). Results were presented as those associated with the strategy with the median cost-effectiveness. Supplementary analyses were undertaken assessing the change in cost-effectiveness when only patients with the most severe prognoses taking NBT were provided with bDMARD treatment. The costs per quality-adjusted life-year (QALY) values were compared with reported thresholds from the UK National Institute for Health and Care Excellence of £20,000 to £30,000 (US$24,700 to US$37,000).Results.In the primary analyses, the cost per QALY of a bDMARD strategy was £41,600 for patients with severe RA and £51,100 for those with moderate to severe RA. Under the supplementary analyses, the cost per QALY fell to £25,300 for those with severe RA and to £28,500 for those with moderate to severe RA.Conclusion.The cost-effectiveness of bDMARD in RA in England is questionable and only meets current accepted levels in subsets of patients with the worst prognoses.

Decision-analytic models (DAMs) used to evaluate the cost effectiveness of interventions are pivotal sources of evidence used in economic evaluations. Parameter estimates used in the DAMs are often based on the results of a regression analysis, but there is little guidance relating to these. This study had two objectives. The first was to identify the frequency of use of regression models in economic evaluations, the parameters they inform, and the amount of information reported to describe and support the analyses. The second objective was to provide guidance to improve practice in this area, based on the review. The review concentrated on a random sample of economic evaluations submitted to the UK National Institute for Health and Clinical Excellence (NICE) as part of its technology appraisal process. Based on these findings, recommendations for good practice were drafted, together with a checklist for critiquing reporting standards in this area. Based on the results of this review, statistical regression models are in widespread use in DAMs used to support economic evaluations, yet reporting of basic information, such as the sample size used and measures of uncertainty, is limited. Recommendations were formed about how reporting standards could be improved to better meet the needs of decision makers. These recommendations are summarised in a checklist, which may be used by both those conducting regression analyses and those critiquing them, to identify what should be reported when using the results of a regression analysis within a DAM.

Economic evaluation conducted in terms of cost per quality-adjusted life-year (QALY) provides information that decision makers find useful in many parts of the world. Ideally, clinical studies designed to assess the effectiveness of health technologies would include outcome measures that are directly linked to health utility to calculate QALYs. Often this does not happen, and even when it does, clinical studies may be insufficient for a cost-utility assessment. Mapping can solve this problem. It uses an additional data set to estimate the relationship between outcomes measured in clinical studies and health utility. This bridges the evidence gap between available evidence on the effect of a health technology in one metric and the requirement for decision makers to express it in a different one (QALYs). In 2014, ISPOR established a Good Practices for Outcome Research Task Force for mapping studies. This task force report provides recommendations to analysts undertaking mapping studies, those that use the results in cost-utility analysis, and those that need to critically review such studies. The recommendations cover all areas of mapping practice: the selection of data sets for the mapping estimation, model selection and performance assessment, reporting standards, and the use of results including the appropriate reflection of variability and uncertainty. This report is unique because it takes an international perspective, is comprehensive in its coverage of the aspects of mapping practice, and reflects the current state of the art. Keywords: economic evaluation, health utility, mapping, quality of life.

Background Many studies have found a statistical association between breastfeeding and childhood adiposity. This paper investigates whether breastfeeding has an effect on subsequent childhood body mass index (BMI) using propensity scores to account for confounding. Methods We use data from the Millennium Cohort Study, a nationally representative UK cohort survey, which contains detailed information on infant feeding and childhood BMI. Propensity score matching is used to investigate the mean BMI in children breastfed exclusively and partially for different durations of time. Results We find statistically significant influences of breastfeeding on childhood BMI, particularly in older children, when breastfeeding is prolonged and exclusive. At 7 years, children who were exclusively breastfed for 16 weeks had a BMI 0.28 kg/m2 (95% confidence interval 0.07 to 0.49) lower than those who were never breastfed, a 2% reduction from the mean BMI of 16.6 kg/m2. Conclusions For this young cohort, even small effects of breastfeeding on BMI could be important. In order to reduce BMI, breastfeeding should be encouraged as part of wider lifestyle intervention. This evidence could help to inform public health bodies when creating public health guidelines and recommendations.

Previous methods of empirical mapping involve using regressions on patient or general population self-reported data from datasets involving two or more measures. This approach relies on overlap in the descriptive systems of the measures and assumes it is appropriate to use different measures on the same population, which may not always be the case. This paper presents a feasibility study for a new approach to mapping between preference-based measures (PBM) using general population visual analogue scale (VAS) values as a common yardstick. We use data from a valuation study of 502 members of the UK general population, where, using ranking and VAS tasks, interviewees simultaneously valued health states defined by three of six PBM: EQ-5D (generic), SF-6D (generic), HUI2 (generic for children and adults), AQL-5D (asthma specific), OPUS (social care specific) and ICECAP (capabilities). Regression techniques are used to estimate the relationship between these VAS values and the original value set (i.e. 'tariff'). These results are subsequently used to estimate the relationship between all six PBM to enable 'value-based mapping' between measures. This new method of mapping potentially has a useful role in evidence synthesis and cross programme comparisons in studies using different measures.

Objective To estimate the cost-effectiveness of primary angioplasty compared with thrombolysis for acute ST elevation myocardial infarction. Design Cost analysis of UK observational database, incorporated into decision analytical model. Methods Patients receiving treatment within a comprehensive angioplasty service were compared with control patients receiving thrombolysis-based care. The treatment costs and delays to treatment of thrombolysis and angioplasty were estimated. These estimates were then incorporated into an existing model of cost-effectiveness that synthesises evidence from 22 randomised trials to estimate health outcomes measured by quality-adjusted life years (QALYs). Main outcome measures Costs from a health service perspective and outcomes measured as quality adjusted. Results The mean cost of the initial treatment was £3509 for thrombolysis at control sites, £5176 for angioplasty in usual working hours at National Infarct Angioplasty Project sites and an additional £245 if undertaken out of hours. Angioplasty-based care had an incremental cost of £4520 per QALY gained and 0.9 probability of being cost-effective at a threshold of £20 000 per QALY gained. This probability was >0.95 if patients were directly admitted to the cardiac catheter laboratory, 0.75 if admitted via the emergency department or coronary care unit and 0.38 if transferred to the angioplasty centre from another hospital. Conclusions Overall, primary angioplasty-based care is highly likely to be cost-effective at an assumed threshold of £20 000 per QALY gained. It is more likely to be cost-effective if patients are admitted directly to the cardiac catheter laboratory rather than via other hospital departments, or if transferred from another hospital.

In this article, we describe the aldvmm command for fitting adjusted limited dependent variable mixture models to either UK or U.S. tariff EQ-5D data. We present and explain the command and postestimation command through examples. The aldvmm command requires use of Stas Kolenikov's simulated annealing package (simann()), which can be easily installed by typing net install simann.pkg, from( http://web.missouri.edu/∼kolenikovs/stata ).

OBJECTIVES: This paper estimates informal care need using the health of the patient. The results can be used to predict changes in informal care associated with changes in the health of the patient measured using EQ-5D. METHODS: Data was used from a prospective survey of inpatients containing 59,512 complete responses across 44,494 individuals. The number of days a friend or relative has needed to provide care or help with normal activities in the last 6 weeks was estimated using the health of the patient measured by EQ-5D, ICD chapter and other health and sociodemographic data. A variety of different regression models were estimated that are appropriate for the distribution of the informal care dependent variable, which has large spikes at 0 (zero informal care) and 42 days (informal care every day). RESULTS: The preferred model that most accurately predicts the distribution of the data is the zero-inflated negative binomial with variable inflation. The results indicate that improving the health of the patient reduces informal care need. The relationship between ICD chapter and informal care need is not as clear. CONCLUSIONS: The preferred zero-inflated negative binomial with variable inflation model can be used to predict changes in informal care associated with changes in the health of the patient measured using EQ-5D and these results can be applied to existing datasets to inform economic evaluation. Limitations include recall bias and response bias of the informal care data, and restrictions of the dataset to exclude some patient groups.

OBJECTIVE The aim of this study was to estimate the cost-effectiveness of combination DMARDs with short-term glucocorticoids in early active RA using data from the 2-year Combination of Anti-Rheumatic Drugs in Early RA (CARDERA) trial. METHODS CARDERA enrolled 467 patients with active RA of

Background: Analysts often need to estimate health state utility values as a function of other outcome measures. Utility values like EQ-5D have several unusual characteristics that make standard statistical methods inappropriate. We have developed a bespoke approach based on mixture models to directly estimate EQ-5D. An indirect method, “response mapping”, first estimates the level on each of the five dimensions of the EQ-5D descriptive system and then calculates the expected tariff score. These methods have never previously been compared.\ud \ud Methods: We use a large observational database of patients diagnosed with Rheumatoid Arthritis (n=100,398 observations). Direct estimation of UK EQ-5D scores as a function of Health Assessment Questionnaire (HAQ), pain and age was performed using a limited dependent variable mixture model. Indirect modelling was undertaken using a set of generalized ordered probit models with expected tariff scores calculated mathematically. Linear regression was reported for comparison purposes.\ud \ud Results: The linear model fits poorly, particularly at the extremes of the distribution. Both the bespoke mixture model and the generalized ordered probit approach offer improvements in fit over the entire range of EQ-5D. Mean average error is 10% and 5% lower compared to the linear model respectively. Root mean squared error is 3% and 2% lower. The mixture model demonstrates superior performance to the indirect method across almost the entire range of pain and HAQ.\ud \ud Limitations: There is limited data from patients in the most extreme HAQ health states.\ud \ud Conclusions: Modelling of EQ-5D from clinical measures is best performed directly using the bespoke mixture model. This substantially outperforms the indirect method in this example. Linear models are inappropriate, suffer from systematic bias and generate values outside the feasible\ud range.

Objective. This study aims to provide robust estimates of EQ-5D as a function of the HAQ and pain in patients with RA.\ud \ud Method. Repeated observations were made of patients diagnosed with RA in a US observational cohort (n = 100 398 observations) who provided data on HAQ, pain on a visual analogue scale and the EQ-5D questionnaire. We used a bespoke statistical method based on mixture modelling to appropriately reflect the characteristics of the EQ-5D instrument and to compare this with results from standard multiple regression.\ud \ud Results. EQ-5D can be predicted from summary HAQ and pain scores. We identify four different classes of respondents who differ in terms of disease severity. Unlike the multiple regression, the mixture model exhibits very good fit to the data and does not suffer from problems of bias or predict values outside the feasible range.\ud \ud Conclusion. It is appropriate to model the relationship between HAQ and EQ-5D but only if suitable statistical methods are applied. Linear models underestimate the quality-adjusted life year benefits, and therefore the cost-effectiveness, of therapies. The bespoke mixture model approach outlined here overcomes this problem. The addition of pain as an explanatory variable greatly improves the estimates. Reimbursement agencies rely on these types of analyses when formulating policy on the use of new drug therapies. Clinicians as well as economists should be concerned with these issues.

Background. Different preference-based measures (PBMs) used to estimate quality-adjusted life years (QALYs) provide different utility values for the same patient. Differences are expected since values have been obtained using different samples, valuation techniques, and descriptive systems. Previous studies have estimated the relationship between pairs of PBMs using patient self-reported data. However, there is a need for an approach capable of generating values directly on a common scale for a range of PBMs using the same sample of general population respondents and valuation technique but keeping the advantages of the different descriptive systems. Methods. General public survey data ( n = 501) in which respondents ranked health states described using subsets of 6 PBMs were analyzed. We develop a new model based on the mixed logit to overcome 2 key limitations of the standard rank-ordered logit model—namely, the unrealistic choice pattern (independence of irrelevant alternatives) and the independence of repeated observations. Results. There are substantial differences in the estimated parameters between the 2 models (mean difference 0.07), leading to different orderings across the measures. Estimated values for the best states described by different PBMs are substantially and significantly different using the standard model, unlike our approach, which yields more consistent results. Limitations. Data come from an exploratory study that is relatively small both in sample size and coverage of health states. Conclusions. This study develops a new, flexible econometric model specifically designed to reflect appropriately the features of rank data. Results support the view that the standard model is not appropriate in this setting and will yield very different and apparently inconsistent results. PBMs can be compared using a common scale by implementation of this new approach.

Previous methods of empirical mapping involve using regressions on patient or general population self-reported data from datasets involving two or more measures. This approach relies on overlap in the descriptive systems of the measures and assumes it is appropriate to use different measures on the same population, which may not always be the case. This paper presents a feasibility study for a new approach to mapping between preference-based measures (PBM) using general population visual analogue scale (VAS) values as a common yardstick. We use data from a valuation study of 502 members of the UK general population, where, using ranking and VAS tasks, interviewees simultaneously valued health states defined by three of six PBM: EQ-5D (generic), SF-6D (generic), HUI2 (generic for children and adults), AQL-5D (asthma specific), OPUS (social care specific) and ICECAP (capabilities). Regression techniques are used to estimate the relationship between these VAS values and the original value set (i.e. 'tariff'). These results are subsequently used to estimate the relationship between all six PBM to enable 'value-based mapping' between measures. This new method of mapping potentially has a useful role in evidence synthesis and cross programme comparisons in studies using different measures.

Objectives Health utility data generated by using the EuroQol five-dimensional (EQ-5D) questionnaire are right bounded at 1 with a substantial gap to the next set of observations, left bounded, and multimodal. These features present challenges to the estimation of the effect of clinical and socioeconomic characteristics on health utilities. Our objective was to develop and demonstrate an appropriate method for dealing with these features. Methods We developed a statistical model that incorporates an adjusted limited dependent variable approach to reflect the upper bound and the large gap in feasible EQ-5D questionnaire values. Further flexibility was then gained by adopting a mixture modeling framework to address the multimodality of the EQ-5D questionnaire distribution. We compared the performance of these approaches with that of those frequently adopted in the literature (linear and Tobit models) by using data from a clinical trial of patients with rheumatoid arthritis. Results We found that three latent classes are appropriate in estimating EQ-5D questionnaire values from function, pain, and sociodemographic factors. Superior performance of the adjusted limited dependent variable mixture model was achieved in terms of Akaike and Bayesian information criteria, root mean square error, and mean absolute error. Unlike other approaches, the adjusted limited dependent variable mixture model fits the data well at high EQ-5D questionnaire levels and cannot predict unfeasible EQ-5D questionnaire values. Conclusions The distribution of the EQ-5D questionnaire is characterized by features that raise statistical challenges. It is well known that standard approaches do not perform well for this reason. This article developed an appropriate method to reflect these features by combining limited dependent variable and mixture modeling and demonstrated superior performance in a rheumatoid arthritis setting. Further refinement of the general framework and testing in other data sets are warranted. Analysis of utility data should apply methods that recognize the distributional features of the data.