A patient-based utility measure of health for clinical trials of cancer therapy based on the European Organization for the Research and Treatment of Cancer Quality of Life Questionnaire

United States Utility Algorithm for the EORTC QLU-C10D, a Multiattribute Utility Instrument Based on a Cancer-Specific Quality-of-Life Instrument

BACKGROUND

The EORTC QLU-C10D is a multiattribute utility measure derived from the cancer-specific quality-of-life questionnaire, the EORTC QLQ-C30. The QLU-C10D contains 10 dimensions (physical, role, social and emotional functioning, pain, fatigue, sleep, appetite, nausea, bowel problems). The objective of this study was to develop a United States value set for the QLU-C10D.

METHODS

A US online panel was quota recruited to achieve a representative sample for sex, age (≥18 y), race, and ethnicity. Respondents undertook a discrete choice experiment, each completing 16 choice-pairs, randomly assigned from a total of 960 choice-pairs. Each pair included 2 QLU-C10D health states and duration. Data were analyzed using conditional logistic regression, parameterized to fit the quality-adjusted life-year framework. Utility weights were calculated as the ratio of each dimension-level coefficient to the coefficient for life expectancy.

RESULTS

A total of 2480 panel members opted in, 2333 (94%) completed at least 1 choice-pair, and 2273 (92%) completed all choice-pairs. Within dimensions, weights were generally monotonic. Physical functioning, role functioning, and pain were associated with the largest utility weights. Cancer-specific dimensions, such as nausea and bowel problems, were associated with moderate utility decrements, as were general issues such as problems with emotional functioning and social functioning. Sleep problems and fatigue were associated with smaller utility decrements. The value of the worst health state was 0.032, which was slightly greater than 0 (equivalent to being dead).

CONCLUSIONS

This study provides the US-specific value set for the QLU-C10D. These estimated health state scores, based on responses to the EORTC QLQ-C30 questionnaire, can be used to evaluate the cost-utility of oncology treatments.

Quality of Life After Curative Resection for Gastric Cancer: Survey Metrics and Implications of Surgical Technique

Gastric cancer is one of the most common cancers worldwide, and radical gastrectomy is an integral component of curative therapy. With improvements in perioperative morbidity and mortality, attention has turned to short- and long-term post-gastrectomy quality of life (QoL). This article reviews the common psychometric surveys and preference-based measures used among patients following gastrectomy. It also provides an overview of studies that address associations between surgical decision-making and postoperative health-related QoL. Further attention is focused on reported associations between technical aspects of the operation, such as extent of gastric resection, minimally-invasive approach, pouch-based conduits, enteric reconstruction, and postoperative QoL. While there are several randomized studies that include QoL outcomes, much remains to be explored. The relationship between symptom profiles and preference-based measures of health state utility is an area in need of further research.

Health Technology Assessment Challenges in Oncology: 20 Years of Value in Health

BACKGROUND

Oncology treatments have changed from chemotherapies to targeted therapies and more recently immuno-oncology. This has posed special challenges in the field of health technology assessment (HTA): capturing quality of life (QOL) associated with toxicity due to chemotherapy, crossover upon progression in targeted therapy trials, and survival extrapolation for immuno-oncology drugs.

OBJECTIVES

To showcase 20 years of Value in Health (ViH) publications in oncology.

METHODS

A review was undertaken of oncology articles published in ViH from May 1998 to August 2018. Full-length articles published in ViH with the keywords "oncology," "cancer," "h(a)ematology," and "malignancy" were included for review. Conference abstracts were excluded.

RESULTS

Four major themes were identified: (1) QOL and the development of multiple functional assessment of cancer therapy tools and mapping instruments; (2) analysis of clinical evidence using indirect comparisons, network analyses, and adjustment for crossovers; (3) modeling, Markov models, partitioned survival models, and extrapolation methods; and (4) financial implications and how to deal with uncertainty, introduction of conditional reimbursement, managed entry, and risk share agreements.

DISCUSSION

This review article highlights the important role ViH has played in disseminating HTA research in oncology. A few key issues loom on the horizon: precision medicine, further development and practical application of new QOL measures, methods for translating clinical evidence, and exploration of modeling techniques. For a better understanding of the complex interplay between access and financial risk management, ViH will no doubt continue to promote pioneering research in HTA and oncology.

Patient Experience-based Value Sets: Are They Stable?

BACKGROUND

Although societal preference weights are desirable to inform resource-allocation decision-making, patient experienced health state-based value sets can be useful for clinical decision-making, but context may matter.

OBJECTIVE

To estimate EQ-5D value sets using visual analog scale (VAS) ratings for patients undergoing knee replacement surgery and compare the estimates before and after surgery.

METHODS

We used the Patient Reported Outcome Measures data collected by the UK National Health Service on patients undergoing knee replacement from 2009 to 2012. Generalized least squares regression models were used to derive value sets based on the EQ-5D-3 level using a development sample before and after surgery, and model performance was examined using a validation sample.

RESULTS

A total of 90,450 preoperative and postoperative valuations were included. For preoperative valuations, the largest decrement in VAS values was associated with the dimension of anxiety/depression, followed by self-care, mobility, usual activities, and pain/discomfort. However, pain/discomfort had a greater impact on VAS value decrement in postoperative valuations. Compared with preoperative health problems, postsurgical health problems were associated with larger value decrements, with significant differences in several levels and dimensions, including level 2 of mobility, level 2/3 of usual activities, level 3 of pain/discomfort, and level 3 of anxiety/depression. Similar results were observed across subgroups stratified by age and sex.

CONCLUSIONS

Findings suggest patient experience-based value sets are not stable (ie, context such as timing matters). However, the knowledge that lower values are assigned to health states postsurgery compared with presurgery may be useful for the patient-doctor decision-making process.

Quality of Life Following Sentinel Node Biopsy for Primary Cutaneous Melanoma: Health Economic Implications

BACKGROUND

Sentinel node biopsy (SNB) is commonly performed in contemporary melanoma management, however there is a paucity of long-term quality of life (QoL) estimates required for economic evaluation of this treatment.

METHODS

A single-center, prospective, cross-sectional study of adults with American Joint Committee on Cancer stage I/II/IIIA melanoma of the limbs, trunk, or neck who had undergone wide excision and SNB, but not complete regional node dissection, was undertaken. Limb volume was measured using perometry, with lymphedema defined as a ≥10% volume increase in the ipsilateral limb compared with the contralateral limb. The Functional Assessment of Cancer Therapy-Breast (FACT-B) questionnaire measured QoL. Associations between patient and treatment characteristics were assessed using linear regression.

RESULTS

Among 694 patients (median time from SNB of 37 months), 14 (2%) had objectively measured lymphedema (i.e. an increase in limb volume of ≥10%). Of 687 stage I/II patients with complete QoL data, the mean weighted QoL was 0.745 (standard deviation 0.04) on a 0-1 scale (i.e. death to full health). In multivariable analysis, weighted QoL was 0.0004 higher for each year of increasing age (p = 0.001); 0.011 lower for females (p = 0.001), 0.018 lower following post-SNB limb trauma (p = 0.002); 0.252 lower for patients who perceived a large increase in limb size (p = 0.015); and 0.027 lower with self-reported difficulty in walking, running, or climbing stairs (p = 0.043).

CONCLUSIONS

Our data suggest that very few patients treated at our institution had lymphedema in the long-term following SNB, with weighted QoL strongly associated with perceived rather than actual changes in limb size.

A Systematic Review of the Literature on the Development of Condition-Specific Preference-Based Measures of Health

BACKGROUND

Health state utility values (HSUVs) are required to calculate quality-adjusted life-years (QALYs). They are frequently derived from generic preference-based measures of health. However, such generic measures may not capture health attributes of relevance to specific conditions. In such cases, a condition-specific preference-based measure (CSPBM) may be more appropriate.

OBJECTIVE

This systematic review aimed to identify all published accounts of developing CSPBMs to describe and appraise the methods used.

METHOD

We undertook a systematic search (of Embase, MEDLINE, PsycINFO, Web of Science, the Cochrane Library, CINAHL, EconLit, ASSIA and the Health Management Information Consortium database) to identify published accounts of CSPBM development up to July 2015. Studies were reviewed to investigate the methods used to design classification systems, estimate HSUVs, and validate the measures.

RESULTS

A total of 86 publications were identified, describing 51 CSPBMs. Around two-thirds of these were QALY measures; the remainder were designed for clinical decision making only. Classification systems for 33 CSPBMs were derived from existing instruments; 18 were developed de novo. HSUVs for 34 instruments were estimated using a 'composite' approach, involving statistical modelling; the remainder used a 'decomposed' approach based on multi-attribute utility theory. Half of the papers that described the estimation of HSUVs did not report validating their measures.

CONCLUSION

Various methods have been used at all stages of CSPBM development. The choice between developing a classification system de novo or from an existing instrument may depend on the availability of a suitable existing measure, while the choice between a decomposed or composite approach appears to be determined primarily by the purpose for which the instrument is designed. The validation of CSPBMs remains an area for further development.

QLU-C10D: a health state classification system for a multi-attribute utility measure based on the EORTC QLQ-C30

PURPOSE

To derive a health state classification system (HSCS) from the cancer-specific quality of life questionnaire, the EORTC QLQ-C30, as the basis for a multi-attribute utility instrument.

METHODS

The conceptual model for the HSCS was based on the established domain structure of the QLQ-C30. Several criteria were considered to select a subset of dimensions and items for the HSCS. Expert opinion and patient input informed a priori selection of key dimensions. Psychometric criteria were assessed via secondary analysis of a pooled dataset comprising HRQOL and clinical data from 2616 patients from eight countries and a range of primary cancer sites, disease stages, and treatments. We used confirmatory factor analysis (CFA) to assess the conceptual model's robustness and generalisability. We assessed item floor effects (>75 % observations at lowest score), disordered item response thresholds, coverage of the latent variable and differential item function using Rasch analysis. We calculated effect sizes for known group comparisons based on disease stage and responsiveness to change. Seventy-nine cancer patients assessed the relative importance of items within dimensions.

RESULTS

CFA supported the conceptual model and its generalisability across primary cancer sites. After considering all criteria, 12 items were selected representing 10 dimensions: physical functioning (mobility), role functioning, social functioning, emotional functioning, pain, fatigue, sleep, appetite, nausea, bowel problems.

CONCLUSIONS

The HSCS created from QLQ-C30 items is known as the EORTC Quality of Life Utility Measure-Core 10 dimensions (QLU-C10D). The next phase of the QLU-C10D's development involves valuation studies, currently planned or being conducted across the globe.

Mapping Functions in Health-Related Quality of Life: Mapping from Two Cancer-Specific Health-Related Quality-of-Life Instruments to EQ-5D-3L

Background. Clinical trials in cancer frequently include cancer-specific measures of health but not preference-based measures such as the EQ-5D that are suitable for economic evaluation. Mapping functions have been developed to predict EQ-5D values from these measures, but there is considerable uncertainty about the most appropriate model to use, and many existing models are poor at predicting EQ-5D values. This study aims to investigate a range of potential models to develop mapping functions from 2 widely used cancer-specific measures (FACT-G and EORTC-QLQ-C30) and to identify the best model. Methods. Mapping models are fitted to predict EQ-5D-3L values using ordinary least squares (OLS), tobit, 2-part models, splining, and to EQ-5D item-level responses using response mapping from the FACT-G and QLQ-C30. A variety of model specifications are estimated. Model performance and predictive ability are compared. Analysis is based on 530 patients with various cancers for the FACT-G and 771 patients with multiple myeloma, breast cancer, and lung cancer for the QLQ-C30. Results. For FACT-G, OLS models most accurately predict mean EQ-5D values with the best predicting model using FACT-G items with similar results using tobit. Response mapping has low predictive ability. In contrast, for the QLQ-C30, response mapping has the most accurate predictions using QLQ-C30 dimensions. The QLQ-C30 has better predicted EQ-5D values across the range of possible values; however, few respondents in the FACT-G data set have low EQ-5D values, which reduces the accuracy at the severe end. Conclusions. OLS and tobit mapping functions perform well for both instruments. Response mapping gives the best model predictions for QLQ-C30. The generalizability of the FACT-G mapping function is limited to populations in moderate to good health.

Using existing data to identify candidate items for a health state classification system in multiple sclerosis

PURPOSE

In multiple sclerosis (MS), the use of preference-based measures is limited to generic measures such as Health Utilities Index Mark 2 and 3, the EQ-5D and the SF-6D. However, the challenge of using such generic preference-based measures in people with MS is that they may not capture all domains of health relevant to the disease. Therefore, the main aim of this paper is to describe the development of a health state classification system for MS patients. The specific objectives are: (1) to identify items best reflecting the domains of quality of life important to people with MS and (2) to provide evidence for the discriminative capacity of the response options by cross-walking onto a visual analog scale of health rating.

METHODS

The data come from an epidemiologically sampled population of people with MS diagnosed post-1994. The dataset consisted of 206 items relating to impairments, activity limitations, participation restrictions, health perception and quality of life. Important domains were identified from the responses to the Patient Generated Index, an individualized measure of quality of life. The extent to which the items formed a uni-dimensional, linear construct was estimated using Rasch analysis, and the best item was selected using the threshold map.

RESULTS

The sample was young (mean age 43) and predominantly female (n = 140/189; 74%). The P-PBMSI classification system consisted of five items, with three response levels per item, producing a total of 243 possible health states. Regression coefficient values consistently decreased between response levels and the linear test for trend were statistically significant for all items. The linear test for trend indicated that for each item the response options provided the same discriminative ability within the magnitude of their capacity. A scoring algorithm was estimated using a simple additive formula. The classification system demonstrated convergent validity against other measures of similar constructs and known-groups validity between different clinical subgroups.

CONCLUSION

This study produced a health state classifier system based on items impacted upon by MS, and demonstrated the potential to discriminate the health impact of the disease.

Predicting SF-6D from the European Organization for Treatment and Research of Cancer Quality of Life Questionnaire scores in patients with colorectal cancer

OBJECTIVES

To develop a mapping model for estimating six-dimensional health state short form (SF-6D) utility scores from the European Organization for Research and Treatment of Cancer Quality of Life Questionnaires (QLQ-C30 and QLQ-CR29) scores in patients with colorectal cancer (CRC), with and without adjustment for clinical and demographic characteristics.

METHODS

Ordinary least squares regression models were applied to a cross-sectional data set of 216 patients with CRC collected from a regional hospital in Hong Kong. Item responses or scale scores of cancer-specific (QLQ-C30) and colorectal-specific health-related quality-of-life (QLQ-CR38/CR29) data and selected demographic and clinical characteristics of patients were used to predict the SF-6D scores. Model goodness of fit was examined by using exploratory power (R(2) and adjusted R(2)), Akaike information criterion, and Bayesian information criterion, and predictive performance was evaluated by using root mean square error, mean absolute error, and Spearman's correlation coefficients between predicted and observed SF-6D scores. Models were validated by using an independent data set of 56 patients with CRC.

RESULTS

Both scale and item response models explained more than 67% of the variation in SF-6D scores. The best-performing model based on goodness of fit (R(2) = 75.02%), predictive ability in the estimation (root mean square error = 0.080, mean absolute error = 0.065), and validation data set prediction (root mean square error = 0.103, mean absolute error = 0.081) included variables of main and interaction effects of the QLQ-C30 supplemented by QLQ-CR29 subset scale responses and a demographic (sex) variable.

CONCLUSIONS

SF-6D scores can be predicted from QLQ-C30 and QLQ-CR38/CR29 scores with satisfactory precision in patients with CRC. The mapping model can be applied to QLQ-C30 and QLQ-CR38/CR29 data sets to produce utility scores for the appraisal of clinical interventions targeting patients with CRC using economic evaluation.

Comparison of generic, condition-specific, and mapped health state utility values for multiple myeloma cancer

OBJECTIVE

Resource allocation informed by cost-utility analysis requires that the benefits be comparable across patient groups and interventions. One option is to recommend the use of one generic utility measure, but this raises the issue of comparability when the preferred measure is inappropriate or unavailable. Many cancer trials do not include generic measures such as the EuroQol five-dimensional (EQ-5D) questionnaire and instead include condition-specific measures and use these to generate utility estimates. We analyze the comparability of generic, condition-specific, and mapped utility values for a multiple myeloma cancer patient data set.

METHODS

Generic EQ-5D, condition-specific EORTC-8D, and EQ-5D utility values mapped from the EORTC QLQ-C30 were compared by using psychometric and statistical analysis to determine discrimination across severity groups, responsiveness, and agreement.

RESULTS

Generic, condition-specific, and mapped utility estimates were responsive over time and show discriminative validity. The EQ-5D had higher responsiveness and detected a greater change across severity groups and treatment periods than did the EORTC-8D but has a higher proportion of responses at full health (12.8%). Differences in the EQ-5D and the EORTC-8D were due at least in part to differences in the classification system. Mapped EQ-5D estimates had a smaller SD and do not reflect the severe range of health states reported by using the EQ-5D.

CONCLUSIONS

Our findings suggest that condition-specific EORTC-8D or mapped EQ-5D utility estimates are broadly comparable to directly obtained EQ-5D utilities for a multiple myeloma patient data set. However, EORTC-8D estimates captured changes in quality of life for patients in mild health states that were not captured by the EQ-5D, but estimated lower utility gains than did the use of the EQ-5D directly.

Evaluation of content on EQ-5D as compared to disease-specific utility measures

PURPOSE

The goal of this study was to appraise the extent of unique content on disease-specific preference-based measures (DSPMs) when contrasted with the EQ-5D using published studies and to inform whether EQ-5D could be inadequate as a utility measure in its content coverage for a given disease-specific application.

METHODS

A structured search of published literature was performed using PubMed and EMBASE/Medline database from Jan 1, 1990 to Mar 31, 2011. Articles were eligible for inclusion if algorithms were developed to convert components from disease-specific measures into utility scores.

RESULTS

Of 1,029 articles identified, 50 studies satisfied the inclusion criteria. The most frequent conditions where DSPMs were developed included cancer (12 studies), coronary artery disease (4 studies), osteoarthritis, rheumatoid arthritis (3 studies of each), obesity, and stroke (2 studies of each). Most studies involved mapping items or scores from disease-specific non-preference-based measures onto a preference-based measure of health such as the EQ-5D. A substantial number of DSPMs appeared to include unique content not covered by EQ-5D dimensions.

CONCLUSIONS

Several conditions were identified as potential areas where the richness of the EQ-5D descriptive system could be enhanced. It is yet unclear whether added dimension(s) would contribute unique explained variance to a utility score. Given the resources required to rigorously develop a utility measure, the need for such measures should be carefully vetted.

Mapping the Functional Assessment of Cancer Therapy-general or -Colorectal to SF-6D in Chinese patients with colorectal neoplasm

OBJECTIVES

To map Functional Assessment of Cancer Therapy-General (FACT-G) and Functional Assessment of Cancer Therapy-Colorectal (FACT-C) subscale scores onto six-dimensional health state short form (derived from short form 36 health survey) (SF-6D) preference-based values in patients with colorectal neoplasm, with and without adjustment for clinical and demographic characteristics. These results can then be applied to studies that have used FACT-G or FACT-C to predict SF-6D utility values to inform economic evaluation.

METHODS

Ordinary least square regressions were estimated mapping FACT-G and FACT-C onto SF-6D by using cross-sectional data of 537 Chinese subjects with different stages of colorectal neoplasm. Mapping functions for SF-6D preference-based values were developed separately for FACT-G and FACT-C in four sequential models for addition of variables: 1) main-effect terms, 2) squared terms, 3) interaction terms, and 4) clinical and demographic variables. Predictive performance in each model was assessed by the R(2), adjusted R(2), predicted R(2), information criteria (Akaike information criteria and Bayesian information criteria), the root mean square error, the mean absolute error, and the proportions of absolute error within the threshold of 0.05 and 0.10.

RESULTS

Models including FACT variables and clinical and demographic variables had the best predictive performance measured by using R(2) (FACT-G: 59.98%; FACT-C: 60.43%), root mean square error (FACT-G: 0.086; FACT-C: 0.084), and mean absolute error (FACT-G: 0.065; FACT-C: 0.065). The FACT-C-based mapping function had better predictive ability than did the FACT-G-based mapping function.

CONCLUSIONS

Models mapping FACT-G and FACT-C onto SF-6D reached an acceptable degree of precision. Mapping from the condition-specific measure (FACT-C) had better performance than did mapping from the general cancer measure (FACT-G). These mapping functions can be applied to FACT-G or FACT-C data sets to estimate SF-6D utility values for economic evaluation of medical interventions for patients with colorectal neoplasm. Further research assessing model performance in independent data sets and non-Chinese populations are encouraged.

Mapping for economic evaluation

INTRODUCTION/BACKGROUND

Mapping provides a statistical algorithm that allows the estimation of utilities and consequently calculation of QALYs in clinical studies where preference-based measures are not implemented.

SOURCES OF DATA

Reviews of the mapping literature were utilized.

AREAS OF AGREEMENT

Mapping requires similar populations between the estimation and study data sets, with a high degree of overlap between the target and base measures being desirable. The National Institute for Health and Clinical Excellence recognizes mapping as a method to provide utility information. Areas of controversy Issues surrounding mapping include the descriptive system of the measure, the appropriate econometric method and model specification.

GROWING POINTS

There is a need for further research into the issue of over-prediction for severe health states and uncertainty around the estimated utility scores.

AREAS TIMELY FOR DEVELOPING RESEARCH

Mapping continues to be an important area of research for economic evaluation, in particular validation of mapping functions.

Deriving a preference-based measure for cancer using the EORTC QLQ-C30

OBJECTIVE

The European Organization for Research and Treatment of Cancer Core Quality of Life Questionnaire (EORTC QLQ-C30) is one of the most commonly used measures in cancer care but in its current form cannot be used in economic evaluation because it does not incorporate preferences. We address this gap by estimating a preference-based measure for cancer from the EORTC QLQ-C30.

METHODS

Factor analysis, Rasch analysis, and other psychometric analyses were undertaken on a clinical trial dataset of 655 patients with multiple myeloma to derive a health state classification system amenable to valuation. Second a valuation study was conducted of 350 members of the UK general population using time trade-off. Mean and individual-level multivariate regression models were fitted to derive preference weights for the classification system.

RESULTS

The health state classification system has eight dimensions (physical functioning, role functioning, social functioning, emotional functioning, pain, fatigue and sleep disturbance, nausea, constipation, and diarrhea) with four or five levels each. Regression models have few inconsistencies (0 to 2) in estimated preference weights and small mean absolute error ranges (0.046 to 0.054).

CONCLUSIONS

It is feasible to derive a preference-based measure from the EORTC QLQ-C30 for use in economic evaluation. Future research will extend this to other countries and replicate across other patient groups.

Use of a disease-specific instrument in economic evaluations: mapping WOMAC onto the EQ-5D utility index

OBJECTIVE

To map the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) onto the EuroQol 5 Dimension (EQ-5D) utility index in patients with knee osteoarthritis (OA).

METHODS

A consecutive sample of patients (n=258) diagnosed with knee OA completed both the WOMAC and the EQ-5D. Regression models with the ordinary least squares (OLS) or the censored least absolute deviations as the estimator were used to establish the mapping function. The WOMAC was represented as explanatory variables in four ways: 1) total score; 2) domain scores (i.e., pain, stiffness, and physical function); 3) domain scores plus pair-wise interaction terms to account for possible nonlinearities; and 4) individual item scores. Goodness-of-fit criteria included the mean absolute error (the primary criterion) and the root mean squared error, and were obtained using an iterative random sampling procedure. Prediction precision was evaluated at individual patient level and at the group level.

RESULTS

The model using the OLS estimator and the WOMAC domain scores as explanatory variables had the best fit and was chosen as the preferred mapping model. The prediction error at the individual level exceeded the maximal tolerance value (i.e., the minimally important difference of the EQ-5D) in about 16% of the patients. At the group level, the width of the 95% confidence interval of prediction errors varied from 0.0176 at a sample size of 400 to 0.0359 at a sample size of 100.

CONCLUSIONS

EQ-5D scores can be predicted using WOMAC domain scores with an acceptable precision at both individual and group levels in patients with mild to moderate knee OA.

Mapping the QLQ-C30 quality of life cancer questionnaire to EQ-5D patient preferences

BACKGROUND

Although cancer-specific Health-Related Quality of Life (HRQOL) are commonly included in randomized clinical trials or other prospective non-randomized clinical studies, it is rare that preference-based instruments are used that allow the calculation of a utility weight suitable for estimating quality-adjusted life-years gained.

OBJECTIVE

To develop a mapping algorithm to transform the EORTC QLQ-C30 questionnaire responses into EQ-5D derived utilities.

STUDY DESIGN

Retrospective data analysis of a multicentre, multicountry prospective clinical trial in breast cancer patients.

METHODS

Regression analysis of individual pairs of EQ-5D and QLQ-C30 scores.

RESULTS

A model that explained 80% of the variance was developed to estimate EQ-5D Utilities from QLQ-C30 scores at individual level. From this reliable group level means and deviations can be derived.

CONCLUSIONS

Mapping from QLQ-C30 scores to EQ-5D-derived utilities when only QLQ-C30 data are available has been shown to be possible with good accuracy. Validation of the proposed algorithm in other external clinical datasets should be encouraged.

A median model for predicting United States population-based EQ-5D health state preferences

OBJECTIVE

The D1 model that was developed to predict US societal preferences for EQ-5D health states addressed several important conceptual and statistical issues. However, it has been criticized for being too complex, failing to account for the nonnormal distribution of health state values, and the transformation of preferences for worse-than-death health states before estimation. This research was conducted to develop an improved model for predicting median preferences for EQ-5D health states for the US population.

METHODS

Probability-weighted least absolute deviations regression was used to fit models to the time trade-off data collected in the US Valuation of the EQ-5D Health States study. No transformation was applied to the values for states considered worse than death. Several model specifications that differed with respect to explanatory variables were evaluated using two-sample cross-validation.

RESULTS

The best-fitting model included only fixed effects for moderate or severe problems in each of the 5 EQ-5D dimensions and excluded a constant. This specification yielded rank correlations between observed and predicted values and median observed and predicted values of 0.635 and 0.991, respectively, as well as a median absolute error of 0.026. The predicted median preferences ranged from 1.00 for full health, to -0.81 for the worst possible health state.

CONCLUSIONS

Due to its simplicity and robustness, a median model is superior to other models for predicting US population preferences for EQ-5D health states. The predictions of this model are suggested for use in applications that require US societal health state values.