Mapping the EQ-5D-5L from the Spanish national health survey functional disability scale through Bayesian networks.

PURPOSE: Preference-based measures are valuable tools for evaluating therapeutic interventions and for cost-effectiveness studies. Mapping procedures are useful when it is not possible to collect these kind of measures. The objective of this study was to evaluate which mapping method is the most appropriate to estimate the EQ-5D-5L index from the Spanish National Health Survey functional disability scale. METHODS: The sample, formed by 5708 older adults (aged 65 years or older), was drawn from the Spanish National Health Survey ("Encuesta Nacional de Salud en España," ENSE in Spanish 2011-2012). The predictions of EQ-5D-5L index were performed with response mapping using Bayesian network (BN), ordered logit (Ologit), and multinomial logistic (ML). The following direct methods were used: ordinary least squares (OLS) and Tobit regression. The intraclass correlation coefficient (ICC), absolute error (MAE), mean squared error (MSE), and root-mean squared error (RMSE) were calculated to compare all models. The predictions of response models were obtained through the expected value method. RESULTS: BN model showed the highest ICC (0.756, 95% confidence interval, CI 0.733-0.777) and lowest MAE (0.110, 95% CI 0.104-0.115). OLS was the model with worse accuracy results with lowest ICC (0.621, 95% CI 0.553-0.681) and highest MAE (0.159, 95%CI: 0.145-0.173). CONCLUSION: Indirect mapping methods (BN, Ologit, and ML) had a better accuracy than the direct methods. The response mapping approach provides a robust method to estimate EQ-5D-5L scores from the functional disability scale.

Obtaining EQ-5D-5L utilities from the disease specific quality of life Alzheimer's disease scale: development and results from a mapping study.

PURPOSE: The Quality of Life Alzheimer's Disease Scale (QoL-AD) is commonly used to assess disease specific health-related quality of life (HRQoL) as rated by patients and their carers. For cost-effectiveness analyses, utilities based on the EQ-5D are often required. We report a new mapping algorithm to obtain EQ-5D indices when only QoL-AD data are available. METHODS: Different statistical models to estimate utility directly, or responses to individual EQ-5D questions (response mapping) from QoL-AD, were trialled for patient-rated and proxy-rated questionnaires. Model performance was assessed by root mean square error and mean absolute error. RESULTS: The response model using multinomial regression including age and sex, performed best in both the estimation dataset and an independent dataset. CONCLUSIONS: The recommended mapping algorithm allows researchers for the first time to estimate EQ-5D values from QoL-AD data, enabling cost-utility analyses using datasets where the QoL-AD but no utility measures were collected.

Simultaneous multioutcome synthesis and mapping of treatment effects to a common scale.

OBJECTIVES: A new method is presented for both synthesizing treatment effects on multiple outcomes subject to measurement error and estimating coherent mapping coefficients between all outcomes. It can be applied to sets of trials reporting different combinations of patient- or clinician-reported outcomes, including both disease-specific measures and generic health-related quality-of-life measures. It is underpinned by a structural equation model that includes measurement error and latent common treatment effect factor. Treatment effects can be expressed on any of the test instruments that have been used. METHODS: This is illustrated in a synthesis of eight placebo-controlled trials of TNF-α inhibitors in ankylosing spondylitis, each reporting treatment effects on between two and five of a total six test instruments. RESULTS: The method has advantages over other methods for synthesis of multiple outcome data, including standardization and multivariate normal synthesis. Unlike standardization, it allows synthesis of treatment effect information from test instruments sensitive to different underlying constructs. It represents a special case of previously proposed multivariate normal models for evidence synthesis, but unlike the former, it also estimates mappings. Combining synthesis and mapping as a single operation makes more efficient use of available data than do current mapping methods and generates treatment effects that are consistent with the mappings. A limitation, however, is that it can only generate mappings to and from those instruments on which some trial data exist. CONCLUSIONS: The method should be assessed in a wide range of data sets on different clinical conditions, before it can be used routinely in health technology assessment.

Mapping between the Roland Morris Questionnaire and generic preference-based measures.

OBJECTIVES: The Roland Morris Questionnaire (RMQ) is a widely used health status measure for low back pain (LBP). It is not preference-based, and there are currently no established algorithms for mapping between the RMQ and preference-based health-related quality-of-life measures. Using data from randomized controlled trials of treatment for LBP, we sought to develop algorithms for mapping between RMQ scores and health utilities derived using either the EuroQol five-dimensional questionnaire (EQ-5D) or the six-dimensional health state short form (derived from Medical Outcomes Study 36-Item Short-Form Health Survey) (SF-6D). METHODS: This study is based on data from the Back Skills Training Trial in which data were collected from 701 patients at baseline and subsequently at 3, 6, and 12 months postrandomization using a range of outcome measures, including the RMQ, the EQ-5D, and the Short Form 12 item Health Survey (SF-12) (from which SF-6D utilities can be derived). We used baseline trial data to estimate models using both direct and response mapping approaches to predict EQ-5D and SF-6D health utilities and dimension responses. A multistage model selection process was used to assess the predictive accuracy of the models. We then explored different techniques and mapping models that made use of repeated follow-up observations in the data. The estimated mapping algorithms were validated using external data from the UK Back Pain Exercise and Manipulation trial. RESULTS: A number of models were developed that accurately predict health utilities in this context. The best performing model for RMQ to EQ-5D mapping was a beta regression with Bayesian quasi-likelihood estimation that included 24 dummy variables for RMQ responses, age, and sex as covariates (mean squared error 0.0380) based on repeated data. The model selected for RMQ to SF-6D mapping was a finite mixture model that included the overall RMQ score, age, sex, RMQ score squared, age squared, and an interaction term for age and RMQ score as covariates (mean squared error 0.0114) based on repeated data. CONCLUSIONS: It is possible to reasonably predict EQ-5D and SF-6D health utilities from RMQ scores and responses using regression methods. Our regression equations provide an empirical basis for estimating health utilities when EQ-5D or SF-6D data are not available. They can be used to inform future economic evaluations of interventions targeting LBP.

Mapping the University of Washington Quality of life questionnaire onto EQ-5D and HUI-3 indices in patients with head and neck cancer.

BACKGROUND: There is no mechanism to predict health utility (HU) values from the University of Washington Quality of Life Questionnaire (UWQoL) scores. We sought to develop a mapping algorithm capable of using UWQoL data to approximate HU scores. METHODS: Outpatients with head and neck cancer completed the UWQoL, EQ-5D, and the Health Utilities Index-Mark 3 (HUI-3). Results of the UWQoL were mapped onto both EQ-5D and HUI-3 scores using ordinary least-squares regression models. Two-part models were explored. The predictive power of the model was assessed using 10-fold cross-validation. RESULTS: A total of 209 patients were recruited. The reduced model converting UWQoL data into EQ-5D scores performed best (adjusted R2 = 0.628, root mean square error = 0.076). Both models demonstrated construct validity by discriminating between clinical indices of disease severity. CONCLUSIONS: The abovementioned algorithms enable researchers to perform health economic evaluations with existing UWQoL data in cases where prospectively collected HU values are not available.

Mapping the EORTC QLQ-C30 and QLQ-H&N35, onto EQ-5D-5L and HUI-3 indices in patients with head and neck cancer.

BACKGROUND: We sought to develop mapping functions that use EORTC responses to approximate health utility (HU) scores for patients with head and neck cancer (HNC). METHODS: In total, 209 outpatients with HNC completed the EORTC QLQ-C30 & QLQ-H&N35 (EORTC), EQ-5D-5L and the HUI-3. Results of the EORTC were mapped onto both EQ-5D-5L and HUI-3 scores using ordinary least squares regression and two-part models. RESULTS: The OLS model mapping EORTC onto the EQ-5D-5L performed best (adjusted R2 = .75, 10-fold cross-validation RMSE = 0.064, MAE 0.050). The HUI-3 model mapping onto EORTC through OLS was more limited (adjusted R2 = .5746, 10-fold cross cross-validation RMSE = 0.168, MAE 0.080). The EQ-5D-5L model was able to discriminate between certain clinical indices of disease severity on subgroup analysis. CONCLUSION: The EORTC to EQ-5D-5L mapping algorithm has good predictive validity and may enable researchers to translate EORTC scores into HU scores for head and neck patients with cancer.

Translating CESD-20 and PHQ-9 Scores to PROMIS Depression.

This study examined the accuracy of depression cross-walk tables in a sample of people with multiple sclerosis (MS). The tables link scores of two commonly used depression measures to the Patient Reported Outcome Measurement Information System Depression (PROMIS-D) scale metric. We administered the 8-item PROMIS-D (Short-Form 8b; PROMIS-D-8), the 20-item Center for Epidemiologic Studies Depression Scale (CESD-20), and the 9-item Patient Health Questionnaire (PHQ-9) to 459 survey participants with MS. We examined correlations between actual PROMIS-D-8 scores and the scores predicted by cross-walks based on PHQ-9 and CESD-20 scores. Intraclass correlation coefficients were used to assess correspondence. Consistency in severity classification was also calculated. Finally, we used Bland-Altman plots to graphically examine the levels of agreement. The correlations between actual and cross-walked PROMIS-D-8 scores were strong (CESD-20 = .82; PHQ-9 = .74). The intraclass correlation was moderate (.77). Participants were consistently classified as having or not having at least moderate depressive symptoms by both actual and cross-walked scores derived from the CESD-20 (90%) and PHQ-9 (85%). Bland-Altman plots suggested the smaller differences between actual and cross-walked scores with greater-than-average depression severity. PROMIS cross-walk tables can be used to translate depression scores of people with MS to the PROMIS-D metric, promoting continuity with previous research.

Mapping algorithms from QLQ-C30 to EQ-5D utilities: no firm ground to stand on yet.

AIM: Over the last years several mapping or cross-walking algorithms for deriving utilities from QLQ-C30 scores have been published. However their external predictive accuracy has not yet been systematically compared. METHODS: We tested the external validity of previously published mapping algorithms to transform the European Organization for Research and Treatment of Cancer QLQ-C30 questionnaire responses to EQ-5D derived Utilities. RESULTS: When applied to different data sets, the currently published mapping showed a large variation between algorithms of the values of the mapped utilities, a low accuracy of the mapping compared to the observed EQ-5D utilities and no consistent performance between competing algorithms. DISCUSSION: Therefore direct mapping from QLQ-C30 profiles to EQ-5D utilities using published algorithms should be viewed cautiously.