A non-linear beta-binomial regression model for mapping EORTC QLQ- C30 to the EQ-5D-3L in lung cancer patients: a comparison with existing approaches

Abstract

BackgroundThe performance of the Beta Binomial (BB) model is compared with several existing models for mapping the EORTC QLQ-C30 (QLQ-C30) on to the EQ-5D-3L using data from lung cancer trials.MethodsData from 2 separate non small cell lung cancer clinical trials (TOPICAL and SOCCAR) are used to develop and validate the BB model. Comparisons with Linear, TOBIT, Quantile, Quadratic and CLAD models are carried out. The mean prediction error, R2, proportion predicted outside the valid range, clinical interpretation of coefficients, model fit and estimation of Quality Adjusted Life Years (QALY) are reported and compared. Monte-Carlo simulation is also used.ResultsThe Beta-Binomial regression model performed ‘best’ among all models. For TOPICAL and SOCCAR trials, respectively, residual mean square error (RMSE) was 0.09 and 0.11; R2 was 0.75 and 0.71; observed vs. predicted means were 0.612 vs. 0.608 and 0.750 vs. 0.749. Mean difference in QALY’s (observed vs. predicted) were 0.051 vs. 0.053 and 0.164 vs. 0.162 for TOPICAL and SOCCAR respectively. Models tested on independent data show simulated 95% confidence from the BB model containing the observed mean more often (77% and 59% for TOPICAL and SOCCAR respectively) compared to the other models. All algorithms over-predict at poorer health states but the BB model was relatively better, particularly for the SOCCAR data.ConclusionThe BB model may offer superior predictive properties amongst mapping algorithms considered and may be more useful when predicting EQ-5D-3L at poorer health states. We recommend the algorithm derived from the TOPICAL data due to better predictive properties and less uncertainty.Electronic supplementary materialThe online version of this article (doi:10.1186/s12955-014-0163-7) contains supplementary material, which is available to authorized users.

The validation of published utility mapping algorithms: an example of EORTC QLQ-C30 and EQ-5D in non-small cell lung cancer

“…the mappings appear to have been a good fit to the data on which they were based on, but not to the dataset we had available ( Table 2). This poor fit is also reflected in the validation by Woodcock & Doble where MAE values for the betabinomial model by Khan and Morris [13] was higher (0.212) than both the validation reported here (0.176) and in the original dataset (0.10).…”

“…The literature search covered all publications to 17 July 2018. Seven publications were identified, of which three met the inclusion criteria of mappings between the QLQ-C30 and EQ-5D derived using lung cancer patients and scored using UK tariffs ( [12][13][14]. The PRISMA diagram is presented in the electronic supplementary material.…”

Bayesian statistical models to estimate EQ‐5D utility scores from EORTC QLQ data in myeloma

“…In conclusion, we found that mean EQ‐5D utility weights can be accurately estimated using a TPM regression mapping algorithm from the EORTC QLQ‐C30/QLQ‐MY20. Whilst previous models for mapping the EORTC QLQ‐C30 to the EQ‐5D exist, this is the first model to our knowledge to explicitly consider a myeloma subgroup and to include the MY‐20 data. Such a model will be of significant use to investigators conducting economic evaluations, by generating preference‐based utility weights in patients with myeloma.…”

Does Selecting Covariates Using Factor Analysis in Mapping Algorithms Improve Predictive Accuracy? A Case of Predicting EQ-5D-5L and SF-6D Utilities from the Women’s Health Questionnaire

“…This is an issue only for the EQ-5D-5L, which can have negative utility values. Following previous practice, this limitation can be overcome by setting all negative utility values to 0 provided the proportion of such values is small [62]. This was however not necessary because there were no negative EQ-5D-5L utility scores in our data.…”