Derivation of Utility Values from European Organization for Research and Treatment of Cancer Quality of Life-Core 30 Questionnaire Values in Lung Cancer

• Published in: Journal of thoracic oncology 2010-12, Vol.5 (12), p.1953-1957
• Main Authors: Jang, Raymond W., Isogai, Pierre K., Mittmann, Nicole, Bradbury, Penelope A., Shepherd, Frances A., Feld, Ronald, Leighl, Natasha B.
• Format: Article
• Language: English
• Subjects: Adult; Aged; Aged, 80 and over; Carcinoma, Non-Small-Cell Lung - psychology; Carcinoma, Non-Small-Cell Lung - therapy; Derivation of utility values; Economic evaluation; EORTC QLQ-C30; EQ-5D; Female; Humans; Linear Models; Lung Neoplasms - psychology; Lung Neoplasms - therapy; Male; Middle Aged; Non-small cell lung cancer; Quality of Life; Surveys and Questionnaires
• ISSN: 1556-0864; 1556-1380
• DOI: 10.1097/JTO.0b013e3181f77a6a

Cancer clinical trials frequently incorporate quality of life (QoL) measures but rarely patient utility. Utility information is required for cost utility evaluations of novel cancer therapies. We assessed the feasibility of converting QoL data into utility scores using the European Organization for Research and Treatment of Cancer Quality of Life-Core 30 questionnaire (EORTC QLQ-C30) and the EQ-5D in patients with non-small cell lung cancer (NSCLC). Outpatients with all different disease states of NSCLC attending a major Canadian cancer center completed the QLQ-C30 and EQ5D on a single visit. Results of the QLQ-C30 summary scores were mapped to predict EQ-5D utility scores using linear regression. Backward variable elimination using the Akaike Information Criterion was used to reduce the full model that included all QLQ-C30 summary scores to examine which QLQ-C30 dimensions best predict a patient's utility score. To test the predictive power of the model, 10-fold cross-validation was used. A total of 172 patients participated in the study. Median age of the sample was 66 years (range, 32–85 years); 46.5% were men. The cross-validation estimate of mean utility score was 0.76 (SD: 0.20), which was the same as the actual mean utility score. Of the 15 QLQ-C30 dimensions, 4 functional dimensions (physical, role, emotional, and social) and the pain symptom dimension were predictive of patient utility scores. Our study demonstrates the feasibility of deriving utility scores from prospective QoL data. Validation of the QLQ-C30 predictors found in this study could further the ability to estimate cost utility of therapies for economic evaluations.