Related Experiment Video
Updated: Oct 9, 2026
A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
Developing thresholds for clinical importance for the EORTC Quality of Life Questionnaire - Elderly 14 (QLQ-ELD14) to
Lisa M Storz1, Maike Thumfart1, Juan I Arraras2
1Health Outcomes Research Unit, University Hospital of Psychiatry II, Medical University of Innsbruck, Innsbruck, Austria.
Introduction:
To support the clinical interpretation of individual patients' health-related quality of life, this study aimed to define thresholds for clinical importance (TCIs) for the seven scales of the European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire Elderly-14 (EORTC QLQ-ELD14).
Materials And Methods:
In this prospective, cross-sectional study, older patients (≥70 years) with mixed cancer diagnoses completed the validated EORTC QLQ-ELD14 alongside an anchor questionnaire. These anchors assessed the clinical importance of scale scores relying on three criteria: limitations in daily life, the need for help/care, and the worries of patients and their partner/family. By categorizing patients based on these anchors, receiver operating characteristic (ROC) curve analysis was used to develop TCIs.
Results:
Data from 251 patients across seven European countries (mean age: 77.2 years; 53.8% female) were analysed. For five of the seven generated TCIs, the EORTC QLQ-ELD14 scales exhibited sensitivity values ranging from 0.75 to 0.91 and specificity values ranging from 0.74 to 0.90 (area under the curve >0.80). The Maintaining Purpose scale showed fair discrimination (area under the curve [AUC] = 0.73). The diagnostic accuracy of the Family Support scale was insufficient to define a meaningful threshold (AUC = 0.60).
Discussion:
This study provides TCIs for six of the seven EORTC QLQ-ELD14 scales with sufficient diagnostic accuracy. However, the TCI for the Maintaining Purpose scale is not suitable for interpretation at the patient level.
