Abstract
METHODS
We recruited adult cancer patients from Austria, the Netherlands, Poland and the UK. No restrictions were placed on diagnosis or type or stage of treatment. Patients completed the QLQ-C30 and three anchor items reflecting potential attributes of clinically important levels of PF, EF, PA and FA. We merged the anchor items assessing perceived burden, limitations in daily activities and need for help into a dichotomous external criterion to estimate thresholds for clinical importance using Receiver Operator Characteristic (ROC) analysis.
CONCLUSIONS
We were able to estimate thresholds for clinical importance for four QLQ-C30 scales. When used in daily clinical practice, these thresholds can help to identify patients with clinically important problems requiring further exploration and possibly intervention by health care professionals.
RESULTS
In our sample of 548 cancer patients (mean age 60.6 years; 54 % female), the QLQ-C30 scales showed high diagnostic accuracy in identifying patients reporting burden, limitations and/or need for help related to PF, EF, PA and FA. All areas under the curve were above 0.86.
BACKGROUND
The EORTC QLQ-C30 is one of the most widely used quality of life questionnaires in cancer research. Availability of thresholds for clinical importance for the individual questionnaire domains could help to increase its interpretability. The aim of our study was to identify thresholds for clinical importance for four EORTC QLQ-C30 scales: Physical Functioning (PF), Emotional Functioning (EF), Pain (PA) and Fatigue (FA).