Identifying health-related quality of life cut-off scores that indicate the need for supportive care in young adults with cancer

Abstract

Purpose Using patient-reported outcomes in routine cancer care may improve health outcomes. However, a lack of information about which scores are problematic in specific populations can impede use. To facilitate interpretation of the European Organisation for Research and Treatment of Cancer Core Questionnaire (EORTC QLQ-C30), we identified cut-off scores that indicate need for support by comparing each scale to relevant items from the Supportive Care Needs Survey (SCNS-LF59) in a young adult (YA) population. Methods We conducted a cross-sectional survey amongst YAs with cancer ages 25–39 at diagnosis. Participants completed the EORTC QLQ-C30 and SCNS-LF59. Patient, clinician and research experts matched supportive care needs from the SCNS-LF59 to quality of life domains of the EORTC QLQ-C30. We evaluated the EORTC QLQ-C30 domain score’s ability to detect patients with need using receiver operator characteristic (ROC) analysis, calculating the area under the ROC curve and sensitivity and specificity for selected cut-offs. Cut-offs were chosen by maximising Youden’s J statistic and ensuring sensitivity passed 0.70. Sensitivity analyses were conducted to examine the variability of the cut-off scores by treatment status. Results Three hundred and forty-seven YAs took part in the survey. Six experts matched SCNS-LF59 items to ten EORTC QLQ-C30 domains. The AUC ranged from 0.78 to 0.87. Cut-offs selected ranged from 8 (Nausea and Vomiting and Pain) to 97 (Physical Functioning). All had adequate sensitivity (above 0.70) except the Financial Difficulties scale (0.64). Specificity ranged from 0.61 to 0.88. Four of the cut-off scores differed by treatment status. Conclusion Cut-offs with adequate sensitivity were calculated for nine EORTC QLQ-C30 scales for use with YAs with cancer. Cut-offs are key to interpretability and use of the EORTC QLQ-C30 in routine care to identify patients with supportive care need.

Young Adult German Breast Cancer Patients Participating in a Three-Week Inpatient Mother–Child Rehab Program Have High Needs for Supportive Care

“…Lidington et al [ 22 ] identified cut-offs for the EORTC QLC-C30 scales to indicate which young adult cancer patients are probably in need of supportive care (age range: 25–39 years). Therefore, they allocated relevant items from the Supportive Care Needs Survey (SCNS-LF59) to each EORTC QLQ-C30 scale.…”

“…The SCNS-LF59 covers different need domains: psychological needs, health system and information needs, physical and daily living needs, patient care and support needs, and sexuality needs as well as four additional items. Lidington et al proposed cut-offs with adequate sensitivity (i.e., >0.7) for nine EORTC QLQ-C30 scales, namely for global health status/QOL, for four functional (physical, role, emotional, and social functioning) and for four symptom scales/items (fatigue, nausea and vomiting, pain, and insomnia) [ 22 ].…”

“…A commonly used QOL questionnaire among cancer patients is the European Organisation for Research and Treatment of Cancer Questionnaire-Core 30 (EORTC QLQ-C30) [ 15 ]. Different approaches have been suggested to find suitable cut-offs for this questionnaire, [ 16 , 17 , 18 , 19 , 20 , 21 , 22 ]. With the aim of focusing on the individual patient, Snyder et al [ 20 , 21 ] identified EORCT QLQ-C30 scores that require the attention of a physician in terms of unmet patient needs.…”

“…Given the large differences between younger and older patients with regard to clinical characteristics, treatments, QOL, and their societal roles [ 5 , 6 , 7 , 8 ], it remained questionable whether the thresholds derived from older and age-mixed cohorts, respectively, would apply to younger patients as well. Only recently, Lidington et al [ 22 ] reported cut-off scores for nine EORTC QLQ-C30 scales that indicate the need for supportive care in young adults with cancer (age range: 25 to 39 years). Using a very similar method to Snyder et al [ 20 , 21 ], they found quite similar cut-off scores in general, but a remarkably lower score for emotional functioning.…”

“…Using a very similar method to Snyder et al [ 20 , 21 ], they found quite similar cut-off scores in general, but a remarkably lower score for emotional functioning. In addition, they were able to identify cut-offs for three further scales [ 22 ].…”