Background: The ‘Worse-Stable-Better’ (W-S-B) question was introduced to capture patient-perceived change in University of Washington Quality of Life (UW-QOL) domains. Methods: 202 head and neck cancer patients in remission prospectively completed UW-QOL and Patients Concerns Inventory (PCI). For each UW-QOL domain, patients indicated whether over the last month things had worsened (W), remained stable (S) or were better (B). Results: 202 patients at 448 attendances selected 1752 PCI items they wanted to discuss in consultation, and 58% (1024/1752) of these were not covered by the UW-QOL. UW-QOL algorithms highlighted another 440 significant problems that the patient did not want to discuss (i.e. the corresponding items on the PCI were not selected). After making allowance for UW-QOL algorithms to identify 'significant problems' and PCI selection of corresponding issues for discussion there remained clear residual and notable variation in W-S-B responses, in particular to identify patients with significant problems that were getting worse, and patients without significant problems that wanted to discuss issues that were getting worse. Changes in mean UW-QOL scores were notably lower for those getting worse on the W-S-B question, typically by 10 or more units a magnitude that suggests clinically important changes in score. Conclusions: The W-S-B question adds little questionnaire burden and could help to better identify patients who might benefit from intervention. The results of this study suggest that the UW-QOL with the W-S-B modification should be used together with the PCI to allow optimal identification of issues for patient-clinician discussion during routine outpatient clinics.