Objectives: Palliation of terminal conditions such as malignant dysphagia must take into account individual preferences for aggressive or nonaggressive care, with a focus on quality of life. Despite this, there are very few data on patients' preferences for palliative therapy. This study is designed to quantitatively determine individual preferences for palliation of malignant dysphagia using health state utilities (HSU).
Methods: HSU were measured using three methods: time trade-off (TTO), visual analog scale (VAS), and the EQ-5D. Patients with esophageal cancer were asked to rate their own state of health and of three standardized scenarios of local, regional, and metastatic disease.
Results: Fifty patients with esophageal cancer were enrolled. Using the TTO method, the utilities of their own health state were 0.80 (95% CI 0.59-0.99) for localized, 0.54 (0.37-0.70) for regional, and 0.52 (0.32-0.71) for metastatic cancer showing no significant difference in mean utility scores for the three staging groups. VAS and EQ5D gave statistically similar values to TTO. Patients consistently rated their own utility better than the utility of standardized scenarios with similar stage and prognosis. Independent of their staging, patients with high dysphagia scores rated their utility worse than patients with low dysphagia scores.
Conclusions: These results confirm the perceived poor state of health of patients with esophageal cancer and are substantially lower than previous estimates in operated patients. Cost-effectiveness models must take into account significant differences between patients' assessment of their own state of health, and that of a "societal" perspective of others with a similar disease. All three methods provided similar estimates. Given the ease of use of VAS and EQ-5D, these methods may be preferable to TTO.