The association between symptom burdens and utility in Chinese cancer patients

Objectives: This study explored the relationship between the M. D. Anderson Symptom Inventory (MDASI), an instrument measuring the severity of symptoms common to patients with cancer, and utility derived from the SF-36. Methods: Cancer patients from Tianjin Cancer Hospital in China (n = 249) completed a demographic questionnaire and Chinese versions of the MDASI and SF-36. Using a published algorithm converting SF-36 scores to standard gamble (SG) utilities, we examined the association between utility and individual symptoms using Spearman's rank correlation, and explored the association between utility and aggregate symptom scores through multivariate regression analyses. Results: The mean SG utility was 0.81 (SD = 0.11); utilities were significantly but moderately correlated with the majority of symptoms, especially those of distress, sadness, fatigue, and pain. Regression models showed a significantly negative association between the total symptom score and the utility. After controlling for sociodemographics, cancer stage and performance status, a significantly negative association between the total symptom scores and utility was found in the multivariate analyses. We also found the total number of severe symptoms to be a stronger predictor of "disutility." Conclusions: Symptom measures were significantly albeit moderately associated with utility derived from the SF-36 scores, suggesting that a full study with rigorously collected utilities is worth exploring.