On the receiving end. VI. Which dimensions of quality-of-life scores carry prognostic information?

Abstract

Quality of life (QL) of patients with advanced malignancy is an important consideration when assessing the impact of the disease and its treatment (I, 2). QL scores are also powerful independent predictors of survival duration, at least in breast cancer (3-51, melanoma (6) and lung cancer (7). Current methods permit simple, reliable and valid measurement of important aspects of QL (1, 3, 8, 9). The reason for the association between QL scores and prognosis is uncertain. A simple explanation might be that the QL score provides a better measure of the extent of the underlying disease, and thus its prognosis, than other clinically recorded parameters. Even this explanation would, of course, be sufficient to justify the measurement of QL. Alternatively, there is a possibility that better perceived QL actively prolongs survival. If true, this could justify interventions aimed at prolonging survival by improving QL. Preliminary studies have demonstrated that psychosocial intervention can improve aspects of QL (IO), and may improve survival (II). The Spitzer QL Index (QLI) (121 was designed as a concise measure of QL. It measures five separate dimensions of QL: physical performance, activities of daily living, health, support by others and outlook, using a three-point scale for each. The authors have recently analysed the prognostic associations of the components of the QLI obtained at the time of entry to two independent multicentre randomized clinical trials of treatment for metastatic breast cancer and malignant melanoma. Patients in each trial also completed linear analogue self assessment (LASA) scales measuring physical well-being (PWB), mood, pain, nausea and appetite. This paper now presents a comparison of the prognostic significance of various aspects of QL as assessed by patients and physicians.