Unsolved problems in evaluating the quality of life of cancer patients.

BACKGROUND: The current means of evaluating the quality of life (QoL) of cancer patients, generally employ a multidimensional approach involving validated self-administered questionnaires. In measuring QoL, as well as in interpreting the results of an assessment, the patient, questionnaire, and setting of administration may all give rise to problems.
DESIGN: Based on our experience and review of a large number of published papers, the most relevant problems in evaluating QoL of cancer patients are highlighted.
RESULTS: The difficulties that patients have with the most frequently used psychometric questionnaires are demonstrated by the fact that more than 10% of patients who gave their consent to be enrolled in a longitudinal study measuring QoL did not fill out the first questionnaire. This percentage is higher among older patients having a low level of education. in patients with disseminated disease, and among those with a low performance status. Moreover, in longitudinal studies, there is a very high incidence of drop-outs, often correlated with the progression of disease, so that the results are sometimes inconsistent.
CONCLUSIONS: Quality of life, depending not only on the type and stage of disease, and the administered treatment, but also on patient characteristics, should be assessed only in a randomised clinical trial. In any case, the most relevant problem is connected to the selection bias due to drop-outs. Reasons for this should be sought in the apparent complexity of the questionnaire and inadequate information given to the patient. Simpler instruments, which can be filled out in a shorter period of time, and specific training by research nurses could help minimize this excessive phenomenon.