Validation of a multidimensional evaluation scale for use in elderly cancer patients

BACKGROUND. Although aging is one of the most important risk factors for cancer, elderly patients tend to be excluded from cancer clinical trials, only on the basis of chronologic age. Performance Status (PS) has been used widely to select adult patients for entry into clinical trials, but it does not include a comprehensive evaluation of various age-related factors in the elderly. This study was designed to assess the reliability and validity of a multidimensional geriatric assessment protocol for elderly patients with cancer. METHODS, Thirty consecutive elderly patients (greater than or equal to 65 years), diagnosed with hematologic neoplasia or solid tumors and undergoing chemotherapy or radiotherapy, were given a specifically structured multidimensional questionnaire (MACE) three times during one week by two different physicians. MACE was intended to collect information on demographics, socioeconomic status, cognitive status, depression, physical performance, disability, and tumor characteristics. In parallel with MACE, information was collected by means of the Sickness Impact Profile (SIP). RESULTS. Both for inter-rater and test-retest reliability, the values of the intraclass correlation coefficient (ICC) were generally higher than 0.7. Disability, cognitive status, depressive symptoms, and the number of days spent in bed sick in the last two weeks were markedly correlated with the global, physical, and social SIP score. Disability alone explained 70% of the variance in the SIP global score, 83% of the variance in the SIP physical score, and 45% of the variance in the SIP psychosocial score. MACE proved to be applicable in a reasonable amount of time (around 30 minutes) for a medical oncology ward. CONCLUSIONS, These data indicate that this structured evaluation of functional status is feasible and reliable. MACE is therefore proposed as a clinical research tool to avoid arbitrary decisions on patient selection for enrollment in clinical trials, to favor uniform monitoring of treatment, and to allow a better comparison of results. (C) 1996 American Cancer Society.