THE DEVELOPMENT AND PSYCHOMETRIC VALIDATION OF A BRAIN CANCER QUALITY-OF-LIFE QUESTIONNAIRE FOR USE IN COMBINATION WITH GENERAL CANCER-SPECIFIC QUESTIONNAIRES

A self-report questionnaire module consisting of 24 items, comprising 5 scales and 7 single items, has been developed for measuring health-r elated quality of life in patients with brain cancer, Module developme nt proceeded through several stages, including a listing of patient, f amily and health care professional concerns, the writing of items, fie ld testing in 105 patients with brain cancer and subsequent item reduc tion and scale construction after multitrait scaling analysis and asse ssment of internal consistency (Cronbach's coefficient alpha), The fin al version of the module exhibits reasonable test-retest stability ove r a period of one week, Differences in the responses between patients with recently-diagnosed and recurrent cancer and between patients with a Karnofsky Performance Score (KPS) of 50-70 and 80-100 were in the e xpected direction, indicating that the module of questions is responsi ve to differing conditions, Patients with either mental confusion, mot or deficit or dysphasia indicated problems in several domains and sing le items as compared to patients without these neurological deficits, Thus, differences in the responses to the items in the brain cancer mo dule appear to reflect differences in neurological status, In addition , deteriorating neurological status was accompanied by a marked increa se in emotional distress, future uncertainty and motor dysfunction. A comparison of the responses in the module with the KPS and with a modi fied Barthel Activities of Daily Living Index (BADLI) shows moderate c orrelations, primarily with scales and items that pertain to motor dys function, while other scales (such as emotional distress, visual disor der and communication deficit) and most single items are not associate d with the KPS or BADLI, Since the emotional distress scale of the mod ule was found to be highly correlated with the emotional function scal e of the EORTC QLQ-C30, it could be omitted when the module is used in combination with the QLQ-C30. This would reduce the module to a total of 20 items with four scales and seven single items. The intention is to combine this module of questions with other core or general qualit y-of-life questionnaires when studying patients with brain cancer in c linical trials.