Aims and background: To obtain proof of external validity of the visual analogue scale and re-evaluate the use of this instrument in assessing cancer patients' quality of life.
Methods: Consecutive patients attending 79 Italian medical oncology and radiotherapy centers over a period of 1 week were asked to fill out both a questionnaire concerning the presence of 19 problems and a 100-mm linear visual analogue scale evaluating their quality of life. Quality of life was rated as "good" and "bad" when given a score of 70-100 and 0-30, respectively. Multifactorial logistic models were used where good and bad quality of life were correlated with explanatory variables including patient and disease characteristics and the presence or absence of the 19 problems.
Results: Gender, level of education, treatment setting, Karnofsky performance status, disease extent, and the presence of 12 out of 19 problems were found to be correlated with good quality of life. A similar pattern of correlations was found with bad quality of life.
Conclusions: Due to the difficulties in attaining reliable assessment of quality of life using psychometric questionnaires, the further proof of validity obtained in this study allows us to propose the re-evaluation of the role of the uniscale in measuring the quality of life of cancer patients.