Comment on: Does use of alternative medicine predict survival from cancer? Eur J Cancer 2003, 39, 372–377

Abstract

Risberg and colleagues [ 1 Risberg T. Vickers A. Bremnes R.M. Wist E.A. Kaasa S. Cassileth B.R. Does use of alternative medicine predict survival from cancer?. Eur. J. Cancer. 2003; 39: 372-377 Abstract Full Text Full Text PDF PubMed Scopus (65) Google Scholar ] present data on cancer survival in users and non-users of alternative medicine. Several aspects of this study severely undermine its validity: 1.Out of 911 patients addressed only 421 were included in the regression analysis. An attrition rate of 54% leaves room for substantial selection bias and there was no analysis of the non-respondents. 2.Among the study patients, alternative medicine (AM) users had initially more severely progressed cancer than non-users: more AM users had distant metastases (65% versus 53%), fewer AM users had curative treatments (32% versus 50%). More severely diseased cancer patients tend to have a shorter survival. Additionally, the distribution of different cancer forms between AM users and non-users was not provided. 3.After attempting to correct for baseline differences, the authors found no statistically significant difference in survival between AM users (eight different AM techniques listed) and non-users. A significant difference was found in only one group (Eastern Cooperative Oncology Group (ECOG) 0) out of six reported subgroups (ECOG 0, 1, 2 and 3, patients with metastatic disease, patients with palliative care).