Measuring comorbidity in older cancer patients.

The aim of this article was to provide oncology researchers with adequate tools and practical advice to integrate comorbidity into clinical studies. Open research questions are also discussed. Commonly used comorbidity indexes were identified and a detailed literature search was done by MEDLINE and cross-referencing. Expert opinion was sought on each index. A common scheme exploring the description of the index, clinical experience, metrological performance, easiness of use, cross-compatibility and preservation of data was followed. The actual indexes are included in the Appendix. Four commonly used indexes were identified: the Charlson Comorbidity Index (Charlson), the Cumulative Illness Rating Scale (CIRS), the Index of Coexistent Disease (ICED), and the Kaplan-Feinstein index. The Charlson is the most commonly used whereas the performance of the first two indexes is best characterised. Most studies are retrospective and focus on mortality as an outcome and a base of grading. All indexes are easy to use and require a maximum of 10 min to be filled. Inter-rater and test-retest reliability is generally good. Little is known about other outcomes and the way various diseases cumulate in influencing prognosis. Thus, several reliable indexes are available to measure comorbidity in cancer patients. They show that globally comorbidity is a strong predictor of outcome. Since little is still known about the importance of individual comorbidities for various outcomes and the way comorbidity cumulates in influencing cancer treatment, a wide integration of comorbidity in prospective studies is essential.