Validation of the Charlson comorbidity index in patients with head and neck cancer: a multi-institutional study.

Comorbid conditions are medical illnesses that accompany cancer. The impact of these conditions on the outcome of patients with head and neck cancer is well established. However, all of the comorbidity studies in patients with head and neck cancer reported in the literature have been performed using the Kaplan-Feinstein index (KFI), which may be too complicated for routine use. This study was performed to introduce and validate the use of the Charlson comorbidity index (CI) in patients with head and neck cancer and to compare it with the Kaplan-Feinstein comorbidity index for accuracy and ease of use. Study design was a retrospective cohort study. The study population was drawn for three academic tertiary care centers and included 88 patients 45 years of age and under who underwent curative treatment for head and neck cancer. All patients were staged by the KFI and the CI for comorbidity and divided into two groups based on the comorbidity severity staging. Group 1 included patients with advanced comorbidity (stages 2 or 3), and group 2 included those with low-level comorbidity (stages 0 or 1). Outcomes were compared based on these divisions. The KFI was successfully applied to 80% of this study population, and the CI was successfully applied in all cases (P < 0.0001). In addition, the KFI was found to be more difficult to use than the CI (P < 0.0001). However, both indices independently predicted the tumor-specific survival (P = 0.007), even after adjusting for the confounding effects of TNM stage by multivariate analysis. Overall, the CI was found to be a valid prognostic indicator in patients with head and neck cancer. In addition, because comorbidity staging by the CI independently predicted survival, was easier to use, and more readily applied, it may be better suited for use for retrospective comorbidity studies.