Factors associated with use of preoperative chemoradiotherapy for rectal cancer.

Abstract

e14079 Background: Studies of patients with stage II/III rectal cancer have demonstrated that preoperative (preop) chemoradiotherapy (CRT) improves local control, reduces toxicity and is associated with more sphincter preserving surgeries than postoperative (postop) CRT. While preop CRT has been recommended as the standard of care, many still receive postop CRT. We determined the proportion and associated characteristics of those who received preop CRT among our sample. Methods: We identified patients with stage II/III rectal cancer who had primary site resection and CRT in the population-based cohort studied by the Cancer Care Outcomes Research and Surveillance Consortium, and used data from patient surveys and abstracted medical records to construct patient variables (age, gender, marital status, race, education, income, insurance status, stage, comorbidity, patient treatment preferences, research site, types and sequence of providers seen) and determine sequence of CRT and surgery. Multivariable logistic regression was used to model the association between predictors and receipt of preop CRT. Results: Of the 220 patients meeting inclusion criteria, 64% received preop CRT and 36% received postop CRT. A multivariate model showed those visiting a medical oncologist (MO) or radiation oncologist (RO) prior to receiving any treatment had a 91% (95% CI, 86% to 97%) probability of receiving preop CRT, whereas those who did not had a 30% (95% CI, 19% to 41%) probability. Subjects visiting a RO prior to a surgeon had a 91% (95% CI, 79% to 100%) probability of receiving preop CRT, whereas those visiting a surgeon first had a 65% (95% CI, 57% to 73%) probability. No other patient characteristics remained significant in the multivariate model. Conclusions: Early visits to an oncologist and sequencing of provider visits appear to be more important predictors of receiving preop CRT than the extensive list of patient characteristics evaluated in this study. Further evaluation of provider characteristics, referral patterns and health system processes that determine treatment should be undertaken to help design targeted interventions to reduce variation from recommended care.