Are Surveillance Practices Following Colorectal Cancer Resection Adequate? A Population-Based Analysis

Abstract

Purpose: Current practice guidelines for standard surveillance after colon resection for carcinoma include colonoscopy. Analysis of large databases may give insight into adequacy of postoperative surveillance and identify potential health care disparities. Methods: A linked Surveillance Epidemiology and End Results(SEER)-Medicare database was examined over an 8-year period to establish surveillance patterns utilized in patients following colectomy and/or proctectomy for cancer. Patients age 65 years and older who underwent resection for local or regional stage cancers between 1992 and 1999 were included. The use of colonoscopy, flexible sigmoidoscopy and barium enema was identified using ICD-9 and CPT-4 codes in inpatient and outpatient data. Procedures within 1 and 3 years after diagnosis were included if the patient survived through the end of that follow up period. Differences in procedure utilization between groups were analyzed with chi square testing. Results: A total of 35,428 eligible patients survived through 1 year and 28,192 patients survived through 3 years (74% colon, 26% rectum). The mean age was 76.8 years, 55% were women, and 6.3% were black. Postcolectomy surveillance within 1 year included colonoscopy (59.7%), flexible sigmoidoscopy (18.9%), and barium enema (14.5%). Corresponding 3 year rates were 81.4%, 24.9% and 17.6%, respectively. Factors associated with significant differences in colonoscopy use at 3 years included age group (86% in 65-69 to 58% in >90), geographic region (range 75% to 85%) and to a lesser extent, race (82% white, 77% black) and tumor site (82% colon, 79% rectum) (p < 0.0001 for all comparisons). Gender was not associated with colonoscopy use. Similar findings were observed for 1 year colonoscopy patterns, with rates among geographic regions ranging from 49% to 66% (p < 0.0001). Conclusions: In this population-based cohort, approximately 60% of postcolectomy cancer patients have obtained recommended colonoscopic screening at 1 year, and 81% at 3 years. There was little variation according to gender or race, and potentially appropriate decrements in the very old. However, there were significant unexplained geographic disparities noted, which may reflect practice patterns and/or access to care.