Patterns of Repeat Colorectal Cancer Screening and Follow-Up of Abnormal Results in an Integrated Safety-Net Health Care System

Background: Colorectal cancer (CRC) screening with stool-based tests, including fecal immunochemical test (FIT), can reduce CRC mortality if patients with normal test results repeat testing annually and those with abnormal results receive timely follow-up with colonoscopy. However, most studies focus on one discrete step rather than longitudinal adherence to screening (ie, undergoing repeat screening and timely follow-up). We examined patterns of repeat CRC screening and follow-up of abnormal results over a 4-year period at Parkland Health & Hospital System, a large, integrated safety-net health care system in Dallas County, Texas. Methods: Eligible patients were aged 50–60 years, completed an index FIT between January 1, 2010, and December 31, 2010, and were followed through the end of the study period in 2014. Patients with a history of CRC, colectomy, or prior CRC screening with colonoscopy or sigmoidoscopy were excluded. We defined longitudinal adherence as a composite outcome of: 1) FIT screening with normal results in each calendar year; 2) crossover to endoscopy screening after one or more normal FITs; or 3) colonoscopy within 180 days following abnormal FIT. We also characterized inconsistent screening as being up-to-date with screening in none, one, two or three out of four possible calendar years. Results: A total of 4,826 patients were included in the analysis. Most were Hispanic (41.1%) or non-Hispanic black (38.1%), female (65.1%), and uninsured or receiving medical assistance/charity insurance (80.2%). Across the study period, 834 (17.3%) showed longitudinal adherence by completing repeat on-schedule screening with some combination of FIT and endoscopy. Some patients (n = 940, 19.5%) completed endoscopy after one or more normal FITs, and 367 (7.6%) had an abnormal FIT result at some point during the study period. A decreasing proportion of patients were up-to-date with screening recommendations in one (n = 1,740, 36.1%), two (n = 1,221, 25.3%) or three (n = 973, 20.2%) out of four calendar years. Conclusion: Less than 20% of patients in a safety-net health care system show longitudinal adherence to the CRC screening process, which may reduce the effectiveness of screening. Efforts to improve timely receipt of repeat FIT and follow-up colonoscopy after abnormal FIT may increase the proportion of minority and low-income patients up-to-date with ongoing screening.