Background: The fecal immunochemical test (FIT) is widely used for colorectal cancer (CRC) screening, but there are no published randomized trials or well-designed observational studies on its effectiveness and whether effectiveness varies by cancer location or demographic groups. We evaluated whether FIT screening is associated with a reduced risk of CRC death.
Methods: This was a nested case-control study in a racially, ethnically, and socioeconomically diverse cohort of 2,127,128 screening-eligible people who were members of two large California integrated health systems during 2012-2017. In that period, we identified people 52-85 years old who died from CRC (case patients) and matched them individually to persons who were alive and CRC-free (control persons) in a 1:8 ratio based on age, sex, health plan enrollment duration, and medical service area. The primary exposure was completion of at least one mailed screening FIT in the 5 years prior to the diagnosis date of case patients (which served as the reference date for selecting control persons), accounting for other CRC screening exposures over a 10-year window. The primary study outcome was CRC death, overall. We also examined associations, separately, according to tumor location and race/ethnicity. We used unconditional logistic regressions adjusting for matching variables and selected confounders, including socioeconomic status.
Results: Among the 2,127,128 members, we identified 1,097 case patients and matched them to 10,105 randomly selected control persons from eligible persons. Overall, 503 (45.9%) case and 5,125 (50.7%) control persons completed FIT, among whom, the cumulative positive rate in the 5 years prior to the reference date was 24.5% and 7.1%, respectively. Rates of follow up colonoscopy among those with abnormal result were 75.6% and 88.3%, respectively. In regression analyses, completion of FIT screening lowered the risk of CRC death overall (adjusted odds ratio (aOR) 0.67, 95% confidence interval [CI]:0.59-0.76). The association was stronger for cancers arising in the left colon/rectum (aOR=0.56, CI: 0.47-0.68) than for those in the right colon (aOR=0.86, CI:0.71-1.03). The association with receipt of FIT screening and overall CRC mortality was present in all race/ethnicity groups that were examined: Non-Hispanic White (aOR=0.68, CI:0.58-0.81), Non-Hispanic Black (aOR=0.67, CI:0.46-1.00), Hispanic/Latino (aOR=0.68, CI:0.50-0.94), and Non-Hispanic Asian people (aOR=0.34, CI:0.21-0.55).
Conclusions: In a program of proactive mailed FIT screening outreach, completing at least one FIT within five years prior to a CRC diagnosis was associated with a lower risk of death from CRC with a stronger association in the left than the right colon; the lower risk of CRC mortality overall associated with use of FIT was observed across all the racial/ethnic groups examined.

Table 1: Characteristics of the study population, 2012-2017

Table 2: Association between Completion of Mailed FIT and Risk of Death from Colorectal Cancer Overall and by Tumor Location (unconditional)

Wei Zhao

Appearances