Prevalence of Advanced, Precancerous Colorectal Neoplasms in Black and White Populations: A Systematic Review and Meta-analysis

Thomas F Imperiale,Priya R Abhyankar,Timothy E Stump,Thomas W Emmett
DOI: https://doi.org/10.1053/j.gastro.2018.08.020
IF: 29.4
Gastroenterology
Abstract:Background & aims: Colorectal cancer incidence and mortality are higher in black vs white populations. The reasons for these disparities are not clear, yet some guidelines recommend screening black persons for colorectal cancer starting at 40-45 years of age. We performed a systematic review and meta-analysis to compare the prevalence of advanced adenomas (AAs) and advanced precancerous colorectal neoplasms (ACNs) between asymptomatic black and white screen-eligible adults. Methods: We searched Ovid MEDLINE, PubMed, Embase, and the Cochrane Library to identify articles (published from 1946 through June 2017) that reported prevalence values of AA or ACN in average-risk black and white individuals undergoing screening colonoscopy. Two authors independently assessed study quality and risk for bias using a modified validated quality assessment instrument. In accord with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, 2 authors independently abstracted descriptive and quantitative data from each study. We performed a random-effects meta-analysis to determine risk differences and odds ratios (ORs). Results: Of 1653 articles, we identified 9 studies for analysis that included 302,128 individuals. Six of the 9 studies were of high methodologic quality, and had a low risk for bias. In these 9 studies, the overall prevalence values for AA and ACN did not differ significantly between black (6.57%) and white (6.20%) screened individuals (OR 1.03; 95% confidence interval [CI] 0.81-1.30). In a subgroup of 5 studies, the prevalence of proximal AA and ACN was significantly higher in black (3.30%) than in white (2.42%) screened individuals (OR 1.20; 95% CI 1.12-1.30). Excluding the largest study did not affect overall prevalence (OR 0.99; CI 0.73-1.34) but did eliminate the difference in prevalence of proximal AA or ACN (OR 1.48; 95% CI 0.87-2.52). Conclusions: In this meta-analysis, we found the overall prevalence of AA and ACN did not differ significantly between average-risk black and white persons, indicating that the age at which to begin colorectal cancer screening need not differ based on race alone.
What problem does this paper attempt to address?