Self-reported rectal bleeding in a United States community: prevalence, risk factors, and health care seeking.

Abstract

Little data on rectal bleeding in the U.S. population are available. We therefore sought to assess the prevalence of different types of rectal bleeding, their association with potential risk factors including other colonic symptoms, and predictors of health care seeking in a U.S. community. We used a crossectional survey by mail, applying a previously validated self-report symptom questionnaire. Our population comprised an age- and gender-stratified random sample of Olmsted County, Minnesota residents aged 20-64 yr. In total, 1643 responded (77%). Rectal bleeding was reported by 235 subjects (age- and gender-adjusted prevalence, 15.5 per 100; 95% confidence interval [CI], 13.6-17.4); 218 found blood on wiping, 74 noted blood coating the stools, and 46 reported dark blood mixed in the stools. The prevalence of rectal bleeding was significantly higher in younger persons (18.9%, 20-40 yr vs 11.3% > 40 yr; p < 0.001). By stepwise logistic regression analysis, constipation (odds ratio [OR] = 3.03; 95% CI, 2.09-4.41) and diarrhea (OR = 1.90; 95% CI, 1.25-2.84) were independent predictors of rectal bleeding. Among those with rectal bleeding, 13.9% (95% CI, 9.6-19.1%) had visited a physician for bowel problems in the prior yr; only a history of abdominal surgery was an independent predictor of physician visits but this explained just 15.9% of the deviance. In otherwise healthy young and middle-aged persons, approximately one in seven have a history of rectal bleeding and this is more frequent in younger people; only a minority seek health care and this is not related to symptom status.