The role of colonoscopy in evaluating hematochezia: a population-based study in a large consortium of endoscopy practices

Abstract

Data on the role of colonoscopy in hematochezia are almost exclusively derived from clinical experience in tertiary care practice. To characterize the patient population who received colonoscopy for hematochezia in a consortium of diverse gastroenterology practices. Retrospective analysis. Clinical Outcomes Research Initiative Database, 2002 to 2008. Adults undergoing colonoscopy for the indication of hematochezia. Demographics, comorbidity, practice setting, adverse events, and colonoscopy procedure characteristics and findings. Age-stratified analyses and analyses of inpatient- versus outpatient-performed colonoscopies were also performed. A total of 966,536 colonoscopies were performed during the study period, 76,186 (7.9%) were performed for evaluation of hematochezia. The majority of patients were white non-Hispanic men younger than 60 years old who underwent colonoscopy at a community practice site (79.1%) and had a low-risk American Society of Anesthesiologists (ASA) score (81.5%), in whom colonoscopy reached the cecum (94.8%), and serious adverse events were rare. Colonoscopy findings were hemorrhoids (64.4%), diverticulosis (38.6%), and polyp or multiple polyps (38.8%). From the overall cohort, 38.3% were 60 years of age and older. The older age cohort had significantly more white non-Hispanic females, high-risk ASA scores, incomplete colonoscopies, and unplanned events. Colonoscopy findings demonstrated significantly higher rates of diverticulosis, polyp or multiple polyps, mucosal abnormality/colitis, tumor, and solitary ulcer (P < .0001). There were 3941 (5.2%) who underwent inpatient-performed colonoscopy. One third of this cohort (32.6%) was defined as having a high ASA score. Retrospective database review. These results describe patient populations and characterize colonoscopy findings in individuals presenting with hematochezia primarily in a community practice setting.