Complications of colonoscopy in Japan: An analysis using large-scale health insurance claims data

Abstract

Background and aim: In Japan, screening colonoscopy for colorectal cancer is expected to reduce colorectal cancer mortality, although its complication rate has not been sufficiently examined. The aim of this study is to analyze severe complications due to colonoscopy.

Methods: As a study population, we retrospectively used commercially anonymized health insurance claims data covering 5.71 million patients from January 2005 to August 2018. We extracted patients who received colonoscopy with lesions resection or without treatment. Main outcomes were rates of hemorrhage, perforation, fatal events, and their risk factors.

Results: Among 341 852 colonoscopy without treatment in 260 128 patients (mean age: 49.6 ± 11.7 years), the rates of hemorrhage, perforation, and fatal events were 0.0059% (95% confidence interval [CI] 0.0031-0.0085), 0.0032% (95% CI 0.0011-0.0052), and 0.00029% (95% CI 0-0.0012), respectively. Regarding hemorrhage, compared with the rate for patients <50 years old (0.0050%), the rates for those 50-59, 60-69, and ≥70 years old were 0.0095% (P = 0.17), 0.0031% (P = 0.17), and 0%, respectively. Regarding perforation, compared with patients <50 years old (0.0056%), the rates for those 50-59, 60-69, and ≥70 years old were 0%, 0.0015% (P = 0.99), and 0.0102% (P = 0.99), respectively. A multivariate analysis for risk factors showed no significant findings for hemorrhage and perforation without treatment. Among 123 087 colonoscopy with lesions resection in 102 058 patients (mean age: 53.7 ± 9.3 years), the rates of hemorrhage, perforation, and fatal events were 0.136% (95% CI 0.1157-0.1572), 0.033% (95% CI 0.0228-0.0437), and 0.00081% (95% CI 0-0.0035), respectively.

Conclusions: The analysis using health insurance claims data demonstrated the safety of colonoscopy.

Keywords: Big-data; Colonoscopy; Complications; Hemorrhage; Perforation.