Procedure Delays and Time of Day Are Not Associated With Reductions in Quality of Screening Colonoscopies.

Abstract

There have been conflicting results from studies to determine whether factors unrelated to endoscopist skill, such as fatigue, affect the quality of screening colonoscopy. We studied the effects of human and system factors on screening colonoscopy withdrawal time and likelihood of detecting an adenoma in a large cohort of patients. We performed a retrospective analysis of operation and quality improvement data in colonoscopies performed at single academic medical center from November 2012 through February 2014. We collected data from the Northwestern Medicine Enterprise Data Warehouse on endoscopy procedure reports, patient demographics, and pathology reports of all patients undergoing endoscopy. We identified all screening colonoscopies during the study period and determined whether an adenoma was identified in each screening colonoscopy procedure. Our study included data from 7004 screening colonoscopies of patients 50-75 years old performed by endoscopists who performed at least 100 screening colonoscopies during the study period (n = 18). Approximately 27% of procedures began on time; the median colonoscope insertion time was 5.9 minutes (interquartile range, 4.0-8.6). In multivariable logistic regression analysis adjusting for covariates and endoscopist-level clustering, adenoma detection was not associated with procedure delay (P = .48), hour of day (P = .40), or performing the second of 2 colonoscopy blocks in 1 day (P = .88). Adenoma detection was associated with insertion time overall (P = .006), but there was no consistent directional relationship across insertion quintiles. Procedure delays and measured factors associated with fatigue, including time of day and multiple procedure blocks, do not reduce the odds of detecting an adenoma. Adenoma detection varies widely among providers, so efforts to improve adenoma detection should focus mainly on optimizing physician skill.