316 Colonoscopic Withdrawal Patterns in a Tertiary Academic Medical Center

Abstract

INTRODUCTION: Withdrawal time is an important quality measure in colonoscopy as numerous studies have shown its effect on adenoma detection rate. Withdrawal during colonoscopy can be affected by many patient factors, but also individual endoscopist’s preferences and practices. We sought to characterize withdrawal patterns in our academic center for quality improvement purposes. METHODS: Inclusion criteria: 1,114 patients at a tertiary academic center who underwent screening and surveillance colonoscopies. Exclusion criteria: patients with missing procedure timestamps, incomplete or aborted procedures, history of gastrointestinal malignancy, prior colon resection, genetic syndromes, or inflammatory bowel disease. For cecal intubation time and withdrawal time, log-transformation was applied to normalize the data. Linear regression was used to assess the association of cecal intubation time with withdrawal time and adjusted for age, gender, and body mass index (BMI). The same analysis was performed by indication (screening and surveillance) and by endoscopist. RESULTS: After adjusting for age, gender, and BMI, increased cecal intubation time was associated with longer withdrawal time for all patients (n = 1144, P = 0.049). Subgroup analysis showed that this pattern held true for patients undergoing screening (n = 488, P = 0.033) but not surveillance (n = 656, P = 0.545). Further subgroup analysis excluding procedures that had polyps removed to control for time spent resecting polyps during withdrawal showed similar results (Table 1). When stratified by endoscopist, two of seven faculty members exhibited an association of longer cecal intubation time with increased withdrawal times. Increased withdrawal time was not associated with prior knowledge of positive family history of colon cancer (P = 0.729), nor personal history of cancer (P = 0.464). CONCLUSION: Longer cecal intubation times are associated with increased withdrawal times in our center particularly with screening colonoscopies, but variability appears to exist between faculty members. Bias against a patient’s family history of colon cancer or personal history of cancer does not appear to affect withdrawal time. Further studies including other academic centers should be done to validate these findings.