S0329 Patient Factors and Adenomatous Polyp Detection Rates for Gastrointestinal Medicine (GI) Fellows

INTRODUCTION: Screening colonoscopies are crucial for prevention of colon cancer. Adenoma detection rate (ADR) is a widely accepted benchmark on quality for screening colonoscopy. It is defined as the proportion of complete screening colonoscopies performed, where one or more adenomas are detected. Goal ADRs have been established at greater than 30% for males, 20% for females, and 25% combined. These rates were established as a result of disease prevalence seen on large scale studies. Our study analyzes ADRs and patient factors of colonoscopies performed by GI fellows. METHODS: Retrospective review of 795 patient colonoscopy reports, after exclusion, was performed at an urban academic hospital. Exclusion criteria comprised those with poor bowel prep, inpatient diagnostic colonoscopies, history of colon surgery, colonoscopies performed by more than one fellow and unavailable pathology results. Patient factors studied include age, height, body mass index (BMI), gender, race, anesthesia type (conscious vs general). All adenomatous polyps were confirmed by pathology. RESULTS: For our patient population, median age was 58 years, 49.8% were male, 46.3% Caucasian, and 48.6% African American. Overall ADR in males, females and combined were 48.6%, 36.7% and 42.1% respectively. According to logistic regression models, age at colonoscopy and gender are significantly associated with ADR. When age increased by 10 years, odds of detecting adenomas increased by 65%. Odds of detecting adenoma is about 43.4% lower in females when compared to males. There appears to be a higher ADR in taller patients with heights being grouped by tertials, although not statistically significant (P = 0.062). Odds of detecting adenoma is expected to be 50% higher for 2nd year GI fellows compared to 1st year fellows. There was no significant difference in ADR based on BMI, race or type of anesthesia. CONCLUSION: Only few studies have evaluated patient factors in colonoscopies done by fellows in training, under attending supervision. Overall ADR was above the quality standard both combined and separated by gender. As suggested by prior studies, age and male gender increase odds of detecting adenomas. Although previous studies support a direct relationship between BMI and increased ADRs, our data did not show significant difference. ADR for fellows was higher than the established quality metrics, substantiating that fellows with close attending supervision can perform at or above the national standard.Table 1Table 2