783 Menopausal Hormone Therapy Is Associated With Increased Risk of Lower Gastrointestinal Bleeding

Abstract

G A A b st ra ct s significant differences in adenoma detection at follow up. Methods: Data were collected prospectively in a randomized placebo controlled chemoprevention trial (Vitamin D/Calcium) from May 2004 to Nov 2008. Subjects (45-75 yrs) had at least 1 colorectal adenoma removed and none remaining after complete colonoscopy. Trial eligibility required that subjects have either 3 yr or 5 yr follow up, as determined by the endoscopist at the qualifying exam. Subjects with 1 to 2 small non-advanced adenomas on baseline were included. Data collected included patient (age, gender, race/ethnicity, family history, BMI, smoking, study center), baseline endoscopist (age, gender, specialty), and baseline and follow up exam findings. Comparisons were made between subjects with recommended follow-up of 3 vs 5 yrs. A t-test was used to compare continuous variables while Chi Square was used to compare proportions or categorical data. We fit a generalized linear model to evaluate predictors of 3 vs. 5 yr surveillance interval. We also used generalized linear models to compute risk ratios for outcome findings at the follow up exam (e.g. adenoma, advanced adenoma) comparing the recommended 5 to 3 yr intervals. Results: Of the 2259 enrolled subjects, 1560 had LRAs on baseline colonoscopy and had a follow up exam during the trial. 38.1 % (n=594) of the subjects had a 3 yr recommended follow up interval. There were no differences between the groups with regards to subject age, gender, smoking, activity level, BMI, ETOH, bowel prep, endoscopist age or gender. After adjustment for subject age, gender and clinical center, significant factors contributing to exams being recommended at a 3 vs a 5 yr interval were family history of CRC, having 2 adenomas on baseline, and having serrated lesions at baseline (Table 1). There were no significant differences in exam findings on follow up for the 3 vs 5 yr group (Table 2). Conclusions: Endoscopists frequently recommend shorter intervals for LRAs than those recommended by guidelines. Both clinical and endoscopic factors such as family history and the number of serrated lesions found appear to influence the decision making process. However neoplasia detection was similar between the two groups and thus the results support current surveillance guidelines that recommend follow up intervals of at least 5 yrs for LRAs. Table 1. Factors contributing to exam being recommended at a shorter interval (3 versus 5 years)