Use of Modifiable and Non-modifiable Factors to Predict Bowel Preparation Quality Prior to Colonoscopy

Abstract

Listen

Translate article

Introduction: Predicting the quality of bowel preparation prior to starting colonoscopy can have several advantages. The study was designed to evaluate the association of patient dependent factors in predicting the quality of bowel preparation. Methods: The prospective study enrolled subjects reporting for outpatient screening colonoscopy. Patients with gastrointestinal bleed were excluded. All subjects were prescribed a gallon of polyethylene glycol (PEG) as bowel prep regimen the day before procedure. Each participant was asked questions regarding their age, sex, weight, height, race, co-morbidities, medication use in last one week, diet in last 24 hours (breakfast on the day of procedure, dinner the night before, and compliance with liquid diet), compliance with the bowel prep agent, bowel habits and the Bristol stool form scale (BSFS). The endoscopist determined the quality of bowel prep using Boston bowel prep score (BBPS). Ordinal logistic regression model in SPSS software was used to analyze the data. Results: Five hundred forty-three subjects were enrolled in the study. The mean age was 58.4 years with 64.9% female and 35.1% male. Based on composite BBPS subjects were divided into 3 groups; poor-n=41 (7.6%), sub-optimal-n=194 (35.9%), and adequate-n=306 (56.6%). Females (vs. Males), patients who had no dinner (vs. solid dinner) the day before procedure, and patients who drank all or more than half (vs. less than half) of the prescribed golytely solution had higher odds of good quality bowel prep (p<0.05). Compliance with clear liquid diet for 24 hours before colonoscopy and bowel movement frequency (=>1/day) showed positive association with quality of bowel prep (p>0.05). Age (>60), history of stroke, use of TCA, and opioid in one week before colonoscopy had negative association with quality of bowel prep (p>0.05). Other variables like history of use of laxatives or fiber supplements in one week before colonoscopy, HTN or DM2, BMI and BSFS did not showed any association. Conclusion: Knowledge of these non-modifiable factors (age, sex, race, comorbidity, bowel habits, and stool morphology) and modifiable factors (BMI, medications, diet before colonoscopy, and compliance with bowel prep agent) will give the clinician an opportunity to identify patients with poor bowel prep characteristics before colonoscopy and plan necessary interventions.