Real-World Effectiveness of Split-dose 2-day 8-liter PEG-ELS in Patients After Prior Failed Split-Dose 1-Day 4-Liter Bowel Preparation

Introduction: Up to 30% of patients have inadequate bowel preparation at colonoscopy after a standard purgative regimen. In such patients, the use of split-dose 2-day 8 liter (L) PEG-ELS (2DBP) has been poorly studied. The aim of this study is to assess the effectiveness of a split-dose 2DBP in patients with a prior inadequate preparation. Methods: We performed a retrospective study of a single endoscopy unit in a Veterans Affairs medical center from 3/10/2015 to 4/30/2021. Using electronic endoscopy and pharmacy records, we identified all outpatients with inadequate bowel preparation on outpatient index colonoscopy with a 1-day bowel prep (BP) who had repeat colonoscopy with a split-dose 2DBP. Data were collected on age, gender, BMI, DM, narcotic use, tricyclic antidepressant use, hypertension, constipation, dementia, Charlson Comorbidity Index (CCI), tobacco use, cirrhosis, index bowel regimen (BR) and bowel preparation adequacy (adequate/inadequate, excellent, good, fair, poor, or BBPS). Adequate bowel preparation was categorized as excellent/good, BBPS 6 or greater and BBPS of 2 in all segments of the colon. A logistic regression model was used identify patient factors associated with inadequate preparation after a split-dose 2DBP. Results: Among 10,160 outpatients, we identified 1,100 with inadequate preparation on their index colonoscopy. Of these, 400 patients had repeat colonoscopy with a split-dose 2DBP. After 2DBP, 83.5% (N=334) achieved an adequate bowel preparation. On univariate analysis, patients with a higher CCI, history of constipation, HTN, Diabetes, active narcotic prescription at time of exam and 2DBP with PEG-ELS as compared to SF-PEG-ELS were more likely to have an inadequate prep on repeat colonoscopy. Utilizing multivariable analysis, controlled by age and gender, only those with a higher CCI (P=.0039), an active narcotic prescription at time of repeat colonoscopy (P=.04) and a rural address predicted inadequate colonoscopy bowel preparation with an ROC curve of 0.72. Conclusion: In this difficult-to-prep population, split-dose 2DBP will succeed in greater than 80% of patients. We found that success is less likely in patients with a higher CCI, active narcotic prescription, and a rural address. This represents the first study to assess of the effectiveness of split-dose 2DBP to our knowledge.