Validation of a Novel Clinical Prediction Scoring System for Improved Bowel Preparation Prior to Outpatient Colonoscopy

Colonoscopy is a common endoscopic procedure which the UVA Gastroenterology & Hepatology department performs on a daily basis for both screening/surveillance and diagnostic/therapeutic intent. As a safety net hospital, the institution also offers open access colonoscopy as a means of allowing outside primary care and urgent care providers a way to have patients receive necessary endoscopy services. However, these patients are not all seen in clinic prior to their procedure but are all prescribed a specific bowel prep by a medical professional who screens the referrals.

Given that the patient population described above are often not well-known to UVA providers, they may be prescribed prep regimens that are not ideal for their comorbidities. Not infrequently, these patients show up to their colonoscopies with an inadequate bowel prep, leading to either cancelled or incomplete procedures, increased healthcare and personal financial costs (such as missed wages from taking off work), suboptimal endoscopy resource utilization, and delay in or missed polyp (or potentially cancer) detection.

The investigators seek to use a novel scoring system designed to predict patients at risk for suboptimal preps (and hence patients that would benefit from an extended prep) and apply it in uniform fashion to a patient population most at risk for suboptimal preps. Because obtaining written consent is not practical given that these patients may not be seen by a UVA provider prior to their endoscopic procedure, the investigators will plan to verbally consent via the telephone in an all-inclusive manner; there will be no randomization but rather the selection of prep to be determined by the novel scoring system. Subjects be prescribed either a standard split-dose GoLytely prep (or SuPrep) or an extended split-dose GoLytely prep. The investigators predict that utilizing this new system will lead to a decreased number of inadequate bowel preps and will increase polyp detection. The investigators will work with a UVA biostatistician to analyze the data and use Chi Square, student's T tests, and logistic regression models to assess significance and help validate the model.