Background: High-risk patients, (ie, with inflammatory bowel disease or a personal/ family history of colorectal cancer or polyps) require regular screening colonoscopy. The tolerability of bowel preparation agents may deter these individuals from returning as needed for screenings, which may have negative consequences for long-term health.

Aims: To evaluate real-world rates of incomplete screening colonoscopies, repeat screenings, and adverse events of various bowel preparation agents in high-risk patients.

Methods: Data from the Truven Health Analytics Commercial and Medicare Supplemental MarketScan databases were extracted for high-risk patients (identified by study period clinical codes) who had at least one colonoscopy between July 1, 2012 – June 30, 2015, were >18 years old, and who could be observed 6 months before and 3 months after the index screening (first screening colonoscopy in the study). Data were stratified by bowel preparation class, including over-the-counter (OTC), low volume (LV), high volume (HV) and other prescription agents. The proportions of individuals with an incomplete screening (identified by screening colonoscopy claims with modifier code 22, 52, 53, 73, or 74), repeat screening (within 30 or 90 days from the index event, a measure of bowel preparation tolerability), or hospitalization due to hyponatremia (ICD-9 codes: 276.0, 276.1) or dehydration (ICD-9 codes: 276.50, 276.51, 276.52) within 10 days from the index event were reported. Multivariate logistic regression, adjusting for age, sex, illness burden and medication use, was conducted.

Results: Of the 804,263 individuals, 56.7% used an OTC agent, 2.9% used sodium picosulfate, magnesium oxide, and citric acid (P/MC), 29.9% used another LV agent, and 9.0% an HV prescription agent. Mean age was 57.0 years, and 53% were female. P/MC had lower incomplete screening rates (0.81%) than all other LV, all HV, or OTC agents (1.07%, 0.98%, and 0.84%, respectively) (Table 1). Patients using P/MC had significantly lower odds of an incomplete screening than patients using other LV… or HV… agents, but not OTC agents. Patients using P/MC who had an incomplete screening had 2.0, 2.8 and 1.6 times the odds of returning for a repeat screening compared to individuals using other LV…, respectively. There were no reports of hospitalization due to hyponatremia, and one case of dehydration in the P/MC group, similar to rates seen with other prescription agents. OTC preparations were associated with higher hospitalization rates.

Conclusions: P/MC was associated with lower rates of incomplete colonoscopy in a high-risk population and higher rates of follow up within 30/90 days, suggesting the preparation was less of a deterrent for initial and repeat screenings. The safety of P/MC was comparable to other prescription agents.