The Prediction of Inadequate Colon Capsule Cleansing: A Cohort Selection Guided by CC-CLEAR

INTRODUCTION: In order to optimize the rate of adequate cleansing in colon capsule, it may be important to identify risk factors that can predict a suboptimal colon preparation. AIM: To define predictive factors for inadequate bowel preparation in colon capsule, according to CC-CLEAR (Colon Capsule CLEansing Assessment and Report). METHODS: Retrospective, single center, cohort study. Patients' demographics, data, and quality of bowel preparation, according to CC-CLEAR, were collected retrospectively. A univariate analysis tested the association between covariables and the outcome, inadequate cleansing. The statistically significant variables were included in multivariable logistic binary regression, and a receiver operating characteristic curve (ROC) assessment was performed. RESULTS: We included 167 consecutive colon capsules. Sixty-eight percent (n = 114) of patients were female, with a mean age of 64 years. The main indication for colon capsule was previous incomplete colonoscopy, in 158 patients (94.6%). The colon capsules cleansing was graded as good or excellent in 96 patients (57.5%) and as inadequate in 71 (42.5%), according to CC-CLEAR. The variables inadequate previous colon cleansing (OR adjusted 41.72 [95% CI 12.57–138.57], p value < 0.001); chronic laxative (OR adjusted 4.86 [95% CI 1.08–21.79], p value = 0.039); antidepressant (OR adjusted 5.00 [95% CI 1.65–15.16], p value = 0.004), and impaired mobility (OR adjusted 5.54 [95% CI 1.17–26.31], p value = 0.031) were independently associated with the outcome inadequate cleansing, after adjusting for confoundment. The model presented an excellent discriminative power towards the outcome variable (AUC ROC 0.937 [CI 95% 0.899–0.975], p value < 0.001). CONCLUSION: A previous inadequate colon cleansing, the use of chronic laxative and antidepressant, or impaired mobility are predictors of inadequate colon capsule cleansing, as assessed by the CC-CLEAR. These 4 predictors come together as a model enabling an accurate categorization of the patients at major risk of inadequate bowel preparation for capsule colonoscopy, with an excellent discriminative power and performance, which seems useful for the selection of patients for tailored optimization of the colon cleansing protocol.