Application of machine-learning model to optimize colonic adenoma detection in India.

AIMS: There is limited data on the prevalence and risk factors of colonic adenoma from the Indian sub-continent. We aimed at developing a machine-learning model to optimize colonic adenoma detection in a prospective cohort. METHODS: All consecutive adult patients undergoing diagnostic colonoscopy were enrolled between October 2020 and November 2022. Patients with a high risk of colonic adenoma were excluded. The predictive model was developed using the gradient-boosting machine (GBM)-learning method. The GBM model was optimized further by adjusting the learning rate and the number of trees and 10-fold cross-validation. RESULTS: Total 10,320 patients (mean age 45.18 ± 14.82 years; 69% men) were included in the study. In the overall population, 1152 (11.2%) patients had at least one adenoma. In patients with age > 50 years, hospital-based adenoma prevalence was 19.5% (808/4144). The area under the receiver operating curve (AUC) (SD) of the logistic regression model was 72.55% (4.91), while the AUCs for deep learning, decision tree, random forest and gradient-boosted tree model were 76.25% (4.22%), 65.95% (4.01%), 79.38% (4.91%) and 84.76% (2.86%), respectively. After model optimization and cross-validation, the AUC of the gradient-boosted tree model has increased to 92.2% (1.1%). CONCLUSIONS: Machine-learning models may predict colorectal adenoma more accurately than logistic regression. A machine-learning model may help optimize the use of colonoscopy to prevent colorectal cancers. TRIAL REGISTRATION: ClinicalTrials.gov (ID: NCT04512729).

Partially covered versus uncovered pyloro-duodenal stents for unresectable malignant gastric outlet obstruction: Randomized controlled study.

OBJECTIVES: The aim of the current study was to compare the efficacy of partially covered duodenal stent (PCDS) vs. uncovered duodenal stent (UCDS) in patients suffering from unresectable primary malignant gastric outlet obstruction (GOO). METHODS: This was a prospective international randomized controlled study conducted in 10 high-volume institutions. Consecutive patients suffering from malignant GOO were recruited. The primary outcome measurement was the reintervention rate. Secondary outcomes included technical and clinical success, 30-day adverse events, 30-day mortality, causes of stent dysfunction, and the duration of stent patency. RESULTS: Between March 2017 and October 2020, 115 patients (59 PCDS, 56 UCDS) were recruited. The 1-year reintervention was not significantly different (PCDS vs. UDCS = 12/59, 20.3% vs. 14/56, 25%, P = 0.84). There was a trend to fewer patients with tumor ingrowth in the PCDS group (6/59 [10.2%]) vs. 13/56 [23.2%], P = 0.07). There were no significant differences in the technical success (100% vs. 100%, P = 1), clinical success (91.5% vs. 98.2%, P = 0.21), procedural time (21.5 [interquartile range [IQR] 17-30] vs. 20.0 [IQR 15-34.75], P = 0.62), hospital stay (4 [IQR 3-12] vs. 5 [IQR 3-8] days, P = 0.81), 30-day adverse events (18.6% vs. 14.3%, P = 0.62), or 30-day mortality (6.8% vs. 5.2%, P = 1.00). CONCLUSION: The use of PCDS was associated with a lower risk of tumor ingrowth but did not improve on reintervention rates or stent patency. Both kinds of stents could be used in this group of patients.

Small Bowel Capsule Endoscopy: Experience from a single large tertiary care centre.

Background and study aims Capsule endoscopy (CE) has transformed examination of the small bowel (SB), once considered a dark continent. The present study aimed to describe the indications, diagnostic yield, practical issues and complications of CE in one of the largest tertiary center in India. Patients and methods This retrospective analysis from a prospectively maintained database, conducted from January 2013 to June 2021 included 1155 CEs performed during this period. Patient medical records were reviewed for indications, results, and complications of CE. Results A total of 1154 patients (809 males and 345 females), mean age 53 years (range 6-87 years), one capsule got stuck in the esophagus, were included in the study. Active SB bleeding had no effect on SB transit time (324.7±161 minutes, n = 137 patients with active bleed vs 310.6±166.9 minutes, n = 1017 patients without active bleed; P = 0.35). The indication and diagnostic yield (DY) of CE were potential overt SB bleed (68.6% & 43.9%), potential occult SB bleed (8.2% and 40%), chronic diarrhea (7.9% and 28.4%), abdominal pain (6.5% and 21.3%), anemia (5.9% and 57.9%), and suspected/known case of Crohn's disease (2.3% & 56.5%) respectively. The DY for patients with age ≥60 years was similar to those with age < 60 years (61.9% vs. 51.8% respectively; P = 0.4). 21 patients (1.8%) had capsule retention of which six (0.5%) had to be referred for surgery. Conclusions CE is a safe and effective investigation with ever increasing range of indications. Potential SB bleed remains the most common indication for CE with high detection rate.