Development and Validation of a Machine Learning-based, Point-of-Care Risk Calculator for Post-ERCP Pancreatitis and Prophylaxis Selection.

BACKGROUND AND AIMS: A robust model of post-ERCP pancreatitis (PEP) risk is not currently available. We aimed to develop a machine learning-based tool for PEP risk prediction to aid in clinical decision-making related to periprocedural prophylaxis selection and post-procedural monitoring. METHODS: Feature selection, model training, and validation were performed using patient-level data from 12 randomized controlled trials. A gradient-boosted machine (GBM) model was trained to estimate PEP risk and the performance of the resulting model was evaluated using the area under the receiver operating curve (AUC) with 5-fold cross-validation. A web-based clinical decision-making tool was created, and a prospective pilot study was performed using data from ERCPs performed at the Johns Hopkins Hospital over a one-month period. RESULTS: A total of 7389 patients were included in the GBM with an 8.6% rate of PEP. The model was trained on twenty PEP risk factors and 5 prophylactic interventions (rectal non-steroidal anti-inflammatory drugs [NSAID], aggressive hydration, combined rectal NSAID and aggressive hydration, pancreatic duct [PD] stenting, and combined rectal NSAID and PD stenting). The resulting GBM model had an AUC of 0.70 (65% specificity, 65% sensitivity, 95% negative predictive value, 15% positive predictive value). A total of 135 patients were included in the prospective pilot study, resulting in an AUC of 0.74. CONCLUSIONS: This study demonstrates the feasibility and utility of a novel machine learning-based PEP risk estimation tool with high negative predictive value to aid in prophylaxis selection and identify patients at low risk who may not require extended post-procedure monitoring.

Capsule endoscopy for small bowel bleed: Current update.

Small intestine, hitherto an obscure area for endoscopists before 2000, is now easily evaluated non-invasively using capsule endoscopy and invasively by device-assisted enteroscopies. Major advances in understanding the causes and management of small bowel diseases have been in obscure gastrointestinal (GI) bleed, currently re-named as small bowel bleed, after the discovery of capsule endoscopy. The current article is a narrative review of the technology of capsule endoscopy, its advantages and limitations, future perspective and Indian studies on its utility in patients with small bowel bleed. Till date, eight large series reporting 2319 patients with obscure GI bleed (1554 overt and 765 occult) undergoing capsule endoscopy have been reported from India. Overall yield of capsule endoscopy to detect lesions in these studies varied from 43.5% to 90%. The major causes detected in various studies for small bowel bleed include vascular malformation, portal hypertensive enteropathy, ulcer, stricture, tumor, polyps, etc. Hookworm can cause both occult as well as overt small bowel bleed as shown mainly from India. Capsule endoscopy has also been quite safe in patients with small bowel bleed as despite 0.6% to 15% retention of imaging capsule in Indian studies, development of clinically evident small bowel obstruction has rarely been reported. The major limitations of capsule endoscopy include lack of maneuvrability and therapeutic capability. Research is in progress to overcome some of the limitations of the current capsule endoscopy system. It is concluded that discovery of capsule endoscopy has brought a new paradigm in GI endoscopy and explored a hitherto unexplored area of GI tract, i.e. small bowel that continued to be a black box for the endoscopists.

Post-infection irritable bowel syndrome following Coronavirus disease-19: A systematic review and meta-analysis.

BACKGROUND AND OBJECTIVES: Persistent gastrointestinal (GI) symptoms and functional gastrointestinal disorders (FGIDs) are increasingly being recognized after Coronavirus disease-19 (COVID-19). Though quite a few studies addressed irritable bowel syndrome (IBS) following COVID-19, the disorders' prevalence varies greatly. We evaluated, (i) overall frequency of post-COVID-19 IBS, (ii) relative risk of development of IBS among COVID-19 patients compared to healthy controls using systematic review and meta-analysis techniques. METHODS: Literature search was performed for studies on GI symptoms and FGIDs after COVID-19 using electronic databases (Medline, Scopus, Cochrane Central Register of Controlled Trials, Google Scholar and Web of Science) till April 28, 2023. We included studies reporting IBS after COVID-19 with any duration of follow-up and any number of subjects. Studies on pediatric population and those not providing relevant information were excluded. Relative risk of development of IBS using Rome criteria among COVID-19 patients compared to healthy controls was calculated. Analysis was done using MedCalc (Applied Math, Mariakerke, Belgium, version 7.2) and Comprehensive Meta-Analysis version 3.3.070 (Biostat Inc. Englewood, NJ 07631, USA). RESULTS: Of the available studies, 13 (four case-control) reporting on IBS after COVID-19 met inclusion criteria. Among 3950 COVID-19 patients and 991 controls, 7.2% of COVID-19 patients and 4.9% of healthy controls developed IBS. Of the four case-control studies reporting post-COVID-19 IBS, patients with COVID-19 were 2.65 (95% confidence interval [CI] 0.538 to 13.039) times more likely to have post-COVID-19 IBS as compared to healthy controls. CONCLUSIONS: Patients with COVID-19 are more likely to develop post-COVID-19 IBS than healthy controls. The heterogeneity of studies, different criteria used by various studies to diagnose post-COVID-19 IBS and some studies not meeting the six-month follow-up duration of the Rome criteria for diagnosing IBS are limitations of this systematic review.

Definitive Role of Contrast-enhanced Endoscopic Ultrasound (EUS) in the Diagnosis of Hepatocellular Carcinoma With Non-diagnostic Cross-sectional Imaging: A Case Report.

The diagnosis of hepatocellular carcinoma is usually centered around cross-sectional imaging (CSI) modalities. However, in some instances focal hepatic lesions may be missed on CSI. Endoscopic ultrasound (EUS) has an evolving role in hepatology and have been shown to be useful in diagnosing focal lesions with advantages of tissue acquisition. We report a case hepatitis B-related cirrhosis presenting with acute decompensation, wherein EUS was used to identify HCC and perform tissue acquisition as CSI was non-diagnostic.

Determinants, profile and outcomes of hepatitis A virus-associated severe acute liver injury in adults.

BACKGROUND AND OBJECTIVES: Hepatitis A virus (HAV)-related hepatitis is witnessing an epidemiological transition with increasing trends in adults. While uncomplicated hepatitis remains common, evidence suggests it to be a growing cause for acute liver failure (ALF). In between the two extremes exists severe acute liver injury (s-ALI) which has a propensity to transition to ALF. We aimed at describing the clinical profile of patients with HAV-related s-ALI and identifying potential predictors of progression to ALF. METHODS: This was a single-center retrospective analysis of adult patients admitted with HAV-related s-ALI between April 2022 and December 2023. Demographic and laboratory parameters were compared between patients with only s-ALI and those with ALF. Predictors of progression from s-ALI to ALF were identified using logistic regression. RESULTS: Forty-three patients satisfied criteria of s-ALI, of which 33 (76.7%) had only s-ALI, while 10 (23.3%) had ALF. Patients with s-ALI had lesser leukocytosis (6.3 ± 3 vs. 13.2 ± 4.8), less incidence of acute kidney injury (9.1% vs. 40%) and lower model for end-stage liver disease (MELD) (20 [18-24.5] vs. 31.5 [26-42]), arterial lactate (2.1 [1.3-3.1] vs. 6.3 [5.2-8.0]), arterial ammonia (94 [72-118] vs. 299 [188-573]), procalcitonin (0.5 [0.28-1.25] vs. 3.2 [1.2-6.1]) and ferritin (482 [213-1633] vs. 5186 [1341-11,053]) compared to HAV-ALF (p < 0.05 for all). Three patients (9.09%) with s-ALI progressed to ALF of whom one (3%) died. Baseline ammonia levels (unadjusted odds ratio [OR] 1.03 [1.01-1.06]) and leukocyte count (OR 1.00 [1.00-1.01]) tended to be associated with ALF progression, although none was significant after multi-variable adjustment. Ammonia levels had an area under receiver operating curve of 0.816 (0.64-0.93) (p = 0.009) (cut-off of 144 µmol/L). Additional comorbidities did not impact overall outcomes. CONCLUSION: HAV presents as s-ALI in young adults, with almost one in 10 progressing to ALF. Baseline ammonia may be an important predictor of progression even in s-ALI, but mandates larger well-designed studies.

Alcoholic Liver Disease/Non-Alcoholic Fatty Liver Disease Index for Classification of Patients with Steatotic Liver Disease.

Background: Steatotic liver disease (SLD) encompasses metabolic dysfunction-associated steatotic liver disease (MASLD) and alcohol-associated liver disease (AALD) at extremes as well as an overlap group termed MASLD with increased alcohol intake (Met-ALD). The Alcoholic Liver Disease/Non-Alcoholic Fatty Liver Disease Index (ANI) was proposed to differentiate ALD from non-alcoholic fatty liver disease (NAFLD). We analysed the performance of the ANI in differentiating within the SLD spectrum. Methods: In a cross-sectional study at a tertiary care center, 202 adults (>18 years) who were prospectively diagnosed with SLD defined by magnetic resonance imaging-proton density fat fraction >6.4% were enrolled. Alcohol consumption (AC) was recorded according to thresholds for significant AC: 140-350 g/week (or 20-50 g/day) for females and 210-420 g/week (or 30-60 g/day) for males. The ANI was calculated, and area under the receiver operating characteristic curve (AUROC) was generated. Results: Of 202 patients (47 years [interquartile range, IQR, 38 to 55], 23.75% females, 77% obese, 42.1% diabetic, 38.1% hypertensive, 28.7% statin use), 40.5% were ever-alcohol consumers; 120 (59%), 50 (24.7%), and 32 (15.8%) were MASLD (ANI, -3.7 [IQR, -7 to -1.6]; Met-ALD, - 1.45 [IQR, -2.4 to 0.28]; and AALD, 0.71 [IQR, -1.3 to 4.8], respectively; P<0.05 for all). The AUROC of the ANI for MASLD and AALD was 0.79 (0.72 to 0.84; cut-off <-3.5) and 0.80 (0.74 to 0.86; cut-off >-1.49), respectively. The ANI outperformed aspartate transaminase/alanine transaminase (AST/ALT) ratio (AUROC=0.75 [0.69 to 0.81]) and gamma glutamyl transpeptidase (GGT) (AUROC=0.74 [0.67 to 0.80]). Addition of GGT did not improve model performance (AUCdiff=0.004; P=0.33). Conclusion: AC is common in MASLD. The ANI distinguishes MASLD and AALD, with individual cut-offs within the intermediate zone indicating Met-ALD. ANI also outperforms AST/ALT ratio or GGT.