Berry, Parul

INTRODUCTION: Recurrent Clostridioides difficile infection (rCDI) is associated with substantial morbidity, healthcare utilization, and costs. US FDA-approved microbiota-based therapies, fecal microbiota spores, live-brpk (VOS) and fecal microbiota, live-jslm (RBL), reduce recurrence risk, but cost-effective placement within the rCDI treatment pathway remains uncertain. METHODS: We performed a US payer-perspective cost-effectiveness analysis using decision-tree modeling according to CHEERS 2022 guidelines. Adults with rCDI entered the model after the first or second recurrence and were followed for <1 year. Effectiveness inputs were derived from ECOSPOR III/IV (VOS), PUNCH CD3/CD3-OLS (RBL) trials, and published literature for standard-of-care (SoC) comparators. Outcomes included costs (2025 USD), quality-adjusted life-years (QALYs), and incremental cost-effectiveness ratios (ICERs). Probabilistic and deterministic sensitivity analyses were performed using a willingness-to-pay threshold (WTP) of $150,000/QALY. ICERs were interpreted across $50,000, $100,000, $120,000, and $150,000/QALY. RESULTS: After the first recurrence, RBL was cost-effective vs SoC alone ($25,415/QALY). Early initiation remained favorable compared with delayed use for VOS ($27,239/QALY) and RBL ($26,704/QALY). After the second recurrence, VOS remained cost-effective vs SoC (ICER $53,983/QALY), whereas RBL exceeded the WTP threshold (ICER $171,496/QALY). Exploratory cross-trial VOS vs RBL comparisons yielded ICERs of $124,636/QALY after first recurrence and $137,350/QALY after second recurrence, indicating threshold-sensitive product-to-product comparisons. DISCUSSION: Earlier initiation of microbiota-based therapies after the first rCDI provides greater health benefits at acceptable cost-effectiveness thresholds compared with delayed use, supporting earlier integration into rCDI treatment strategies. VOS vs RBL comparisons should be interpreted as exploratory indirect comparisons pending comparative real-world or head-to-head data.

Background: Clostridioides difficile infection (CDI) treated with bezlotoxumab (BEZ) has been demonstrated to have a lower recurrence rate than placebo in clinical trials. However, real-world data on BEZ’s effectiveness remain limited and heterogeneous. Objectives: To evaluate the real-world effectiveness of BEZ in preventing CDI recurrence through a single-center retrospective cohort and a meta-analysis. Design: A retrospective cohort study and a meta-analysis. Methods: A retrospective study of patients treated with BEZ from 2017 to 2021 was performed at the Mayo Clinic. Demographics, CDI diagnostics, and several prespecified risk factors were analyzed. A literature search was conducted in August 2024 utilizing the Cochrane Central Register of Controlled Trials, Embase, Medline, Scopus, and Web of Science Core Collection. Studies reporting CDI resolution rates with BEZ were included. The random-effects model described by DerSimonian and Laird was used to calculate weighted pooled resolution rates (WPR) with 95% confidence intervals (CI). We assessed heterogeneity with the inconsistency index ( I 2 ) statistic. Results: Across 28 studies and our retrospective cohort, 2639 CDI patients were analyzed. Among 1786 patients treated with BEZ, 1450 achieved clinical resolution (WPR: 81.6%, 95% CI 77.2–85.6%), with significant heterogeneity ( I ² = 77.3%). In subgroup analysis, the pooled relative risk of recurrence was 0.56 (95% CI 0.36–0.88; p < 0.01) for BEZ with SoC compared to SoC alone. A WPR of 83.3% (95% CI 75.5%–91.1%) for patients receiving BEZ with SoC was observed when compared with a WPR of 70.8% (95% CI 62.7%–78.8%) in patients receiving SoC alone on subgroup analysis. In our cohort of 47 patients, the CDI resolution rate was 72.3% (34/47). Conclusion: Our retrospective study and meta-analysis demonstrate the real-world efficacy of BEZ in reducing CDI recurrence. The higher recurrence rates in our cohort likely reflect the high-risk nature of the population, including a greater proportion of immunocompromised patients.

ObjectivePatients with rheumatoid arthritis (RA) frequently visit emergency departments (EDs), but their clinical characteristics and admission factors are poorly understood. Our study investigates the epidemiology and outcomes of ED visits among patients with RA using the Nationwide Emergency Department Sample (NEDS).MethodsThis cross‐sectional study used the 2019 NEDS data to identify RA‐related ED visits using International Classification of Disease, Tenth Revision codes M05.X or M06.X. RA‐related ED visits were defined as encounters in which RA was recorded in any diagnostic position. Demographics, clinical features, and comorbidities were compared between RA and non‐RA ED visits. Racial variations were assessed, and multivariable logistic regression identified factors associated with inpatient admission.ResultsWe identified 905,811 (0.8%) ED visits for adults aged ≥18 years with RA. Compared to non‐RA visits, RA ED visits had a higher proportion of patients aged ≥65 years, women, White patients, and Medicare‐insured individuals with a greater comorbidity burden. Admission rates were 46% for RA visits versus 16% for non‐RA visits. Black and Hispanic patients with RA were younger than White patients and more likely to belong to the lowest income quartile. Older age, male sex, and comorbidities were associated with higher admission odds, whereas Black race, lowest income quartile, and Medicaid coverage correlated with lower odds of admission. Septicemia was the most common primary ED diagnosis in patients with RA.ConclusionPatients with RA visiting the ED were older, had a higher comorbidity burden, and were three times more likely to be admitted than patients without RA. Black patients and those in the lowest income quartile had lower odds of admission, highlighting potential disparities and the need for targeted interventions to improve health equity.

Large language models (LLMs) transform healthcare by assisting clinicians with decision-making, research, and patient management. In gastroenterology, LLMs have shown potential in clinical decision support, data extraction, and patient education. However, challenges such as bias, hallucinations, integration with clinical workflows, and regulatory compliance must be addressed for safe and effective implementation. This manuscript presents a structured framework for integrating LLMs into gastroenterology, using Hepatitis C treatment as a real-world application. The framework outlines key steps to ensure accuracy, safety, and clinical relevance while mitigating risks associated with artificial intelligence (AI)-driven healthcare tools. The framework includes defining clinical goals, assembling a multidisciplinary team, data collection and preparation, model selection, fine-tuning, calibration, hallucination mitigation, user interface development, integration with electronic health records, real-world validation, and continuous improvement. Retrieval-augmented generation and fine-tuning approaches are evaluated for optimizing model adaptability. Bias detection, reinforcement learning from human feedback, and structured prompt engineering are incorporated to enhance reliability. Ethical and regulatory considerations, including the Health Insurance Portability and Accountability Act, General Data Protection Regulation, and AI-specific guidelines (DECIDE-AI, SPIRIT-AI, CONSORT-AI), are addressed to ensure responsible AI deployment. LLMs have the potential to enhance decision-making, research efficiency, and patient care in gastroenterology, but responsible deployment requires bias mitigation, transparency, and ongoing validation. Future research should focus on multi-institutional validation and AI-assisted clinical trials to establish LLMs as reliable tools in gastroenterology.

Background: Fecal microbiota transplantation (FMT) is a successful therapy for Clostridioides difficile infection (CDI). FMT from overweight donors is speculated to influence the recipient’s body mass index (BMI) after administration for CDI. Objectives: We investigated changes in the recipient’s BMI after FMT in relation to the donor’s BMI. Design: We conducted a retrospective cohort study involving patients who underwent FMT for recurrent CDI at Mayo Clinic between 2012 and 2019. Methods: We analyzed demographic and donor data for patients undergoing FMT at Mayo Clinic (2012–2019). Recipient BMI (pre- and post-FMT) and donor BMI were extracted from medical records. Mixed-effects linear regression was used to evaluate the impact of donor BMI, donor BMI category, recipient baseline BMI, time before and after FMT, and interactions between these variables on overall BMI change and BMI change per month. Kaplan–Meier curves were used to assess BMI changes (⩾5 units) based on the last recorded post-FMT BMI. Results: We analyzed data from 401 patients with recorded BMI measurements before and after FMT. The median age of the recipients at the time of FMT was 59.1 years (interquartile range (IQR): 40.5–70.1 years), with 61.6% being female. The median BMI for recipients prior to FMT was 26.7 kg/m² (IQR: 22.7–31.6 kg/m²), while the median BMI of the donors was 24.5 kg/m2 (IQR: 23.9–27.5 kg/m2). Stool from donors with a normal BMI was used for 58.2% of recipients, while 41.8% received stool from pre-obese donors. Donor BMI data were missing for 3.2% of recipients. Donor BMI was not significantly associated with changes in recipient BMI; for each 1-unit increase in donor BMI, a 0.01-unit monthly increase was observed (95% confidence interval: −0.0003, 0.02; p = 0.11). The log-rank test for BMI increases (⩾+5) and decreases (⩽−5) revealed no significant differences among the donor BMI groups (Chi-squared = 4.4, p = 0.1 for increases, Chi-squared = 2, p = 0.4 for decreases). Conclusion: The lack of impact of donor BMI on BMI changes post-FMT suggests that these changes are more dependent on the recipient’s metabolic profile. Prospective, controlled trials are required to analyze these results more comprehensively.