Faculty Bibliography

Over the past year, the emergence of state-of-the-art large language models (LLMs) in tools like ChatGPT has ushered in a rapid acceleration in artificial intelligence (AI) innovation. These powerful AI models can generate tailored and high-quality text responses to instructions and questions without the need for labor-intensive task-specific training data or complex software engineering. As the technology continues to mature, LLMs hold immense potential for transforming clinical workflows, enhancing patient outcomes, improving medical education, and optimizing medical research. In this review, we provide a practical discussion of LLMs, tailored to gastroenterologists. We highlight the technical foundations of LLMs, emphasizing their key strengths and limitations as well as how to interact with them safely and effectively. We discuss some potential LLM use cases for clinical gastroenterology practice, education, and research. Finally, we review critical barriers to implementation and ongoing work to address these issues. This review aims to equip gastroenterologists with a foundational understanding of LLMs to facilitate a more active clinician role in the development and implementation of this rapidly emerging technology.

IntroductionOlder adults with IBD are at higher risk for postoperative complications as compared to their younger counterparts, however factors contributing to this are unknown. We assessed risk factors associated with adverse IBD-related surgical outcomes, evaluated trends in emergency surgery, and explored differential risks by age.MethodsUsing the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) database, we identified adults ≥18 years of age who underwent an IBD-related intestinal resection from 2005-2019. Our primary outcome included a 30-day composite of mortality, readmission, reoperation, and/or major postoperative complication.ResultsOverall, 49,746 intestinal resections were performed with 9,390 (18.8%) occurring among older adults with IBD. Nearly 37% of older adults experienced an adverse outcome as compared to 28.1% among younger adults with IBD (p<0.01). Among all adults with IBD, the presence of preoperative sepsis (aOR, 2.08; 95%CI 1.94-2.24), malnutrition (aOR, 1.22; 95%CI 1.14-1.31), dependent functional status (aOR, 6.92; 95%CI 4.36-11.57), and requiring emergency surgery (aOR, 1.50; 95%CI 1.38-1.64) increased the odds of an adverse postoperative outcome, with similar results observed when stratifying by age. Further, 8.8% of surgeries among older adults were emergent, with no change observed over time (p=0.16).DiscussionPreoperative factors contributing to the risk of an adverse surgical outcome are similar between younger and older individuals with IBD, and include elements such as malnutrition and functional status. Incorporating these measures into surgical decision-making can reduce surgical delays in older individuals at low-risk and help target interventions in those at high risk, transforming care for thousands of older adults with IBD.

Background and study aim Cold snare endoscopic mucosal resection (CS-EMR) can reduce some of the risks associated with electrocautery use during colon polyp resection. Data regarding efficacy have yielded variable results. We conducted a systematic review and meta-analysis to estimate the pooled efficacy and safety rates of CS-EMR. Patients and methods We conducted a literature search of multiple databases for studies addressing outcomes of CS-EMR for colon polyps from inception through March 2023. The weighted pooled estimates with the 95% confidence interval (95% CI) were calculated using the random effects model. I2 statistics was used to evaluate heterogeneity. Results 4137 articles were reviewed, and 16 studies met the inclusion criteria. 2584 polyps were removed from 1930 patients and48.9% were females. 54.4% were adenomas, 45% were sessile serrated lesions (SSLs), and 0.6% were invasive carcinoma. Polyp recurrence after CS-EMR was 6.7% (95% CI: [2.4-17.4%], I2=94%). The recurrence rate for polyps ≥ 20 mm was 12.3% (95% CI: [3.4-35.7%], I2= 94.%), 17.1% (95% CI: [4.6-46.7%], I2= 93%) for adenomas, and 5.7% (95% CI: [3.2-9.9%], I2= 50%) for SSLs. The pooled intraprocedural bleeding rate was 2.6% (95% CI: [1.5-4.4%], I2=51%), the delayed bleeding rate was 1.5% (95% CI: [0.9-2.8%], I2=20%) and no perforations or post-polypectomy syndromes were reported with estimated rates of 0.6% (95% CI: [0.3-1.2%], I2=0%) and 0.6% (95% CI: [0.2-1.3%], I2=0%), respectively. Conclusion CS-EMR demonstrated an excellent safety profile for colon polyps with variable recurrence rates based on polyp size and histology. Large prospective studies are needed to validate these findings.

BACKGROUND:Colorectal cancer among adults aged <50 years [early-onset colorectal cancer (EOCRC)] is projected to be the leading cause of cancer-related death by 2030. Although evidence-based guidelines for colorectal cancer screening now recommend beginning screening at age 45, the needs of many at-risk young adults are potentially being overlooked. Unanswered questions also remain regarding the effects of EOCRC on quality-of-life and psychosocial outcomes. This qualitative study explored the lived experiences and perceptions of a sample of adult EOCRC survivors in the United States through one-on-one interviews. METHODS:An EOCRC advocate survivor team member led 27 structured virtual interviews using a 10-question interview guide. Data were analyzed using a 9-step inductive approach. RESULTS:Participants were geographically diverse. Most were women (66.6%) who self-identified as non-Hispanic White (85.2%). The mean age at interview was 40.19 ± 5.99; at diagnosis, 33.93 ± 5.90. Six overarching themes emerged: signs and symptoms, risk factors, system-level factors, quality of life, social support, and reflection. CONCLUSIONS:The specific needs of individuals in this younger population of patients with colorectal cancer should be considered during treatment and future interventions and throughout survivorship. IMPACT:While the reasons for the increasing incidence of EOCRC are currently unknown, the lived experiences and perceptions of EOCRC survivors noted in this study highlight specific needs of this population that can inform educational materials, comprehensive care, future research, and policy change.

INTRODUCTION/BACKGROUND:Adenoma detection rate (ADR) is a quality metric that has been emphasized by multiple societies as improved ADR leads to reduced interval colorectal cancer (CRC). It is postulated that increased withdrawal time (WT) can lead to higher ADR. Multiple randomized controlled trials (RCTs) were undertaken to evaluate this. We performed a systematic review and meta-analysis of RCTs to analyze the impact of higher WT on ADR during colonoscopy. METHODS:The following databases were comprehensively searched through November 8, 2022: Embase, MEDLINE, Cochrane, Web of Science, and Google Scholar. Only RCTs were eligible for inclusion. We applied the random effects model using the DerSimonian Laird approach and calculated risk ratio (RR) for binary outcomes and mean difference (MD) for continuous outcomes. 95% CI and P values were generated. RESULTS:A total of 3 RCTs with 2159 patients were included of which 1136 patients were included in the 9-minute withdrawal group (9WT) and 1023 patients in the 6-minute withdrawal group (6WT). The mean age range was 53.6 to 56.8 years and the male gender was 50.7%. The overall ADR was significantly higher for 9WT (RR=1.23; 95% CI, 1.09-1.40; P<0.001). The overall adenoma per colonoscopy (APC) was also higher for the 9WT group (MD: 0.14; 95% CI, 0.04-0.25; P=0.008). CONCLUSION/CONCLUSIONS:The 9-minute withdrawal time improved ADR and APC compared with the 6-minute withdrawal. Given the high-quality evidence, we recommend that clinicians at least perform a 9-minute withdrawal to achieve higher quality metrics including ADR to reduce interval CRC.

INTRODUCTION:To investigate the impact of procedure-related and endoscopist-related factors on the effectiveness of a computer-aided detection (CADe) device in adenomas per colonoscopy (APC) detection. METHODS:The SKOUT clinical trial was conducted at 5 US sites. We present prespecified analyses of procedure-related and endoscopist-related factors, and association with APC across treatment and control cohorts. RESULTS:There were numeric increases in APC between SKOUT vs standard colonoscopy in community-based endoscopists, withdrawal time of ≥8 minutes, for endoscopists with >20 years of experience, and endoscopists with baseline adenoma detection rate <45%. DISCUSSION:The application of CADe devices in clinical practice should be carefully evaluated. Larger studies should explore differences in endoscopist-related factors for CADe.