Explore the vast range of titles available.

Early identification of patients who require onward referral to social care can prevent delays to discharge from hospital. We introduce an explainable machine learning (ML) model to identify potential social care needs at the first point of admission. This model was trained using routinely collected data on patient admissions, hospital spells and discharge at a large tertiary hospital in the UK between 2017 and 2023. The model performance (one-vs-rest AUROC = 0.915 [0.907 0.924] (95% confidence interval), is comparable to clinician’s predictions of discharge care needs, despite working with only a subset of the information available to the clinician. We find that ML and clinicians perform better for identifying different types of care needs, highlighting the added value of a potential system supporting decision making. We also demonstrate the ability for ML to provide automated initial discharge need assessments, in the instance where initial clinical assessment is delayed and provide reasoning for the decision. Finally, we demonstrate that combining clinician and machine predictions, in a hybrid model, provides even more accurate early predictions of onward social care requirements (OVR AUROC = 0.936 [0.928 0.943]) and demonstrates the potential for human-in-the-loop decision support systems in clinical practice.

Objective This review assesses the progress of NLP in gastroenterology to date, grades the robustness of the methodology, exposes the field to a new generation of authors, and highlights opportunities for future research. Design Seven scholarly databases (ACM Digital Library, Arxiv, Embase, IEEE Explore, Pubmed, Scopus and Google Scholar) were searched for studies published between 2015 and 2023 that met the inclusion criteria. Studies lacking a description of appropriate validation or NLP methods were excluded, as were studies ufinavailable in English, those focused on non-gastrointestinal diseases and those that were duplicates. Two independent reviewers extracted study information, clinical/algorithm details, and relevant outcome data. Methodological quality and bias risks were appraised using a checklist of quality indicators for NLP studies. Results Fifty-three studies were identified utilising NLP in endoscopy, inflammatory bowel disease, gastrointestinal bleeding, liver and pancreatic disease. Colonoscopy was the focus of 21 (38.9%) studies; 13 (24.1%) focused on liver disease, 7 (13.0%) on inflammatory bowel disease, 4 (7.4%) on gastroscopy, 4 (7.4%) on pancreatic disease and 2 (3.7%) on endoscopic sedation/ERCP and gastrointestinal bleeding. Only 30 (56.6%) of the studies reported patient demographics, and only 13 (24.5%) had a low risk of validation bias. Thirty-five (66%) studies mentioned generalisability, but only 5 (9.4%) mentioned explainability or shared code/models. Conclusion NLP can unlock substantial clinical information from free-text notes stored in EPRs and is already being used, particularly to interpret colonoscopy and radiology reports. However, the models we have thus far lack transparency, leading to duplication, bias, and doubts about generalisability. Therefore, greater clinical engagement, collaboration, and open sharing of appropriate datasets and code are needed. Key Messages • What is already known on this topic– NLP can accurately detect polyp mentions in colonoscopy reports; however, no systematic review has yet been performed across clinical gastroenterology and hepatology. • What this study adds– An overview of NLP applied to gastroenterology up to 2023 highlighting areas of current strength and opportunities for future focus in the age of the large-language model. • How this study might affect research, practice, or policy— This study helps inform future priorities for NLP research in Gastroenterology and Hepatology while focusing on increased transparency and bias reduction within the field

Background and AimAccurate diagnosis of invasion depth for T1 colorectal cancer is of critical importance as it decides optimal resection technique. Few reports have previously covered the effects of endoscopic morphology on depth assessment. We developed and validated a novel diagnostic algorithm that accurately predicts the depth of early colorectal cancer.MethodsWe examined large pathological and endoscopic databases compiled between Jan 2015 and Dec 2018. Training and validation data cohorts were derived and real-world diagnostic performance of two conditional interference tree algorithms (Models 1 and 2) was evaluated against that of the Japan NBI-Expert Team (JNET) classification used by both expert and non-expert endoscopists.ResultsModel 1 had higher sensitivity in deep submucosal invasion than that of JNET alone in both training (45.1% vs. 28.6%, p < 0.01) and validation sets (52.3% vs. 40.0%, p < 0.01). Model 2 demonstrated higher sensitivity than Model 1 (66.2% vs. 52.3%, p < 0.01) in excluding deeper invasion of suspected Tis/T1a lesions.ConclusionWe discovered that machine-learning classifiers, including JNET and macroscopic features, provide the best non-invasive screen to exclude deeper invasion for suspected Tis/T1 lesions. Adding this algorithm improves depth diagnosis of T1 colorectal lesions for both expert and non-expert endoscopists.

Introduction In the management of acute hospital admissions during the COVID‐19 pandemic, safe patient cohorting depends on robust admission diagnostic strategies. It is essential that screening strategies are sensitive and rapid, to prevent nosocomial transmission of COVID‐19 and maintain patient flow. Methods We retrospectively identified all COVID‐19 positive and suspected cases at our institution screened by reverse transcription polymerase chain reaction (RT‐PCR) between 4 April and 28 June 2020. Using RT‐PCR positivity within 7 days as our reference standard, we assessed sensitivity and net‐benefit of three admission screening strategies: single admission RT‐PCR, composite admission RT‐PCR and CXR and repeat RT‐PCR with 48 h. Results RT‐PCR single‐test sensitivity was 91.5% (87.8%–94.4%) versus 97.7% (95.4%–99.1%) (p = 0.025) for RT‐PCR/CXR composite testing and 95.1% (92.1%–97.2%) (p = 0.03) for repeated RT‐PCR. Net‐benefit was 0.83 for single RT‐PCR versus 0.89 for RT‐PCR/CXR and 0.87 for repeated RT‐PCR at 0.02% threshold probability. Conclusion The RT‐PCR/CXR composite testing strategy was highly sensitive when screening patients at the point of hospital admission. Real‐world sensitivity of this approach was comparable to repeat RT‐PCR testing within 48 h; however, faster facilitating improved patient flow. RT‐PCR/CXR composite testing was highly sensitive when screening patients at the point of hospital admission for COVID‐19 infection. Real‐world sensitivity of this approach was comparable to repeat RT‐PCR testing within 48 h; however, faster facilitating improved patient flow.

IntroductionThe age-old question of solving patient flow in a hospital is still a mystery. With the number of people presenting to the hospital increasing year after year, we need to find new and innovative ways to reduce the length of stay (LOS) in the hospital and the need for an inpatient episode.To protect patient safety and ensure a positive outcome for the care episode, a pathway was initiated where gastroenterology patients could be discharged earlier or reviewed as outpatients to avoid an admission altogether but still have specialist clinical oversight. We aimed in this project to evaluate the flow, cost, and LOS effects of this change.MethodsReal-time data was captured using Excel for the patients referred to the service by the gastroenterology advanced clinical practitioner. Criteria were set to determine what was considered admission avoidance and outpatient clinic avoidance. Basic descriptive statistics were calculated, and a survey using GatherIt© software was given to patients to complete anonymously and securely. Economic and financial data were obtained from NHS England websites and local trust databases.ResultsIn 3 years, 148 patients came through the pathway; the demographics of those included ranged from 18 - 84, and 60% were female. 56 admissions and 100 clinic appointments with consultants, etc, were saved. The value of this to the NHS is more than £124,000 in monetary terms, but it also improved flow through the front door and made care more patient-centric. In the survey, 86.86% of patients felt the service was either very good or good, and the rest thought it was satisfactory. The national LOS in the UK rose by 13% in 2022. The local trust data shows a reduction of 2% in LOS for gastroenterology in 2024.ConclusionsThis service change saved substantial resources, increased hospital flow, and reduced length of stay. It could be implemented in any centre with a gastroenterology specialist nurse.

IntroductionThere is an increasing prevalence of patients presenting with complex disorders of gut brain interaction (DGBI). Anecdotally there has been an increase in patient distress associated with their symptoms and consequent requests for escalation in treatments. Current evidence supports a bio-psycho-social model for the management of DGBIs, acknowledging the impact of psychological and social factors on the development of symptoms. Concurrent psychiatric co-morbidities including anxiety and depression or eating disorders can further complicate the presentation in this group of patients.Management of DGBIs benefits from multi-disciplinary team (MDT) involvement including physicians, dietitians, psychologists, psychiatrists and pain management clinicians (Basnayake et al 2022). The aim of this strategy is to improve patient’s symptom management whilst reducing iatrogenic harm. To support this approach, we developed a ‘Functional GI MDT’ (FGI MDT) with core MDT members from the specialties described above. This provides a regular forum to discuss holistic patient management.The aim of this study was to outline the demographics and diagnoses and the referral source of patients discussed in the FGI MDT alongside the clinician perceived benefit of the FGI MDT.MethodsRetrospective data was collected for all patients discussed at the FGI MDT between 01/11/2023 – 01/11/2024 including age, gender, diagnosis and the presence of psychiatric co-morbidities. A questionnaire was sent to all core members of the MDT and all clinicians that had referred a patient to the FGI MDT within this time frame.Results146 MDT discussions took place in 78 patients. 4 patients were discussed >5 times, 3 patients were discussed 4 times. The average age was 28.4 with a range 17-68. 92% were female. The DGBI’s discussed included: functional dyspepsia (43.6%), abdominal pain (15.4%), IBS (10.3%), cyclical vomiting syndrome (5.1%), constipation (3.8%), functional nausea and vomiting syndrome (2.6%) and other (19.2%). 82.1% had a co-existent mental health diagnosis. 53.4% referrals were from a gastroenterologist, 35.6% from a dietitian and 11% from a pain specialist.Feedback was received from 13 clinicians who attended the MDT. 100% of respondents found the MDT helpful, felt that it improved patient care and reduced medical interventions. Further feedback included: ‘Group opinions on complex cases empowers and supports the treating clinician in managing these patients’, ‘I feel the MDT substantially contributes to the reduction in stress managing functional GI patients which helps at a number of levels’, ‘The MDT empowers clinicians to pursue safe, evidence based multi-modality treatments in patients with DGBIs’.ConclusionsA significant proportion of patients discussed at the FGI MDT were female with a co-existent mental health disorder. This likely represents the patient cohort presenting with more complex DGBIs. A significant number of patients present with functional dyspepsia with significant foregut symptoms. Alongside the need for MDT management f, there is commonly self-doubt about missed pathology and how best to manage this group of patients. The FGI MDT has been perceived by clinicians to improve patient care whilst also providing a supportive environment that supports and empowers decision making whilst reducing iatrogenic harm. We hope that this model of care can be expanded to support local trusts in the management of complex DGBIs.

Introduction and AimsEndoscopic ultrasound guided transmural drainage (ETD) followed by endoscopic transluminal necrosectomy (ETN) is the evidence based preferred modality of treatment for symptomatic pancreatic fluid collections (PFC). EUS guided insertion of a lumen apposing metal stent (LAMS) facilitates improved drainage of fluid and improves efficacy of ETN. There is recognised risk associated with the procedure, primarily including bleeding, stent displacement and buried stent. Recent ESGE guidelines on the management of acute necrotising pancreatitis describe the use of imaging prior to drainage and at a 4 week interval, primarily to quantify the solid component in the collection. No definitive imaging protocols are established. In our institution, a protocol was developed to reduce the risk of adverse events associated with drainage. This included pre-intervention arterial phase CT and if identified, prophylactic embolisation of underlying pseudoaneurysm. In addition, all patients underwent CT at 4-5 weeks post stent insertion to determine efficacy of drainage and quantify residual component to determine benefit of long term plastic stents. We sought to assess the impact of the protocol on reducing LAMS associated adverse events.MethodsWe evaluated our practice over a two year period between November 2018 and 2020. Prospectively collected data was reviewed retrospectively for the rates of technical success, clinical success and adverse event.ResultsA total of 56 ETD procedures were performed on 52 patients. The majority of patients in the cohort were male (70.6%) with a mean age of 58 years. All patients underwent an arterial phase CT prior to ETD. Nine patients (17.3%) required embolisation of a previously unrecognised pseudoaneurysm prior to ETD. All procedures were technically successful (100%). Thirty five (67.3%) patients underwent a single ETN and 10 (19.2%) had multiple ETN procedures. Twenty two (62.8%) patients had a 20mm lumen diameter stent inserted and the remainder 15mm. Forty eight patients (92%) achieved complete resolution of collection with a single stent. Four patients (8%) required either an additional stent (multi-gated approach) or additional percutaneous drain. Stent dislodgement occurred in 4 (7.6%) patients during ETN. The median duration of LAMS placement was 44 days. No procedural or delayed LAMS related complications occurred.ConclusionAppropriate pre-procedural cross-sectional imaging facilitates identification and treatment of underlying pseudoaneurysm in this complex patient group. Post procedure interval imaging enables quantification of the residual collection to determine the benefit of long term plastic stents or additional drainage procedure. Our experience suggests adherence to a rigorous imaging protocol may reduce the risk of complication associated with LAMS deployment.

IntroductionPrior to Covid-19, demand for secondary care appointments continued to rise year on year suggesting unsustainable future post-pandemic demand. Now is thus the right time to invest in triage and clinical pathway innovation.MethodsA new fully-integrated digital triage system was built at our institution allowing for document upload and electronic triage. Data pertaining to referral time, triage decision, outpatient appointments and direct-to-test was extracted from the backend to plot empirical cumulative distribution functions, interquartile ranges and allow statistical comparison using the Kruskal-Wallis’ test.ResultsWe analysed the first 704 luminal Gastroenterology referrals through the new triage system with the following sub-specialty classifications: Iron deficiency anaemia (IDA) – 200, Upper gastrointestinal symptoms (UGI) – 152, Inflammatory bowel disease (IBD) – 116, Irritable bowel syndrome (IBS/Functional) – 95, Lower gastrointestinal symptoms/change in bowel habit alone (LGI/CIBH) – 59, Coeliac – 27, Surgical – 25, Complex Functional – 12, Intestinal failure (IF/Nutrition) – 12, Hepatology – 4. 664 (95%) of referrals were accepted with 179 (27%) being sent direct to test. Of these only 42 (23.5%) had a subsequent clinic appointment booked, vs 436 (90%) for those not going direct to test. In addition, sending patients direct to test increased the proportion of subsequent routine clinic appointments from 55% to 70%. Median timelag from referral to grading was four days with grading taking a single day and appointments occurring 17 days later on average. Direct-to-test was most common amongst patients in the UGI (52.6%) and IBD (50%) sub-cohorts. This was significantly different vs other groups at the (p<0.05) level.Abstract PTH-32 Figure 1Subspecialty Referrals vs Direct-To-Test NumbersConclusionsUsing a system as described here substantially improves data capture and efficiency. Direct to test reduces both need for clinic appointments and the urgency of subsequent appointments. IBD and UGI are the subspecialties most likely to benefit from direct to test approaches. IDA could be another suitable specialty and the plan is to address this in the future.Characters2414

IntroductionIron deficiency anaemia (IDA) and iron deficiency without anaemia (IDWA) are associated with worse quality of life in patients with inflammatory bowel disease (IBD). The prevalence of iron deficiency in IBD varies greatly in recent studies. We aimed to identify the prevalence of iron deficiency in patients with IBD at our institution.MethodsPatients with diagnoses of IBD were identified by searching electronic records at UHS, with confirmation by manual chart review by a clinician. Patients with a confirmed diagnosis of IBD within the city and in contact with the UHS IBD team in 2022 were included. Pathology results were extracted for 12 months over 2022 for haemoglobin, ferritin, transferrin saturation, CRP and calprotectin. Anaemia was defined as Hb <130 g/L in males and <120 g/L in females. Absolute iron deficiency was defined as ferritin <11µg/L in females and <24µg/L in males. Iron deficiency in inflammation was defined as ferritin above the previous cuts-off and <100µg/L with CRP > 8mg/L within ten days or calprotectin >250mcg/L within 60 days. For diagnosis of iron deficiency anaemia (IDA) or iron deficiency without anaemia (IDWA), ferritin was tested within 30 days of Hb. Statistical comparisons were made using Chi-squared.Results694 patients met the inclusion criteria and were tested for ferritin and Hb within 30 days of each other. 307 (44.24%) were male. 331 (47.69%) had a diagnosis of Crohn’s disease, 328 (47.26%) had ulcerative colitis (UC), and 35 (5.04%) had IBD unclassified. Table 1 shows the detected prevalence of iron deficiency by diagnosis. 34 patients had tests showing IDWA and IDA occurring at different times over the 12 months. Women had a significantly higher IDWA prevalence than men 25.58% v 18.89% (p = 0.045).Abstract P132 Table 1Number of patients with iron deficiency by diagnosis Diagnosis n IDA IDWA All iron deficiency Crohn’s 331 44 (13.29%) 77 (23.26%) 105 (31.72%) Ulcerative colitis 328 44 (13.41%) 72 (21.95%) 99 (30.18%) IBD Unclassified 35 5 (14.2%) 8 (22.86%) 12 (34.29%) p-value 0.987 0.922 0.834 There were no significant differences in the number of female patients with iron deficiency in ulcerative colitis v Crohn’s disease.Conclusions Unlike previous studies, there is a similar prevalence of iron deficiency in patients with Crohn’s disease and UC. Overall, nearly a third of patients with IBD were iron deficient, within the range of prevalence found in previous studies.

IntroductionCollecting structured patient feedback is challenging, particularly during the pandemic with many virtual appointments. Our electronic IBD-patient feedback covers outpatient (OP), endoscopy and flare-line experiences.MethodsIBD patients provide anonymous feedback at the time-of-service contact. GATHER, a survey platform hosted by our institution, collects anonymous information via QR codes (scan QR codes for surveys), electronic links or handheld tablet. Demographics, disease characteristics and medication were noted in all 3 surveys. The OP survey collated clinic type/modality and feedback on individual health care professionals based on an adapted Royal College of Physicians questionnaire as well as preferences for future appointments. Endoscopy surveys gathered information on referral pathway, endoscopy type, treatment advice, length of wait and pre-test information. Flare line surveys allowed individual feedback on IBD nurses, assessed response time and outcomes. Patients’ attitudes regarding use of our online portal My Medical Record (MyMR) were explored. All surveys allowed sign up for MyMR. Patients could leave individual comments.ResultsSince September 2021, 425 patients responded. Figure 1 outlines the findings of the surveys.Abstract P58 Figure 1ConclusionElectronic surveys are well accepted by our IBD patients and provides useful demographic data. It gives patients the option to inform the service of their preferences for future appointments and allows clinicians to get personal patient feedback for appraisals. Furthermore, it provides feedback on new services such as direct access endoscopy service and the acceptability of patient directed online healthcare (MyMR). Patient-centred feedback enables the user to help shape their future local IBD service.