BMJ Open Gastroenterology最新文献

Objective: American Association for the Study of Liver Disease guidelines recommend regular testing of alkaline phosphatase (ALP) among patients with primary biliary cholangitis (PBC) to monitor disease progression and response to treatment with ursodeoxycholic acid (UDCA), but previous studies have shown that adherence to recommended testing intervals is low. We used data from the Fibrotic Liver Disease (FOLD) Consortium to evaluate rates of adherence among US routine care patients.

Methods: PBC cases were confirmed with chart abstraction. Patients from three FOLD sites (Henry Ford Health (Detroit, MI), Kaiser Permanente-Southern California (Los Angeles, California) and Kaiser Permanente-Northwest (Portland, Oregon)) were observed from 1 January 2018 through 31 December 2021. We divided our evaluation of adherence to monitoring guidelines into two segments: (1) the first 12 months post-UDCA initiation; and (2) >12 months post-UDCA initiation.

Results: A total of 1756 patients were identified for the 2018-2021 period; 67 patients did not receive UDCA and were excluded from the sample. A total of 1689 patients were included in one or both segments (segment 1: 740, segment 2: 1689). Only 52% of patients received appropriate ALP testing to ascertain response to UDCA after roughly 1 year of treatment; rates were significantly higher among patients with specialist care compared with those without (54% vs 45%, p=0.001). For the period following the first year of UDCA treatment, the observed monitoring rate was 67%, where hepatology or gastroenterology specialist care was associated with significantly higher rates of monitoring (65%-78%) compared with those without care from a specialist (30%-57%, p<0.0001) with the same level of comorbidity.

Conclusion: In a large US PBC cohort, there were concerning levels of non-adherence to recommended biochemical monitoring. Receipt of care from a specialist was associated with higher rates of monitoring. Strategies to increase rates of biochemical testing are needed.

Objective: We aimed to quantify the per-procedure costs of acquiring, maintaining/repairing and reprocessing reusable gastrointestinal endoscopes by observing practices in a large National Health Service (NHS) hospital.

Methods: We conducted a bottom-up micro-costing analysis to capture the costs of reusable gastrointestinal endoscopes using a detailed resource-use data sheet and observations at the University Hospital Coventry and Warwickshire (UHCW). The data sheet drew on the published literature and NHS decontamination guidance. Cost categories included (1) measuring personnel time for reprocessing endoscopes, (2) reprocessing materials and (3) acquisition and maintenance/repair of endoscopy and reprocessing equipment. Data were obtained through observation and interviews with staff. Costs were calculated using the data collected at UHCW and cross-checked with data from two other NHS Trusts, manufacturers and the literature.

Results: Staff time for reprocessing averaged 35 min per procedure (£23.57: 22% of the total cost). The reprocessing materials' cost per procedure was £16.41 (15% of the total cost). Total capital acquisition cost per procedure was £46.9 (44% of the total cost), including endoscopy capital (£37.4) and reprocessing capital (£9.5). Total maintenance/repair cost per procedure was £20.46 (19% of the total cost). These led to a total cost of £107.34 per endoscopy procedure.

Conclusions: Some observed values were slightly lower than but generally comparable to similar studies. We identified the key drivers of costs, led by capital costs. The results of our study could be used in economic evaluations involving reusable gastrointestinal endoscopes. The methodology can inform the cost evaluation of medical devices that require intensive reprocessing.

Objective: Pan-intestinal capsule endoscopy (PCE) offers a safer, more effective alternative to colonoscopy for detecting potentially haemorrhagic lesions (PHL) in suspected mid-lower gastrointestinal bleeding (MLGIB), though it is limited by time-consuming review and missed lesions. We compared the diagnostic performance of artificial intelligence-assisted PCE (AI-PCE) versus conventional reading PCE (CR-PCE) and colonoscopy.

Methods: We retrospectively analysed 100 prospectively enrolled patients undergoing PCE for suspected MLGIB using an externally validated convolutional neural network. Diagnostic performance of AI-PCE, CR-PCE and colonoscopy was evaluated against a consensus reference standard. Accuracy metrics (sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV)) were assessed overall and by lesion type and intestinal segment.

Results: AI-PCE detected PHL in 60% of patients versus 42% with CR-PCE (p<0.01). Lesions included vascular (51% vs 33%, p<0.01), ulcers/erosions (16% vs 7%, p=0.012), protuberant (5% vs 4%, p=1.0) and active bleeding (7% vs 7%, p=1.0). AI-PCE achieved higher sensitivity than CR-PCE (95% vs 67%, p<0.0001) with comparable specificity (97% vs 97%), PPV (98% vs 98%) and superior NPV (92% vs 63%, p=0.0015). For the small bowel, AI-PCE outperformed CR-PCE in sensitivity (96% vs 59%, p<0.0001) and NPV (97% vs 76%, p=0.0010). In colon, AI-PCE also showed greater sensitivity (90% vs 68%, p=0.027) and NPV (94% vs 86%, p = 0.066). Compared with colonoscopy, AI-PCE was markedly more sensitive (90% vs 32%, p<0.0001) with higher PPV (100% vs 65%, p<0.001) and NPV (94% vs 65%, p<0.0001).

Conclusion: AI-PCE significantly improves diagnostic accuracy over conventional reading and colonoscopy, offering superior sensitivity without compromising specificity, and may establish a new standard for PCE in MLGIB.

Objective: Colon capsule endoscopy (CCE) is a recognised diagnostic tool, but there is little research exploring patient experience of this relatively new technology. We aimed to understand the patient experience of CCE and explore similarities to and differences from colonoscopy and CT colonography (CTC).

Methods: We conducted a structured patient experience survey exploring preprocedural, procedural and postprocedural elements of CCE, alongside colonoscopy and CTC, using descriptive statistics. Consenting patients were recruited from the NHS England CCE pilot, referred either on a suspected colorectal cancer or a 3-year postpolypectomy surveillance pathway.

Results: 927 of 1937 patients (48%) responded to the survey invitation. 486 had CCE as their index procedure, 399 colonoscopy and 42 CTC. Two per cent of CCE patients found the procedure painful compared with 21% of colonoscopy and 12% of CTC patients (p<0.001). The CCE procedural information was easily understood by 81% of patients compared with 92% having colonoscopy (p<0.001). There was no significant difference in the bowel preparation experience with 20% of CCE and 16% of colonoscopy patients experiencing severe or more discomfort (p=0.439). However, 19% of CCE patients felt the bowel preparation would put them off a future CCE compared with 8% of colonoscopy patients (p<0.001). This was not wholly explained by the need for further investigations. Using regression analysis, we found that high-quality preprocedural information, tolerability of bowel preparation, procedural comfort and investigative closure were predictors of patient satisfaction with CCE. 74% of patients were satisfied with CCE in diagnosing or reassuring them compared with 91% in colonoscopy and 80% in CTC (p<0.001).

Conclusions: CCE was similarly or better tolerated than colonoscopy and CTC throughout the patient journey, with significantly less pain experienced. A future CCE clinical service should ensure that the patient is well informed and optimise the likelihood of the investigative closure.

Objective: Natural language processing (NLP) can identify cohorts of patients with inflammatory bowel disease (IBD) from free text. However, limited sharing of code, models, and data sets continues to hinder progress. The aim of this study was to evaluate multiple open-source NLP models for identifying IBD cohorts, reporting on document-to-patient-level classification, while exploring explainability, generalisability, fairness and cost.

Methods: 15 algorithms were assessed, covering all types of NLP spanning over 50 years of NLP development. Rule-based (regular expressions, spaCy with negation), and vector-based (bag-of-words (BoW), term frequency inverse document frequency (TF IDF), word-2-vector), to transformers: (two sentence-based sBERT models, three bidirectional encoder representations from transformers (BERT) models (distilBERT, BioclinicalBERT, RoBERTa), and five large language models (LLMs): (Mistral-Instruct-v0.3-7B, M42-Health/Llama-v3-8B, Deepseek-R1-Distill-Qwen-v2.5-32B, Qwen-v3-32B, and Deepseek-R1-Distill-Llama-v3-70B). Models were comparatively evaluated based on full confusion matrices, time/environmental costs, fairness, and explainability.

Results: A total of 9311 labelled documents were evaluated. The fine-tuned DistilBERT_IBD model achieved the best performance overall (micro F1: 93.54%), followed by sBERT-Base (micro F1: 93.05%); however, specificity was an issue for both: (67.80-64.41%) respectively. LLMs performed well, given that they had never seen the training data (micro F1: 86.47-92.20%), but were comparatively slow (18-300 hours) and expensive. Bias was a significant issue for all effective model types.

Conclusion: NLP has undergone significant advancements over the last 50 years. LLMs appear likely to solve the problem of re-identifying patients with IBD from clinical free text sources in the future. Once cost, performance and bias issues are addressed, they and their successors are likely to become the primary method of data retrieval for clinical data warehousing.

Objective: Metabolic dysfunction-associated steatotic liver disease (MASLD) presents considerable variability in disease progression and treatment outcomes. We aimed to determine whether specific patterns of liver inflammatory flares are correlated with distinct treatment responses.

Methods: We conducted an analysis of a well-characterised prospective cohort involving treatment-naïve MASLD patients from January 2015 to November 2023 at The First Affiliated Hospital of Sun Yat-sen University. Participants underwent a standardised 48-week lifestyle modification programme, with follow-up extending through December 2024. Liver fat content (LFC) was assessed using MRI-based proton density fat fraction (MRI-PDFF), whereas liver stiffness measurements (LSMs) were performed using two-dimensional shear wave elastography at baseline and after 48 weeks.

Results: Participants were stratified by alanine transaminase (ALT) and liver fibrosis status: normal ALT/no fibrosis (n=149), elevated ALT/no fibrosis (n=264), normal ALT/fibrosis (n=91) and elevated ALT/fibrosis (n=178). While MRI-PDFF (≥30% LFC decline) and ALT responses (≥17 U/L decrease) did not differ between groups, the elevated ALT/fibrosis group exhibited a significantly higher probability of LSM response (≥1 fibrosis stage improvement) than in the normal ALT/ fibrosis group (53.4% vs 31.9%, p=0.001; OR=2.53, 95% CI: 1.31 to 4.85, p=0.006). Receiver operating characteristic analysis revealed that the cut-offs for weight loss (8.55% vs 4.94%, p=0.023) and LFC reduction (39.85% vs 20.57%, p=0.062) associated with LSM response were higher in patients with normal ALT/fibrosis than in those with elevated ALT/fibrosis.

Conclusion: MASLD patients with liver fibrosis and persistently normal ALT levels exhibited a less favourable treatment response to fibrosis than those with elevated ALT levels, necessitating more substantial reductions in steatosis and weight to achieve the desired outcomes.

Objective: Ischaemic colitis is the most prevalent form of ischaemic enteritis and represents a major cause of acute lower gastrointestinal bleeding. Although the American College of Gastroenterology's clinical guidelines recommend colonoscopy after ischaemic colitis to screen for colorectal cancer, the actual detection rate of neoplastic lesions in patients without suspected malignancies on CT remains unclear. This study aimed to assess the efficacy of colonoscopy in detecting colorectal neoplasms after the resolution of ischaemic colitis.

Methods: This retrospective, single-centre, observational study included patients diagnosed with ischaemic colitis at the Isesaki Municipal Hospital in Japan between 2014 and 2023. Patients with CT-confirmed ischaemic colitis without a suspicion of colorectal cancer were eligible. Clinical data, colonoscopic findings and histopathological results were extracted from medical records. Comparative analyses were conducted between patients who underwent complete colonoscopy and those who did not.

Results: Among the 418 patients diagnosed with ischaemic colitis, 396 underwent CT imaging, and 116 underwent subsequent complete colonoscopy. Colorectal polyps were identified in 34.5% (40/116) of the patients, with 75 lesions predominantly located in the right-sided colon. Invasive colorectal carcinoma was detected in 3.4% (4/116) of the patients, along with an additional case of intramucosal carcinoma. Notably, one invasive adenocarcinoma was located proximal to the site of the ischaemic injury. Between the complete colonoscopy and incomplete/no colonoscopy groups, the patients in the incomplete/no colonoscopy group were significantly older, had a higher proportion of poor performance status and were more likely to have used saline laxatives.

Conclusion: Colonoscopy after ischaemic colitis revealed a non-negligible prevalence of colorectal neoplasms even in the absence of CT findings suggestive of malignancies. These results underscore the importance of colonoscopy after recovery, particularly in patients without a poor performance status. Further prospective, multicentre studies are warranted to validate these findings and optimise postischaemic colitis management strategies.

Objective: Upadacitinib is the first Janus kinase inhibitor and oral advanced therapy licensed for Crohn's disease (CD). Following NICE approval in 2023, real-world data on outcomes are limited. The effectiveness and safety of upadacitinib in a cohort of patients with CD was assessed.

Methods: A multicentre retrospective cohort analysis across 19 UK hospitals. Adult patients with active CD who started upadacitinib between April 2023 and October 2023 were included. Outcomes were reviewed over 24 weeks. The primary endpoint was clinical remission (Harvey Bradshaw Index (HBI) <4) at 12 and 24 weeks. Biochemical remission (faecal calprotectin <200 μg/g and C-reactive protein ≤5) and endoscopic remission (Simple Endoscopic Score for Crohn's Disease ≤3) were assessed at the same intervals. Adverse events (AEs) were recorded until 24 weeks or drug withdrawal.

Results: 312 patients were included, with a minimum follow-up of 12 weeks. The cohort had difficult-to-treat disease; 64% failing 3 or more biologics, 51% exhibiting penetrating or stricturing disease and 41% requiring prior resection. 50% (113/227) of patients achieved clinical remission at 12 weeks and 45% (77/172) at 24 weeks. Patients with colonic disease had higher remission rates at 24 weeks compared with other disease locations. At 24 weeks, 51 patients (16%) had discontinued upadacitinib. Treatment persistence was 90.3% at 12 weeks and 84.1% at 24 weeks. 28% had AEs, with 18% experiencing serious AEs and 16.6% requiring hospitalisation.

Conclusion: This is a large real-world study reporting outcomes in patients with CD treated with upadacitinib. Our data demonstrated good short-term effectiveness and tolerance in a clinically refractory population.

Objective: Colorectal cancer (CRC) is one of the most frequently diagnosed cancers and a leading cause of cancer-related deaths worldwide. While inflammatory bowel disease (IBD) is a well-established risk factor for CRC, the potential link between other autoimmune diseases and CRC is unclear. In light of the growing prevalence of autoimmune diseases and their recognised link to various malignancies, this study seeks to investigate whether different autoimmune diseases are associated with CRC.

Methods: A total of 20 146 patients with an initial diagnosis of CRC and 100 730 propensity score-matched cancer-free individuals were identified from the Disease Analyzer database (IQVIA). Univariable conditional logistic regression models were used to examine whether each autoimmune disorder was associated with subsequent CRC diagnosis.

Results: Only IBD was significantly associated with CRC (OR 1.53; 95% CI 1.33 to 1.75). Type 1 diabetes, rheumatic diseases, autoimmune thyroiditis, and multiple sclerosis did not show a significant association with CRC. Psoriasis showed a non-significant trend towards an association with CRC (OR 1.11; 95% CI 0.97 to 1.27). Coeliac disease was not associated with the development of CRC (OR 1.06; 95% CI 0.69 to 1.64). A sex-stratified analysis revealed that the association between IBD and CRC was similar in both women (OR 1.48; 95% CI 1.22 to 1.81) and men (OR 1.57; 95% CI 1.29 to 1.89). No significant sex differences for any other autoimmune disease were observed.

Conclusion: The presence of IBD, but not any other autoimmune diseases, was significantly associated with a subsequent CRC. This finding serves to emphasise the significance of routine screening for patients suffering from IBD.