Introduction

Proton pump inhibitors (PPIs) are extensively prescribed due to the widespread diagnosis of gastro-oesophageal disease (GORD). ^1,2 Overutilisation of PPIs is common, and growing concerns over adverse events, inappropriate prescribing and unnecessary costs ^3,4 are driving exploration into the implementation of holistic management, including deprescription strategies¹.

Aims & Methods

We aimed to quantify healthcare professionals’ (HCPs) preferences on the management of GORD in the UK, and identify attributes perceived as enhancing or detracting from quality of care.
Attributes were derived from qualitative data gathered through 3 focus groups with gastroenterologists, pharmacists and general practitioners (GPs), and interviews with 4 GORD patients.
A discrete-choice experiment (DCE) was developed and administered via online survey (Jan–Apr 2025) to 38 HCPs. Respondents chose between paired hypothetical GORD management scenarios, varying by symptom severity, appointment frequency, treatment approach, information delivery method, and cost. Preferences were modelled using a multinomial conditional logit with effects coding. Marginal rates of substitution (MRS) were calculated to quantify trade-offs.

Results

74% of respondents were pharmacists (n=28) and 26% GPs (n=10). All had managed GORD patients in the past year, and 81% expressed concern over long-term PPI use.
A holistic approach combining PPIs with lifestyle modification was strongly preferred over PPI monotherapy (β=0.4244, p<0.001), representing the most influential driver of care quality. Lifestyle modification alone was also favoured over PPI monotherapy (β=0.1125), though this did not reach statistical significance (p=0.127).
Respondents expressed significant aversion to providing information via pamphlets compared to GP-led counselling (β=–0.1836, p=0.046), while no preference emerged between pharmacist- and GP-led counselling (β=0.0512, p=0.573). Neither appointment burden nor out-of-pocket costs significantly influenced decision-making.
Symptom severity was a notable driver: HCPs selected scenarios involving moderate symptoms over mild (β=0.2353, p=0.005), indicating increased urgency and prioritisation for moderate disease.
Analysis of MRS between attributes suggested that HCPs are willing to accept 4.4 additional appointments to manage moderate symptoms, rather than mild, and nearly 8 additional appointments to deliver combined holistic lifestyle and pharmacological therapy. This highlights a willingness to absorb appointment-related resource demands in favour of more comprehensive, effective care.

Conclusion

Concerns around long-term PPI use support a shift toward active GORD management and PPI deprescribing, with structured follow-up to monitor treatment response.
This interim analysis highlights a clear preference among HCPs for multimodal GORD management, particularly combining lifestyle modification with pharmacological treatment, and reinforces the importance of interactive clinician[BA1] -led patient education, which can promote compliance and improve outcomes. The willingness to accept 4.4 additional appointments reflects the value placed on ongoing monitoring and personalised care. However, this raises the question on the impact this may have on health services, and whether the integration of self-care into multimodal approaches developed with HCPs could be an alternative solution to be investigated. [LH2] These findings highlight the need for a system-wide shift toward guideline-informed, stepped-care models.
A further analysis on patient population preferences is under review.

References

1. Savarino E, Anastasiou F, Labenz J, Hungin APS, Mendive J. Holistic management of symptomatic reflux: rising to the challenge of proton pump inhibitor overuse. Br J Gen Pract. 2022; 72(724):541-544.
2. Hungin APS, Scarpignato C, Keefer L, Corsetti M, Anastasiou F, Muris JWM, Mendive JM, Kahrilas PJ. Review article: rethinking the "ladder" approach to reflux-like symptom management in the era of PPI "resistance" - a multidisciplinary perspective. Aliment Pharmacol Ther. 2022; 55(12):1492-1500.
3. Plehhova K, Wray J, Aluko P, Sutton S, McArdle J, Dawson A, Coyle C, Stevens RM. Prescribing practices for proton pump inhibitors among primary care physicians in England: an evaluation. BJGP Open. 2025.
4. Sosio J, McKee N, [BA1] [LH2] Gold R, Singson A, Boroda K. S579 PPI Stewardship: A Quality Improvement Tool to Improve Long-Term Prescribing Practices. Am J Gastroenterol. 2024; 119(10S): S400.

Disclosure

This study was funded by Reckitt Benckiser Healthcare Ltd.
K. Plehhova, and E. Horvath are employees of Reckitt.

Introduction

Artificial intelligence (AI) characterization (CADx) systems for colonic polyps exhibit high diagnostic accuracy. However, they may only support, not replace, clinical decision-making and little is known of this human-AI interaction – even less so with trainee endoscopists. This issue is difficult to explore, but data from a study investigating CADx-supported trainee endoscopists compared to AI-blinded experts could give insight.

Aims & Methods

This analysis draws on data from a prospective observational study comparing the diagnostic performance of CADx-supported trainee endoscopists (using GI Genius®, Medtronic) to expert endoscopists blinded to AI support. Two key scenarios were explored: (a) instances in which the CADx system produced a ‘no prediction’ output, and (b) cases where trainees actively overruled AI-generated diagnoses. These were assessed to examine the impact of AI absence and human-AI discordance on diagnostic accuracy.

Results

The baseline cohort included 11 trainees performing 225 colonoscopies, mean age 63.8 years (SD 12.7), 48.9% male patients. Baseline ADR was 57.8%. Of 630 resected lesions (median size 3 mm, IQR 2;4), 40.0% were in the rectosigmoid, and 53.6% were non-adenomatous. CADx failed to provide a prediction in 19.2% (n=121) of cases. These ‘no prediction’ polyps were smaller (median 3 mm, p=0.048) and more often located in the right colon (72% vs. 58%, p=0.005. In the absence of AI support, trainee diagnostic accuracy declined significantly by 12.8% (63.6% vs. 76.4%, p=0.007). Interestingly, expert endoscopists—despite being blinded to AI—also performed worse on these polyps (64.7% vs. 73.9%, p=0.045), suggesting an inherent difficulty in classifying these lesions.
Trainees overruled AI suggestion leading to discordant diagnoses in 9.0% (57/630) of polyps, comprised of 20 (35.1%) non-adenomatous, and 37 (64.9%) adenomatous lesions. Trainees incorrectly overruled AI in 59.6% (34/57), which mostly consisted of downgrading lesions to NICE 1. Discordant polyps did not differ in size (3 mm vs. 3 mm, p=0.413), but were more often located proximally (77.2% vs. 58.1%, p=0.005). The difference in false negative rate (FNR) between trainees and AI was 62.16% (81.07% vs. 18.92%, p<0.0001), demonstrating a risk of missed adenomas when trainees override AI. Blinded expert endoscopists achieved identical overall diagnostic accuracy in those polyps as the AI system (59.6%). However, their error profiles differed – In this subset, blinded experts and AI performed equally in overall accuracy (59.6%), although with differing error profiles: AI showed a numerically lower FNR (18.9% vs. 35.1%, Δ16.2%, p=0.17), offset by a higher false positive rate (80.0% vs. 50.0%).

Conclusion

CADx systems generate ‘no prediction’ outputs in a substantial proportion of polyps, significantly impairing trainee diagnostic performance. These cases may represent inherently more challenging lesions. Despite CADx’s high accuracy, trainees overrule AI in almost every tenth polyp, most often incorrectly downstaging polyps. Structured training on appropriate use of CADx, emphasizing both its capabilities and limitations, may enhance clinical effectiveness. Further research on human-AI interaction and trust is warranted, as these factors will be critical for successful integration of CADx into routine endoscopic practice.

Disclosure

The original study (AC-CADx) received technical support by Medtronic. Medtronic was not involved in the design of the study or analysis of the data, neither in the original study nor this exploratory analysis.