Development and implementation of novel virtual triage and exploration of attitudes towards the potential use of artificial intelligence in the irritable bowel syndrome (IBS) dietetic pathway

Background: Dietary management is integral to the irritable bowel syndrome (IBS) pathway. Triage facilitates the decision-making process for the right dietetic intervention; however telephone triage is time intensive. Digital advances provide an opportunity to target waiting times and clinical capacity. The aim of this work was to develop and implement a novel semi-automation virtual triage, assess its impact in the IBS pathway and to investigate attitudes towards the use of artificial intelligence (AI) in triage and dietetic healthcare.Methods: The Consolidated Framework for Implementation Research (CFIR) provided a structure to develop and implement virtual triage into the IBS pathway. A digital triage questionnaire was developed using experience-based co-design. The efficacy of virtual triage was compared to telephone triage for waiting times from referral to triage, clinicians’ time taken to triage and clinical capacity. Using qualitative interviews, views on AI in virtual triage and the IBS pathway were collected from 3 patients and 2 dietitians who had experience of the newly developed virtual triage process. An exploratory survey in 7 gastroenterology dietitians was used to assess attitudes and experience of AI in clinical practice.Results: A digital questionnaire was developed and embedded into the IBS pathway for virtual triage. Following implementation, 643 patients received virtual triage with 83% completing the digital questionnaire. From telephone triage to virtual triage, mean waiting times reduced from 56.6 days to 17.5 days, mean clinician time to triage decreased from 20 minutes/patient to 11 minutes/patient and clinical capacity increased from 400 to 1000 appointments/year (all pConclusions: Virtual triage increases clinical capacity and reduces waiting times without increasing clinician burden. Attitudes towards AI shows interest, however there is a need for validation to determine confidence and acceptability for both clinicians and patients in terms of problem solving and healthcare efficiency.