Digital-Supported Versus Standard Consent for Advanced Therapeutic Endoscopy

Overview

This study will evaluate whether a digital-supported informed consent process improves patient understanding before advanced therapeutic endoscopy procedures. Patients undergoing advanced therapeutic endoscopy are often asked to make decisions about procedures that may be complex and involve important risks, benefits, and alternatives. Standard informed consent is usually provided through discussion with the treating physician. However, this process can vary between clinicians and may be affected by time pressures in busy endoscopy units. In this study, eligible adult patients scheduled for elective advanced therapeutic endoscopy will be randomly assigned to one of two consent pathways. One group will receive standard physician-led informed consent. The other group will receive a digital-supported consent pathway that includes a standardized procedure-specific educational video, an offline artificial intelligence-based question-answering tool, and final confirmation with the treating physician before the procedure. The digital tool is designed to provide plain-language information about the procedure, including why it is being done, expected benefits, common and serious risks, alternatives, what to expect before and after the procedure, and the patient's right to ask questions or decline the procedure. The AI question-answering tool will use only investigator-approved educational content and will not provide individualized medical advice. Questions requiring patient-specific medical judgment will be directed back to the treating clinical team. The main outcome of the study is patient comprehension of the consent information, measured after the assigned consent process and before the procedure. The study will also evaluate patient anxiety, satisfaction, confidence in decision-making, usability of the consent process, workflow outcomes, and consent-related safety or ethical concerns.

This is a pragmatic multicenter randomized controlled trial evaluating digital-supported informed consent compared with standard physician-led informed consent for adults undergoing elective advanced therapeutic endoscopy. The study will be conducted in three sequential phases. In the first phase, the digital consent tool will be developed through an iterative co-design process involving the study team, a specialist engineer, and at least one patient partner. The tool will include a standardized procedure-specific educational video and an offline artificial intelligence-based question-answering interface. The AI component will be developed using only investigator-approved educational materials, standardized consent language, and curated frequently asked questions. It will not browse the internet, retrieve external information in real time, or provide individualized medical advice. In the second phase, a pilot study will assess feasibility, usability, acceptability, and operational performance of the digital consent pathway. The pilot will assess whether participants can access and complete the video and AI interface, whether the digital tool functions as intended, whether the survey-based outcome assessment workflow is practical, and whether refinements are needed before the full randomized trial. The pilot phase is not intended to provide definitive comparative effectiveness results. In the third phase, the finalized intervention will be evaluated in a multicenter, parallel-group randomized controlled trial. Eligible participants will be randomized in a 1:1 ratio to either standard physician-led informed consent or the digital-supported consent pathway. Randomization will be centralized and concealed using a secure web-based system, with stratification by site and potentially by major procedure category. Participants assigned to the standard consent arm will receive the usual physician-led informed consent process according to local clinical practice. Participants assigned to the digital-supported arm will receive access to a procedure-specific digital consent package, ideally before the procedure. The digital package will include a standardized educational video explaining the nature of the procedure, the reason it is being performed, expected benefits, major and common risks, reasonable alternatives, what to expect before and after the procedure, and the right to ask questions or decline the procedure. After viewing the video, participants may ask questions through a secure offline AI-based interface. The tool will provide plain-language responses grounded only in pre-approved study materials and will redirect individualized or high-risk questions to the treating clinical team. In both study arms, final procedural consent remains the responsibility of the treating physician. In the digital-supported arm, the physician will conduct a final confirmation discussion before the procedure to review unresolved questions, address individualized risks, benefits, and alternatives, confirm patient understanding, and complete standard institutional consent documentation. The primary outcome is patient comprehension of informed consent information, assessed immediately after completion of the assigned consent pathway and before the procedure. The comprehension assessment will address key domains of informed consent, including the nature of the procedure, indication, risks, benefits, alternatives, right to ask questions or refuse, and basic expectations before and after the procedure. Secondary outcomes include pre-procedural anxiety, satisfaction with the consent process, decisional confidence or self-efficacy, perceived usability and acceptability, perceived opportunity to ask questions, preference for future consent format, physician time spent on consent, staff time where measurable, consent-related delays, same-day cancellation or deferral due to unresolved consent concerns, and need for additional pre-procedural contact related to consent. Digital process outcomes will include video completion rate, AI tool utilization rate, number and categories of patient questions, frequency of questions escalated to clinician review, and completion of the intervention before arrival. Safety and ethical quality outcomes will include withdrawal of procedural consent after the assigned intervention, patient-reported misunderstanding, complaints related to the consent process, protocol deviations involving consent, and any adverse event judged related to misunderstanding of instructions or procedural expectations. The study is designed to minimize collection of identifiable information. Study outcomes will be collected primarily through secure survey links, and the research dataset will not include direct identifiers such as names, dates of birth, medical record numbers, addresses, email addresses, telephone numbers, or IP addresses. Any retained AI interaction data will be anonymized and will not be linked to participant identifiers.