Intelligent Navigation using AI to bust waiting times for urgent healthcare (INA)

Long waiting times and avoidable delays in urgent care (UC) negatively impact patient outcomes, decrease staff morale and diminish public satisfaction with the NHS. The demand for UC is increasing, driven by an aging population, increasing health inequalities, and rising multimorbidity. Many patients seek UC because they are waiting for scheduled treatments or admission. To address these challenges, a transformational approach to patient assessment and prioritisation is required. Currently, UC triage assesses presenting symptoms and cannot consider background medical history or broader health risks. Patient-specific data is only accessible post-assessment. Real time analysis of shared primary and secondary care patient data could enable earlier risk stratification during triage, leading to more effective, faster care. Intelligent Navigation is a significant advancement over current NHS111 triage, which relies on rigid decision trees navigated by humans. It integrates a regulated medical device (Visiba Triage) with the Johns Hopkins ACG (Adjusted Clinical Groups) system, to prioritise care based on the patient’s medical history. By harnessing AI to analyse health records alongside symptom assessment waiting times can be reduced by ensuring that the most urgent patients are seen first, and by providing clinical information so that multiple touchpoints where there is no added value are removed. This can free up an estimated 29m GP appointments per year and release administrative capacity at the NHS front door. Objectives: We will undertake a comprehensive mixed-method evaluation of AI-based intelligent triage as a way to reduce UC waiting times, creating a replicable blueprint for AI-powered NHS UC navigation that can be scaled by: 1. Deploying Intelligent Navigation for UC Buckinghamshire, Oxfordshire and West Berkshire Integrated Care Board area for 12 months; 2. Using a mixed-method research design to evaluate this deployment and impacts on waiting times; 3. Establishing a robust clinical governance framework for Intelligent Navigation; 4. Using project learning to develop implementation guidance for national scale-up Methods: Patients using NHS111 will be invited to opt into Intelligent Navigation alongside assessment. We will examine effectiveness and cost-effectiveness using monthly service performance data and deidentified routinely collected administrative, demographic and clinical data. Our primary outcome is total assessment time reduction. Statistical analyses will include pre and post intervention descriptive measures. The primary outcome will be analysed using a multivariable linear regression model, and we will use transformation and non-parametric analysis if total call time deviates significantly from a normal distribution. We will use surveys to explore patient and staff experience and implementation, conduct 50 semi-structured interviews with staff and stakeholders and 9 workshops with diverse groups patients. We will analyse relevant documents and conduct two 90 minute online expert workshops with staff and stakeholders to explore implementation challenges. We will synthesise findings to identify learning and overarching conclusions. Our patients & communities panel will input directly throughout. Anticipated outcomes include measurable improvements in UC access, significantly reduced waiting times, enhanced patient safety and navigation, increased system efficiency and improved clinical outcomes.