NHS England has published the telephone-queue data behind one of the headline claims supporting its national AI triage plans, allowing closer scrutiny of whether moving patients towards structured digital triage is reducing pressure on GP access channels.

NHS England released the dataset on 4 September after receiving a request for the figures used in its July AI-rollout announcement. The records come from Wealden Ridge Medical Partnership’s SurgeryConnect reporting system and cover December 2024 to April 2026, spanning periods before and after AI triage integration.

The earlier NHS announcement said the Sussex practice recorded a 29% reduction in the number of people queuing on the telephone after introducing the system while maintaining patient satisfaction. The result helped support plans to extend AI triage through the NHS App to more than 200,000 patients within 12 months and ultimately to all app users by April 2028.

Techopia examined the wider NHS rollout in July, when integrated triage and clinical notetaking were placed among the priorities for new technology spending. The publication of the source data now makes one of the programme’s claimed operational results more open to independent examination.

Call queues measure only part of the workflow

Wealden Ridge is a rural practice serving around 23,000 patients across four sites. Its same-day access process directs urgent appointment requests through AI-assisted triage whether a patient starts through the NHS App, online, by telephone, or at reception, with information then reviewed as part of the clinical workflow.

Call queues are a useful operational measure because one objective of digital triage is to reduce the morning rush in which large numbers of patients compete for telephone access at the same time. Structured digital requests can spread demand and provide information before a decision is made about whether a patient needs a GP appointment, pharmacy, emergency department, community service, or another route.

A lower telephone queue does not establish that total demand has fallen, however. Contacts can shift from voice to digital channels, while staff may encounter a different workload reviewing structured submissions after the visible queue has reduced.

The newly published dataset is specifically telephony evidence. NHS England says it contains the data and statistics underpinning the July press notice, but it should not be treated as proof of clinical outcomes or productivity improvements that the figures were not designed to measure.

National rollout raises a wider evidence requirement

As the service expands, evaluation will need to connect access metrics with what happens after a patient enters the system. Useful measures include review times, accuracy of prioritisation, the number of patients redirected between services, clinical outcomes, workload for reception and clinical staff, and whether patients who use non-digital routes receive equivalent access.

Human clinical involvement remains important in the current model. The triage tool adapts questions according to the patient’s responses and provides structured information that can help direct care or assist clinicians with prioritisation, rather than replacing every clinical judgement with an autonomous decision.

That design limits some risks associated with automated clinical decision-making, although productivity still depends on the whole workflow. A structured digital form that takes longer to review than the process it replaces may improve information quality without increasing capacity.

Publication of the Wealden Ridge figures is therefore useful because public-sector AI claims can otherwise move ahead of the evidence beneath them. Operational data gives clinicians, researchers, and procurement teams something more concrete with which to distinguish an improvement from a change in where demand appears.

The NHS has already committed to a much larger deployment timetable, so evidence requirements will become harder as the service reaches practices with different patient populations, staffing models, telephone systems, and local service capacity. A 29% reduction in one telephone queue supports further examination; demonstrating that AI triage improves access across the NHS will require a broader set of operational and clinical measures.