Evaluating the impact of artificial intelligence triage in online consultations to reduce delays in urgent primary care: interrupted time series analysis and quantitative process evaluation

The University of Manchester requires access to NHS England data for the purpose of the following research project:

Evaluating the impact of artificial intelligence (AI) triage in online consultations to reduce delays in urgent primary care: interrupted time series analysis and quantitative process evaluation.

PATCHS is an online consultation system (OCS) which has been available to GP practices in England and Wales since 2020, to help manage patient queries and improve access to services. It acts as a "total triage" system, allowing patients to submit a range of requests online, which are then reviewed and prioritised by the practice staff. Patients access PATCHS through their GP practice's website or the NHS App. Via the AI-Powered Triage, the system uses artificial intelligence (AI) to help practice staff triage the requests. its purpose is to provide an efficient and convenient way for patients to contact their GP practice for various health needs. The AI can identify urgent cases, assign requests to the most appropriate clinician (e.g., GP, nurse practitioner, or pharmacist), and highlight requests that may require a face-to-face consultation.

The aims of the research project are to identify the following:

• Whether patients registered with practices that use the PATCHS AI have higher rates of Accident & Emergency (A&E) events and/or being admitted to hospital, what is their proportion compared to patients in practices that use PATCHS without AI.

• What are the patient-level characteristics of patients who attend A&E and/or being admitted to hospital in practices that use PATCHS in terms of age range, gender, ethnic groups, level of social deprivation, region in England.

• What are the most common modes of arrival (will be defined using the 'ARRIVAL_MODE' field) and causes of admission (will be defined using the ECDS 'DIAGNOSIS' fields), discharge destinations and discharge statuses (the facility which the patient has been streamed to). Data will be defined using the ‘DISCHARGE_DESTINATION_VALID_APPROVED’ and ‘DISCHARGE_STATUS_VALID_APPROVED’ data fields, respectively).

• What are the patient-level characteristics of injuries in terms of geographical location of injury, injury intent (unintentional / self-inflicted / assault), type of activity at time of injury (e.g. leisure / work), physical activity being undertaken by individual at time of injury, mechanism of injury, and whether drugs and alcohol were likely to have contributed to the injury.

The following NHS England Data will be accessed:

• Emergency Care Data Set (ECDS) & HES Accident & Emergency (A&E) data – necessary to identify whether patients registered with practices that use the PATCHS AI have higher rates of A&E events, what is their proportion compared to patients in practices that use PATCHS without AI. It is also necessary to identify patient-level characteristics of patients who attend A&E in practices that use PATCHS.

• Hospital Episode Statistics Admitted Patient Care (HES APC) necessary to identify whether patients registered with practices that use the PATCHS AI have higher rates of admission to hospital, what is their proportion compared to patients in practices that use PATCHS without AI. It is also necessary to identify patient-level characteristics of patients being admitted to hospital in practices that use PATCHS.

• Civil Registrations of Death dataset – necessary to calculate patient-level rates of deaths to identify whether practices that use PATCHS AI have higher rates of deaths and what is their proportion compared to patients in practices that use PATCHS without AI.

The level of the Data will be pseudonymised.

The Data will be minimised as follows:

• Patients registered with up to 1,100 practices GP who use a particular online consultation software system (the PATCHS) and those practices that do not.

• Data limited to the years 2019-2026.

• Events limited to when patients are registered with specific 1,100 selected GP practices.

The University of Manchester is the research sponsor and the controller as the organisation responsible for ensuring that the Data will only be processed for the purpose described above.

The lawful basis for processing personal data under the UK GDPR is:

Article 6(1)(e) - processing is necessary for the performance of a task carried out in the public interest or in the exercise of official authority vested in the controller.

The lawful basis for processing special category data under the UK GDPR is:

Article 9(2)(j) - processing is necessary for archiving purposes in the public interest, scientific or historical research purposes or statistical purposes in accordance with Article 89(1) based on Union or Member State law which shall be proportionate to the aim pursued, respect the essence of the right to data protection and provide for suitable and specific measures to safeguard the fundamental rights and the interests of the data subject.

This processing is in the public interest because it hopes to help members of the public and GP practices understand what AI is and how they can use it to benefit both patients and staff.

The funding is provided by National Institute of Health Research (NIHR) Health and Social Care Delivery Research (HSDR). The funding covers the study described but is not specifically limited to the study described. Funding is in place until 30/04/2026.

The funders will have no ability to suppress or otherwise limit the publication of findings.

The project has an independent Study Steering Committee (SSC) group consisting of a chair and three members, including a lay member.

Data will be accessed by substantive employees of University of Manchester.

Interviews were conducted with seven patients and workshops held with 10 GP practice staff to gather feedback on the study plans and they helped refine the purpose of the research (for the overall project and including aims related to this study). As part of the successful NIHR funding application, which included three PPIE co-applicants, the study team conducted extensive public and staff involvement to shape the study design and objectives. This included interviews with seven patients and two workshops with ten GP staff. Participants supported the project’s relevance, especially given the pressures on primary care and the rise of online consultations. This activity was for the overall project, including the aims related to this study.

Concerns were raised about AI bias, especially for non-native English speakers, and data confidentiality. In response, the study team incorporated measures to address health inequalities and privacy, such as help articles explaining the AI system and monthly engagement seminars. These materials were co-developed with patients and will be refined throughout the project.

Ongoing PPIE includes two dedicated groups: a Patient and Carer Involvement group (6 public members) and a Primary Care Staff Involvement group (6 staff members). Over five meetings have been held with each group to discuss project developments and emerging findings.

The PI is a part-time employee of PATCHS Health as Chief Medical Officer and shareholder in the company. The PI's role in the company is to provide medical and clinical expertise to ensure that the system is safe, efficient, and helpful for both patients and healthcare staff.

PATCHS is a commercial product provided by PATCHS Health company, which is a private limited company. The system is funded by the NHS for use by GP practices, with a focus on providing a solution to improve healthcare delivery. It is a data-driven business that uses anonymised data to improve its service, including the development of its AI capabilities. PATCHS is a commercial product from a private company, but its funding model and data usage are structured to align with NHS objectives.

No data will flow to NHS England for the purposes of this Data Sharing Agreement (DSA).

NHS England will grant access to the Data via the Secure Data Environment (SDE). The SDE is a secure data and research analysis platform. It allows approved researchers with approved projects access to pseudonymised data and industry-leading analytics tools.

NHS England will provide the relevant records from Hospital Episodes Statistics (HES APC), Emergency Care Dataset (ECDS) and the Civil Registrations of Death to University of Manchester via the NHS England Secure Data Environment (SDE).

The Data will contain no direct identifying data items. The Data will be pseudonymised and individuals cannot be reidentified through linkage with other data in the possession of the recipient.

The Data will not be transferred to any other location.

The Data will be stored on servers at NHS England via the SDE.

SDE users can request exportation of aggregated analysis results (suppressed and summarised according to the NHSE SDE Disclosure Control rules) subject to review and approval by the NHS England SDE Output Checking team. The SDE Output Checking team will ensure that no output contains information which could be used either on its own or in conjunction with other data to breach an individual's privacy.

The Data will be accessed by authorised personnel via remote access.

The Controller(s) must confirm and provide evidence upon audit by NHS England that access via any remote device complies with the data security obligations within this DSA and the Data Sharing Framework Contract.

For remote access:

- Remote access will only be from secure locations situated within the territory of use (as further restricted elsewhere within the DSA if so done) stated within this DSA;

- Access controls granting users the minimum level of access required are in place;

- Remote access is only via secure connections (e.g., VPNs or secure protocols) to protect data;

- Multifactor authentication (MFA) is required for remote access;

- Device security, including up-to-date software and operating systems, antivirus software, and enabled firewalls are utilised for the remote access;

- All remote access is undertaken within the scope of the organisation’s DSPT (or other security arrangements as per this DSA) and complies with the organisation’s remote access policy.

The above applies in addition to any condition set out elsewhere within the DSA (e.g. who may carry out processing, and for what purpose).

The Data will not leave England/Wales at any time.

Access is restricted to employees of University of Manchester who have authorisation from the Principal Investigator.

All personnel accessing the Data have been appropriately trained in data protection and confidentiality.

The Data will not be linked with any other data.

There will be no requirement and no attempt to reidentify individuals when using the Data.

The expected outputs of the processing will be:

• A report of findings to the funder NIHR HSDR annually and at the end of the project.

• Submissions to peer reviewed journals. One paper submission is expected towards the end of the project.

• Presentations in departmental seminars to students and members of staff at the University of Manchester, including the PPIE group.

• Presentations at academic conferences, including: the Society for Academic Primary Care (SAPC) scientific conference July 2026, and the Health Services Research UK (HSR UK) conference July 2027.

The outputs will not contain NHS England Data and will only contain aggregated patient-level information with small numbers suppressed as appropriate in line with the relevant disclosure rules for the dataset(s) from which the information was derived.

The outputs will be communicated to relevant recipients through the following dissemination channels:

• Journals.

• Webinars open to students and members of staff at the University of Manchester, including the PPIE group.

• Social media.

• Posters displayed at academic conferences.

• Press/media engagement.

The outputs will be aimed at wider and diverse audiences to optimise the potential public benefits from the use of the data, mainly via lay and technical and summaries and outputs outlined above and by advertising the findings via different media outlets e.g. online blogs, newspapers and local radio with the support of the Media Relations officers at the University of Manchester. The study team will also disseminate the findings to local policy makers via the support of Policy@Manchester team. Importantly, the study team will consult with the PPIE group on the planned dissemination strategy and on potential public benefits of the use of the data.

Target date for production and dissemination: January 2028.

It is hoped this research study will contribute to evidence-based decision-making for policy-makers, local decision-makers such as doctors, and patients to inform best practice to improve the care. If it is concluded that AI system reduces care delays and hence A&E and hospital admissions, patients who need urgent and emergency help will receive it sooner

It is hoped the research will help the NHS and companies that make online consultation systems decide whether they should use AI. Through detailed training within GP practices on how the module works and PPIE engagement, the research aims to help members of the public and GP practices understand what AI is and how they can use it to benefit both patients and staff. Regardless of whether AI Triage is effective, evidence generated from this project will be used to create help guides and toolkits on how to use AI Triage safely and effectively.

Yielded Benefits is not a requirement for new applications.