Haematology Automate. Novel Intelligent Automation to save GP time and reduce referral pressures in secondary care.

GPs often struggle with high workload (Royal College of GPs; 2023). Time pressures mean they often do not have the time to thoroughly consult the guidelines that should inform their clinical decisions when pathology results are returned with abnormal or borderline results. To ensure that patients are treated in the best possible way, GPs therefore often refer patients to secondary care, even if that isn't the best course of action.

An audit of data from Imperial College NHS Trust showed 41% of routine clinic referrals and 87% of Advice and Guidance referrals to general haematology clinics that are triaged are returned to the referrer just with advice. However, lack of resource in secondary care also means many referrals are given a clinic appointment automatically without triage. This takes up consultant time and contributes to the large waiting lists seen in many hospitals.

In this project, GP Automate will develop a new type of software using Intelligent Automation, that will automatically provide guidance on the best course of action that a GP should take if abnormal pathology results are received. We will develop decision trees based on local and national guidelines, and with the help of expert haematologists to work out when GPs make referrals to secondary care that aren't needed.

We will then develop automated prompts that appear as notes on a patient's electronic health record precisely at the point that the abnormal/borderline result is seen. This would prevent uncertainty for time-limited GPs by bringing the guidance to them, thus facilitating better decision making and ensuring patients receive the right care quickly.

In this project we will focus on automation of haematology results, but the proposed technology is potentially disruptive, in the future enabling automated clinical decision support for numerous specialities throughout the NHS.