AI for Prospective Lung Cancer ID

  • Research type

    Research Study

  • Full title

    Prospective Identification of Patients at Risk of Lung Cancer using Artificial Intelligence

  • IRAS ID

    349748

  • Contact name

    William Ricketts

  • Contact email

    william.ricketts@nhs.net

  • Sponsor organisation

    Barts Health NHS Trust

  • Duration of Study in the UK

    1 years, 0 months, 1 days

  • Research summary

    This study aims to enhance the early detection of lung cancer by using electronic health records (EHR) and a predictive model to identify high-risk individuals more effectively. Currently, the diagnosis of lung cancer heavily relies on general practitioners recognising specific symptom presentations. This subjective approach can lead to significant delays in diagnosis and treatment, with many patients potentially being overlooked. Moreover, existing lung cancer screening programs in the UK are constrained, particularly for non-smokers and older individuals over 75, due to strict eligibility criteria and associated costs.

    In a previous study (IRAS: 353181), an AI predictive model was developed using existing patient data collected from primary and secondary care EHRs and census records. In this study, high-risk patients according to the developed model will be invited for a chest X-ray to detect any lung abnormalities. Should the chest X-ray results indicate potential malignancy, patients will be adopted into the typical diagnostic pathway for lung cancer, which includes a chest CT and subsequent biopsies.

    High risk individuals will be identified from a cohort comprised of patients treated at Barts Health NHS Trust, aged between 40 and 80 years old who have attended a primary and secondary care facility since 2015, excluding patients who have opted out of their data being used for research through a type 1 opt out.

    Following the prospective phase of the study, health economic analyses will be conducted to determine the financial implications of implementing the predictive model in routine clinical practice. This analysis will compare the costs associated with the current care model against those of the proposed early detection strategy, factoring in potential savings from earlier diagnoses and improved patient outcomes.

  • REC name

    London - Queen Square Research Ethics Committee

  • REC reference

    25/LO/0683

  • Date of REC Opinion

    6 Nov 2025

  • REC opinion

    Further Information Favourable Opinion