Medium and long-term outcome prediction after cardiac interventions

  • Research type

    Research Study

  • Full title

    Medium and long-term outcome prediction after cardiac interventions

  • IRAS ID

    350787

  • Contact name

    P D Fudulu

  • Contact email

    daniel.fudulu@bristol.ac.uk

  • Sponsor organisation

    University of Bristol

  • Duration of Study in the UK

    4 years, 10 months, 30 days

  • Research summary

    As many as 24000 adult patients need a heart operation and more than 110000 patients need cardiac percutaneous procedures in the UK annually. Heart surgery is associated with relevant procedural risk; therefore, weighing the risks and benefits in each case is paramount. For this purpose, doctors and patients need to rely upon risk stratification models, which can inform on the likelihood of complications and long-term benefits following the surgery. No tailored information can be provided to patients on long term consequences (i.e. mortality, risk of rehospitalisation, pacemakers, arrhythmias ) because such a tool has not been developed yet. This is because while data on in-hospital mortality are available for risk modelling through the national cardiac audit (i.e. NACSA), which collects hospital data, no routine linkage exists between the NACSA and administrative datasets (i.e. ONS, HES), which collects information longitudinally (long term). Also, most prediction tools focus on short-term outcomes and very few on long-term outcome prediction. We propose to link the NACSA dataset with ONS and HES administrative dataset to inform clinicians and patients on long term consequences of heart surgery and to develop a risk stratification tool that includes not only hospital mortality but also long term outcomes. Access to cardiac interventional dates will enable us to develop decision tools to guide clinicians about the best approaches – percutaneous or surgical in high-risk patients and patient groups where there is uncertainty over the best type of intervention.

    A similar initiative of linking the National Congenital Heart Disease Audit (NCHDA) dataset with paediatric and adult intensive care data sets and administrative hospital episodes (HES) and mortality registry data was successfully performed in the paediatric population by the LAUNCHES group to understand the longitudinal healthcare trajectories for patients with congenital heart disease.

    Sample size calculation

    No formal sample size calculations need to be performed for this method of analysis. These population figures consider the total number of patients that would have been undergone cardiovascular interventions during the time period specified. Any increase in sample size will improve the sensitivity and specificity of our prediction model and hence we hope to incorporate data on all available patients.

  • REC name

    Wales REC 7

  • REC reference

    26/WA/0017

  • Date of REC Opinion

    5 Feb 2026

  • REC opinion

    Favourable Opinion