Attempting to model referral patterns from emergency departments

  • Research type

    Research Study

  • Full title

    Decision support algorithms for emergency departments: Developing machine learning classifiers to predict departure points in emergency departments

  • IRAS ID

    275577

  • Contact name

    Neil White

  • Contact email

    nmw@ecs.soton.ac.uk

  • Sponsor organisation

    University of Southampton

  • Duration of Study in the UK

    1 years, 0 months, 30 days

  • Research summary

    Using data collected by the University Hospitals Southampton Emergency Department (ED), the focus of the work will be investigating whether data collected during a given ED episode (typically lasting less than 4 hours) can be used with machine learning methodologies to predict the episode outcome (i.e., was the patient discharged, admitted to hospital or did they die?).
    The study will be an entirely retrospective study, using only historical data, and data will include patient observations (e.g., blood pressure), patient laboratory tests (e.g., blood tests), patient information (e.g., date or year of birth, gender, ethnicity) and episode information (e.g., date of episode, duration of episode). The primary aim will be to validate the machine learning models we will build and compare the predictions to conventional methods.

  • REC name

    N/A

  • REC reference

    N/A