LTSS: Accurately predicting the road to recovery

The patient’s roadmap can be uncertain, but like any journey, the patient and the clinician have an idea where they need to go - hopefully a healthy recovery! We’ve been working on a tool that feels like a GPS for doctors, giving an accurate prediction of how long a patient is likely to stay in hospital, and the risks along the road to a safe, timely discharge.

Built and trained with real patient data, our Long Term Stay Scoring (LTSS) tool now gives doctors and nurses unmatched insight into the length of stay for patients; supplying a Machine Learning Estimated Discharge Date (ML EDD) directly to their Electronic Patient Record. LTSS is constantly updated, in real time, reflecting every risk factor to the patient’s journey, as it arises.

Our prediction power has been independently verified to predict the length of stay for any hospital patient to within an average of 6 hours - this is four times as accurate as clinicians when they’re asked to predict a patient Estimated Discharge Date.

Here at Polygeist Health, we don’t think AI should ever replace or displace doctors - but it can give them clearer, smarter insights to use as they map out the twists and turns of individual care pathways.

For the first time, the LTSS tool provides your clinicians with a robust evidence-base to make admission decisions, particularly when those patients might be better treated in a community care setting without admitting. This one use-case alone could save as many as 7000 patient lives and over 3m bed days each year in the NHS - or £2.28m per year for an average Trust.

Once patients are admitted, LTSS provides highly accurate real-time prediction of each patient’s length of stay, as well as their unique, personalised risk-factor combinations and the likelihood of mortality.

A fundamental feature of this tool is the surfacing of key insights to the right person at the right time, without fuss. Working seamlessly and continuously in the background to deliver insights directly into any EPR, our tool was built with busy ED clinicians in mind, minimising the cognitive burden to maximise the impact on individual patient care.

Early Discharge Saves Lives

Our analytical assessment, matched with recently published peer-reviewed evidence suggests that clinicians could safely reduce the average stay by between one and five days per long-staying patient.

Over 7000 lives per year in England can be saved simply by identifying those high risk patients and discharging five days earlier; mortality of discharged patients would be reduced by 10% and readmissions would go down by 50% over 12 months.

Polygeist Health have been working hand-in-hand with our friends at Gloucestershire Hospitals NHS Foundation Trust and NHS England since 2021 to develop LTSS, and we are extremely excited about the mission impact it will have.

If you’d like to know more about the technology, the science, and the benefits, click the button by my name at the top of the page to arrange a quick briefing!