Stix

  • Research type

    Research Study

  • Full title

    The Stix project

  • IRAS ID

    239010

  • Contact name

    Cecilia Lindgren

  • Contact email

    cecilia.lindgren@wrh.ox.ac.uk

  • Sponsor organisation

    University of Oxford

  • Duration of Study in the UK

    4 years, 11 months, 30 days

  • Research summary

    As part of clinical investigations, urine is a surprisingly effective place to look for signals of possible disease conditions. It has been known since ancient times that sweet urine is a sign that a person might be diabetic, but there are many health conditions where urine can be useful for diagnosis.
    There are different ways of examining urine, from direct visual or smell inspection, to dipstick testing, to full laboratory chemical analysis. Tests where values fall outside the range of normal values can indicate the presence of several diseases, including urinary tract infections (UTIs), diabetes, kidney disease, liver disease, preeclampsia, kidney stones, eating disorders and gestational diabetes. Together, these cover some of the most common reasons people seek healthcare services in this country and around the world. By far the most common urine test performed is the urine dipstick. In the Oxford University Hospitals NHS Foundation Trust (OUH) alone, >40,000 urine dipstick tests are purchased annually.
    While urine dipstick tests have been a cheap and widely used tool for frontline clinical assessments for the past 120 years, the recording and interpretation of the tests remains inconsistent. Given the wide use of urine dipsticks, they are, collectively, a significant time and cost burden on healthcare systems - the ability to perform them at low cost, high speed, anywhere, by anyone, and with high accuracy would be an enormous potential cost efficiency improvement.
    Surveys of general practitioners in the UK identify the direst clinical need for improvements in how urine tests are performed and interpreted. We will be using machine learning approaches to capture data from dipstick test and evaluate these algorithms in terms of clinical utility. The view is to evaluate if the algorithms developed from the data in the Stix project could be implemented in a new application for future clinical testing.

  • REC name

    West of Scotland REC 5

  • REC reference

    21/WS/0061

  • Date of REC Opinion

    7 May 2021

  • REC opinion

    Favourable Opinion