Assessment of a diagnostic assistant app in Accident and Emergency

  • Research type

    Research Study

  • Full title

    Assessment of DemDx, a differential diagnosis tool, and its accuracy at triaging patients in an A&E setting

  • IRAS ID

    229534

  • Contact name

    Hannah Noone

  • Contact email

    H.Noone@warwick.ac.uk

  • Sponsor organisation

    University of Warwick

  • Duration of Study in the UK

    0 years, 1 months, 18 days

  • Research summary

    Diagnostic error is a significant problem in emergency medicine, where initial clinical assessment and decision making is often based on incomplete clinical information. Traditional computerised diagnostic systems have been of limited use in the acute setting, mainly due to the need for lengthy system consultation. Recent years has seen numerous attempts at creating viable solutions to this problem; Isabel and Babylon are two such examples. These apps possess the ability to generate differential diagnoses based on a patient's presenting symptoms and demographics. Despite these efforts, there is no system routinely used in the NHS setting as an assistant to healthcare professionals.

    DemDx, an iOS app, represents another modality of app which allows the user to generate differential diagnoses based on clinical features. It is not currently in NHS use and has been designed as an educational tool but has the potential to be considered for clinical use at a later stage if its utility can be demonstrated. No patient data is ever stored on the app.

    This is a prospective data collection project. An assessment of the ability of DemDx to generate accurate differential diagnosis in patients presenting to A&E in a UK healthcare setting (UHCW, Coventry, UK).

    Population: all adults (>18) attending to A&E majors department at UHCW after which exclusion criteria will be applied. The aim is to assess the accuracy of DemDx’s differential diagnosis algorithm in an A&E setting. This research project will assess the accuracy of a new differential diagnosis generator in the context of its potential utility in a healthcare setting.

    The study will last one month and the researcher will aim to see 5 patients a day, totalling to 100 patient sample size across the 4 weeks of data collection.

  • REC name

    West Midlands - Black Country Research Ethics Committee

  • REC reference

    17/WM/0397

  • Date of REC Opinion

    27 Jul 2018

  • REC opinion

    Further Information Favourable Opinion