Assessment of an artificial intelligence method for determining the result of a COVID-19 lateral flow test

ISRCTN	ISRCTN30075312
DOI	https://doi.org/10.1186/ISRCTN30075312

Submission date: 14/05/2021
Registration date: 16/06/2021
Last edited: 20/07/2023

Recruitment status: No longer recruiting
Overall study status: Completed
Condition category: Infections and Infestations

Prospectively registered

Protocol

Statistical analysis plan

Results

Individual participant data

Plain English Summary

Background and study aims
Rapid lateral flow tests help to find cases of COVID-19 in people who may have no symptoms but are still infectious and can give the virus to others.
The test usually involves taking a sample from your tonsils (or where they would have been) and from your nose, using a swab. You can get a result in 30 minutes.
Lateral flow devices (LFD) can be used to detect COVID-19 infection. Low viral loads as well as human factors can make interpretation of these devices inconsistent between users. We aimed to develop an artificial intelligence reader to interpret lateral flow devices for COVID-19

Who can participate?
Any health care workers invited by NHS Test and Trace can participate

What does the study involve?
Taking a photo of a lateral flow device after testing for COVID-19 as part of the self report process from NHS Test and Trace

What are the possible benefits and risks of participating?
Benefits: To assist improvements in the accuracy of LFD interpretation
Risks: None known

Where is the study run from?
University of Birmingham

When is the study starting and how long is it expected to run for?
November 2020 to March 2021

Who is funding the study?
NHS Test and Trace (Department of Health and Social Care) (UK)

Who is the main contact
Professor Andrew Beggs, a.beggs@bham.ac.uk

Contact information

Prof Andrew Beggs
Scientific

University of Birmingham
Vincent Drive
Birmingham
B15 2TT
United Kingdom

ORCID ID	0000-0003-0784-2967
Phone	+44 (0)121 414 7458
Email	a.beggs@bham.ac.uk

Study information

Study design	Observational diagnostic accuracy study
Primary study design	Observational
Secondary study design	Diagnostic accuracy study
Study setting(s)	Other
Study type	Diagnostic
Participant information sheet	No participant information sheet available
Scientific title	Machine learning for determining lateral flow device results for COVID-19 in an asymptomatic population: A diagnostic accuracy study
Study hypothesis	That use of an AI reader of lateral flow devices is at least as accurate as an expert read from an asymptomatic test site human reader
Ethics approval(s)	Public Health England Ethics Committee (exempt as observational with anonymised data only)
Condition	Detection of SARS-CoV-2 infection using a lateral flow device
Intervention	1. Participants decide to take a lateral flow test 2. They take the COVID-19 lateral flow test as per the manufacturer’s instructions for use 3. They then log onto the NHS Test and Trace website to return their test result 4. They are then invited to take a photo of the lateral flow test they have used, and record their personal details (name, address, date of birth, NHS number) and whether they are reporting the test as negative or positive 5. The result is not returned to the user but is stored for comparison purposes 6. The end user then finishes the process The whole process takes 35 minutes from end-to-end and there is no follow-up after this.
Intervention type	Device
Pharmaceutical study type(s)	Not Applicable
Phase	Not Applicable
Drug / device / biological / vaccine name(s)	AI reader, lateral flow devices
Primary outcome measure	Accuracy of AI reader compared to human readers measured using the test result (i.e COVID lateral flow device) as reported by the user of the test and as determined by the machine learning algorithm at a single time point
Secondary outcome measures	There are no secondary outcome measures
Overall study start date	01/11/2020
Overall study end date	31/03/2021

Eligibility

Participant type(s)	Health professional
Age group	Mixed
Sex	Both
Target number of participants	100,000
Total final enrolment	115436
Participant inclusion criteria	1. Any health care worker invited by NHS Test and trace who is willing to participate in the study 2. Aged 18 years or above or adolescents aged 12 - 17 years (self-test and report with adult supervision) or children under 12 years (should be tested and reported by an adult) 3. Without any common COVID-19 symptoms 4. Able (in the Investigators’ opinion) and willing to comply with all study requirements
Participant exclusion criteria	1. Did not agree with privacy statement 2. Any common COVID-19 symptoms 3. Any other significant disease or disorder which, in the opinion of the Investigator, may either put the participants at risk because of participation in the study, or may influence the result of the study, or the participant’s ability to participate in the study
Recruitment start date	12/03/2021
Recruitment end date	31/03/2021

Locations

Countries of recruitment

England
United Kingdom

Study participating centres

University of Birmingham

Vincent Drive
Birmingham
B15 2TT
United Kingdom

University of Durham

Stockton Road
Durham
DH1 3LE
United Kingdom

Sensyne Health

Schrödinger Building, Heatley Road, Oxford Science Park
Oxford
OX4 4GE
United Kingdom

NHS Test and Trace

c/o Department of Health and Social Care, Victoria Street
London
SW1H 0EU
United Kingdom

Sponsor information

Department of Health and Social Care
Government

NHS Test and Trace
Victoria Street
London
SW1H 0EU
United Kingdom

Phone	+44 (0)207 210 4850
Email	robert.bananathy@nhs.net
Website	https://www.gov.uk/government/organisations/department-of-health
"ROR"	https://ror.org/03sbpja79

Funders

Funder type

Government

Department of Health and Social Care
Government organisation / National government

Alternative name(s): Department of Health & Social Care, DH
Location: United Kingdom

Results and Publications

Intention to publish date	01/07/2021
Individual participant data (IPD) Intention to share	No
IPD sharing plan summary	Data sharing statement to be made available at a later date
Publication and dissemination plan	Preprint then publication in a high impact journal.
IPD sharing plan	The current data sharing plans for this study are unknown and will be available at a later date.

Study outputs

Output type	Details	Date created	Date added	Peer reviewed?	Patient-facing?
Protocol file	version v1.0	04/03/2021	08/07/2021	No	No
Preprint results		21/06/2021	12/04/2022	No	No
Results article		18/10/2022	20/07/2023	Yes	No

Additional files

ISRCTN30075312_PROTOCOL_v1.0_04Mar2021.pdf: Uploaded 08/07/2021

Editorial Notes

19/07/2023: Publication reference added.
12/04/2022: Preprint reference added.
08/07/2021: Uploaded protocol Version 1.0, 04 March 2021 (not peer reviewed).
01/06/2021: Trial's existence confirmed by Department of Health and Social Care.