Healthy.io: Smartphone albuminuria urine self-testing
Patients with diabetes and high blood pressure are benefiting from pioneering artificial intelligence (AI) that turns a smartphone camera into a clinical-grade tool to detect early kidney disease.
Healthy.io is a recipient of phase 4 funding via the AI in Health and Care Awards.
Core themes: Primary care, remote monitoring, chronic disease.
AI used: Healthy.io uses colourimetric analysis, computer vision, and AI to transform the smartphone camera into a clinical-grade medical device.
Overview
Urine testing is one of the most frequently used diagnostic tools in primary care. As well as identifying infections, it can help monitor for chronic kidney disease (CKD) by looking at the ratio of two proteins in the urine: albumin and creatinine. The albumin to creatinine ratio (ACR) helps identify possible kidney disease and guide patient management and care.
With the home test kit and mobile app, Healthy.io’s solution empowers patients to self-test at home with clinical grade results. Fully integrated to the Electronic Medical Record (EMR), real-time results are available for clinician review and follow-up. Shifting testing to the home increases uptake, improves quality, reduces workload in primary care, and creates savings.
How does the AI product work?
The test is for people with diabetes, high blood pressure or other risk factors for chronic kidney disease (CKD). The technology consists of an ACR test kit and a mobile application containing image recognition software which scans and analyses the urine reagent strip. The ACR test kit is sent to the patient in the post and contains all single-use equipment needed for in vitro ACR testing (testing outside the body). It includes:
- a urine collection cup
- an individually wrapped ACR urine reagent strip
- an absorbent pad to remove excess liquid from the reagent strip
- a colour-board to enable image recognition.
The patient also receives a text message from their healthcare provider inviting them to download the ACR app. The app can be installed on Android and iOS smartphones that meet the minimum hardware criteria. The app guides the patient through each step of the testing process using video, audio and text guidance.
To do the test, the user collects a urine sample using the urine collection cup. They then dip the reagent strip into the sample for 1 second, remove any excess liquid with the absorbent pad and place it on the colour-board. The user then waits 60 seconds before scanning the colour-board. The software algorithm analyses the scanned image. Once the test is complete, the app automatically and securely sends the analysed results to the referring clinician through the patient's electronic medical record (EMR).
Impacts and outcomes
Evidence from a single-arm observational study involving 2,196 people with diabetes who were previously non-compliant with standard ACR testing shows that the technology can help improve ACR test compliance and that most people prefer home testing.
According to an independent evaluation by the York Health Economics Consortium, if rolled out nationally Healthy.io’s solution has the potential to save more than 11,000 lives and save the NHS at least £660 million over 5 years.
The service has enabled us to identify and prioritise follow-up care for people who may have early-stage chronic kidney disease that could have otherwise gone undetected. It allows us to offer our patients a new way of engaging with their care that is more convenient for both patients and staff.
Dr David Lipscomb, diabetes clinical lead at Sussex Community Foundation NHS Trust.
Contact information
For more details about the AI in Health and Care Award or its projects please contact: ailab@nhsx.nhs.uk
For more information about Healthy.io, visit the Healthy.io website.
