Testing the accuracy of an innovation in diverse settings: A multi-country study assessing a smartphone blood pressure application

Spread the Message

Read Time:2 Minute, 14 Second

21 April 2023 Departmental news

In a pioneering study, researchers have tested the accuracy of an algorithm-based smartphone application to ensure it works across diverse settings – urban and rural, as well as pregnant and general population.

The app, OptiBP, estimates blood pressure by reading optical pulse waves through the camera of a smartphone. OptiBP was identified by the WHO compendium of innovative health technologies for low-resource settings as an innovation with potential, and recognized by Grand Challenges Canada as a technology to incorporate into the health system during routine antenatal care.

Hypertensive disorders of pregnancy are a leading cause of maternal morbidity and mortality. As such, monitoring blood pressure is recommended as an essential part of all antenatal care in the WHO recommendations on antenatal care for a positive pregnancy experience. Importantly, the WHO Recommendations on digital health interventions for health system strengthening highlights the need to rigorously evaluate and generate evidence to promote the appropriate integration and use of technologies.

“The ability to estimate blood pressure through a phone app has the potential to help address barriers to the diagnosis and treatment of blood pressure during pregnancy,” said Ӧzge Tunçalp, one of the lead authors of the study and Medical Officer at the UN Special Programme on Human Reproduction (HRP) and WHO. “Innovations like OptiBP need to ensure that they are designed, tested and used in an equitable way that maximizes their impact and brings us closer to delivering health for all.”

While the study found the app to be accurate in most readings, this research demonstrates the importance of testing such apps in a wide range of settings, among different populations and health systems. It is now known that machine learning and algorithm development that lack diversity in the training data can lead to inconsistencies, inaccuracies and biases. The inclusion of the end user from the beginning of development is the best approach.

“Innovations like OptiBP can help save lives only if they are accurate,” said Tigest Tamrat, Scientist, HRP and WHO and another lead author of the study. “We need to be ahead of the curve in evaluating these tools for effectiveness and accuracy in as many contexts as possible.”

This study was done in collaboration across the HRP, along with Ifakara Health Institute, South African Medical Research Center, JiVitA Bangladesh, and Digital Health and Innovations at WHO’s Science Division, and conducted in both Asia and Africa.

“This is an example of an emerging approach to research that WHO is taking in order to ensure technologies are up to the life-saving tasks they promise,” said Garrett Mehl, Unit Head, Digital Health Innovations, WHO and co-author of the study.