Bridging digital health innovation with governance and patient safety

Editor's note: This is the second article in a four-part series on connected diagnostics. Read the first article here.

Digital innovation is transforming health care—it is improving service delivery and health outcomes for many people and communities around the world.

In the diagnostics space, connected diagnostics—diagnostic systems that are digitalized to adapt to and integrate with the broader health care ecosystem—are improving the accuracy, efficiency, and accessibility of medical diagnostic technologies and data. However, fast-moving new technologies like these are often developed and implemented before strong governance is in place to ensure safety and privacy policies can catch up.

Understanding the current state of digital health governance and connected diagnostics, including how they impact one another, will be key for meaningful change in public health.

There has been significant progress in digital health governance over the last five years as stakeholders have worked to rapidly develop and implement new policies around the world.

These policies and programs cover a range of governance areas, from efficacy of digital health tools to market access and approvals. All of these, however, attempt to ensure that digital health systems can appropriately respond to health system needs and priorities with transparency and accountability.

For instance, in 2019, Tanzania launched a National Digital Health Strategy, with support from PATH’s Data Use Partnership (DUP). To prepare for launch and implementation, DUP supported the government of Tanzania in establishing an inaugural governance unit and appointing a National Digital Health Steering Committee.

This national approach provides a model for digital transformation and data use that prioritizes governance, workforce development, and digitization of primary health care.

As countries continue advancing digital health transformation, it can be valuable to share experiences and lessons learned. PATH’s Data Use Acceleration and Learning (DUAL) project, in partnership with Cooper/Smith, collects, synthesizes, and shares findings from five countries, including Tanzania.

Global guidance can help inform better country-level digital health governance. For example, the Principles for Digital Development is a community-driven set of living guidance to help digital development practitioners successfully and appropriately apply digital technology in development programs.

The Global Health Data Governance Principles are a resource for policymakers, advocates, and other stakeholders who are working toward equitable, human rights–based approaches to health data governance.

These two examples supplement other important global governance efforts, such as the World Health Organization’s data principles, Digital Investment Principles, as well as policy recommendations mapped out in the Broadband Commission for Sustainable Development’s Future of Virtual Health and Care, the individual rights–based framework to digital transformation of global health in the Lancet and Financial Times Commission on governing health futures 2030, and more.

Despite concerted efforts to establish appropriate governance for digital health, the speed of innovation outpaces their development and implementation.

During the COVID-19 pandemic, an explosive growth in virtual health care solutions were made possible through connected diagnostics. The UK’s NHS Pathways telemedicine software, for example, logged a million additional virtual doctor appointments each week in 2021, compared to 2020.

Connected diagnostics are strengthening disease surveillance as they help detect outbreaks and manage interventions. Many local and global programs are leveraging this emerging digital functionality and interoperability. For instance, the government of India’s Pradhan Mantri Ayushman Bharat Health Infrastructure Mission creates and links a network of public health laboratories at the local, regional, and national levels.

Connected diagnostics offer vast potential for generating and sharing data. This is unlocking new opportunities for advanced data analytics. Artificial intelligence (AI)—in which computer systems demonstrate human-like intelligence such as perception, speech recognition, inference, and decision-making—is further expanding the capabilities of connected diagnostics.

Machine learning (ML) is a subset of AI that is able to improve its performance or “learn” using algorithms and statistical models. ML plays a large and growing role in health care today, with applications across predictive diagnoses, clinical care guidance, as well as various efforts in health system strengthening.

Digital readers that apply ML, for example, use computer vision models to read and classify diagnostic results. This capability supports health care workers who may have limited experience or a large workload.

AI and ML can also help avoid medical supply stockouts and manage critical inventory, or they can conduct epidemiological modeling to guide public health interventions.

Digital health developers such as Audere, C2Sense, Dimagi, SystemOne, and many others are developing software to advance connected capabilities. To help train the models and improve performance, PATH developed a diagnostic data repository that contains certified and high-quality data.

The repository includes high-resolution images of results from rapid diagnostic tests (with varying sample concentrations and environmental settings) and associated metadata. To support continued development, we make this repository available to digital health developers and partners committed to supporting resource-constrained markets.

The intersection of evolving governance and innovative technological development presents various challenges, particularly relating to existing health inequities.

The data provided to train and develop ML can introduce bias that can affect how well these technologies work for patients. And data used for AI/ML in connected diagnostics often don’t meet standards for quality (e.g., timeliness, accuracy, completeness) and availability.

Privacy and security are essential considerations with regard to health data, and clear communications and guidance for protecting them are still needed.

Multifaceted digital tools often also come with challenges of complex ownership and intellectual patent rights.

Finally, global governance policies are still under development, and they do not always incorporate input from all sectors, including the private industries that develop technologies. When they are applied, these new tools require careful monitoring and evaluation to ensure appropriate use and fit.

In order to address these challenges and realize the potential of connected diagnostics for improving global health, collaboration will be critical. It will also be especially important to ensure equity by supporting and elevating representation from patients, managers, and health workers from resource-constrained health systems. Both public- and private-sector partners must expand and adopt digital health governance appropriately.

Digital health developers must prioritize transparency and evidence of fit, performance, and sustainability for the intended uses of new connected diagnostics technologies, and public partners must remain aware of developers’ nuanced business needs. Finally, continuous learning and experience among health care workers, ministries of health, and regulatory bodies will ensure a cohesive, holistic approach to the full cycle of development, implementation, and evaluation of connected diagnostics.

Read the next article in the series here.

Learn more about PATH’s work in diagnostics.