Are Universal EHRs Key to Healthcare Value, Trust, and AI Adoption?

Healthcare providers in the United States generally don’t harbor very warm feelings for their electronic health records (EHRs).

Despite efforts from vendors and regulators to improve the experience of interacting with these foundational health IT systems, dissatisfaction and frustration with usability issues and fragmented information are still causing users’ blood to boil.

It may seem counterintuitive, therefore, to suggest that expanding the industry’s reliance on EHRs is actually the key to making EHR use less stressful and more useful – but that is exactly what the 2018 Future Health Index, commissioned by Philips, seems to indicate.

“Universal EHR” structures, in which every citizen’s electronic health record is connected to a single national system, are tied to higher trust in the healthcare system and higher value for patients, the international assessment showed.

Unified electronic records have also prompted national governments to address issues of privacy and security, data integrity, and health information exchange that are left up to the private sector in other countries – with varying degrees of success.

Half of the 16 countries included in the index have universal EHRs, while the other half have let a combination of free market forces and regulatory guidance dictate the course of health IT adoption.

The eight countries with universal EHRs saw an average overall “Value Measure” of 47.29, the report says.  The Value Measure is a combined score based on access to care, patient satisfaction, and an efficiency ratio score.

In contrast, the countries without universal EHR coverage, including Brazil, India, the Netherlands, Saudi Arabia, South Africa, Sweden, the UK and the US, saw an average value measure of just 39.67.

Healthcare professionals and patients in universal EHR countries - Australia, China, France, Germany, Italy, Russia, Singapore and Spain – also expressed a significantly higher degree of trust in their health systems as a whole. 

“Healthcare professionals and the general population in countries with ‘universal’ EHRs are also more likely to see health system integration as a goal worth pursuing,” the report states.

“In Spain, for example, 96 percent of healthcare professionals and 85 percent of the general population believe integration is important, versus 75 percent of healthcare professionals and 70 percent of the general population in the US.”

The interest in integration could be a little bit of a chicken-and-egg question: Spanish providers and patients may prioritize integration because they feel it is expected with a nationalized system, while US patients and professionals are more aware of the functional limitations of the current piecemeal approach to interoperability and data exchange.

The report points out that individuals who view their health system as more integrated are much more likely to trust that system than those who have a less positive view of data exchange capabilities, the report added.

Seventy-nine percent of participants who think their health environment is highly integrated expressed trust in the system, compared to just 47 percent of integration skeptics, indicating that previous experiences may have a significant impact on future optimism – or lack thereof.

The countries with more universal approaches to technology also rank higher on international assessments of health system quality. 

In 2018, The Lancet published an assessment of health system performance for 195 countries based on data gathered from 2016.

Five out of the eight universal EHR countries ranked in the top 20 performers internationally.  Singapore was ranked 22nd.

In contrast, only two non-universal EHR nations – Sweden and the Netherlands – cracked the top 20.  The United States ranked 29th.

While there are certainly other social, economic, and political reasons for these outcomes, healthcare data and the way it is used to support population health and care quality is doubtless a major contributor to value, outcomes and trust.

"The Future Health Index's latest report shows that the smooth exchange of health information is central to delivering better value in healthcare," said Jan Kimpen, Chief Medical Officer for Philips.

"By connecting people, data and systems we can create a network that allows information to flow seamlessly across care providers, locations and systems. It is through data that the outcomes that define value are tracked, measured and improved. Yet, health systems around the world are still struggling to collect, organize, analyze and use health data in a meaningful way.”

More cohesive EHR infrastructure may support advanced big data analytics capabilities – and the evolution of artificial intelligence – that can help providers extract more value from health data.

All of the included universal EHR countries, except for China, have established policies around data security and protections, and have also implemented regulations to guide data sharing.

“Among the eight countries without ‘universal’ EHRs, the extent of legislation is more mixed,” the report notes. “While the Netherlands, the UK and the US have comprehensive data protection and sharing policies, such legislation is lacking in Brazil and South Africa, which also have comparatively low Value Measures.”

“Creating clearer frameworks for the use, protection and exchange of data could not only provide a sounder platform for EHRs, but advance digitization in other areas.”

AI is likely to be a beneficiary of more comprehensive data sharing frameworks, and providers and patients will subsequently derive value from AI tools.

Access to enough high-quality data for training artificial intelligence models is an ongoing pain point for developers and researchers operating in fragmented health IT environments. 

Data sets are often locked away in proprietary siloes, and organizations are often reluctant to subject data assets to privacy and security risks by sharing them with peers.

Patients, too, are wary of how their data is used and shared for research and development purposes. 

More than half of FHI survey participants rank their health data as the most sensitive information about themselves, topping concerns about their emails and social media data being exposed during a data breach.

Overcoming these barriers may be crucial for combining robust EHR infrastructure with AI to produce value.

“The 2018 FHI identifies AI as a significant contributor to a healthcare system’s ability to provide data-driven care,” says the report. “Singapore, the country with the highest Value Measure, ranked third for spending on healthcare-related AI for therapy planning.”

Sweden, the United States, and the Netherlands – all countries without a universal approach to EHRs – spend the most per capita on artificial intelligence for diagnostics. 

And yet in the United States, at least, this technology is not exactly widespread.  While organizations are quick to predict that AI will take over the healthcare industry within the next decade, executives are not moving very rapidly to make that assertion a reality.

Close to two-thirds of providers ranked AI as a low priority for 2018 in a Center for Connected Medicine poll conducted at the close of 2018, and many questioned how they would achieve ROI if they did pursue machine learning tools.

In a separate recent KLAS survey, just 17 percent of US organizations had adopted AI for clinical decision support around diagnostic imaging.  Fifty-three percent of organizations had no defined strategy for doing so.

Once again, there are additional reasons why AI has not yet found a home in the majority of the US healthcare system.  But data siloes created by patchwork EHR and other health IT infrastructure is likely to be a culprit in the slow adoption of machine learning technology.

In order to get ahead of these challenges, the FHI report suggests that provider organizations focus on creating integrated workflows that do not enhance existing irritations with electronic health record systems. 

Collaborating with developers and experts outside of the healthcare industry, where user-centered design has been applied more successfully, could help to craft more intuitive and effective workflows.

Senior leadership should actively seek out opinions and education about how to create a data-driven culture that prioritizes technical integration, data governance, and feedback on how to best employ AI and other health IT strategies to produce impactful results.

“It is clear that experts and practitioners from various parts of the healthcare spectrum are virtually united in recognizing the potential of connecting people, data and systems to create a network that allows information to flow seamlessly across care providers, locations and systems,” the report concludes.

“They see how it can address pressing systemic constraints and deliver more effective care, and they are working to advance these technologies both within their own organizations and more broadly. Importantly, there is also a high level of agreement on the steps needed to drive change and make integrated care a more universal reality – and these steps are broadly applicable regardless of the technology or solution in question.”