The Graveyard of Interoperability Initiatives in the Past & How We Can Drive the Future
In the last several decades, there have been at least 100 initiatives and companies that have attempted to solve interoperability in healthcare. Yet the fax machine still rules the industry — what went wrong, and how can the industry solve this?
In the United States alone, millions of healthcare visits occur every day, and each visit generates data. This leads to an overwhelming amount of healthcare data — most of which is never used again.
Despite healthcare data having the potential to improve the effectiveness of care, identify public health trends, and spur the innovation of new therapies, most of the United States’ healthcare data goes unused due to its fragmented nature. Across the healthcare system, one individual’s health data sits in approximately 100 different systems, all siloed and kept separate from one another. This fragmentation of healthcare data prevents us from answering the most pressing questions in healthcare today.
This is not a new issue, and many companies, non-profits, and agencies have worked to solve it. Although progress has been made, we are still a long way from having our healthcare systems be connected and fully interoperable.
The Fax Machine, Mail, and the State of Health Information Exchange Today
Despite great technological advancements, most data in healthcare is still generally exchanged by paper or fax machines. As much as 75% of all medical communications are done through fax machines, with patient and provider requests for records often arriving on paper and being sent back through fax. In other cases, records are taken from an electronic health record, printed, and then faxed, adding more paper into the equation. In the last year, within Datavant’s Ciox division, we fulfilled 15 million requests for health records that were requested via mail and fax; over 100 million pages were sent over fax. Unfortunately, the fax machine has remained supreme because attempts at interoperability have not reached the universal adoption that faxes provide.
Although government regulations have mandated that healthcare providers go digital, initial mandates did not ensure that healthcare systems were connected with one another. As such, we have a healthcare system that is extremely digital, but not connected.
Imagine a world where the top wireless cellular networks (think Verizon, Sprint, AT&T) were all independent networks and you couldn’t make calls from one to the other. In addition to being extremely inconvenient, essential communications would not happen only because networks were not connected. This is roughly where we are with healthcare today. We have many small networks and many players in the ecosystem, but not many of them are connected.
Approaches to Health Data Exchange
So what approaches have we taken towards the exchange of health data to date?
Electronic Health Records (EHRs) Platforms: EHRs are the single biggest technology that has changed the landscape of healthcare data. EHRs allow healthcare data to be stored and accessed digitally, providing accurate, up-to-date, and complete information about patients at any one point of care. Currently, EHRs are missing the connectivity piece. Any individual health system uses an average of 16 different EHRs, many of which do not have the ability to exchange data between them. Attempts have been made to connect EHRs together, but often, only EHRs from the same company can be connected, or they must join a specific network to be interoperable.
Secure Direct Messaging: Direct messaging is a set of technologies that allow patient data sharing between providers. Established by the Office of the National Coordinator for Health Information Technology (ONC), direct messaging connects two EHR systems together to pass limited amounts of encrypted patient information between the two. Direct messaging has been traditionally difficult to integrate, and most of the industry has moved to more modern approaches like HL7 messaging and APIs.
Health Information Exchanges: A Health Information Exchange (HIE) allows for patients and providers to access their medical information electronically. Exchange of medical data within an HIE can happen provider-to-provider through directed exchanges, through queries of the network for all relevant clinical data, and directly through requests by patients. Currently, there are many models and vendors for health information exchanges, but most are established regionally. Nationally, the National Health Information Network is used as an HIE.
Patient Record Aggregation: More and more companies are becoming available that aggregate patient records for patients. Many companies have built applications that automatically aggregated records across providers. These applications may be developed by EHR vendors, or by third-party vendors who connect to EHRs and HIEs to aggregate data from various providers. To be successful, these applications need to build expansive networks of providers to ensure that coverage is high enough for proper aggregation.
Interoperability Solutions Companies: Over the years since HITECH and the proliferation of EMRs, a number of companies have cropped up to help connect providers, payers, patients, and other requesters and holders of data. Most of those organizations have struggled to reach critical mass, frequently adding to the fragmentation instead of solving it. Many of these companies have utilized standards-based APIs which allow systems to talk to one another securely and under defined parameters. The proliferation of the FHIR standard has catapulted the industry forward allowing scalability amongst applications connecting to EMRs.
Government Mandates: Throughout the history of the healthcare industry, the government has been an indispensable resource in pushing new standards for healthcare. Due to the American Recovery and Reinvestment Act mandating the use of EHRs, approximately 85% office-based practices have EHRs, and 95% of hospitals. Most recently, the 21st Century Cures Act interoperability rule requires providers to share health data electronically. This rule ultimately puts the onus on technology vendors to ensure that their systems used by providers allow for the easy sharing of health data.
Challenges to Health Data Exchange
Due to the complexity of the issue, there have been hundreds of attempts to make health data exchange easy– businesses, non-profits, regulations. We have tried to standardize, but now there are over 40 different standards. The following five bullets outline why complexity still remains today.
Technical and Legal Complexity: Although the adoption of EHRs is high, the average hospital utilizes approximately 16 different EHR systems. The duplicative nature of these systems means that even within a hospital, systems cannot exchange data easily. There will not be a one-size-fits-all API for the industry given the complexity of how data is stored, governed, and used. Instead, a viable solution needs to integrate separately into thousands of institutions and EHRs. Additionally, not all state and local governments have the same requirements for patient consent in the release of information. The industry needs a solution that is flexible and can be configured to meet the demands of these complexities.
Economic Incentives: Many regulations and initiatives tried previously have put the cost and burden of data exchange fully on providers, despite the many other beneficiaries of this exchange in the healthcare ecosystem. Data flow maps directly to the flow of payments in healthcare. The biggest opportunity to drive change in how data flows is to make the data accessible and ensure that incentives across the health system are aligned to discourage the hoarding of data.
Risk Tolerance: While some health systems may lean into more progressive data sharing, others see risk in oversharing information. This creates a patchwork of granular requirements and limitations on how scalable any single solution can be. For example, one health system might release up to one year of records when requested by a patient, but another might release three.
Lack of a Ubiquitous, Neutral Network: In order for data exchange in healthcare to be successful, there must be a critical mass using a solution to become a foundational infrastructure. We need an interoperability solution that works across disparate systems, rather than many different solutions built on top of many different solutions. EHR vendors may favor exchanging with other hospitals that have the same software, but a patient’s care journey often spans different vendor solutions. Being neutral and working with all systems is key to getting the full picture of a patient’s care. As such, vendors solving the problem of data exchange must be neutral to work with all partners across the healthcare ecosystem.
Data Access and Control: Increasing data exchange by nature means that healthcare data is going to be moving more fluidly. We need to ensure that data is protected in transit and at rest, that we control who has access to what data, and that the right data is going to the right, authorized party. This requires a solution that leaves control of the movement of the data in the provider’s and patient’s hands. This includes determining when patient consent is needed to move information and obtaining that consent.
The Road Ahead
One can think of health information exchange as the highways we drive on each day. In the 1950s, cars were just starting to become an integral part of US society. Although we had roads that connected cities and towns, they were often small two-lane roads, and not all cities and towns were connected.
Over time, we built out an interstate highway system that connected cities and towns and allowed individuals to quickly move between places. Now, it is much easier to get from place to place quickly. In the healthcare ecosystem, we are at a similar crossroads. We have electronic health record systems (EHRs) and other tools that are beginning to connect the dots. However, these connections are not everywhere. Like the highways before, we need to build out every mile of our connection system to ensure we can move medical records and health information to wherever they are needed.
The future relies on our ability to build meaningful networks, obtain the proper patient consent, and build analytics on top of data.
Build a network (or interconnect the existing networks): For health data exchange to work, the ecosystem needs a “network of networks” that can connect all different systems present.
Obtaining and manage proper patient consent, data security, and privacy: Any solution to data exchange must consider the requirements to obtain proper patient consent and authorization before releasing patient information. Proper solutions must have fine-grained settings of permissions for access privileges to any set of data. The patient and provider must have control over their data.
Manage incentives: There are large incentives for providers to optimize data exchange with their technology solutions, improve care, improve patient experience, coordinate with public health authorities, and even improve connectivity with payers. A network solution can start from these unified incentives as a carrot for interoperability, and then interoperability mandates can be used as a stick in limited circumstances.
Unlock an ecosystem: True interoperability is achieved when we not only can access the data necessary, but we can use the information. As data continually moves around the healthcare ecosystem, solutions will need to be built to ensure that providers, caregivers, and payers can derive actionable insights from the data. This will require analytics platforms to be built on top of APIs, where the APIs retrieve the data and the analytics platforms ensure that data can be understood and used.
What Free Exchange of Healthcare Data will Unlock
Easier exchange of healthcare data will unlock a number of use cases that will ultimately improve the healthcare system in the United States and improve patient outcomes. The following section highlights three of these use cases.
Care Coordination and Continuation of Care
When health data exchange becomes instantaneous and easy, care across multiple providers can be coordinated for the benefit of the patient. This information can be used to continue care between two providers, or coordinate care across different specialists. Additionally, having data on hand can reduce medical errors due to a lack of information between doctors.
Value-Based Care Provided by Payers
The medical records of members in any one insurance plan can be used to understand what treatments and what types of care are actually effective. Using care that is effective and efficient for the patient ultimately drives down costs for payers and providers, all while ensuring that the patient gets the best care possible.
By using data obtained through interoperable systems to inform value-based care, my care as a patient will be informed by all those who have come before me, and my care and outcomes will inform the next patient’s treatment. This consistent refinement of effective and value-based care relies on data. Providers must have all the puzzle pieces to accurately diagnose and select the best treatment options.
Giving Patients Full Access to their Records
By law, patients are allowed to have access to their medical records. However, obtaining these records is often a long and arduous process, and is often frustrating for the patient. By building interoperable systems, we are streamlining patient access to their full medical records.
Conclusion
Success for interoperability will require breaking through walled gardens and vanquishing the mountains of fax machines and paper. Once we build the digital connections necessary to provide the information exchange ecosystem this will lead to better health and health care. From there, we can stop talking about interoperability and instead talk about what we can do with patient information to improve health and healthcare.
Special thanks to David Shulkin, Abel Kho, Doug Fridsma, Alex Muir, Shahir Kassam-Adams, & Shannon West for their review and contributions to this post.
