Updates to Federal Regulations for Health IT Interoperability and What it Means for AI

May 4, 2020 | Interoperability, Policy

Introduction

The vast majority of physician practices and hospitals have moved on from paper medical records to electronic health records (EHRs). This is a fundamental step towards fulfilling the promise of interoperable health information technology (IT). Interoperability refers to the different IT systems and software applications being able to access and exchange data accurately, seamlessly, and in a timely manner, and to use the shared information to optimize the health of individuals and populations. For example, with full interoperability, lab results could be integrated automatically from one facility into the EHR system of another.

Health IT systems and software available in the US health care industry are largely developed in silos. Integrations work to an extent; however, the intent of interoperability is to have a holistic view of patients despite the variance in technologies. As part of the effort to achieve the full promise of health IT, the Office of the National Coordinator for Health Information Technology (ONC) and Centers for Medicare & Medicaid Services (CMS) finalized separate but related federal regulations addressing health IT interoperability. These regulations advance provisions from the 21st Century Cures Act (Cures Act) passed by Congress in December 2016 and help clarify standards to better ensure that different health IT systems can effectively communicate with each other. The aim is to impact patients, health care providers and Health IT developers by:

  • Providing patients with convenient access to their cost and clinical data through computers, cell phones, and mobile applications
  • Supporting providers ability to seamlessly share relevant clinical data on patients in order to ensure the best medical decisions possible are being made
  • Advancing innovation in the health IT industry to maximize the potential of modern technology

The ONC regulation—”21st Century Cures Act: Interoperability, Information Blocking, and the ONC Health IT Certification Program”—includes significant changes to the health IT certification program that will require developers to update their technology. Rules are also set for how the health care industry can prevent information blocking among health care providers, health IT developers, and health information networks. Highlights of key provisions are provided below:

  • Credentialing for Health IT developers – Updates certification requirements for health IT. The changes include a small number of new certification criteria, as well as revisions and removal of several existing criteria. The new requirements promote the ability to establish secure, standards-based interoperability between certified health IT systems and make it easier for patients to access their own electronic health information on their smartphones. Criteria that were removed were done so to limit burden and increase flexibility to both providers and developers to increase innovation. It is expected that many of the removed criteria for credentialing (e.g., ability to access medication lists) will still be captured by health IT systems as they are critical for clinical care, even though they are no longer required.
  • Exporting Electronic Health Information (EHI)  Supports the ability of patients to export their EHI across certified health IT platforms, as well as health care providers to export EHIs of an entire patient population to another health IT system (such as when a provider chooses to transition to another health IT system).
  • API Conditions and Maintenance of Certification. Requirements are set for patients to securely and easily obtain their EHI through application programming interfaces (APIs) at no additional cost when electronically accessed (e.g., by using a smartphone application).
  • Deterring Information Blocking  Mandates secure data sharing between payers, providers, and vendors for all “reasonable and necessary” healthcare activities. Organizations who continue to block information—i.e., interfere with access, exchange, or use of EHI—may face fines of $1 million or more. Examples of information blocking include imposing fees that make exchanging EHI cost prohibitive or contractual arrangements that limit how information can be shared. The regulation also details eight common sense information blocking exceptions that identify reasonable and necessary activities that do not constitute information blocking. For example, it may not be considered information blocking if limiting access to data would prevent harm to a patient or reduce the risk of security breaches to EHI.
  • Core Data Standards – Adopts the S. Core Data for Interoperability (USCDI) standard for the type of data (including “clinical notes” among other data important for clinical care) and constituent data elements that would be required to be exchanged in support of interoperability. The USCDI will help improve the flow of EHI and ensure that the information can be effectively understood when received. The USCDI standards will be updated and expanded over time.

The CMS regulation—known as “Interoperability and Patient Access”—advances efforts to improve access to the clinical, encounter, claims, and other types of data that can be shared among patients, health insurance plans, and federal agencies by choosing standards for data formats and elements as well as an API. These standards are known as the Fast Healthcare Interoperability Resources (FHIR). The regulation also describes how CMS will discourage information blocking, capture more electronic addresses for providers, and require hospitals to send admission, discharge, and transfer notifications electronically.

The regulations were finalized on March 9, 2020 and will begin to go into effect May 2020. Compliance for certain provisions, including information blocking, will start to be required in six months, along with compliance reporting by EHR vendors every six months. EHR vendors will have up to 24 months to provide customers with API technology supporting the new standards. EHI export requirements must be met within 36 months. Note that CMS has extended the implementation timeline for certain provisions in recognition that hospitals are on the front lines of the COVID-19 public health emergency. This includes the requirement for hospitals around electronic transmission of admission, discharge, and transfer notifications.

Implications for AI: Innovation & Interoperability are Linked

There has been a lot of progress in creating a connected health care system in which data is exchanged freely and made easily accessible to patients, insurers, providers, and Health IT developers. However, the task of having a truly interoperable health care system still remains incomplete. For example, almost all hospitals are currently able to send out clinical records to community-based physicians; however, only 70 percent can easily consume and assimilate incoming information into their EHRs. Moreover, even though most hospitals and physicians have EHI systems, the extent of their data sharing is generally constrained to providers within their network. These new federal interoperability regulations have the potential to break down many of the remaining barriers to the kind of data sharing that can revolutionize clinical decision-making and administrative processes on a nationwide scale. As health care data continues to standardize and capacity expands to seamlessly move data across health care networks, the use of innovations such as artificial intelligence (AI) techniques and algorithms will become increasingly useful in terms of filtering and presenting data that will help with patient profiling and clinical decision-making. In particular, the ability to export EHI across health information networks along with new and enhanced APIs will make it easier for health IT developers to access data from organizations across the health care industry. Moreover, deterring information blocking will pave the way for AI algorithms to access more large datasets for learning and training. Standardizing the type of data being shared, including how data is organized, will simplify the process of data cleaning, making it possible to implement AI programs more quickly and on a larger scale. Such interoperability will make it possible for risk adjustment validation to routinely analyze all medical records for patients across time and all of the different provider types (including for ambulatory, inpatient and post-acute care services, as well as information for drugs and medical equipment). Already, there is much that can be done with AI. As interoperability becomes the standard across all health care information systems, the efficiencies to be gained from AI will only grow. Interoperability improvements will facilitate greater adoption of AI solutions, such as the RaLytics platform and its medical record ingestion engine. With access to more data (and cleaner data), the accuracy and value of AI applications will only improve. The next several months and years of interoperability enhancements will be pivotal in paving the way for AI advancement and adoption.