What is the Standard Health Record Collaborative (SHRC)?
The SHR Collaborative is an open-source, health data interoperability effort based at The MITRE Corporation in Bedford, MA. The Collaborative's central focus is to establish standards for the structure and content of health record information.
The SHRC is prototyping the Standard Health Record (SHR) by leveraging existing medical record models, distilling from them a central, necessary set of structures and content. Initial SHR implementations will prioritize patient identification, primary care and emergency care providers. The SHR is patient and provider-centric, and addresses the dynamic data needs of providers, patients, and caregivers by including data specifications for many areas related to social determinants of health.
The SHRC focuses on the needs of the patient, including the need to manage their own data. The SHR Collaborative holds that patients must have access to their data in an understandable, portable, and transparent manner. By using an open source model, the SHR actively promotes industry innovation by leveraging big-data at the patient and population level to increase patient engagement, support public health and research needs, and ensure providers consistently have the right information, at the right time, for the right person.
What makes a health record a Standard Health Record (SHR)?
Primarily, it conforms to the US national standard for the content of electronic health records being developed through the Collaborative. The initial specification of the SHR can be found here. Health records conforming to the initial SHR specification contain all information critical to patient identification, emergency and primary care.
Future extensions of the SHR specification will be driven by multidisciplinary input and governance. Additionally, in alignment the patient-centric nature of the SHR, patients will have the ability to contribute information and manage access to part or all of their record.
Widespread SHR implementation will result in interoperability across care settings as individuals and providers gain digital access to their SHR. Key benefits of clinical interoperability include improved care coordination, reduction of medical errors, minimization of waste, fraud, and abuse, and decreased costs that accompany healthier lives, improving healthcare access, quality, and uniformity. SHR provides the foundation for collection, communication, and aggregation of patient data, accelerating secondary uses in public health, disease surveillance, post-approval monitoring, and patient-centered outcomes research.
Is the SHR an open source project on GitHub?
Yes. The development of the SHR and all resources are open source and have been since the very start of the effort; you can find the SHR on GitHub. If you are interested in participating in the development of the SHR please email shr-project-list@lists.mitre.org or enter an issue on the SHR GitHub page.
Is the SHR technical approach different than what has been done previously for Health Data Interoperability in the US?
Yes. To explain why the SHR approach is different requires some background.
Twenty years ago, the healthcare industry turned towards interoperability standards to improve healthcare information exchange. However, the very lack of data consistency was not acknowledged and has led to exchange standards that are purposefully left extremely flexible, allowing enough freedom to model the information in virtually any EHR system. Even new standards like HL7’s FAST Healthcare Interoperability Resources (FHIR®) have made a deliberate decision to remain flexible and ambiguous on data specificity for this same reason. While FHIR provides a stronger foundation then previous healthcare exchange standards, problems of inconsistent implementation and semantic interoperability remain.
Today there are over 1,500 different EMR/EHR health IT system products certified as part of HHS Meaningful Use. Most of those health IT products are themselves highly configurable, to facilitate flexibility in divergent clinical settings. This explains why, even after the nearly universal adoption of health IT technology has occurred, the primary mechanism of communication between clinicians remains the fax machine, which is reliant on manual human interpretation of text on paper.
In our view, the real problem lies much deeper than the flexibility of information exchange standards. In addition to capturing too much information as free text, the fundamental problem is that today’s health IT systems contain semantically incompatible information. Because of the great variety of the data models of EMR/EHR systems, transferring information from one health IT system to another frequently results in the distortion or loss of information, blocking of critical details, or introduction of erroneous data. This is unacceptable in healthcare.
The approach of the Standard Health Record (SHR) is to standardize the health record and health data itself, rather than focusing on exchange standards. When the health record and data is standardized, exchange and aggregation of patient information will become trivial (as in every other industry that has gone digital).
What do we think will make SHR successful when past efforts have failed?
MITRE and the SHR open source community recognize that achieving consistency in healthcare data is a difficult problem to solve. There will be resistance to change and building consensus will be a challenge. MITRE and the SHR open source community’s influence on health policy and health standards will be as important as the SHR technical innovation. This is why the SHR effort invests heavily in engagement with healthcare thought leaders and healthcare alliances. The SHR community is connecting with medical associations, standards organizations, and vendor consortia such as Commonwell Health Alliance, the Argonauts, the National Association for Trusted Exchange (NATE), the Carequality Collaborative, The Sequoia Project, and the Healthcare Services Platform Consortium.
The SHR Collaborative will also build on existing standards and borrow from similar initiatives outside of the US. For example, the ONC Common Clinical Data Set (CCDS), which mandates several types of information be exchanged during transitions of care, is a natural building block for the SHR. England has published standards for the clinical structure and content of patient records. Both England and US standards fall short of defining the structured data necessary for a computable SHR, so we are addressing that today. Other existing data models, such as the Federal Health Information Model (FHIM), the OMOP Common Data Model, the Observational Health Data Sciences and Informatics (OHDSI) “Common Data Model”, and FHIR resources provide important and valuable inputs. But none of these models, nor any similar healthcare model that we are aware of, fully defines the content of a useful, complete standard health record.
With FHIR establishing a solid foundation for basic exchange, the time is right to try and solve this critical problem. MITRE and the SHR open source community can provide a target and start working towards nationwide adoption of the SHR.
Is anyone else standardizing health data or is the SHR project alone?
This is not the first effort to standardize health data. Internationally, many of the top healthcare systems in the world have defined or are in the process of developing a standard health record. For example, a standard health record (the Summary Care Record) has been successfully adopted in England, and it has provided very positive results. Many developed countries have some form of a standard health record. The US is behind in this respect.
While international efforts will help to inform the Standard Health Record, there are many areas where the SHR project can leapfrog those older developments. For instance, many of the internationally defined standard health records focus on standardizing narrative data sections alone and do not address fine-grained structured data. We will use FHIR to access and exchange the contents of the SHR down to individual data elements (e.g. current patient core temperature in degrees F). There are other groups whose work can be leveraged to accelerate the SHR, including the Clinical Information Modeling Initiative (CIMI). The CIMI project is developing a detailed specification for exchanging laboratory health information that may also leverage the SHR.
Will SHR address only “traditional electronic health records” or will SHR be much broader in scope and include non-hospital health data?
The plan is to initially focus on electronic health record data for patient identification, emergency care, and primary care. After those initial domains have been developed, the Standard Health Record will be expanded to cover other domains such as precision care, outpatient care, genomics, social determinants of health, and patient generated data. Early SHR work will establish a repeatable process to “standardize health data” that will be extended to new use cases as an evolutionary strategy to standardize the majority of modern health data.
The SHR community will look to the national healthcare community at large to help identify the key health data domains to add next. However, we want to begin with a manageable yet impactful set of initial domains to prove the SHR concept and define the technical representation. We will continue to work through the consensus process, better understand and develop solutions to policy, privacy, patient consent, and security issues around the SHR.
How is SHR different from FHIR®?
FHIR is an exchange standard, not a content specification. FHIR defines high level classes of things, such as “person,” “immunization,” and “observation”, that might be exchanged between healthcare systems. It provides no information about what particular persons, immunizations, or observations might be important. In contrast, the SHR lists specific things that should be part of every health record, for example, specific persons (e.g., an emergency contact), specific immunizations (e.g., the date of the last tetanus shot), and specific observations (e.g., the patients fasting glucose).
The SHR Collaborative pinpoints those specific elements by leveraging existing medical record models and physician expertise to account for the data needs of primary and emergency care providers. The SHR also addresses the dynamic needs of patients and their caregivers over the course of life by including a recommended structure and content for documenting many variables related to social determinants of health.