ONC Interoperability Outcomes Survey

Dear Path Bytes:

This message is sent on behalf of the API Technical Standards Committee.

The Office of the National Coordinator for Health Information Technology (ONC) is establishing benchmarks for measuring interoperability in medicine. It is an admirable effort to set up outcomes that would be used as a measuring tool of success or failure. Your input could be critical to make sure Pathology related outcomes are well represented among the proposed benchmarks.

Below is a list of suggested outcomes that we collected from a prior discussion in this mailing list. If you feel passionately about a specific outcome, please follow the link to the ONC feedback page and submit your request.  It does not matter if you created it or someone else did, please take a few minutes, and add your voice to this discussion.

Please follow the link for the instructions on how to submit your proposed outcome measure.

https://www.healthit.gov/topic/interoperability/health-interoperability-outcomes-2030

If you feel even more inclined to dig deeper and spend a bit more time thinking about it, please do so!  The suggested outcomes below could be further generalized, made more concrete, or could be split into several related outcomes. 

This message also serves as an introduction of a new GitHub “organization” (https://docs.github.com/en/organizations/collaborating-with-groups-in-organizations/about-organizations) for the Association of Pathology Informatics. We created a repository for content related to the topic of interoperability. The starting point is the list of suggested responses to the ONC survey; these have been placed in that repository for ongoing discussion. It is an experiment! We hope that our GitHub organization will become a platform where a complex theoretical question related to interoperability and standards in Pathology can have verbal and programmatic representation. Software examples could illustrate, clarify, and emphasize the point. Another great way to have a discussion on GitHub is to use “issues” - please feel free to submit comments or suggestions as issues here: https://github.com/assoc-path-informatics/interop/issues

Please explore the GitHub site at the link below and subscribe for changes and announcements:

https://github.com/assoc-path-informatics/interop/

List of suggested outcomes:

Because of interoperability, before/by 2030 …

… healthcare providers will be able to order a laboratory test from any laboratory using the same test name.

… pathologists will guide therapeutic management choices same day, not a day or weekend later.

… there's a robust marketplace of medical software applications, breaking the Epic/Cerner oligopoly, so that clinicians and their institutions can choose from a wide range of highly usable apps that use patient-derived data to make it easier and more efficient to care for patients in high quality ways.

… laboratory results will be able to be shared in a manner that uniquely identifies the method, device, and laboratory that produced it, including whether the result is traceable to an international reference material.

… laboratory reporting for notifiable diseases will be entirely electronic at the national, state, and local level.

… all data elements in a pathologist's synoptic cancer report will be available as for downstream use by EHRs and cancer registries as discrete, machine-readable, and SNOMED-encoded elements.

… a patient will be able to send a DICOM file (or something else??) of their digital pathology images to any hospital in the United States and that hospital's pathologist will be able to review the images.

… telepathology for either primary diagnosis or for expert consultation in difficult cases will become routine.

…  a patient getting laboratory testing will be able to query their insurance provider and get a reasonable estimate of the cost to them of this testing.

… there will be large-scale monitoring of commercial assay results by test kit and reagent/calibrator lot such that real lot-lot variability is known by vendor. New lots with problems are recognized by result median and distribution shifts less than a week after deployment, and fixes are verified in a similar time frame.

… many more tests will produce results that are traceable to standardized reference materials and reference methods. The data model for tests will include traceability. Results from tests that are standardized in the same way can be displayed and trended together across provider organizations, and processed by the same decision support rules and machine learning models.

… there will be new patient-centric data standards (eg. standard universal patient ID).

… pagers are no longer used as a primary means of inter-hospital communication by 2030.

… health IT platforms will be designed primarily for communication and collaboration. Business transactions and compliance functionality will be secondary rather than central functions.

… patient data will become portable in a secure manner.

… will be established seamless data querying mechanisms for public health actions.

… will be established real time big data research tools on a national/international level while protecting individual privacy and security.

… there is a robust coding system we would commit to using and which fields in HL7 should be used and populated to make pathology/lab-medicine data as specific as possible and standardized as much as possible.

… the lab interoperability nuances are well articulated and understood on all levels.

… we have a clear definition of interoperability and clear understanding of the goals and priorities of lab interoperability

… all stakeholders of interoperability are identified with clearly stated interests in lab interoperability. (e.g. regulatory agencies (ONC, FDA, CMS), diagnostic device/kit vendors, laboratorians, clinicians, and public health and health services researchers). The superset of the partially overlapping goals of these individual constituencies is identified

… there is a comprehensive statement of the interoperability requirements.

… there are relevant examples of interoperability in pathology that illustrate successes and challenges.

… data model for lab tests is well defined by domain experts and includes all elements that are needed for clearly defined use cases or workflow scenarios.

… it will become clear that existing standards are not adequate (not by a long shot) for common laboratory tests and are unusable for more complicated tests such as pathology and genomic reports.

… there will be one interoperability standard because there is significant cross-over between pathology, genomics and clinical laboratory tests (e.g., cultures on tissue also examined for pathology).
… the standards will be accessed for their safety when used for interoperability. (Ease of data retrieval differs from interoperability safety).

The scenarios below illustrate that seemingly interoperable terminology, semantically similar, may lack the depth of meaning for a particular activity in the context of direct patient care as opposed to data aggregation and analysis where it would be perfectly adequate:

 
  1. Coding a laboratory test for data retrieval has far lower safety implications than putting a code on a test so that it can be automatically charted in a different EHR.
  2. Miscoding a test is annoying for data retrieval and frankly dangerous for interoperability and patient care.
  3. Using a more generic code for a test may cause it to lose meaning such that misinterpretation is likely when the result is charted under a more generic name.
  4. Requiring 30+ codes for a single genomic variant is a maintenance nightmare and at high risk for error and misinterpretation when it is separated from the interpretation.
 
API Technical Standards Committee
https://www.pathologyinformatics.org/api_committees.php