Are Open APIs The Future Of Meaningful Use?
By Tom Giannulli, MS, MD, CMIO, Kareo
CMS has finally conceded to the weaknesses of their EHR Incentive Program, or Meaningful Use (MU). In recent announcements, Acting Administrator Andy Slavitt admitted the program has lacked the focus on interoperability and usability it should have had and that it is time for a change.
The new plan is to find a solution that ties together the incentive programs like MU and PQRS with the move towards value-based reimbursement and a deeper focus on data sharing. Specifically, Slavitt and National Coordinator Karen DeSalvo said the coming changes will be guided principals including:
- rewarding providers for outcomes
- allowing more flexibility to customize HIT
- unlocking EHR data through open APIs
- prioritizing interoperability
Releasing the power of clinical data and breaking down silos through open APIs and allowing free market dynamics to advance state-of-the-art EHR capabilities and the exchange of structured data are good concepts. And they have been applied successfully to other industries. Of particular note is the customer relationship management (CRM) industry and SalesForce.com’s stance on an open API, data exchange, and a third-party marketplace of value-added partners. Their approach has been hugely successful, making them the market leader in their space.
The EHR market is a little different and there are exceptions that apply to this model. First, the total addressable market (TAM) for EHR add-ons is relatively small compared to the CRM Marketplace, and so are the distribution channels to EHR customers. Second, many decisions around deploying software that touches protected health information (PHI) is heavily controlled so bolting on a widget is generally not acceptable — or easy — unless the vendor is vetted and HIPAA-compliant. As a result, data governance and access to the clinical data is not as unfettered as in the CRM example.
These factors can elevate the risk and narrow the reward for vendors investing in this approach. So the market still favors the incumbents and their protected relationships with customers. Despite these barriers, if the API approach is successful, we are back where we started — a point-to-point data exchange solution with serious limitations for aggregated data studies and wide scale frictionless exchange.
Rather than selectively open API ports for limited data exchange, one suggestion is to require every vendor and provider to export encounter summary data at the end of each business day into a narrow network of medical bank repositories as part of a federal law requiring medical data exchange. Once the data is liberated it can be queried by authorized stakeholders, de-identified, aggregated, and used for Big Data analysis related to pharmaceuticals, interventions and outcomes, and to elucidate the genetic basis for disease and effective treatments.
Several benefits arise from this approach:
- Exchange of data will be uncoupled from the silos of origin, unaffected by politics and security risks surrounding opening thousands of unmanaged API ports.
- There is a short time to value, since the CCDA standard clinical data export exists in all modern EHR systems, it is more about will than skill to get this data released.
- Proficiency with Big Data and large scale data security is a manageable issue, vendors like Bank of America, Google, and others mange similar data sets with success. While this is a major concern, it must be addressed in any scenario.
I think it is important to consider such an approach as we re-think our past efforts and root causes for their failure. As we plan for then next 10 years in healthcare IT, we need to be bold and enable data to drive discoveries and improvements to reduce healthcare costs and shape the future of healthcare delivery.