News

API aims to hit clinical software with disruptive innovation

Brian Eastwood

Clinical software is ripe for disruptive innovation. As Dr. Kenneth Mandl sees it, that innovation will arrive in the form of apps that reference a common application program interface (API) and as a result, can be added to or removed from any electronic health record (EHR) system.

    Requires Free Membership to View

Mandl, director at the Intelligent Health Laboratory of the Children's Hospital Informatics Program in Boston, is also co-director of the Substitutable Medical Apps, reusable technology (SMART) initiative, which is developing an open source API. (A Strategic Health IT Advanced Research Projects [SHARP] program grant from the Office of the National Coordinator for Health Information Technology [ONC] provided funding for SMART.)

Once an application has integrated the SMART API, it can interact with and use information from any other clinical software using that API. SMART applications are Web-based -- they support the HTML 5 standard -- and access the SMART API using JavaScript or Representational State Transfer, or REST, calls. These apps, with their access to the data within EHR systems, could do a variety of things -- manage patients' medication, for example, or integrate census data with patient records.

The philosophy is "a new conception of electronic medical records as platforms," Mandl said during a recent presentation before the Massachusetts Health Data Consortium.

Ideally, organizations would be able to adopt new health applications as easily as consumers buy new apps from Apple Inc.'s App Store. "Health Information technology should look more like the iPhone than it does," Mandl said.

Clinical software would be used if it worked properly

For Mandl, the popularity of the iPhone -- and increasingly, the iPad -- is evidence that, contrary to public perception, physicians are not Luddites. Far from it, he said, also citing the prevalence of surgeons using software and even robotics during operations.

Health Information technology should look more like the iPhone than it does.

Dr. Kenneth Mandl, director, Intelligent Health Laboratory, Children's Hospital (Boston) Informatics Program

By that reasoning, if a type of technology is not being used, there's probably a good reason for it. Take computerized physician order entry (CPOE) systems, which are prone to avoidable errors, Mandl said. For example, doctors writing out prescriptions by hand "would never accidentally prescribe the next drug in an alphabetical list."

The EHR Incentive Programs, by encouraging rapid adoption of rigid technology, further complicate matters, Mandl said. To demonstrate meaningful use, providers implementing an EHR system today will have to be using the same system in five years. In contrast, no one will be using the same laptop or smartphone in 2016, he noted.

The better option, then, is to view an EHR system as an extensible platform onto which an organization can add small, inexpensive applications that address specific needs: meeting a single meaningful use requirement, for example, or linking a patient wellness application to a larger clinical data warehouse.

Better clinical software, better data collection

The SMART initiative and the ONC are sponsoring a SMART Apps for Health contest, with the developer of the winning Web application receiving $5,000.

As currently constituted, SMART applications can interact with only three systems: Regenstrief Institute Inc.'s Careweb EHR system, Children's Hospital's Indivo personal health record (PHR) service or Harvard Medical School's i2b2 analytics engine. Several vendors, however -- including Microsoft, Surescripts LLC and Cerner Corp. -- have signed on already as SMART partners, and several others have expressed interest.

As additional vendors and organizations embrace the SMART platform, the ability to easily add -- and remove -- third-party applications that run within EHR systems could help bring about that disruptive innovation, Mandl said.

It could also bring much-needed reform to clinical trials. The health industry could be "collecting data on many, many patients, not [on] small numbers of patients and generalizing them" -- a practice that leads to bias and is all too common in today's clinical trials, Mandl noted. "Every patient should be an opportunity to learn."

Let us know what you think about the story; email Brian Eastwood, Site Editor.


Join the conversation Comment

Share
Comments

    Results

    Contribute to the conversation

    All fields are required. Comments will appear at the bottom of the article.