Wednesday, February 25, 2015

The Argonaut Project Kickoff

Jason returning with the Golden Fleece
I just got off a kickoff call for the Argonaut Project Implementation Program.

This is, simply stated (and to paraphrase John Halamka on the call today), the most promising path towards meaningful healthcare interoperability we've ever known. Finally, we have modern protocols such as REST, OAuth and OpenID that are being applied to healthcare in a scalable way.

The SMART on FHIR platform epitomizes the work done to date in this effort, and the Argonaut Project is intended as a sprint to get the FHIR DSTU 2 deliverables in time for a May ballot. To be clear, this is still an alpha/beta product and not quite ready for public consumption, but the promise is already quite evident.

When I arrived at Duke in the summer of 2013 (and before the SMART project had been converted to work with FHIR), I had a goal to create a standardized platform for interoperability for web and native mobile apps that would work with our Epic implementation.

Starting in 2014, while we already had a functional implementation of our own framework, we realized that we shouldn't recreate the wheel, and that the SMART on FHIR project was being developed the accomplish the same goals. With that in mind we realigned our efforts to push forward with integration of the SMART platform into our Epic EHR.

As of January 2015 we have a functional implementation of the SMART platform here at Duke in our proof-of-concept environment. We currently have an iOS app and a SMART-enabled web app working against this environment with plans for several more current and future app integrations, including apps developed here at Duke.

We're looking forward to be involved in this Pilot Implementation process (see graphic below) and look forward to sharing more in the coming months, including at HIMSS in April where we will demo these integrations.

Never before have we had such a golden opportunity for robust healthcare interoperability!

Anyone can get involved in this open process.

Thursday, February 12, 2015

FDA Updates & Finalizes Mobile App & Device Guidance


This week the FDA released the final version of their guidance on Medical Device Data Systems (MDDS - original draft was released on June 20, 2014) as well as an update to their guidance on Mobile Medical Applications (last updated on September 25, 2013) to make it consistent with the MDDS document.

I've shared my thoughts on this guidance previously.

For those who aren't familiar with this guidance, the FDA divides mobile medical applications into three categories, only the first of which will be regulated:

Apps that are medical devices

These apps will be regulated, and fall into one of two categories, including those that are either intended:
  1. to be used as an accessory to a regulated medical device; or
  2. to transform a mobile platform into a regulated medical device.
The updated Mobile Medical Applications document provides a number of helpful examples of hypothetical apps as well as a list of approved applications that have gone through the 510(k) process.  Most of these fall into one of the following categories:
  • Apps that use an integrated or attached sensor to obtain data used in medical decision making
  • Apps that use the camera to analyze images and present data used in medical decision making
  • Apps that connect to and alter the behavior of other medical devices (physical connection or remote)
  • Apps used to calibrate medical devices
Apps for which FDA intends to exercise enforcement discretion

These are apps that MAY meet the definition of medical device but likely pose lower risk to the public, therefore they will not be regulated unless decides the app poses a significant patient safety risk.

This is a big deal because all clinical decision support apps will fall into this category. Clinical decision support apps represent one of the most promising categories of medical software, including those that can be integrated into a provider's workflow through integration into the EHR via technologies such as the SMART of FHIR framework.

The FDA realizes that the volume of these apps will likely exceed their capacity to regulate them, but the bottom line is that purchasers and consumers of these applications need to be ever vigilant that what they are using is safe, and immediately report any concerns to the FDA.

Apps that are not medical devices

These are apps the FDA has decided it will not regulate. Types of apps include:
  • Apps that provide medical reference material
  • Apps used as educational tools for medical training
  • Apps used for patient education
  • Apps that automate office operations in a health care setting
  • Apps that are generic aids or general purpose products (e.g., apps that magnify images, record audio, facilitate communication, provide turn-by-turn directions)
For further examples of any of these categories of apps, especially if you're planning to develop a medical app, be sure to check out the full document.

Once again I applaud the FDA for helping clear up some of the confusion regarding regulation of mobile medical applications.  It will be expected, though, that these guidelines will continue to evolve as technology matures.

Sunday, February 1, 2015

The Death of Meaningful Use: ONC's Interoperability Roadmap

ONC's vision of the health IT ecosystem as a Learning Health System

This past week the ONC released it's vision for health IT interoperability in a draft 1.0 document entitled "A Shared Nationwide Interoperability Roadmap."  There were a few items that caught my attention in the 166-page document, most notably the following words, nestled deep in the middle of page 50:
"This Roadmap shifts the nation’s focus from meaningfully using specific technologies with specific features to working together as a nation to achieve the outcomes desired from interoperability and a learning health system."
Granted, this is a draft document, but that sure sounds like an official death knell to meaningful use (MU) as we know it.

From my perspective, MU has done quite a bit of good to raise awareness of the importance of health IT to help us speak the same language so we can get the right information to the right person at the right time to make the right decision (sound familiar?).

However, the accompanying MU certification process has bogged-down health systems and stifled innovation through timed incentives and disincentives that has resulted in hospitals scurrying to claim their entitlements at the expense of thoughtful and measured health IT progress.

So it's with great excitement that I read the ONC's draft document that focuses this effort on public-private collaboration to start solving the thorniest issues we've faced, including policies and technologies to promote streamlined and robust interoperability.

The document also touched on a subject near and dear to my heart (p. 10):
"Given the increasing volume of mobile technology usage among consumers and across the care delivery system, approaches to enable "send, receive, find and use" in the near-term must support the flow of electronic health information across both institutional and mobile-based technologies. This means traditional approaches to health IT interoperability will need to become more agile and leverage the experience of modular consumer applications, such as those created by Facebook, Amazon and Apple. These secure, but simple architectures have enabled an ecosystem of applications that allow users to engage with electronic health information across a variety of different platforms and devices and open opportunities for entrepreneurial third parties to thrive."
Through the SMART on FHIR framework, this will soon be a reality in health care.  I'm excited to be leading the initiative at Duke to be the first Epic-based hospital with a functional implementation of SMART.

Furthermore, the document highlights the importance of facilitating the incorporation of patient-generated health data into our EHRs (p. 46):
"There needs to be a greater focus on incorporating patient-generated health data and ensuring the availability of tools for individuals to use this information to manage their health and make more informed health-related decisions."
Our experience with Apple's HealthKit here at Duke has shown us that this idea is a reality, today.  It's never been easier to get high-quality patient data integrated directly into our clinical systems so that providers can quickly act to improve patient outcomes.

This type of technology is the essence of a Learning Health System, which is emphasis of the ONC's draft document.  2015 is going to be an exciting year for healthcare technology and interoperability.

RIP, Meaningful Use.