APIs, Patient Empowerment, and Improving Health Outcomes
By Gajen Sunthara / Presidential Innovation Fellow
For the past year, I have been honored to serve my country as a Presidential Innovation Fellow (PIF) with the Office of the National Coordinator for Health IT (ONC). As my PIF journey comes to an end, I wanted to provide a brief recap of my work and its potential for the future.
At a high-level, my work this year has focused on three components:
- Demonstrating how Application Programming Interface (APIs) can be leveraged to improve data access;
- Creating a prototype personal health record (PHR) powered by data obtained from APIs for patient engagement and empowerment; and
- Demonstrating how patient-mediated health data sharing and donation could take place, especially in the context of precision medicine.
The following three segments break down my work and its relevance to the future.
1. Health Data: Transform, Access and Transport
Right now, health data access is not seamless for patients or providers. In my experience, electronic health record (EHR) systems have been built around documentation, not collaboration. Patient access to their own health data is mainly available through patient portals and most of the patient portals are not configured to make it easy for patients to download their data in electronic standardized format. On the other hand, providers face a similar situation, with patient health data scattered across multiple provider systems that can’t always talk with one another.
Health data access can be simplified using APIs. APIs are the means by which all our apps communicate and exchange information across systems. We can and should embrace what works in other industries. While health data may be “different” than any other industry, the technical means by which we can access and use data is not.
Health data is currently stored in its own respective health system database. Before the information can be exchanged, one way to “standardize” it is to transform it into Fast Healthcare Interoperability Resources (FHIR), open, HL7 standards, and transported via HTTPS a secure internet protocol for APIs.
For me to try out solutions, I built a FHIR server. The FHIR server transforms the data into FHIR specifications and reads/writes the information from the source and destination database.
There are two important use cases for this FHIR server:
- Health system to patients – Business-to-Consumer (B2C).
- Health system to health system – Business-to-Business (B2B).
With B2C, health systems can implement Blue Button functionality in their patient portals enabling access to patients’ data via FHIR APIs or allowing them to download it themselves. FHIR API access provides seamless transmission of health data from a health system to consumers or the app that the consumer chooses. This ensures patients will be in full control of how information flows to the destination or desired apps in a way that typically doesn’t exist with most patient portals today.
FHIR standards that apply to APIs also can be leveraged by business-to-business. FHIR API’s support read and write, which is great for interoperability. Health systems can leverage this simple API to exchange information with other approved provider systems.
My source code for the FHIR server, which provides data transformation, access and transport, is available in GitHub.
2. Patient engagement & empowerment using Personal Health Record (PHR)
Blue Button is all about enabling patients online access to their data. Once the patient is able to access and download their data, s/he still needs a platform or third party app to keep them engaged, help them understand and visualize the data, and become empowered to take action and improve their own health. The huge ecosystem of innovative health applications in the market will play an important role in this engagement and empowerment.
As a patient myself, I would like to see my Patient-Generated Health Data (PGHD) and my customized travel alerts (based on allergies) be proactively provided to me in a single app. PGHD can be gathered from wearable devices such as Fitbit and Apple Watch, and when combined with visuals from my lab results, medications, reminders, schedule, allergies, care team, care plan, immunizations, genomics as well as other types of health data (you get the idea), it becomes an incredibly powerful platform to create a more complete picture of an individual’s health. Combining all these disparate datasets is typically not possible today. To test out how the industry might address this challenge, I created a prototype app called “myHealth”.
An important product feature of this prototype is called “myHealth API” which enables patients to keep track of patient health systems data access points (aka FHIR API endpoints). For a single patient with multiple providers, their care might span across multiple health systems, especially patients with more serious and chronic conditions. This means these patients have multiple patient portals they have to sign into if they want to try and keep track of all their data. The burden of keeping up with multiple portals leads to drop-off in use of the portals and frustration for many patients today. In addition, any new, updated clinical data are not sent out automatically to the patient’s preferred destinations, whether to their caregiver or a mobile app. myHealth API can collect and aggregate patient’s data via API’s and gather data intelligently behind the scenes to create a dynamic dashboard for patients to review and understand a comprehensive snapshot of their health.
myHealth API also supports outbound data sharing capabilities via APIs, where patients would be able to share their own data with 3rd party applications. Imagine if a Type 1 diabetes patient on Continuous Glucose Monitoring (CGM) device tracked their glucose and, similarly, that their provider wanted to monitor their patient population. An app could be created to get patient data in an API format with patient consent and authorization and also enable patient’s to have full control to share the data and discontinue anytime.
3. Patient-mediated health data sharing / donation
The myHealth prototype I created also provides functionality for the patient to collect health data from multiple health systems. It allows users to simply drag and drop to keep them in one place and have full control over their data and privacy.
As a patient, I might switch providers, or I might want to contribute my data to research and science. myHealth also provides functionality for the patient to not only share their data with each provider, but also share data and genomics information for research initiatives under the Precision Medicine Initiative (PMI). This wealth of data can then be utilized for research purposes.
Imagine one open source PHR app platform for patients, which can connect to any health system via FHIR standards, securely exchange health data with APIs, connect to mHealth applications/sensors, and provide genomics data. This PHR can offer widgets/apps like ecosystem, where patients can control apps created by range of different stakeholders from health systems to tech companies. This scenario, which I hope isn’t too far into the future, will help make personalized medicine a reality for many patients with the help of the open source community.
Patients should be at the center of the health care system. We have one of the most sophisticated data-driven health care systems in the world. As we shifted focus from a, piece meal, fragmented, Fee-for-Service model, to a value-based, data driven pay for performance and outcomes model, bigger and better quality data will become increasingly important. These data provide us with expanded tools to educate and empower patients to become actively involved in managing their health.
In order to improve outcomes and cut costs across the system (better health and smarter spending) we need to provide tools and the proper platform to free the data from health systems and help empower patients. Innovative third party apps can help to put patients in the driver’s seat, giving patients full control over their health information flow, help them engage with dynamically changing PGHD, and share health data for research and science to improve the lives of themselves and others.
This post was originally published on the Health IT Buzz and is syndicated here with permission.