It’s the ultimate catch-22. There’s more emphasis than ever before on sharing knowledge — particularly as IT environments become more complex — and yet, the barriers that prevent the flow of data stand as tall as ever.
The solution? Applying a little “positive disruption,” says Sandy Aronson, Executive Director of IT at Partners Personalized Medicine, a division of Partners HealthCare. Recently, PPM partnered with Persistent Systems to develop an open-source platform to more effectively facilitate data exchange among providers and enable a new generation of decision support apps. The goal is to “to dramatically reduce the costs of building, deploying and sharing focused IT interventions to improve clinical decision making,” says Aronson.
In this interview, he talks about the “enormous opportunity” the industry has to leverage tools to improve care delivery, why Partners chose to collaborate with Persistent, and the unique perspective PPM offers through its experience in the genetics space. Aronson also talks about the challenges that still exist, why they’re using SMART and FHIR, and how he landed with Partners.
Kate Gamble: Talk about the Partners HealthCare Personalized Medicine. What the key areas of focus, and what role do you play in moving these initiatives along?
Sandy Aronson: Partners HealthCare Pesonalized Medicine (PPM) was created in 2001 to proactively deploy infrastructure our clinicians would need to effectively integrate advancing genomic techniques into clinical practice. Partners realized that genetics and genomics had the potential to fundamentally improve the way we care for patients, but would also require substantial new infrastructure to achieve broad adoption. PM supports research, clinical laboratory, educational, and information technology infrastructure in support of this goal.
We have established IT-enabled continuous learning processes that support the clinical variant interpretation that lies at the core of genetic testing. We are now working to develop an expanded infrastructure capable of enabling clinicians to more deeply analyze a broad spectrum of data types, including but not limited to genetics at the point of care.
Gamble: When you look at how personalized medicine is being underutilized in the industry, despite the enormous potential it holds for transforming care, is it a point of frustration?
Aronson: It is more an enormous opportunity that has the potential to benefit us all to a degree that is difficult to imagine. So I feel excitement rather than frustration. I think there is the potential to instantiate true continuous learning clinical processes that will accelerate the fundamental rate of improvement in healthcare in a way never before seen in human history.
Gamble: What do you believe are the biggest challenges in advancing personalized medicine – is it mostly the technical aspect, or are there other major obstacles?
Aronson: It depends on what you mean by personalized medicine. If you focus on the genetic testing aspect, then the biggest challenge is genetic test reimbursement. Tests are not as available as they need to be.
You could also define personalized medicine to be the process of gathering multiple different forms of data, including but not limited to genetic data, and combining them for a clinician so they can more finely diagnose patients, and therefore, make better treatment decisions. In this case, perhaps the biggest challenge is on the clinical data side. When we want to introduce new IT support into the EHR ecosystem, we need the ability to more uniformly and securely access more of the clinical data that could aid decision making. This will open up the potential for broader applications of machine learning and other techniques that are difficult to introduce into clinical processes today.
Gamble: Talk about the collaboration between Partners and Persistent Systems to create an open-source platform that will facilitate knowledge exchange between providers. What was the impetus for this initiative, and what are the key goals?
Aronson: Our group’s work in the genetic space to date has taught us that it truly is possible to establish infrastructure that facilitates the exaction of medical knowledge from clinical process, and then feeds that knowledge back into the healthcare system in near real-time for use in patient care. But through that work, we also learned that there are many barriers that slow the construction of this kind of infrastructure. We hope to dramatically reduce the costs of building, deploying and sharing focused IT interventions to improve clinical decision making. If we can do this, we think it will empower both others and ourselves to innovate in a broader way than is currently possible.
Gamble: At what point is the project right now, and what are the next steps?
Aronson: We are nearing the deployment of our first focused app based intervention to assist in platelet management. We will learn from this experience as we expand to other areas.
Gamble: What type of interaction do you have with clinicians? Is there a user group that provides input or participates in testing?
Aronson: Clinicians generate the ideas for interventions and then provide feedback and insights throughout the process. They are critical contributors every step of the way.
Gamble: Talk about the decision to utilize FHIR. Was there any hesitancy go that route, as FHIR still seems to have its share of doubters?
Aronson: We really like FHIR’s flexibility, particularly in the way profiles allow you to enhance and evolve the standard. We will be pragmatic as we move forward, and in some cases may rely on non-FHIR interfaces, but in general we are happy with the decision to use FHIR. We will often build adaptors that link to data sources that do not have native FHIR support. These adaptors will provide FHIR based endpoints to make apps built on top of them more portable.
Gamble: We’re starting to see more examples of organizations partnering with vendors to develop solutions, rather than just buying products off the shelf. Do you think this trend will continue? What do you enjoy about partnering with vendors such as Persistent Systems?
Aronson: Absolutely. We are particularly excited that vendors are realizing that open business models — for example, the open-source model we are pursuing with Persistent — provide ways to monetize functionality without constraining innovation.
Gamble: In terms of your career, what is your background, and what made you interested in coming to the health IT side?
Aronson: I graduated with a BA in Computer Science and an MA in Organizational Behavior from Stanford University. I started an IT consulting company while I was in school, then went to Monitor Company, a strategic consulting firm. After Monitor, I joined Sapient enjoyed the ride over the next 4 years as it approached the SP500. Then I spent a year in Taiwan learning Manderin and spending time with my future wife. After that, I came back to the US and started a web-based training company that is now part of Best Software. Then at around 30 I had an early mid-life crisis and began to feel that I was not making as significant a difference in my career as I had hoped. So I went back to school at night to learn biology. I was extremely fortunately that Partners HealthCare ultimately took a chance on hiring me and gave me the opportunity to do what I do now.
Gamble: What do you think it will take for healthcare to move toward in terms of IT adoption and gain some ground in comparison to other industries?
Aronson: The work we are doing with Persistent is squarely focused in exactly this area. We believe a platform is needed that simplifies the sharing of app-based IT interventions across healthcare institutions. Adoption of the platform will follow the creation of apps that are proven to provide both economic and clinical value.