In September of 2013, HIMSS and the U.S. Department of Health and Human Services announced Adam Culbertson’s appointment as an innovator-in-residence. In this interview, Adam shares the details of his work on patient data matching, a problem that, if solved, would revolutionize the use of Big Data.
HIMSS: What is an innovator-in-residence (IIR)?
Adam Culbertson: The IIR program is the creation of the the Office of the Chief Technology Officer's (CTO) office at the U.S. Department of Health and Human Services (HHS) in Washington, DC and funded by a partnering organization, in my case, HIMSS. The program’s goal is to bring creative, energetic people together to tackle tough healthcare technology problems. The IIR Fellow, of which there have been five since the program’s inception in 2012, works to cross boundaries, applying lean start-up methods and design thinking to solve difficult problems within and external to HHS. They leverage rapid ideation and iteration of solutions through real world testing to solve challenges in healthcare.
HIMSS: How did you apply to become an IIR?
AC: I applied online and then began a series of interviews with stakeholders from HIMSS, the Office of the National Coordinator for Health Information Technology (ONC), and the CTO’s office at HHS. Despite the long process, it was worth it to have an opportunity to dramatically impact the future of patient care.
HIMSS: What was your background prior to the IIR program?
AC: I spent approximately two years as a biomedical informatics fellow at the National Institutes of Medicine where I worked on Natural Language processing under the mentorship of Dr. Thomas Rindflesh. We modified and evaluated the system to extract adverse drug events from drug labels from the freely available Daily Med. At the Regenstrief Institute I was exposed to a variety of topics from national experts including Big Data, patient matching, health information exchange, HL7 and LOINC [Logical Observation Identifiers Names and Codes].
HIMSS: What is the main focus of your research as an IIR?
AC: My main focus for the two years of my appointment is patient data matching, also referred to as record linkage. Record linkage deals with the ability to identify specific patients’ data in varied data silos. Sources include hospital electronic medical records, laboratory systems, primary care electronic health records [EHR] and claims data. Matching is complicated by the fact that different providers’ electronic medical records record different data attributes. Errors in data entry and quality further challenge linking patient records.
HIMSS: Why is patient data matching important to healthcare reform?
AC: Patient data matching is essential if we are to realize the value of electronic health records and interoperability. New payment and healthcare delivery models, such as ACOs [Accountable Care Organizations], require accurate dissemination of patient healthcare data across the system. Patient data matching also has broad implications for population health and cost reduction by identifying fraud and abuse. Most importantly patient matching is important to patient safety to ensure the right patient gets the right treatment at the point of service. To achieve this goal we also need to have a discussion on patient privacy and security.
HIMSS: With whom might you collaborate as partners in the development of your solutions and who do you anticipate may become early adopters?
AC: It’s still early in my work, but there are key stakeholders and experts in patient matching with whom I would like to engage. Ideally we want to work with as many stakeholders as possible, but we also need to balance collaboration with the need to accomplish day-to-day tasks. A big part of my work is gaining momentum around the problem and working with partners to find a solution that fits multiple needs. In terms of adoption, the community as a whole implements solutions; the more we can build a movement around the problem the more readily solutions can be adopted, whether the solution is a framework or a product.
HIMSS: What do you think will advance patient data matching the most?
AC: I think the single biggest challenge is to bring all the stakeholders together to find a common vision and direction. At the end of the day technology alone won’t solve the problem. A mentor of mine said, “patient matching is a 90 percent sociopolitical problem and a 10 percent technical one.” Technology plays a significant role in the solution but the social aspect plays a bigger role. The three-legged stool to improve patient matching requires improving data quality, processes and algorithms. Patient matching is such a difficult problem because it is intertwined with many core components of health IT infrastructure; to achieve success you need to look at the problem from a holistic perspective.
As a past Fellow of Biomedical Informatics at both the National Library of Medicine and Regenstrief Institute, Adam brings a wealth of knowledge to his position as HHS IIS. Adam has an MS in Health Informatics and Informatics: Human-Computer Interaction Design.
Article updated April 11, 2018