Health

The Quiet Research Revolution in Health Care

We are reaching an inflection point in health care. In the last 10 years, we have digitized hundreds of billions of health records and are generating health-related data throughout our daily lives, but we have not seen a corresponding improvement in our understanding of disease or our ability to improve patient outcomes. 

Progress has been constrained for a simple reason. Though digitized, the vast majority of this health data is inaccessible to researchers and caregivers. The potential insights are locked away in thousands of data silos.

Breaking down these health data silos in order to transform our health care system is one of the greatest challenges of our time. The improvement in both medical research and patient care could be exponential: not only improved patient outcomes, but an improvement in the rate of improvement in patient outcomes. Innovative organizations across the country have already shown what is possible when even a small fraction of this health information is de-identified, connected and made available in real (or near-real) time.

Here are five powerful examples:

— Connect health data to build a global pandemic detection and response system

Five years ago, an outbreak of Ebola in West Africa caused global panic. Thousands of people died, and countries spent billions of dollars trying to control the disease. The damage easily could have been much worse.

To mitigate this enormous risk, innovative researchers have combined data from dozens of health sources together with geospatial mapping data to build a pandemic risk model. Others are using linked data for epidemiology, gathering data that would allow outbreak response to be tailored to local conditions. When outbreaks do occur, access to relevant, linked health data is critical to coordinating a rapid response: for example, linking diagnostic data for disease carriers back to airline records to understand who might have been on the same plane as the host.

— Connect health data to improve drug safety monitoring

In 2008, the Food and Drug Administration launched the Sentinel Initiative to monitor the safety of FDA-regulated medical products. As a complement to the agency’s existing adverse event reporting system, Sentinel allows the FDA to rapidly access de-identified and connected electronic health data in order to closely monitor the safety of medical products while protecting patient privacy.

Sentinel has quickly become the largest multisite distributed health database in the world focused on medical product safety, and the FDA continues to add new data types and analytical methods.

— Connect electronic health record and claims data to improve our understanding of rare, pediatric conditions

The Patient-Centered Outcomes Research Institute funded a data sharing network of leading academic medical centers, community health clinics, disease-specific registries, health plans and patients in order to conduct large-scale, patient-centered research. Within this national network, researchers connect data from a variety of care settings and databases to rapidly identify patient cohorts and perform population analytics and comparative effectiveness studies, using methods that protect the privacy of individuals.

While many of the PCORnet networks are based in specific geographies such as CAPriCORN in Chicago or REACHnet in Louisiana, others such as PEDSnet link together institutions based on the specific patient populations they serve.

— Connect health data to support clinical diagnostic support tools at the point of care

When a physician at the Lucile Packard Children’s Hospital at Stanford saw a teenage girl that had developed inflammation of the kidneys and pancreas as a complication of lupus, she realized that the patient was at risk of blood clots that could cause organ failure or death. The patient needed anticoagulants, but they would also increase the risk of bleeding. The physician asked around the hospital and got conflicting advice.

Unsure of what to do, the physician then went to Stanford’s de-identified database of electronic health records. Querying data on lupus patients, she was quickly able to see that many of those who developed inflammation of the kidneys and pancreas went on to develop blood clots. The database allowed the physician to quickly identify a pattern and make the decision to administer anticoagulants.

Such decision support tools have the ability to dramatically improve the quality and consistency of patient care. Connecting de-identified electronic data across hospitals and health systems can make such tools even more powerful, allowing physicians to quickly understand disease progression and patient outcomes across relevant cohorts of patients.

— Connect health data to enable faster and more representative recruitment for clinical trials

When PCORI decided that it wanted to run a five-year study to compare the effectiveness of two aspirin doses in preventing heart attack and stroke among 20,000 patients with coronary heart disease, it knew that it needed a new way to recruit and enroll patients in the study. Traditionally, clinical trial populations under-represent minority groups and patients who have limited access to care. Since one of PCORI’s main goals is to involve community organizations and patients in research, the project was designed with a community-based method for recruitment that would allow for engaging a more diverse pool of patients.

The researchers partnered with community organizations to conduct outreach and identify potential study participants. Upon consenting, patients had their information de-identified and linked to an electronic health record database across five different medical centers, as well as corresponding claims data. This allowed for automatic assessment of eligibility based on the patient’s medical history and a faster enrollment process.

Over the next five years, we believe that the research productivity gap is going to close, and that the results will be dramatic. The key is seamless interoperability: the ability to connect relevant health data in real time.

However, technology is not enough. Patients need to understand what is possible and vote with their feet. We encourage you to help us advocate for this change. 

 

Travis May is the founder and CEO of Datavant, which helps organizations protect, link and exchange their health data; Jake Plummer is general manager, health systems and government, at Datavant; Abel Kho is the director of the Center for Health information Partnerships and associate professor of medicine and preventive medicine at Northwestern University; and Jyotishman Pathak is the Frances and John L Loeb Professor of Medical Informatics, chief of the Division of Health Informatics and vice chair of the Department of Healthcare Policy and Research  at Weill Cornell Medicine.

Morning Consult welcomes op-ed submissions on policy, politics and business strategy in our coverage areas. Updated submission guidelines can be found here.

Morning Consult