Health networks show options for national data sharing
- By Stephanie Kanowitz
- Sep 04, 2014
Health information technology and electronic health data have great potential for streamlining and improving the health care system. But challenges are many, including a lack of standards in electronic health record (EHR) formats and how to manage privacy requirements surrounding data that carries personally identifiable information (PII).
Still, significant strides are being made in discovering ways to make health data usable, informative and protected, according to Health Affairs. The policy journal recently highlighted public sector solutions that advance national health data sharing in an article titled, Four Health Data Networks Illustrate the Potential for A Shared National Multipurpose Big-Data Network.
The oldest of the applications, called Mini-Sentinel, has been run by the Food and Drug Administration and is part of FDA’s Sentinel Initiative, launched in 2008 to improve the tracking of reports of adverse events linked to the use of regulated products and other safety issues.
Part of FDA’s efforts to create an active adverse events surveillance system, Mini-Sentinel uses pre-existing electronic healthcare data from multiple sources, including administrative and claims information developed by 18 private health plans. The system accrues data on nearly 50 million people.
Mini-Sentinel consists of a Coordinating Center and experts at 33 organizations. Participating health plans and organizations create data files in a common format, based on how the distributed data will be used, and keep control over them behind firewalls, according to the Health Affairs article.
“To answer many of the FDA’s questions, it is important to have a complete record of individuals’ relevant medical product exposures and outcomes during the period under evaluation,” the journal states. “Because EHRs typically do not satisfy this need, Mini-Sentinel requires insurer data for every individual.”
When further data is required, Mini-Sentinel pulls from full-text records with PII redacted. The system also adds EHR data, e-lab results and registry information as needed.
The system works by sending a query to participating organizations that execute an analytic code on data stored in a common format. To do so, it uses PopMedNet, an open-source query distribution software application that enables simple creation, operation and governance of distributed health data networks.
Data partners refresh their datasets quarterly, and they can access full-text electronic and paper records from ambulatory centers and hospitals for about 90 percent of their members, according to the article.
“Using the distributed querying system, a set of several hundred diagnostic tests for the form, content and logical consistency of each data element is performed each time that new data are extracted from the host systems, transformed into the common data format and loaded for use,” article authors said.
The security requirements are high for Mini-Sentinel. As a public health surveillance activity, it follows privacy rules set forth in the Health Insurance Portability and Accountability Act. Additionally, PopMedNet conforms to the Query Health Initiative in the Standards and Interoperability Framework of the Office of the National Coordinator for Health IT, according to the article.
Another big data system highlighted by Health Affairs, the National Patient-Centered Clinical Research Network (PCORnet), will go live next year. It aims to create a large, highly representative, national network for conducting clinical outcomes research.
PCORnet is being built on a foundation of data derived from EHRs, but will also rely on a coordinating center for network development and analytic tools, the article states. PCORnet’s architecture will be able to send queries to and receive responses from Mini-Sentinel, although members will be able to review requests and results before releasing them.
The third system is the National Institutes of Health’s Collaboratory Distributed Research Network (DRN). Started in 2013, it is creating an infrastructure to improve the conduct of clinical trials being done in clinical care settings. This system eases the use of host organizations’ electronic data without having to transfer protected or proprietary data as often as possible.
“The NIH Collaboratory DRN will also support collaborations by enabling investigators or organizations with research data sets to allow remote collaborators to conduct analyses without transferring the data,” the report states.
The fourth system, ESPnet, has been in use in Massachusetts since 2007, reporting on diseases and health trends based on records from about 20 percent of the commonwealth’s residents. Health care practices in the network pull data nightly and put it into a standard format that the network can read automatically the next day, the journal stated. The system automatically reports on infectious diseases such as the flu and sends ad hoc queries, such as those about variations in obesity across locations and population subgroups, according to the report.
ESPnet uses the same query distribution platform as the other networks and also lets practices maintain control over their data. The system differs from the other three in that it lets public health workers create queries without their having to know a programming language.