Four existing health data networks are demonstrating that national patient databases are an invaluable tool in advancing research on America’s health. An article by DPM researchers and colleagues discusses the experience of four featured programs -- FDA’s Mini-Sentinel (which focuses on medical product safety), the National Patient-Centered Clinical Research Network (comparative effectiveness research), the National Institute of Health (NIH) Distributed Research Network (biomedical research), and ESPnet (public health surveillance).
To date, data from these large networks has been available for research studies on:
- assessing the potential for angioedema, an adverse reaction associated with some antihypertensive drugs used by millions
- identifying risk of intussusception (intestinal blockage) following rotavirus vaccination in infants
- monitoring the occurrence of influenza-like illnesses
Lessons have been learned in developing these initial national networks for use in secondary research. First, the networks are largely composed of data created by health systems or health plans as electronic health records or claims data. The sensitivity of the data requires careful consideration of issues such as privacy, security, ethics and governance. Before it can be used for research, there needs to be a clear understanding of the data’s limitations. Researchers must also understand how the data was collected and coded, and also where there may be gaps. Lastly, plans to maintain and update the networks are also needed to keep them robust for future studies.
According to paper co-author Richard Platt, MD, MS, who has led the coordination of the Mini-Sentinel, Collaboratory, and ESPnet, with the right development and controls in place, large data networks can make a significant contribution to evidence-based practice.
To read the article abstract in Health Affairs, click here.