Cardiovascular registries: Patient data, appropriate use, and technology innovation

Jan. 31, 2018 / By L. Gordon Moore, MD

Sometimes science upends what we think of as common sense.  We might believe that a degenerative tear in the meniscus (cartilage in the knee) ought to be fixed. Then some researchers compare sham (fake) surgery to really fixing these tears and discover similar outcomes.[i]  For this type of meniscus tear (let’s be careful not to over-generalize, there are other types of tears), it appears to make no difference if you do nothing or have surgery.

It is very important that we apply scientific methods to the things we do to people, because sometimes our best intentions may not lead to the outcomes we hope to achieve. To this end, we capture data from patients across the U.S. in national registries: Data repositories that allow researchers to explore the risks and benefits of procedures and therapies.

The American College of Cardiology has the National Cardiovascular Data Registry (NCDR) that tracks many different types of cardiac procedures and therapies. The Implantable Cardioverter Defibrillator (ICD) registry is one that tracks devices and procedures used to reduce the risk of sudden cardiac death.  Information submitted to the registry can help identify the people for whom the device is most appropriate (a.k.a. “appropriate use criteria”).

“The ICD Registry™ is a nationwide quality program that helps participating hospitals measure and improve care for patients receiving implantable cardioverter defibrillators (ICDs) and cardiac resynchronization therapy devices with defibrillator (CRT-Ds). The ICD Registry captures the characteristics, treatments, and outcomes of patients receiving ICDs. Patient-level data is submitted by participating hospitals on a quarterly basis to the American College of Cardiology Foundation’s (ACCF) National Cardiovascular Data Registry (NCDR) which then produces an Outcomes Report of the hospital’s data, with comparison to both a volume peer group (number of ICD patients submitted annually) and the entire ICD registry data set.”

From the NIH National Library of Medicine/Clinical site

Dodson et al ask an important question: Has the existence of this registry helped to improve outcomes?[ii] Their study looked at the first four years following the introduction of the registry and found that the rate of optimal medical therapy and cardiac resynchronization therapy (both demonstrated in prior studies to improve outcomes) increased, and the rate of complications experienced by patients undergoing ICD procedures decreased in the study period.

Knowledge comes at a cost: Highly trained hospital staff can spend hours searching through a chart to compile information and that is only one of many registry groups (Trauma Registry, Cancer Registries, etc).  Improvements in natural language processing makes it possible to automate much of the more cumbersome searches – looking through the written notes to find documentation of a person’s New York Heart Association Heart Failure Classification level, for instance. This automation can both speed up the work of registry abstraction and potentially lead to improved data accuracy (reduced human error).

Common sense and intuition may serve us in some circumstances, but science brings us to a deeper understanding and sometimes (like with the sham knee surgery) surprises us. The existence of these registries is a benefit to all who undergo these procedures. Advancements in technology make it easier to capture the needed information and could lead to improved data accuracy, which may in turn lead to deeper understanding and better outcomes.

L. Gordon Moore, MD, is senior medical director for Populations and Payment Solutions at 3M Health Information Systems.

Natural language processing meets health care.

[i] Sihvonen, Raine, Mika Paavola, Antti Malmivaara, Ari Itälä, Antti Joukainen, Heikki Nurmi, Juha Kalske, and Teppo L.N. Järvinen. “Arthroscopic Partial Meniscectomy versus Sham Surgery for a Degenerative Meniscal Tear.” New England Journal of Medicine 369, no. 26 (December 26, 2013): 2515–24.

[ii] Dodson, J. A., R. Lampert, Y. Wang, S. C. Hammill, P. Varosy, and J. P. Curtis. “Temporal Trends in Quality of Care Among Recipients of Implantable Cardioverter-Defibrillators: Insights From the National Cardiovascular Data Registry.” Circulation 129, no. 5 (February 4, 2014): 580–86.