© Copyright AAMI 2014. Single user license only. Copying, networking, and distribution prohibited. Managing Risk Controlling for Cybersecurity Risks of Medical Device Software Kevin Fu and James Blum About the Authors Editor‘s Note: This article originally appeared in Communications of the ACM (2013;56[10]21–3; doi: 10.1145/2508701). Reprinted with permission. Kevin Fu, PhD, is an associate professor of While computer-related failures are known to Measuring Medical Device Security: computer science play a significant role in deaths and injuries Quantitative or Qualitative? and engineering involving medical devices reported to the U.S. Between years 2006 and 2011, 5,294 recalls and at the University of Food and Drug Administration (FDA),1 there is approximately 1.2 million adverse events of Michigan in Ann Arbor, MI. He also is a Sloan research no similar reporting system that meaningfully medical devices were reported to the FDA’s fellow. E-mail:
[email protected] captures security-related failures in medical Manufacturer and User Facility Device devices. Experience (MAUDE) database.1 Almost 23% of James Blum, MD, Medical device software must satisfy system these recalls were due to computer-related is chief of critical properties, including safety, security, reliability, care and surgical failures, of which approximately 94% presented specialty anesthesia resilience, and robustness, among others. This medium to high risk of severe health conse- at Emory University column focuses on the challenges to satisfying quences (such as serious injury or death) to Hospital in a security property for medical devices: post- patients.1 For security incidents on medical Atlanta, GA, and market surveillance, integrity and availability, devices, no systematic national reporting assistant professor of anesthesiology and regulation and standards.