AMIA Urges More Work on FDA’s Decision Support Guidance

Feb. 6, 2018
In comments submitted to the Food & Drug Administration (FDA) this week, the Bethesda, Md.-based American Medical Informatics Association (AMIA) said new draft guidance is well-timed to initiate a broad discussion over the future of clinical decision support (CDS) and patient decision support (PDS) software.

In comments submitted to the Food & Drug Administration (FDA) this week, the Bethesda, Md.-based American Medical Informatics Association (AMIA) said new draft guidance is well-timed to initiate a broad discussion over the future of clinical decision support (CDS) and patient decision support (PDS) software.

AMIA officials noted that several trends are converging to transform the delivery of knowledge resources through decision support software for patient care and wellness, including cloud-based computing, machine learning, and improved patient access to their clinical data. However, according to AMIA, the FDA’s draft guidance was incomplete, given a lack of discussion regarding how FDA would regulate CDS and PDS still considered medical devices.

AMIA’s comments are in regard to the FDA’s regulatory oversight of CDS software intended for healthcare professionals and PDS software intended for patients and caregivers who are not healthcare professionals.

Section 3060 of the 21st Century Cures Act of 2016 (Cures) compels the FDA to exclude certain software functionalities from the definition of device. This legislative language effectively prohibits the FDA from regulating that software as a device, which then may require pre-market review, clinical trials, and other clearance activities. According to AMIA, some of these excluded functionalities describe what is commonly understood as CDS, and the patient-facing equivalent, PDS.

“However, not all CDS/PDS was excluded from the definition of the device by Cures. This draft guidance tries to articulate where the line between regulated device and unregulated functionality lies per the FDA’s interpretation of Cures,” AMIA officials state.

AMIA supported development of the FDA’s guidance, but the group warned of “lingering confusion among developers and clinicians trying to determine whether specific decision support software is, or is not, considered a device.” According to AMIA, some of the criteria used to determine if a functionality should be excluded is ambiguous. For instance, one criterion requires independent review of the basis for recommendations presented by the CDS/PDS, and FDA describes such review to be independent if the clinician/patient does not “rely primarily on such recommendations, but rather on their own judgement,” AMIA says. Unfortunately, this caveat is founded in the Cures statute, so while ambiguous, it remains part of the equation to determine if the functionality is a device.

To be useful and practical for its intended audiences, AMIA recommended that the FDA articulate why examples of software are categorized as exempt or not, using the criteria established by Cures. AMIA also recommends that the FDA include discussion regarding it’s intended regulatory controls for CDS software considered a device, even if the intended regulatory controls are still in development.

What’s more, AMIA also urges the FDA to include discussion regarding the anticipated literacy levels, and their variance, across intended users of patient decision support software functions.

In 1997, AMIA led a consortium of diverse stakeholders to develop a framework for monitoring and regulating clinical software systems. This consortium recommended that “the FDA focus its regulatory efforts on those systems posing highest clinical risk that give limited opportunities for competent human intervention,” and it established a decision algorithm for FDA to use in its regulation of standalone clinical software systems as medical devices.

Since that time, AMIA notes, cloud computing and machine learning have changed the way decision support software can impact patient care. “CDS and PDS functionalities can be based on a trained neural network, multivariate regressions, or fuzzy logic with perhaps millions of data points, enabling a data-driven approach to decision-making at the bedside. This new paradigm also creates new risks to patient safety, and requires new thinking around regulation,” the group noted.

“Twenty years ago, AMIA led a national dialogue over how to regulate software,” AMIA president and CEO Douglas B. Fridsma, M.D., Ph.D., said in a statement. “Now is the time to revisit old assumptions, and the informatics community is eager to help.”

Sponsored Recommendations

A Cyber Shield for Healthcare: Exploring HHS's $1.3 Billion Security Initiative

Unlock the Future of Healthcare Cybersecurity with Erik Decker, Co-Chair of the HHS 405(d) workgroup! Don't miss this opportunity to gain invaluable knowledge from a seasoned ...

Enhancing Remote Radiology: How Zero Trust Access Revolutionizes Healthcare Connectivity

This content details how a cloud-enabled zero trust architecture ensures high performance, compliance, and scalability, overcoming the limitations of traditional VPN solutions...

Spotlight on Artificial Intelligence

Unlock the potential of AI in our latest series. Discover how AI is revolutionizing clinical decision support, improving workflow efficiency, and transforming medical documentation...

Beyond the VPN: Zero Trust Access for a Healthcare Hybrid Work Environment

This whitepaper explores how a cloud-enabled zero trust architecture ensures secure, least privileged access to applications, meeting regulatory requirements and enhancing user...