US FDA Regulatory Approach to Clinical Decision Support (CDS) Software
The demand for Clinical Decision Support (CDS) tools utilizing digital solutions and AI is swiftly expanding, with their benefits being evident.
In this emerging market, there are various types of CDS tools, such as those that adhere to clinician prescriptions and convey alerts through pop-up notifications; CDS tools that involve the analysis of patients' physiological parameters, aiding doctors in making optimal therapeutic decisions amidst a range of treatment options; AI-based CDS systems employed in medical imaging processing, enhancing early detection of cancer.
Numerous medical device companies are actively developing CDS tools with the goal of enhancing clinical practices and health outcomes.
However, distinguishing between tools subject to US FDA medical device regulation (Device Software Functions) and those exempt from such regulation (Non-Device CDS) can be complex. This article aims to provid an insigth into FDA approach to regulating CDS.
NON-device CDS:
The FDA excludes software functions from the medical device definition (thus exempting them from medical device regulation) if they meet all four of the following criteria:
Not intended to acquire, process, or analyze a medical image or a signal from an in vitro diagnostic device or a pattern or signal from a signal acquisition system.
Intended for the purpose of displaying, analyzing, or printing medical information about a patient or other medical information.
Intended for the purpose of supporting or providing recommendations to an HCP about prevention, diagnosis, or treatment of a disease or condition.
Intended for the purpose of enabling an HCP to independently review the basis for the recommendations that such software presents so that it is not the intent that the HCP rely primarily on any of such recommendations to make a clinical diagnosis or treatment decision regarding an individual patient.
​
​Examples of Non-Device CDS software include:
Software that provides a list of diagnostic and presents treatment order options to an HCP based on clinical guidelines for the care of adult patients presenting with pneumonia symptoms.
Matching patient-specific medical information from records or reports with reference information (e.g., clinical guidelines) that is routinely used in clinical practice.
Contextually relevant information about diseases or conditions, aiding HCPs in selecting treatment options, including drug, device, surgical and/or lifestyle changes for heart failure patients according to disease stage and clinical guidelines.
Software offering a list of preventive, diagnostic or treatment options to an HCP, like FDA-approved chemotherapeutic agents for specific cancer types, identified through medical information analysis and confirmed biopsy results.
Software analysing patient-specific symptoms to provide an HCP with a prioritized list of diagnostic tests to consider.
The examples above fulfill criteria 1-3. Emphasis should be placed on meeting Criterion 4 for classification as Non-Device CDS. The software's labelling and functionality must provide information to enable HCPs to independently assess the basis for presented recommendations, ensuring their reliance on their own judgment.
Examples of Device Software Functions:
Software utilizing image sets (e.g., CT, MRI) to create individual treatment plan for review by an HCP for patients undergoing radiation therapy. This software constitutes a device function. It does not meet Criterion 1 because it is intended to analyze a medical image. It does not meet Criterion 2 because it is not intended to display, analyze, or print medical information. It does not meet Criterion 3 because it generates the treatment plan, which is intended to provide a specific treatment directive.
Software identifying patients possibly suffering from opioid addiction based on analyses of medical information, family history, prescription patterns, and geographical data. This software also falls under device function. The software does not meet Criterion 3 because it provides a specific diagnostic or treatment output or directive.
Software analyzing near-infrared camera images for diagnosing brain hematoma. This software is considered a device function. It does not meet Criterion 1 because it is intended to analyze a signal. It does not meet Criterion 2 because it is not intended to display, analyze, or print medical information. It does not meet Criterion 3 because it offers specific diagnostic outputs.
AI-based software analyzing medical information, including mammography reports, to provide follow-up options for HCPs post a patient's annual mammogram. This is a device function unless HCPs can understand key variables influencing the recommendations, aligning with Criterion 4 for Non-Device CDS.
Software providing clinical recommendations to patients or caregivers, rather than HCPs, qualifies as device CDS software.
It's important to note that MDCG 2019-11 may interpret the regulatory status of CDS software differently. Consequently, while a CDS software might be classified as Non-Device according to the FDA, it could still need to adhere to EU MDR (2017/745) requirements for medical devices.
If you have more questions or you need assistance in defining the regulatory strategy for CDS software don't hesitate to contact us.
​