The FDA Digital Health Center of Excellence is part of the planned evolution of the digital health program with the intent to drive synergy for digital health efforts, align strategy with implementation, prepare the FDA for the digital health future, and protect patients and maintain the FDA standards of safety and effectiveness.
Ultimately, the program works to strategically advance science and evidence for digital health technologies that meets the needs of
stakeholders.
This free in-depth webinar, presented by Matthew DiamondChief Medical Officer, Digital Health Center of Excellence, will cover the digital health landscape and areas of application, goals and outcomes, planned services and launch plan, and the current areas of focus - including AI/ML-Based SaMD.
This presentation originally aired during the 2021 State of Medical Device Virtual Summit.
The FDA Digital Health Center of Excellence and the Advancement of Digital Health Technology
1. www.fda.gov/digitalhealth
THE FDA DIGITAL HEALTH CENTER OF EXCELLENCE
AND THE ADVANCEMENT OF DIGITAL HEALTH TECHNOLOGY
Matthew Diamond, MD, PhD
Chief Medical Officer for Digital Health – Center for Devices & Radiological Health (CDRH), US FDA
December 8, 2020
2. www.fda.gov/digitalhealth 2
Digital Health Center of Excellence (DHCoE)
Digital Health Landscape and Areas of Application (Spectrum)
Need to Coordinate and Synergize
Goals and Outcomes
Planned Services & Launch Plan
Current Areas of Focus
3. www.fda.gov/digitalhealth 3
Digital Health
The convergence of connectivity, data and computing power for healthcare and
related uses across the life of an individual or a patient.
Leveraging computing power, sensors, connectivity and software
Moving health care from the
Clinic to the
Patient
Understanding patient’s
behavior and physiology
“In the wild”
Focusing on prevention for
early/smaller
interventions
Healthy living Prevention Diagnosis TreatmentRecovery Home care
4. www.fda.gov/digitalhealth 4
Digital Health Solutions Landscape
Tele-care: Activity monitoring; Remote medication management; Video Consultation
Healthcare Analytics: Public Health analytics; Care delivery analytics
Health IT Systems: EHR ; E-prescribing systems
Software Solutions: Software as a Medical Device ( Diagnostic and Therapeutic, CADx, CADe); General Wellness Apps
Services: Healthcare systems services; Monitoring services for Chronic disease management; Monitoring services for aging; Post acute monitoring
Wearables: Vital signs monitors ; Sleep Apnea Monitors (PSG); Neurological Monitors; Activity Trackers/ Actigraphy
Moving health care from
the Clinic to the
Patient
Understanding patient’s
behavior and physiology
“In the wild”
Focusing on prevention
for early/smaller
interventions
Diagnostics and Therapeutic Solutions Areas:
Post-Traumatic Stress Disorder ; Generalized Anxiety Disorder; Depression adjunct therapy; Mild Cognitive Impairment; Autism Spectrum Disorder;
ADHD; Neuropsychological Diagnosis and Therapy; Behavioral Therapy
HEALTH AND WELLNESS Solutions
• Improve Cognitive Function
• Promote Exercise & Weight Management
• Stress Management
• Mood and Resilience
• Disability Solutions
• Addressing Isolation
• Grief Counseling
Post Care solution: Activities of Daily Living; Physical Medicine – OT/PT
Patient engagement: routine lab result; appointment reminders; treatment prompts prescription refills; adherence to treatment; Patient Education
4
Healthy living Prevention Diagnosis TreatmentRecovery Home care
5. www.fda.gov/digitalhealth 5
Digital Health Technology
Convergence of
computing power,
connectivity, sensors,
and software used in
healthcare.
Used as a medical product
Incorporated into a medical product
(include a pharmacologic product)
Used to develop a medical product
Used to study a medical product
Used as a companion or adjunct to a medical product,
including diagnostics and therapeutics.
5
Healthy living Prevention Diagnosis TreatmentRecovery Home care Management
6. www.fda.gov/digitalhealth 6
Why a Digital Health Center of Excellence?
• Part of the planned evolution of the digital
health program
• Intent to
• Drive synergy for digital health efforts
• Align strategy with implementation
• Prepare the FDA for the digital health
future
• Protect patients and maintain the FDA
standards of safety and effectiveness
6
7. www.fda.gov/digitalhealth 7
Develop novel,
efficient regulatory
approaches that are
least burdensome
Connect
Build partnerships,
new networks to
accelerate and scale
Share
Gather, simplify,
and share
information to
increase awareness
and understanding
Digital Health Center of
Excellence
FDA’s
Empowering All to Advance Healthcare
Our goal: Empower stakeholders to advance health
care by fostering responsible and high-quality
digital health innovation that meets FDA standards
of safety and effectiveness.
The Digital Health Center of Excellence aims to:
∙ Connect and build partnerships to accelerate
digital health advancements.
∙ Share knowledge to increase awareness and
understanding, drive synergy, and advance best
practices.
∙ Innovate regulatory approaches to provide
efficient and least burdensome oversight .
8. www.fda.gov/digitalhealth 8
Anticipated Outcomes
• Strategically advance science and evidence for digital
health technologies that meets the needs of
stakeholders.
• Efficient access to a highly specialized expertise,
knowledge, and tools to accelerate access to digital
health technology that maintain standards of safety and
effectiveness.
• Aligned regulatory approach to harmonize international
regulatory expectations and industry standards.
• Increased awareness and understanding of digital health
trends.
• Consistent application of digital health technology policy
and oversight approaches.
• Reimagined medical device regulatory paradigm tailored
for digital health technologies.
10. www.fda.gov/digitalhealth 10
Digital Health Center of Excellence Services
✔ Set and lead strategic direction in digital health
✔ Launch strategic initiatives
✔ Establish and promote best practices
✔ Enable efficient, transparent, and predictable product review with
consistent evaluation quality
✔ Build new capacity to oversee and leverage DH technologies
✔ Create more shared resources
✔ Coordinate the development of cross cutting DH policies
11. www.fda.gov/digitalhealth 11
Digital Health Center of Excellence Services
✔ Provide scientific expertise across the Agency
✔ Offer training opportunities for FDA staff
✔ Disseminate shared resources
✔ Foster collaboration across FDA in common interest areas
✔ Facilitate synergies in regulatory science research in digital health
✔ Leverage, share, and avoid duplication of work
✔ Promote and showcase existing work at the Centers
12. www.fda.gov/digitalhealth 12
Digital Health Center of Excellence Services
✔ Provide clarity on regulation
✔ Advance international harmonization on device regulatory policy
✔ Facilitate and build strategic partnerships
✔ Communicate FDA research interests
✔ Advance digital health device international standards
13. www.fda.gov/digitalhealth 13
Phase I: Communication
Fall 2020
• Stakeholder Listening Sessions
• Update and develop resources for
FDA staff
• Begin operationalizing the DHCoE
and outcome measurement
• Amplify current work being done at
FDA in digital health
Phase II: Coordinate
Fall and Winter 2020
• Build strategic partnerships for
policy, regulatory science, and
fellowships
• Develop resources for external
stakeholders
• Create a digital health community
of practice
• Assemble FDA and CDRH advisory
groups
Phase III: Amplify
Winter 2021 onwards
• Continued strategic partnership
building and communication
• Update and implement regulatory
framework for digital health
• Harmonization with other regulators
Following is our roadmap for bringing the benefits of digital health
to all Americans, efficiently and collaboratively:
Raise Awareness and
Engage Stakeholders
Build Partnerships
Build and Sustain
Capacity
Digital Health Center of Excellence Roadmap
14. www.fda.gov/digitalhealth 14
Current Areas of Focus
Software as a
medical device
(SaMD)
Artificial
Intelligence/
Machine Learning
Wearables
Software in a
medical device
(SiMD)
Wireless
Connectivity
Interoperability
Medical Device
Cybersecurity
Virtual Reality/
Augmented
Reality
Real-world
Evidence and
Advanced Clinical
Studies
Advanced
Manufacturing
Patient-Generated
Data
Digital Biomarkers Digital Pathology
16. www.fda.gov/digitalhealth 16
Evolving Digital Health Device World …
Product Development Timeline
• Months to years +
• Less frequent modifications
Weeks to months + (incremental, iterative) and
potentially frequent modifications
Postmarket Data
• Limited availability and access to real world data
(522, PAS, MDRs, MedSun)
Potential for exponential increase in volume of
submissions
FDA Premarket Program Volume:
• Stable (~3,500 510(k) submissions / 2200
pre-submissions)
“Traditional” Device World
Evolving Digital Health Device
World
Potential for high availability and access to rich real
world data (benefits and risks)
The Need for a Tailored Approach
17. www.fda.gov/digitalhealth 17
IDx-DR
FDA News Release
FDA Authorizes Marketing of First Cardiac
Ultrasound Software That Uses Artificial
Intelligence to Guide User
February 7, 2020
FDA News Release
FDA Permits Marketing of Artificial
Intelligence-Based Device to Detect
Certain Diabetes-Related Eye Problems
April 11, 2018
Examples of AI/ML-Based SaMD
Caption Health
18. www.fda.gov/digitalhealth 18
www.fda.gov/digitalhealth
IDx-DR
AI/ML-Based Medical Devices: Opportunities
• Ability of AI/ML systems to learn from the wealth of real world
data and improve their performance
• Development of novel AI/ML devices in all medical fields
• Fundamentally transform the delivery of healthcare
• Earlier disease detection
• More accurate diagnosis
• New insights into human physiology
• Personalized diagnostics and therapeutics
19. www.fda.gov/digitalhealth 19
• Need for large, high-quality,
well-curated data sets
• Explain-ability of these
“black box” approaches
• Identifying and removing bias
QuantX
AI/ML-Based Medical Devices: Challenges
20. www.fda.gov/digitalhealth 20
Tailoring a Regulatory Framework for AI/ML-Based SaMD
Enhance patient
access to high quality
digital medical
products
Maintain a reasonable
assurance of safety and
effectiveness
Enable manufacturers to
rapidly improve software
products with minor
changes
Minimally burdensome
21. www.fda.gov/digitalhealth 21
Continuous Learning for Software as a Medical Device
If adaptations are pre-specified,
and the methods for determining an appropriate adaptation clearly delineated,
then a decision-making framework, as described here,
may be similarly applied for both locked and adaptive algorithms.
“Locked”
Algorithm
with Discrete
Updates
Updates more frequent
and performed by computer
Updates less frequent
and performed by human
Continuously
Adaptive
Algorithm
Spectrum of AI/ML-Based Algorithms
22. www.fda.gov/digitalhealth 22
Data selection and
management
Model training
and tuning
Model monitoring
o Log and track
o Evaluate performance
Model validation
o Performance evaluation
o Clinical evaluation
Deployed Model
New (Live) Data
Data for re-training
Data
for
re-training
Good Machine Learning Practices
AI Model Development
AI Device Modifications
Legend
AI Production Model
AI/ML Workflow
23. www.fda.gov/digitalhealth 23
Data selection and
management
Model training
and tuning
Model monitoring
o Log and track
o Evaluate performance
Model validation
o Performance evaluation
o Clinical evaluation
Deployed Model
New (Live) Data
Data for re-training
Data
for
re-training
Good Machine Learning Practices
AI Model Development
AI Device Modifications
Legend
AI Production Model
2
1
Premarket Assurance
of Safety and
Effectiveness
4
3
Review of
SPS (SaMD
Pre-Specifications) &
ACP (Algorithm Change
Protocol)
Culture
of
Quality
and
Organizatio
nal
Excellence
Real-Wo
rld
Perform
ance
Monitor
ing
Proposed TPLC Approach
Proposed Total Product Lifecycle (TPLC) Approach
AI/ML Workflow
24. www.fda.gov/digitalhealth 24
Good Machine Learning Practices (GMLP)
Data selection and
management
Model training
and tuning
Model monitoring
o Log and track
o Evaluate performance
Model validation
o Performance evaluation
o Clinical evaluation
Deployed Model
New (Live) Data
Data for re-training
Good Machine Learning Practices
GMLP= Good Machine Learning Practices
•Accepted practices in ML/AI algorithm design, development, training, and testing that
facilitate the quality development and assessment of ML/AI-based algorithms
•Based on concepts from quality systems, software reliability, machine learning, and
data analysis, etc
25. www.fda.gov/digitalhealth 25
SPS = SaMD Pre-Specifications:
• WHAT are the proposed types of changes to
the SaMD the sponsor intends to achieve?
• Draws a virtual “region of potential changes”
around the initial specifications and labeling of
the device.
SPS&ACP: A Pre-Determined Change Control Plan
ACP = Algorithm Change Protocol:
• HOW will the changes (pre-specified in the SPS) be
performed and validated?
• Step-by-step delineation of the procedures to be
followed for a specific device and type of change
26. www.fda.gov/digitalhealth 26
Continuous Learning for Software as a Medical Device
Medical software manufacturers are encouraged to leverage the software technology’s capability
of capturing real world performance data to understand user interactions with the SaMD,
and to conduct ongoing monitoring of analytical and technical performance to support future intended uses.
Adapted from “Software as a Medical Device (SaMD): Clinical Evaluation,” www.imdrf.org
27. www.fda.gov/digitalhealth 27
Further Questions or Feedback
DigitalHealth@fda.hhs.gov
www.fda.gov/digitalhealth
Matthew Diamond, MD, PhD
Chief Medical Officer for Digital Health
Digital Health Center of Excellence
Center for Devices and Radiological Health (CDRH), US FDA
Office of Strategic Partnerships & Technology Innovation, Division of Digital Health
matthew.diamond@fda.hhs.gov (301) 796-5386