Life Sciences Horizons Brochure 2025 - Flipbook - Page 14
13
2025 Horizons Life Sciences and Health Care
Using AI for device postmarket surveillance to meet FDA and EU obligations
The regulatory landscape for postmarket surveillance
(PMS) is evolving, with stringent requirements set by
both the FDA and the EU to ensure the ongoing safety
and performance of medical devices. As AI becomes
more integrated into regulatory compliance and
medical device monitoring, its potential to reduce
the burden of monitoring adverse events, streamline
reporting, and enhance patient safety will only grow.
The combination of Natural Language Processing
(NLP), anomaly detection, and predictive analytics
allows for a more proactive, intelligent approach
to medical device PMS, and will facilitate how
medical device manufacturers manage their
products and their regulatory obligations with
better data and analytics.
AI-driven technologies are transforming PMS by automating
adverse event detection, analyzing vast amounts of real-world
data, and predicting potential device failures before they escalate
into widespread issues. AI enhances regulatory compliance
by streamlining data processing and improving accuracy in
reporting, offering a pathway toward greater harmonization in
global postmarket surveillance.
In the U.S., the primary data sources for monitoring device
performance are complaints and real world evidence (RWE),
which may well include sources such as electronic health records
(EHRs), social media discussions, and patient-reported outcomes.
Additionally, for a small number of medical devices, the FDA
imposes an obligation to conduct PMS studies (21 CFR Part 822),
which provides a more methodical look at a device’s performance
in larger populations than were studied in the pivotal clinical trial.
The EU Medical Device Regulation (MDR) and In Vitro
Diagnostic Regulation (IVDR) outline comprehensive postmarket
surveillance (PMS) requirements to ensure ongoing safety and
performance evaluation. These include Postmarket Clinical
Follow-Up (PMCF) to continuously gather clinical data, and
Periodic Safety Update Reports (PSUR) for regular risk-benefit
assessments. Additionally, the EU’s vigilance system mandates
rigorous complaint handling and adverse event reporting,
enabling proactive risk mitigation and enhancing patient
safety. A systematic literature review and the use of RWE play a
crucial role in PMS by providing valuable insights into long-term
device performance, identifying emerging risks, and supporting
regulatory compliance.
In both regions, manufacturers must actively monitor available
data sources and assess any indications of potential safety
concerns. When such concerns arise, they are required to analyze
the data and determine whether corrective action is necessary.
While both regions share a commitment to device safety,
differences in reporting timelines, data collection methodologies,
and compliance frameworks pose challenges for global
manufacturers. Additionally, complaints and RWE data sources
that serve as inputs to PMS are comprised of large data sets that
are inherently messy. Historically, monitoring and analyses have
been performed using largely human driven methods, such as
control limit charting, pareto analysis, run charts, and other signal
detection and trend analysis methods.
As the volume of PMS data grows, traditional monitoring methods
struggle to keep pace with the sheer complexity and speed of
medical device complaint reporting. AI and Machine learning
(ML) are revolutionizing PMS by automating data analysis,
detecting hidden patterns, and predicting potential failures before
they escalate into serious patient safety concerns. AI-powered
surveillance enhances regulatory compliance, reduces response
times, and improves overall device reliability. Different AI-driven
techniques are transforming complaint monitoring.
Read the
Life Sciences
and Health
Care Chapter
of our AI Hub
