Skip to main content
Physiatry '25 Main banner
Final Program


Final Program

914 - Leveraging Advanced Text Mining to Identify Unique Risks Among Peripheral Nerve Stimulation Devices: A MAUDE Database Study

Friday, February 28, 2025
5:00 PM - 6:30 PM MST
Objectives: Peripheral nerve stimulation (PNS) devices are targeted neuromodulation technologies increasingly utilized for the treatment of chronic pain.  Given the substantive differences between devices, it is important that clinicians understand and can communicate potential reported adverse events to patients considering these interventions.  To support this discourse, we completed a retrospective, cross-sectional study of adverse events reported to the FDA’s Manufacturer and User Facility Device Experience (MAUDE) Database, specifically employing advanced text mining techniques to leverage free-text event descriptions to identify unique and shared complications of specific devices.

Design: We analyzed 594 adverse event reports submitted to the FDA’s MAUDE Database between January 1st and December 31st, 2023, involving patients with implanted PNS devices for chronic pain. Primary outcome measures were frequencies of words and phrases used in the free text “Event Description” field for reported adverse events, with devices segmented by premarket approval/notification (PMA/PMN) number.  Term frequency-inverse document frequency (TF-IDF) scores were also computed to identify over-represented words and phrases which suggest aspects of reports unique to each device.

Results: When all reports were analyzed together, hardware components and surgical interventions were described most commonly. Reports for the Curonix Freedom® System most frequently referenced increased pain, erosion, and overstimulation.  Reports for the Nalu™ Neurostimulation System most commonly described hardware complications, namely with its pulse generator and therapy discs.  Mainstay Medical’s Reactiv8® and SPR® Therapeutics’ SPRINT® PNS System had lower overall numbers of reports which most frequently discussed hardware (pulse generator and lead)-related and exit site (skin and surgical glue)-associated factors, respectively. 

Conclusions: This study showcases the power of advanced text mining in uncovering distinct and shared complications among PNS devices. By analyzing device-specific adverse event reports, we provide critical insights into differential risk profiles, empowering clinicians to better inform patients of potential adverse effects during treatment with these devices.