Understanding Variation Among Medical Device Reporting Sources: A Study of the MAUDE Database.

Abstract

Background: Increasing medical device usage raises concerns regarding unexpected, potentially life-threatening events that pose public health risks. Such events are reported to the Food and Drug Administration (FDA), and cataloged in the Manufacturer and User Facility Device Experience (MAUDE) database, a vital tool for post market surveillance that requires information of high quality and integrity, particularly concerning reporting sources.

Purpose: To analyze reporting behavior among different reporters, including manufacturers, distributors, and user facilities, by examining differences in reported factors, namely: (1) device types, (2) product problem attribution, and (3) selection of Device Problem Codes (DPCs) associated with the root causes of events.

Methods: Data spanning from 2005 to 2022 was retrieved from the MAUDE database using Python. Reports were grouped by reporter type and divided according to device type, and the reporter's indication of association with a product problem. Furthermore, events were classified by their respective DPCs, which were manually grouped into four categories: device issues, user issues, clinical issues, or unknown.

Findings: The analysis revealed significant variations among reporters across all examined aspects (P < 0.00001 in all comparisons according to the proportion test). Manufacturers predominantly focused on infusion pumps (10.1%) and Implant, Endosseous, Root-Form (IER) devices (7.6%), with a product problem indication rate of 29.2% in their reports. Device issues codes were the most frequently observed category in their reports, comprising 36.3%, followed by unknown codes (32%) and clinical codes (19.3%). Distributors, on the other hand, primarily reported on IER devices (89%) and exhibited the lowest product problem indication rate at 2.7%. Clinical issues codes predominated in their reports, constituting 85.7%, followed by unknown codes (6.7%). User facilities concentrated on intravascular sets (4.7%), Electrosurgical, Cutting & Coagulation & Accessories (4.2%), and Ventricular (Assist) Bypass (4.1%). Remarkably, their product problem indication rate was the highest at 56.7%. They predominantly reported device issues codes (54.3%), followed by use codes (30.8%), and unknown codes (11.4%) IMPLICATIONS: The notable variation among different reporters underscores the importance of incorporating diverse sources when analyzing the database, particularly in cases where majority of reports originate from manufacturers. Decision-makers must approach database information comprehensively, considering data sources and diverse perspectives to inform regulatory decisions effectively. Developing strategies that encourage various reporters to contribute their unique and complementary viewpoints is advisable.