FDA MAUDE Trends 2020–2025: 14.4M Reports, Diabetes Dominance, Rising Deaths

Executive Summary

The FDA Manufacturer and User Facility Device Experience (MAUDE) database is the agency's primary repository for medical device adverse event reports — capturing every reported death, serious injury, and malfunction linked to a medical device in the United States. Each year, the FDA receives over two million such reports.

Our analysis of the complete MAUDE event dataset from 2020 through 2025 — 14.4 million adverse event reports across six years — reveals a database increasingly dominated by a single device category: diabetes management. Continuous glucose monitoring (CGM) systems and insulin infusion pumps together account for more than half of all reported events, driven primarily by the high-volume reporting practices of a small number of manufacturers.

Setting aside the diabetes device cluster, the data shows a more nuanced picture: death reports rose from 9,347 in 2020 to a peak of 15,262 in 2024 (a 63% increase), before moderating to 13,243 in 2025. Injury reports have nearly doubled from 611,803 to 1,142,840 over the same period. And the reporting volume has grown 84% from 1.57 million events in 2020 to 2.89 million in 2025.

This article quantifies the MAUDE trends by event type, manufacturer, device category, and year — giving regulatory, quality, and safety professionals a data-driven understanding of what the adverse event landscape actually looks like, beyond the headlines.

Data Source and Methodology

All statistics in this article are derived from our direct analysis of the FDA MAUDE database event files (public dataset, partitioned by receive-year), covering events received from January 2020 through December 2025. The dataset includes the following fields per record: report number, event type (Death / Injury / Malfunction / Other / No answer provided), date received, date of event, report source code, patient problems, product code, brand name, generic name, manufacturer, device class, and medical specialty.

Each row represents one device associated with a report. A single adverse event involving multiple devices generates multiple rows. Total records analysed: 14,418,055 across six years.

Event type percentages are calculated against the total for each year. Manufacturer and product code rankings reflect raw event counts, not severity-weighted scores.

The Scale of Reporting: 14.4 Million Events in Six Years

The MAUDE database has grown substantially year over year:

Total (2020–2025): 14,418,055 events

Several patterns stand out:

• Volume has nearly doubled. From 1.57 million events in 2020 to 2.89 million in 2025 — an 84% increase over six years. Part of this growth reflects expanded mandatory reporting requirements and manufacturer compliance programmes, not necessarily a proportional increase in device safety problems.

• The 2022 spike and correction. The 45.3% jump in 2022 (to 2.95 million events) was followed by a 20.6% decline in 2023. This pattern is driven primarily by CareFusion SD (a BD subsidiary), which reported 871,329 events in 2022 — 29.5% of all events that year — before normalising to 154,744 in 2023.

• Death events are rising. Despite representing less than 0.6% of all reports, death events increased from 9,347 (2020) to 15,262 (2024) — a 63% rise. The 2025 figure of 13,243 represents a partial moderation but remains 42% above the 2020 baseline.

• Injury reports are growing faster than malfunctions. From 39.0% of events in 2020 to 43.0% in 2024, injury reports have been climbing as a share of total events, suggesting either increased injury reporting compliance or a genuine shift in the event mix.

The Diabetes Device Cluster: Over Half of All Events

The single most important fact about the MAUDE database is that diabetes management devices — continuous glucose monitors (CGMs), insulin pumps, and automated insulin dosing systems — account for a disproportionate share of all reports. This is not because these devices are inherently more dangerous; rather, it reflects their unique combination of high user volume, continuous wear, and manufacturer reporting practices.

Top Product Codes (2020–2025 Combined)

The diabetes-related product codes (QBJ + FRN + QFG + OZP + OYC + OZO + LZG + QLG + PZE + PQF) collectively account for approximately 40–50% of all MAUDE events in any given year.

Why diabetes devices dominate: CGMs and insulin pumps are worn continuously by millions of patients. Each sensor failure, pump occlusion, or inaccurate reading generates a reportable malfunction. Manufacturers like Dexcom and Medtronic have established systematic reporting programmes that capture every complaint as an individual MAUDE entry. This creates reporting volume that dwarfs other device categories — but the clinical significance of individual events is often modest.

The Dental Implant Anomaly

Product code DZE (Implant, Endosseous, Root-Form) — dental implants — is the single largest product code in MAUDE, with approximately 3.6 million events across six years. The primary reporters are Nobel Biocare and Institut Straumann, two major dental implant manufacturers. The events are overwhelmingly classified as "Malfunction" or "Injury" and typically relate to implant failures, component fractures, or prosthetic complications.

Dental implant events in MAUDE are primarily driven by volume — millions of dental implant procedures annually — rather than acute safety concerns. However, the sheer volume makes DZE the dominant product code in the database.

Top Manufacturers: A Concentrated Reporting Landscape

In each year, a small number of manufacturers generate a large share of total events:

2025 Top 10 Manufacturers

Top 10 combined share in 2025: 54.3% of all 2.89 million events.

The top four manufacturers alone — Dexcom, Nobel Biocare, Tandem, and Straumann — account for 38.5% of all MAUDE events. This concentration reflects two factors: (1) high-volume consumer devices (CGMs and insulin pumps) that generate thousands of reports per day, and (2) dental implant manufacturers reporting every implant-related complaint.

Dexcom's Reporting Volume

Dexcom is consistently the single largest reporter in MAUDE, with events growing from 214,142 in 2020 to 353,869 in 2025 (a 65% increase). This reflects the rapid growth of CGM adoption: as more patients use Dexcom sensors, the absolute number of reported events grows proportionally. Most Dexcom events are classified as "Malfunction" (sensor failure, inaccurate readings, adhesive issues) rather than "Death" or "Injury."

Medtronic's Dual Presence

Medtronic appears in MAUDE through two distinct divisions:

• Medtronic Puerto Rico Operations — primarily insulin pump and diabetes device reports (176,665 events in 2025)
• Medtronic MiniMed — additional insulin pump and CGM reports (72,441 events in 2025)

Combined Medtronic diabetes-related events in 2025: 249,106, making Medtronic the second-largest contributor after Dexcom.

The Reporting Context: Volume vs. Risk

A critical nuance in interpreting MAUDE data is that reporting volume does not equal device risk. Several factors inflate the apparent severity of high-volume product codes:

1. Systematic reporting by manufacturers. Large manufacturers with established complaint-handling processes file an individual MAUDE report for every complaint, even minor ones. A sensor that gives a slightly inaccurate reading and a sensor that fails catastrophically both generate a single MAUDE entry.

2. Continuous-use devices generate more reports. A CGM sensor worn for 10 days has 10 days of exposure time during which any issue can be reported. An implantable pacemaker may function for years without generating a single report.

3. User population size matters. Dexcom and Abbott CGMs are used by millions of patients. A device used by 5 million people will generate more reports than a device used by 50,000 people, even if the per-patient event rate is identical.

4. Duplicate and follow-up reports. MAUDE allows initial reports, follow-up reports, and consolidated reports. The event counts in this analysis include all report types, which may overstate the number of unique adverse events.

A BMJ study published in March 2025 found that medical device manufacturers often delay reporting adverse events to the FDA, with more than half of late reports attributed to just three manufacturers and 13 specific devices. The study, by Everhart et al., analysed MAUDE reporting timeliness and found that late reporting is a systemic issue that may obscure the real-time safety picture.

What the Death Data Shows

Death events are the most closely watched category in MAUDE. While they represent less than 0.6% of all reports, the trend is concerning:

The 2023–2024 spike in death events coincides with increased FDA enforcement activity and heightened public attention to device safety following the Philips Respironics recall and other high-profile events. The 2025 moderation may reflect improved device safety, changes in reporting practices, or simply the natural variability of a relatively small numerator against a large denominator.

Important caveat: MAUDE death reports indicate that a death occurred in temporal proximity to device use — they do not establish causation. A death report filed when a patient dies while connected to a ventilator may reflect the patient's underlying condition rather than a device malfunction.

FDA's Transition from MAUDE to AEMS

The MAUDE database has now been consolidated into the FDA's new Adverse Event Monitoring System (AEMS). FDA launched AEMS on 11 March 2026 as a unified platform for adverse event data across all regulated product categories, and migrated MAUDE's device data into AEMS by the end of May 2026. The move reflects the agency's recognition that MAUDE — in use since 1999 — had become difficult to navigate and analyse at scale.

Key changes AEMS brings:

• Modernised search and retrieval — a unified public dashboard replacing MAUDE's legacy web interface
• Real-time publication — reports appear in near real time rather than monthly batches, so safety signals surface faster
• Improved data quality — better standardisation of event coding and manufacturer identification
• Cross-product analytics — enhanced tools for identifying emerging safety signals across all FDA-regulated product categories

The migration does not change reporting obligations for manufacturers, who continue to file Medical Device Reports (MDRs) under 21 CFR Part 803.

Implications for Industry Professionals

• Benchmark against the right peer group. If your company makes cardiovascular implants, comparing your MAUDE event rate to Dexcom's (a CGM manufacturer) is meaningless. Benchmark within your device category and adjust for units-in-use.

• Watch the death trend. The 63% increase in death reports from 2020 to 2024 is a signal that warrants attention, even after accounting for reporting volume growth. FDA is paying close attention to this metric.

• Reporting practices are a competitive differentiator. Companies with systematic, timely reporting programmes face lower regulatory risk than those with inconsistent or late reporting. The BMJ study on late reporting suggests this remains an industry-wide challenge.

• AEMS is now live. With MAUDE migrated into AEMS (device data by end of May 2026), companies that invest in data infrastructure — clean MDR submissions, consistent coding, timely follow-up — are better positioned to take advantage of real-time signal detection.

Key Takeaways

• 14.4 million adverse event reports across 2020–2025, growing from 1.57M to 2.89M per year (+84%)
• Diabetes devices dominate — CGMs and insulin pumps account for approximately 40–50% of all events
• Dexcom is the largest single reporter with 353,869 events in 2025 (12.2% of total)
• Death events rose 63% from 9,347 (2020) to 15,262 (2024), before moderating to 13,243 in 2025
• Top 10 manufacturers account for 54.3% of all events, driven by CGM and dental implant reporting
• The 2022 spike was driven by CareFusion/BD reporting 871,329 events in a single year
• MAUDE has migrated to AEMS — FDA launched AEMS on 11 March 2026 and moved device data over by end of May 2026, bringing real-time publication and improved analytics

Data Source

Analysis sample: FDA MAUDE database event files (public dataset, partitioned by receive-year), covering events received January 2020 through December 2025. Total records analysed: 14,418,055. Event type classification based on the event_type field (Death, Injury, Malfunction, Other, No answer provided). Manufacturer identification based on the manufacturer field. Product code descriptions cross-referenced with FDA classification database. MAUDE data has since been migrated into the FDA Adverse Event Monitoring System (AEMS), launched 11 March 2026.