Flawed FDA Device Safety Reports Delay Detection of Dangerous Malfunctions

A sweeping analysis of the United States Food and Drug Administration’s medical device safety database has revealed systemic weaknesses in the way manufacturers report adverse events, raising urgent questions about whether the agency can detect emerging hazards in time to protect patients. The study, published in The BMJ, examined six years of summary adverse event reports submitted to the agency’s Manufacturer And User Facility Device Experience database, known as MAUDE, and found widespread incompleteness, undercounting, misclassification, and delay. The findings arrive at a moment of heightened scrutiny for the postmarket surveillance system that is supposed to serve as the nation’s early warning network for dangerous device malfunctions.

Under US regulations, medical device manufacturers are required to report malfunctions and adverse events to the FDA within 30 days. These reports feed into the MAUDE database, which regulators, clinicians, researchers, and the public rely on to identify potential safety signals that may warrant product recalls, label changes, or other corrective actions. For the system to function as intended, the reports must be complete and accurate. Yet earlier programs, most notably the Alternative Summary Reporting program, allowed manufacturers to submit millions of adverse events as bulk summaries rather than as individual case reports, a policy framed as a way to reduce the burden on industry. The FDA ended that program in 2019 amid concerns that it was masking safety signals, but similar summary reporting exemptions remain in use today, and until now their accuracy, completeness, and compliance with FDA rules were largely unknown.

To close that gap, researchers analyzed all adverse event summary reports submitted to MAUDE between July 2019 and June 2025. They looked for four specific types of failure: incomplete identification of the device or use problem involved, undercounting of the number of adverse events, failure to correctly identify the submission as a summary report, and delays in reporting serious adverse events beyond the regulatory deadline. The dataset they examined was substantial. During the six-year study period, 6,308 summary reports were submitted to the FDA, containing a total of 159,503 adverse events. The majority of these events involved high-risk devices, the very category of products where timely safety intelligence matters most.

The results paint a troubling picture of a reporting pipeline riddled with defects. Fifty-seven percent of the summary reports identified no device or use problem at all, leaving regulators without a clear indication of what went wrong. Thirty-eight percent contained insufficient information to evaluate the events described. Eleven percent were incorrectly identified as non-summary reports, meaning they would not be flagged or processed as aggregated submissions. Seven percent incorrectly listed only one event in the number of events field, even when multiple adverse events were involved. Perhaps most alarmingly, 55 percent of the reports exceeded the FDA’s 30-day reporting deadline, meaning that more than half of the aggregated safety information arrived late, in some cases long after the window in which it could have contributed to a rapid regulatory response.

The scale of the problem is growing rather than shrinking. The researchers note that the number of summary reports submitted to the FDA, and the number of adverse events contained within them, more than doubled between 2020 and 2024. That trajectory suggests manufacturers are increasingly relying on bulk summary mechanisms to discharge their reporting obligations, even as the quality of those submissions falls short of the standards the agency itself has set. In a passive reporting system that depends entirely on the diligence of the regulated industry, systematic gaps in data quality translate directly into blind spots in national device surveillance, potentially delaying the recognition of hazards affecting widely implanted or widely used products.

The authors of the study are candid about the limitations inherent in their analysis. Passive adverse event reporting has well-known weaknesses, including incomplete ascertainment of events in the real world and the absence of a denominator of device use against which event rates can be calculated. The researchers also lacked access to publicly available information on the specific requirements governing the summary reporting exemptions, which constrained their ability to assess compliance against detailed regulatory criteria. Even so, they argue that identifying and correcting the flaws documented in the study is vital, particularly given the accelerating volume of summary reporting. Immediate improvements, they write, are needed to better protect patient safety and adhere to the transparency principles outlined by the FDA, adding that making serious adverse event reports fully visible and accessible is critical to patient safety.

The study’s publication was accompanied by a linked editorial from Koji Todaka of Kyushu University Hospital in Japan, who frames the findings as a challenge to the fundamental purpose of the reporting system. The MAUDE database exists to support public scrutiny of medical device safety, and reports that are incomplete, mislabeled, or late undermine that mission, he argues. When the public cannot see an accurate and timely account of adverse events associated with the devices used in their care, trust in postmarket surveillance erodes, and the evidence base for clinical decision-making weakens accordingly.

Todaka proposes a set of reforms aimed at restoring confidence in the system. Regulators, he suggests, should require accurate and timely summary reporting, standardized device identification, and clear links between adverse event reports and device registries. Standardized identification is particularly consequential, because ambiguous device problem descriptions and inconsistent device identifiers make it difficult to aggregate reports across manufacturers and time periods, which is precisely the kind of aggregation needed to detect a safety signal. He also recommends that researchers integrate spontaneous reports, registry data, and linked healthcare records rather than treating any single data source as definitive, reflecting a broader consensus that passive reporting alone cannot carry the full weight of postmarket safety surveillance.

Change may be on the horizon within the FDA itself. The agency is currently transitioning to a new Adverse Event Monitoring System, or AEMS, designed to consolidate existing reporting systems and publish adverse event reports in real time. If the new platform enforces structured data entry, validates event counts, timestamps submissions against the 30-day requirement, and makes serious adverse event reports immediately visible to the public, it could address many of the deficiencies documented in the BMJ analysis. The transition also presents an opportunity to revisit the summary reporting exemptions themselves, tightening the conditions under which bulk submissions are permitted and requiring manufacturers to supply the device and problem information that more than half of current reports omit.

The stakes of this administrative infrastructure are far from abstract. High-risk devices, including implants, cardiac devices, and surgical equipment, are used in millions of patients, and a delayed or obscured safety signal can mean continued exposure to a malfunctioning product for months or years. The BMJ study demonstrates that the mechanisms meant to catch those signals are, in their current form, leaking information at every stage: at submission, in classification, in event counts, and in timing. As the researchers and the accompanying editorial both emphasize, transparency builds trust and evidence improves safety, and both are essential to effective postmarket surveillance and safer medical devices. Whether the FDA’s new monitoring system, combined with stricter oversight of summary reporting, can close the gaps identified in this study will be a defining test of the next chapter of medical device regulation in the United States.

Subject of Research: Accuracy, completeness, and timeliness of summary adverse event reporting to the FDA MAUDE medical device safety database

Article Title: Medical device adverse event reports are often incomplete, inaccurate, and delayed

Article References: Medical device adverse event reports are often incomplete, inaccurate, and delayed. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: medical devices, FDA, MAUDE database, adverse event reporting, postmarket surveillance, patient safety, summary reporting exemptions, BMJ, device regulation, safety signals, transparency, regulatory compliance