One MAUDE report is not one malfunction

A MAUDE search result reads like a list of events: one row, one malfunction, one device. For a defined set of product codes, that reading is wrong, and FDA built the exception into the rule on purpose.

The alternative FDA can grant

21 CFR 803.19 lets FDA grant a manufacturer an exemption, variance, or alternative to the standard medical device reporting requirements, including a change in reporting frequency to quarterly, semiannual, or annual. FDA used that authority in 2018 to create the Voluntary Malfunction Summary Reporting program, after a pilot that ran from 2015, and refreshed it again in 2024 to match the current MDR form and event codes. Under the program, a manufacturer of a device in an eligible product code can report malfunctions — not deaths, not serious injuries — in summary form once a quarter, instead of filing an individual report within 30 days of becoming aware of each one.

Eligibility lives in a different database than the reports do

Whether a product code qualifies is not a fact on the malfunction report itself. It is a designation FDA maintains in the device Product Classification Database, reviewed and updated as FDA adds codes to the program — generally not before a product code has existed for at least two years, so a brand-new classification is not eligible on day one even if its device type resembles one that already is. Death and serious-injury reports keep their individual, 30-day reporting requirement regardless of a product code’s eligibility; the program only ever touches malfunctions.

What one row is actually standing in for

A quarterly summary report does not file one malfunction. It reports the total count of reportable malfunctions FDA received for a single combination of brand name, device model, and device problem code, for that quarter — and the individual narrative text that makes a standard MDR readable, event by event, does not exist for what is bundled inside that total. Search MAUDE for a product code enrolled in the program and a single row can be the record of one malfunction or of several dozen, and the row itself does not look different either way. Counting rows, for an eligible product code, is not counting events.

The report is the unit the database shows you. The malfunction is the unit you actually care about. For most product codes those are the same thing. For this one, by design, they are not.

The practice

Before comparing malfunction volume across product codes, or across quarters for the same one, check the code’s summary-reporting eligibility in FDA’s classification database first. If it is enrolled, a row count is a report count, and the number worth reading is the total inside each summary, not the number of rows returned. Skip that check and a device with a genuinely higher malfunction volume can look quieter than a comparable device that simply is not enrolled in the program — an artifact of which reporting track FDA granted, read as a finding about which device is safer. What a real difference in malfunction volume means for a specific product is a judgment for a regulatory or quality professional with the file in front of them; this is a reading method for the primary source, never a substitute for that judgment.

MAUDE sits outside FDA Radar’s current source coverage — noted, with the reason, on what we monitor and what we don’t. The record is public today; the eligibility flag is the field worth checking before the count is.