All In MedTech

Device safety

How to search MAUDE and read device adverse event reports

MAUDE search guide: how to find MDR reports and device malfunction reports in FDA's database and openFDA, read a report, and avoid common traps.

By the All In MedTech editorial team · Published November 19, 2026 · 7 min read

You can search MAUDE, FDA's database of medical device adverse event reports, in three ways: FDA's own MAUDE search page, the free openFDA device adverse event API, or summary tools such as this site's MAUDE Monitor. The most reliable way to find reports for a type of device is to search by its FDA product code, then narrow by brand, manufacturer, event type and date. Whatever you find, the counts describe what was reported, not how often a device fails or whether it caused harm.

This guide covers each search route, how to read a report, a worked example and the traps to avoid. For background on where the reports come from, start with what is MAUDE.

Before you search: pick the right identifier

MAUDE records carry several ways to identify a device. They are not equally reliable.

FieldHow it is enteredBest for
Product codeThree-letter FDA code for the device typeFinding every report for a device type
Brand nameFree text typed by the reporterA specific product, if you try several spellings
Manufacturer nameFree text, often in several formsA company's reports, with care
Model or catalog numberFree textA specific version of a product
UDI device identifierTaken from the label, where reportedAn exact product, when present
PMA or 510(k) numberEntered by the reporterThe products covered by one FDA decision

Brand and manufacturer names are the usual starting point, but they vary. The same product can appear under a trade name, an abbreviation, a model number or a misspelling, and one company can appear under its full legal name, a subsidiary or a former name. The product code avoids this. If you do not know it, look the device up on /devices or in FDA's Product Classification database. FDA device classes and product codes explains how the codes work.

Option 1: FDA's MAUDE search page

FDA's MAUDE search is the official interface. It offers two modes.

Simple search looks for words in reports received in a selected year. It accepts single words, phrases in quotes and terms joined with "and," and you can enter a UDI.

Advanced search lets you combine fields:

  1. Choose a product problem or product class from the lists, or type a product code.
  2. Add a brand name, manufacturer or model number if you want a specific product.
  3. Pick an event type: death, serious injury, malfunction, other, or no answer provided.
  4. Set the date report received by FDA range.
  5. Optionally filter by summary report (yes or no), report number, PMA or 510(k) number or UDI device identifier.
  6. Run the search and open any record from the results list.

Three limits shape what you see:

  • 500 results at most. FDA's search returns no more than 500 records per query. If your search hits that cap, narrow the dates or add fields. For larger pulls, FDA points users to its downloadable MDR data files.
  • About ten years. The online search covers roughly the last ten years. Older reports, and some reports filed under exemptions and alternative programs before June 2019, are only in the data files.
  • Monthly updates. FDA updates MAUDE monthly with reports received through the end of the previous month.

FDA also offers the Total Product Life Cycle (TPLC) database, which shows, for each product code, premarket decisions, adverse event report counts by device and patient problem, and recalls on one page.

Option 2: openFDA

openFDA provides the same reports through a free device adverse event API. It suits anyone who wants totals rather than lists. You can count reports by event type, by year, by product problem or by manufacturer for any search, and the 500-record cap on FDA's search page does not apply to counts.

Two cautions. openFDA processes its own copy of the data, so totals can differ a little from FDA's search page. And openFDA states plainly that its results should be treated as unvalidated and not used for medical decisions.

Option 3: this site's MAUDE Monitor

The MAUDE Monitor reads openFDA and shows, for reports FDA received from January 1, 2015:

  • totals by event type (malfunction, injury, death and other);
  • a monthly chart of reports received;
  • rankings by product code, manufacturer and brand;
  • the most common device and patient problem terms;
  • a list of reports, each linked to FDA's own record.

The Monitor leaves out the free-text narratives. To read what the reporter wrote, follow the link to FDA's record.

The signals page is arithmetic on report counts, not a statistical test. A jump often reflects how reports were filed, for example a manufacturer clearing a backlog or a recall prompting reports of older events. The methods page sets out exactly how the numbers are made.

Worked example: implantable pacemaker pulse generators

Product code DXY covers implantable pacemaker pulse generators, a Class III device type. In openFDA data last updated November 11, 2026, reports received since January 1, 2015 that list product code DXY total 11,682:

  • Injury: 6,099
  • Malfunction: 4,890
  • Death: 690
  • Other or blank: 3

The most frequent device problem terms were "Adverse Event Without Identified Device or Use Problem" (2,656 reports), over-sensing (1,180), device dislodged or dislocated (999), premature discharge of battery (763), signal artifact or noise (763) and failure to capture (711).

How to read this:

  • The top term says that no device problem was identified. That is common in MAUDE and means the reporter recorded an adverse event without pointing to a device fault.
  • Problem terms can reflect the whole system. A pulse generator report may describe something that involved the lead or the procedure.
  • The 690 death reports record that a death was reported in connection with the device. FDA's stated position is that a report does not by itself show the device caused or contributed to the outcome, and patients with pacemakers often have serious heart disease.
  • The total is small compared with how widely pacemakers are used. Without a count of devices implanted, the number cannot be turned into a failure rate. Summary reporting can also lower counts for malfunctions.

How to read a single report

When you open a report on FDA's site, work through it in this order:

  1. Event type. Death, injury, malfunction or other, as the reporter classified it. A report that mentions a death is coded as a death only if the reporter believed the device was or may have been a factor.
  2. Device fields. Brand, model, lot, product code and manufacturer. Check that this is the device you meant.
  3. Dates. Date of event, date the manufacturer received the information and date FDA received the report. These can be months or years apart.
  4. Problem codes. Device problems and patient problems, from FDA's coded terminology.
  5. Event description. The reporter's narrative. It is often short, and patient details are removed and marked "(b)(6)".
  6. Manufacturer narrative. For manufacturer reports, the firm's evaluation: whether the device was returned, what testing found and the conclusion. Trade secrets are marked "(b)(4)".
  7. Report source and summary flag. Whether the report came from a manufacturer, user facility, importer or a voluntary reporter, and whether it is a summary of several events.

Look out for follow-up reports. Manufacturers file supplements as they learn more, and the record shows the latest information.

Common mistakes

  • Ranking devices by report count. Counts depend on how many devices are in use, how long they have been sold, publicity and how thoroughly firms report. FDA says MDR data should not be used to estimate or compare rates.
  • Counting reports as patients. The same event can be reported by the hospital and the manufacturer, and a summary report can cover many events.
  • Reading the year received as the year of the event. Recalls and lawsuits can bring in reports of older events.
  • Relying on brand names alone. You will miss reports filed under other spellings. Use the product code and then filter.
  • Ignoring the narrative. Coded fields can be misleading on their own. Read a sample of full reports before drawing conclusions.

Taking it further

MAUDE is one piece of a device's safety record. Pair it with the recall history (medical device recalls explained) and the device's FDA decision, which you can find with how to check if a device is FDA cleared or approved. If a pattern in MAUDE looks important, read the underlying reports on FDA's site and look for published studies or registry data, which can measure rates in a way MAUDE cannot.

Frequently asked questions

How do I search the MAUDE database?
Use FDA's MAUDE search page, which has a simple keyword search by year and an advanced search by fields such as brand name, manufacturer, product code, event type and date received.
Why does MAUDE stop at 500 results?
FDA's online search returns at most 500 records per query, so larger searches need narrower filters, FDA's downloadable MDR data files or the openFDA API.
How far back does the MAUDE search go?
FDA's online search covers about the last ten years of reports and is updated monthly, while older reports are in the MDR data files.
What is the best way to search MAUDE for a device type?
Search by FDA product code rather than brand name, because brand names are typed by each reporter and vary widely, while the product code is the same for every device of that type.

For information only, not medical, legal or investment advice. Facts about FDA, CMS and SEC rules are current as of the publication date; check the linked official source for changes.

Keep reading