Devices › Stimulator, autonomic nerve, implanted for epilepsy
Device brand name · Generator
Pulse Gen Model 103: medical device reports filed with FDA
4,518 reports name it, 2008–2026. Manufacturer given most often on reports: Cyberonics. Product code LYJ.
- 4,518
- device reports naming the brand
- 0% of all MAUDE reports · about 244 a year
- 34
- reports, 12 months to August 2026
- 40 in the 12 months before
- 29.8%
- classified as malfunction
- 58.7% across the product code
- 8.5%
- classified as death, as reported
- 384 reports · not verified by FDA
4,518 medical device reports received by FDA name the brand "Pulse Gen Model 103" (generator); the manufacturer given most often on reports is Cyberonics; received from January 2008 to July 2026. Brand names are as the reporter or manufacturer wrote them, so one product can appear under several spellings.
34 reports arrived in the 12 months to August 2026, down 15% from 40 in the 12 months before. FDA releases MAUDE data monthly; reports can arrive months after an event, so the newest months grow in later releases. The event types recorded are injury (61.7%), malfunction (29.8%) and death (8.5%); across all stimulator, autonomic nerve, implanted for epilepsy reports (product code LYJ) death is recorded in 5.6% and malfunction in 58.7%. Event type is the reporter's classification of the report, not a verified outcome.
The device problems coded most often are insufficient device problem information (23%), device operates differently than expected (7.4%) and battery problem (6.9%). The patient problems coded most often are seizures, therapeutic response, decreased and death. FDA states that "MDR data alone cannot be used to establish rates of events, evaluate a change in event rates over time or compare event rates between devices. The number of reports cannot be interpreted or used in isolation to reach conclusions about the existence, severity, or frequency of problems associated with devices." It adds that the reporting system "has limitations, including the potential submission of incomplete, inaccurate, untimely, unverified, or biased data."
Reports by month, five years
Reports by the month FDA received them. FDA releases MAUDE data monthly; reports can arrive months after an event, so the newest months grow in later releases.
By year received
Event type and report source
Event type, as classified on the report
Who filed the report
Grey bar: all 52,004 reports carrying product code LYJ. Event type is the filer's classification and is not verified by FDA.
Device problems coded
| Device problem (FDA code) | Reports | Share | Share across product code |
|---|---|---|---|
| Insufficient device problem information | 1,037 | 23% | 9% |
| Device operates differently than expectedthe device behaved unexpectedly | 334 | 7.4% | 4.5% |
| Battery problema problem with the battery | 311 | 6.9% | 3.4% |
| High impedance | 278 | 6.2% | 18.1% |
| Device displays incorrect message | 111 | 2.5% | 2.1% |
| Failure to interrogate | 99 | 2.2% | 2.5% |
| Migration or expulsion of device | 85 | 1.9% | 1.1% |
| Device inoperablethe device could not be used | 84 | 1.9% | 1.5% |
| Premature discharge of battery | 80 | 1.8% | 1.3% |
| Improper or incorrect procedure or method | 69 | 1.5% | 0.8% |
| Unexpected therapeutic results | 54 | 1.2% | 0.4% |
| Low battery | 45 | 1% | 0.4% |
| Premature end-of-life indicator | 43 | 1% | 1.3% |
| Use of incorrect control/treatment settings | 42 | 0.9% | 2.3% |
| Communication or transmission problemthe device could not communicate or transmit | 40 | 0.9% | 1.6% |
| Extrusion | 32 | 0.7% | 0.3% |
| Programming issue | 27 | 0.6% | 1.3% |
| Loose or intermittent connection | 26 | 0.6% | 0.6% |
| Human-device interface problem | 25 | 0.6% | 0.3% |
| Device remains implanted | 23 | 0.5% | 4.9% |
FDA's device problem codes as entered on the report; one report can carry several. Generic codes such as "no apparent adverse event" are left out.
Patient problems coded
| Patient problem (FDA code) | Reports | Share |
|---|---|---|
| Seizures | 1,497 | 33.1% |
| Therapeutic response, decreased | 608 | 13.5% |
| Deaththe patient died; cause not stated by this term | 363 | 8% |
| Painpain, site not specified | 314 | 6.9% |
| Convulsion, clonic | 159 | 3.5% |
| Apnea | 104 | 2.3% |
| Chest painchest pain | 81 | 1.8% |
| Bradycardiaslow heart rate | 80 | 1.8% |
| Dyspneashortness of breath | 79 | 1.7% |
| Bacterial infectiona bacterial infection | 66 | 1.5% |
| Discomfortdiscomfort | 66 | 1.5% |
| Neck painneck pain | 63 | 1.4% |
| Arrhythmia | 58 | 1.3% |
| Swellingswelling | 58 | 1.3% |
| Falla fall | 57 | 1.3% |
Patient problem codes are present mainly in reports filed since 2019 and are often blank.
In context
| Measure | Pulse Gen Model 103 | Product code LYJ | All MAUDE reports |
|---|---|---|---|
| Reports | 4,518 | 52,004 | 26,136,888 |
| Share of that pool | — | 8.7% | 0% |
| Classified as death, per 1,000 reports | 85 | 56 | 9 |
| Classified as injury | 61.7% | 35.5% | 36.2% |
| Classified as malfunction | 29.8% | 58.7% | 62.3% |
| Filed by the manufacturer | 100% | 99.5% | 94.9% |
Shares are of reports, not of devices in use. FDA states that "MDR data alone cannot be used to establish rates of events, evaluate a change in event rates over time or compare event rates between devices. The number of reports cannot be interpreted or used in isolation to reach conclusions about the existence, severity, or frequency of problems associated with devices." It adds that the reporting system "has limitations, including the potential submission of incomplete, inaccurate, untimely, unverified, or biased data."
Other brands with product code LYJ
| Brand | Reports | Latest 12 months | Classified as death |
|---|---|---|---|
| Programming Software | 7,453 | 1 | 0% |
| Lead Model 302 | 7,259 | 136 | 0% |
| Pulse Gen Model 102 | 5,903 | 14 | 12.2% |
| Pulse Gen Model 106 | 4,023 | 144 | 5.1% |
| Pulse Gen Model 1000 | 3,601 | 549 | 5.7% |
| Lead Model 304 | 3,329 | 272 | 0.1% |
| Bipol Lead Model 300 | 1,955 | 12 | 0.1% |
| Pulse Gen Model Unk | 1,914 | 7 | 13.6% |
| Ncp Pulse Generator | 1,272 | 0 | 30% |
| Lead Model Unknown | 1,253 | 46 | 1.6% |
| Pulse Gen Model 101 | 1,210 | 0 | 12.6% |
| Pulse Gen Model 105 | 1,191 | 7 | 6.7% |
| Vns Therapy Pulse | 1,064 | 0 | 10.7% |
| Pulse Gen Model 102r | 985 | 0 | 10.9% |
| Pulse Gen Model 104 | 864 | 4 | 9% |
Ordered by report count for navigation only. FDA states that "MDR data alone cannot be used to establish rates of events, evaluate a change in event rates over time or compare event rates between devices. The number of reports cannot be interpreted or used in isolation to reach conclusions about the existence, severity, or frequency of problems associated with devices." It adds that the reporting system "has limitations, including the potential submission of incomplete, inaccurate, untimely, unverified, or biased data." FDA also states that "common products may have a higher number of adverse events due to the higher total number of people using the product."
Questions about Pulse Gen Model 103 reports
How many FDA reports name Pulse Gen Model 103?
4,518 medical device reports in the openFDA MAUDE release of 28 Sep 2026, covering reports received through August 2026; 34 in the latest 12 months.
What kinds of events are reported?
injury (61.7%), malfunction (29.8%) and death (8.5%). "Event type" is how the report was classified when filed: death, injury, malfunction or other.
What device problems are coded?
insufficient device problem information (23%), device operates differently than expected (7.4%), battery problem (6.9%), high impedance (6.2%) and device displays incorrect message (2.5%). These are FDA's device problem codes as entered on the report.
Who files these reports?
manufacturer reports (100%). Manufacturers and user facilities are required to report; voluntary reports come from health professionals and the public.
Do these reports show that Pulse Gen Model 103 was at fault?
No. FDA states that "MDR data alone cannot be used to establish rates of events, evaluate a change in event rates over time or compare event rates between devices. The number of reports cannot be interpreted or used in isolation to reach conclusions about the existence, severity, or frequency of problems associated with devices." It adds that the reporting system "has limitations, including the potential submission of incomplete, inaccurate, untimely, unverified, or biased data." FDA adds that "a causal relationship cannot be established between product and reactions listed in a report."
How do I report a problem with a medical device?
Through FDA MedWatch (Form FDA 3500 for health professionals and consumers), or to the manufacturer. See the guide on how to report.
Where this comes from. Counts are built from the openFDA Device Adverse Event (MAUDE) bulk export released 28 Sep 2026, which holds reports FDA received through 31 Aug 2026. Brand, generic and manufacturer names are as written on the reports; product code names come from FDA's device classification. A report counts here when any device it lists carries this brand name after punctuation and capitalisation are normalised. No report narrative, lot number, patient detail or reporter identity is published. See the methodology and sources; FDA's record is authoritative, and corrections are handled within five working days.
Drug data (FAERS) through 30 Jun 2026; device data (MAUDE) through 31 Aug 2026. Not medical advice.