How do you read an adverse event report count?
To read a report count without being misled:
- Treat the count as a numerator. The FDA data records reports filed, never doses given, so no count on this page is a rate.
- Ask who files the report. Owners and veterinarians report by choice; drug makers must pass on what they hear. A watched product draws more reports.
- Separate "did not work" from harm. In dogs 38% of reports say a product failed to work, which the FDA files as an adverse event.
- Ask who gets the medicine. A drug for heart failure or old-age arthritis is given to animals that are already ill, and their deaths appear in its reports.
- Compare a medicine with the all-reports figure for the species, not with another medicine, and take the question to a veterinarian.
What can FDA adverse event counts tell you, and what can they not?
FDA adverse event counts tell you what was reported for a medicine and how those reports break down. They cannot tell you how often an event happens, because the number of animals treated is not recorded.
| Question | Where the answer is | What that means |
|---|---|---|
| How many reports name this medicine? | Counted | A report counts under every active ingredient it names. |
| How many of those are marked serious? | Counted | The flag is missing from about a third of all reports, so this share runs low. |
| How many record a death or euthanasia? | Counted | An outcome in the report, with no finding of cause. |
| Which clinical signs are reported most? | Counted | The 25 most frequent signs per medicine are in the download. |
| How many doses were given? | Not in the data | Without it there is no rate and no safety ranking. |
| Did the medicine cause the event? | Not in the data | A report is a suspicion. The FDA says so in openFDA's guide to reading the results. |
| How many events were never reported? | Not in the data | Reporting by owners and veterinarians is voluntary. |
How does a side effect become a number in the FDA data?
A side effect becomes a number when an owner, a veterinarian or a drug maker files a report with the FDA Center for Veterinary Medicine, which publishes the reports through the openFDA animal and veterinary adverse event interface. This dataset then counts reports for each active ingredient.
768 counts were read from the FDA interface on 11 November 2026, and each one is saved beside the query that produced it. The list of medicines and the measures were fixed before the capture ran.
Why is a report count not a side effect rate?
A report count is not a side effect rate because a rate needs two numbers and the FDA data holds one. The count of reports is recorded; the count of doses given is not.
The same gap applies to every medicine on this page. A product sold for twenty years to millions of dogs will show more reports than one launched last year, whatever their safety.
What do the reports show for flea and tick chews?
Reports for the four isoxazoline flea and tick chews show a seizure or convulsion more often than dog reports overall, and between 32% and 49% of them say the product did not work.
What the difference can mean. Among reports for these four chews, a seizure or convulsion appears in 5.1% (Fluralaner) to 11% (Sarolaner), against 4.4% of all dog reports. That fits the FDA's fact sheet on this drug class, which says neurologic events have been reported. It does not show which chew is safer: the products differ in how long they have been sold, what they are combined with and who reports.
What do the reports show for dog arthritis medicines?
Reports naming bedinvetmab, the arthritis injection sold as Librela, are marked serious in 62% of cases, against 19% for carprofen and 29% for dog reports overall.
What the difference can mean. Bedinvetmab reports record a death or euthanasia in 20% of cases and carprofen reports in 18%. The FDA wrote to veterinarians about the reports it had received (FDA's letter to veterinarians about adverse events reported in dogs treated with Librela). The gap cannot be read as a difference in risk: Librela is new, given in clinics to old dogs, and closely watched, while carprofen has 47,327 reports gathered since 1997.
What do the reports show for dog itch medicines?
Reports naming oclacitinib, the active ingredient in Apoquel, are marked serious in 30% of cases, and 29% say the medicine did not work.
What the difference can mean. Oclacitinib has 28,973 reports and cyclosporine 15,092. A share that sits near the all-reports tick is the ordinary case; a share far from it is a reason to ask a veterinarian why, not an answer.
What do the reports show for cat medicines?
Cat reports record a death or euthanasia more often than dog reports, 9.4% against 5.4%, and far fewer say a product did not work, 19% against 38%.
What the difference can mean. Frunevetmab reports are marked serious in 58% of cases and meloxicam reports in 23%. Both are given to older cats, and the same limits apply as for dogs: no count of doses, and a report is a suspicion.
How does a report share get misread?
A report share gets misread when it is treated as the chance of an outcome. The clearest case in this data is Furosemide, which has the highest share of dog reports recording a death or euthanasia: 26% of 1,525 reports.
The misreading
"Furosemide kills 26% of the dogs that take it."
The fix
26% of the reports naming Furosemide record a death. Furosemide is a diuretic given to dogs in heart failure, so many of the dogs in those reports were gravely ill before the first dose, and a report naming several medicines counts under each. The number of dogs treated is unknown.
The most reported signs in Furosemide reports are general ones, and a report often names other medicines too, so a report alone cannot separate the illness, the other medicines and Furosemide.
The six most reported signs among 1,525 dog reports naming Furosemide. A report can list several signs.
Why do so many reports say a product did not work?
So many reports say a product did not work because the FDA files a suspected lack of effect as an adverse event, and parasite preventives are the most widely used pet medicines. Ivermectin has the highest share: 69% of 168,964 dog reports.
Dog medicines with at least 20,000 reports, ranked by the number saying the product did not work (hatched). Combination products count under each ingredient.
A heartworm preventive failing is a serious matter for the dog. The FDA facts about heartworm disease and the American Heartworm Society heartworm basics explain testing and prevention; a failure report is a reason to test, not proof the product was at fault.
How many dog reports does the FDA receive each year?
The FDA received 54,119 dog reports in 2025, and the peak year in the published record is 2017.
What do the reports show for every dog medicine profiled?
The table lists the 44 dog medicines with at least 1,000 reports, most reported first, with the all-reports row at the top for comparison.
| Active ingredient | Reports | Marked serious | Death or euthanasia | Did not work | Seizure | First year | Peak year |
|---|---|---|---|---|---|---|---|
| All dog reports | 1,001,380 | 29% | 5.4% | 38% | 4.4% | 1987 | 2017 |
| Pyrantel | 236,005 | 38% | 2.3% | 63% | 3.9% | 1987 | 2025 |
| Spinosad | 185,702 | 10% | 2.7% | 22% | 3.5% | 2007 | 2013 |
| Milbemycin Oxime | 179,859 | 14% | 2.7% | 32% | 3.4% | 2002 | 2013 |
| Ivermectin | 168,964 | 34% | 2.3% | 69% | 2.5% | 1987 | 2017 |
| Moxidectin | 76,891 | 47% | 4.3% | 43% | 4.3% | 1997 | 2025 |
| Sarolaner | 70,962 | 51% | 2.1% | 49% | 11% | 2016 | 2024 |
| Fluralaner | 68,773 | 56% | 3.2% | 40% | 5.1% | 2012 | 2017 |
| Afoxolaner | 65,969 | 31% | 3.1% | 37% | 7.2% | 2013 | 2016 |
| Praziquantel | 59,564 | 16% | 3.0% | 47% | 2.8% | 1987 | 2019 |
| Carprofen | 47,327 | 19% | 18% | 6.0% | 5.6% | 1997 | 2000 |
| Lufenuron | 36,856 | 27% | 2.1% | 61% | 3.4% | 1995 | 2016 |
| Selamectin | 29,523 | 7.0% | 2.7% | 58% | 4.3% | 1999 | 2000 |
| Oclacitinib | 28,973 | 30% | 5.0% | 29% | 4.3% | 2013 | 2017 |
| Melarsomine | 22,745 | 41% | 5.8% | 33% | 2.0% | 1995 | 2017 |
| Bedinvetmab | 19,619 | 62% | 20% | 12% | 9.2% | 2023 | 2024 |
| Imidacloprid | 17,831 | 47% | 3.9% | 39% | 4.1% | 2006 | 2019 |
| Cyclosporine | 15,092 | 25% | 4.9% | 7.3% | 3.3% | 1995 | 2012 |
| Lotilaner | 12,797 | 14% | 2.7% | 32% | 6.3% | 2015 | 2025 |
| Trilostane | 11,925 | 46% | 11% | 15% | 3.1% | 2004 | 2019 |
| Maropitant | 11,886 | 51% | 22% | 18% | 6.1% | 2007 | 2025 |
| Deracoxib | 11,012 | 15% | 21% | 4.0% | 4.1% | 2002 | 2004 |
| Nitenpyram | 10,786 | 10.0% | 3.4% | 59% | 3.1% | 2001 | 2017 |
| Gabapentin | 9,028 | 57% | 17% | 17% | 7.0% | 2011 | 2024 |
| Prednisone | 8,848 | 53% | 13% | 22% | 4.3% | 1988 | 2025 |
| Meloxicam | 8,083 | 30% | 16% | 6.2% | 5.3% | 2002 | 2004 |
| Firocoxib | 8,009 | 26% | 19% | 6.0% | 4.7% | 2005 | 2007 |
| Grapiprant | 7,776 | 30% | 10.0% | 17% | 3.9% | 2013 | 2017 |
| Amoxicillin | 6,749 | 35% | 17% | 20% | 6.0% | 1987 | 2025 |
| Febantel | 6,733 | 38% | 3.5% | 23% | 1.6% | 1988 | 2018 |
| Fipronil | 6,219 | 47% | 3.1% | 28% | 6.1% | 2010 | 2013 |
| Pimobendan | 5,696 | 33% | 14% | 6.7% | 5.6% | 2003 | 2023 |
| Clavulanate | 4,777 | 28% | 16% | 22% | 5.8% | 1987 | 2025 |
| Diphenhydramine | 4,727 | 48% | 9.8% | 17% | 4.1% | 1999 | 2013 |
| Enrofloxacin | 4,084 | 55% | 19% | 11% | 9.5% | 1989 | 2024 |
| Doxycycline | 4,015 | 58% | 13% | 46% | 4.6% | 1997 | 2025 |
| Tramadol | 3,769 | 47% | 17% | 15% | 5.4% | 2006 | 2017 |
| Metronidazole | 3,094 | 55% | 20% | 18% | 5.6% | 1998 | 2025 |
| Levothyroxine | 3,014 | 44% | 8.4% | 16% | 6.8% | 1998 | 2012 |
| Cephalexin | 2,680 | 39% | 10% | 15% | 5.5% | 1999 | 2013 |
| Famotidine | 2,644 | 55% | 24% | 16% | 4.9% | 2006 | 2016 |
| Fluoxetine | 2,588 | 23% | 4.3% | 22% | 8.1% | 2004 | 2025 |
| Trazodone | 1,702 | 54% | 11% | 18% | 7.4% | 2011 | 2025 |
| Enalapril | 1,651 | 45% | 16% | 12% | 8.8% | 1994 | 2024 |
| Furosemide | 1,525 | 56% | 26% | 13% | 10% | 1989 | 2025 |
What do the reports show for every cat medicine profiled?
The table lists the 31 cat medicines with at least 1,000 reports.
| Active ingredient | Reports | Marked serious | Death or euthanasia | Did not work | Seizure | First year | Peak year |
|---|---|---|---|---|---|---|---|
| All cat reports | 151,164 | 27% | 9.4% | 19% | 2.6% | 1987 | 2024 |
| Selamectin | 30,168 | 15% | 4.9% | 21% | 2.7% | 1999 | 2019 |
| Fluralaner | 13,791 | 51% | 3.7% | 35% | 1.9% | 2014 | 2024 |
| Praziquantel | 12,556 | 33% | 6.2% | 15% | 2.5% | 1987 | 2024 |
| Spinosad | 11,875 | 6.8% | 3.0% | 16% | 0.9% | 2007 | 2013 |
| Moxidectin | 11,447 | 48% | 4.6% | 19% | 1.9% | 1998 | 2020 |
| Imidacloprid | 9,231 | 50% | 4.9% | 19% | 1.9% | 2006 | 2017 |
| Nitenpyram | 8,753 | 11% | 4.9% | 27% | 2.1% | 2001 | 2017 |
| Cefovecin | 8,686 | 42% | 22% | 15% | 6.0% | 2008 | 2024 |
| Sarolaner | 8,530 | 34% | 2.6% | 19% | 4.4% | 2016 | 2021 |
| Frunevetmab | 8,036 | 58% | 11% | 13% | 2.8% | 2022 | 2024 |
| Emodepside | 5,732 | 48% | 4.9% | 9.7% | 1.7% | 2007 | 2018 |
| Buprenorphine | 5,611 | 34% | 14% | 4.6% | 2.9% | 2006 | 2023 |
| Velagliflozin | 4,798 | 16% | 8.0% | 26% | 0.4% | 2023 | 2025 |
| Bexagliflozin | 4,053 | 22% | 7.8% | 15% | 0.5% | 2023 | 2024 |
| Ketamine | 4,012 | 29% | 26% | 20% | 5.5% | 1987 | 2023 |
| Eprinomectin | 3,856 | 14% | 7.7% | 20% | 4.3% | 1998 | 2024 |
| Maropitant | 3,847 | 51% | 22% | 13% | 3.7% | 2007 | 2025 |
| Meloxicam | 3,654 | 23% | 24% | 2.5% | 2.8% | 2003 | 2009 |
| Robenacoxib | 3,276 | 48% | 17% | 11% | 3.5% | 2011 | 2024 |
| Esafoxolaner | 3,248 | 15% | 7.9% | 21% | 4.4% | 2023 | 2024 |
| Cyclosporine | 3,030 | 36% | 7.1% | 13% | 1.1% | 1997 | 2012 |
| Amoxicillin | 2,970 | 25% | 17% | 16% | 4.1% | 1987 | 2025 |
| Dexmedetomidine | 2,862 | 42% | 18% | 17% | 4.2% | 2007 | 2023 |
| Clavulanate | 2,155 | 20% | 16% | 18% | 3.9% | 1987 | 2009 |
| Pyrantel | 2,023 | 37% | 10% | 11% | 3.1% | 1995 | 2016 |
| Gabapentin | 1,956 | 53% | 14% | 13% | 3.4% | 2012 | 2025 |
| Mirtazapine | 1,878 | 46% | 21% | 16% | 3.5% | 2008 | 2024 |
| Carprofen | 1,711 | 3.7% | 9.6% | 0.6% | 1.1% | 1997 | 2000 |
| Enrofloxacin | 1,644 | 35% | 16% | 3.4% | 8.9% | 1991 | 2001 |
| Methimazole | 1,250 | 47% | 13% | 14% | 4.4% | 1998 | 2025 |
| Ivermectin | 1,095 | 19% | 11% | 8.4% | 4.6% | 1987 | 2013 |
To check what a product contains, use Animal Drugs @ FDA, the approved product database. Current FDA notices are on FDA animal drug product safety information. Europe keeps a separate system, the European Medicines Agency's veterinary reporting system.
Where can you download the data?
The data is free to download and reuse under CC BY 4.0. Each file is listed with its hash.
| File and hash | What it holds | Size |
|---|---|---|
profiles.csvsha256 7999f4af8458 | One row per species and ingredient: reports, and the count and share for each measure | 11 KB |
reactions.csvsha256 9055366d2e59 | The 25 most reported clinical signs for each species and ingredient | 117 KB |
reports-by-year.csvsha256 f9d9da03674b | Reports per year the FDA first received them | 52 KB |
data.jsonsha256 f0d6f31eb52f | Headline figures, the capture hash and the checks the build ran | 5.1 KB |
captures.jsonlsha256 2ce2f4816526 | Every query sent to the FDA interface, with the number it returned and when | 319 KB |
protocol.jsonsha256 0b288b161bd8 | The ingredient rule, the measures and the counting rule, written before the capture | 2.4 KB |
Cite it: Couey, V. W. (2026). Veterinary adverse event report profiles 2026. PetMaxxing. https://petmaxxing.com/research/vet-adverse-event-report-profiles-2026/ (FDA data read 11 November 2026). https://doi.org/10.5281/zenodo.24144162.
Compiled and checked by Vincent Wesley Couey (ORCID 0009-0005-6869-308X). Use of the FDA data follows the openFDA terms of service. This study pairs with Pet Prescription Drug Pricing 2026, which prices 30 of these medicines. To look up one medicine by brand or ingredient, use the Pet Medicine Report Lookup, which adds the FDA approval finding for each species.
How was the dataset built, and what are its limits?
The dataset was built from 768 count queries against the FDA's public interface on 11 November 2026. The 30 most-reported ingredient words in dog reports and the 25 most-reported in cat reports (ingredient strings split on ; , / + and the word and; salt and dosage-form words dropped; counts summed), plus every ingredient in the Pet Prescription Drug Pricing 2026 dataset, each queried for both species.
- Counting rule. A report counts for an ingredient when any drug in the report names it. A combination product counts under each ingredient; a report naming two drugs counts under both. Every figure is a count of reports received by the FDA, not of animals, doses or treated patients.
- Checks. No part exceeds its whole in any of 96 rows. Reports per year add up to the report total in 81 rows; the other 15 fall short by a handful of reports that carry no receive date.
- Seen before the method was fixed. While testing that the queries work, one drug was read in full: bedinvetmab in dogs (reports, serious, died or euthanized, lack of efficacy, seizure, not used per label, top reactions). The all-species totals for outcome, seriousness, label use, top reactions and reports per year were also read. No other ingredient figure was seen.
- Serious flag. The flag is missing from about a third of all reports, and shares here divide by all reports, so they understate the share among reports that carry it.
- Matching. A medicine is matched by ingredient name anywhere in the report's drug list. Spellings that do not contain the name are missed.
- One day. The FDA adds and revises reports. These figures are the source as it stood on 5 October 2026.
Not claimed.
- A rate of adverse events: no count of doses sold or animals treated exists in the source.
- That a drug caused the event: a report is a suspicion filed by a vet, an owner or a manufacturer.
- A safety ranking: heavily sold and heavily watched products draw more reports.
- That share figures are comparable across drugs: products differ in who reports and why (a preventive draws lack-of-efficacy reports, an injectable given in a clinic draws vet reports).
Found a wrong figure? Tell us; we correct it and date the correction.
FAQ
How many adverse event reports has the FDA received for dog and cat medicines?
The FDA's published animal drug adverse event reports held 1,001,380 reports for dogs and 151,164 for cats when this dataset was captured on 11 November 2026 (source last updated 5 October 2026). 29% of dog reports and 27% of cat reports are marked serious.
Does a high number of reports mean a pet medicine is dangerous?
No. A report count is not a rate. The FDA data records how many reports were filed, not how many doses were given or how many animals were treated, so a widely used medicine draws more reports than a rarely used one whatever its safety. A report is also a suspicion filed by an owner, a veterinarian or a manufacturer, not a finding that the medicine caused the event.
Which dog medicines have the most adverse event reports?
By count of reports naming the active ingredient, the five most reported in dogs are Pyrantel (236,005), Spinosad (185,702), Milbemycin Oxime (179,859), Ivermectin (168,964), Moxidectin (76,891). These are among the most widely used parasite preventives, and 38% of all dog reports are complaints that a product did not work.
What does "did not work" mean in an adverse event report?
The FDA files a suspected lack of effect as an adverse event. In dogs 38% of all reports carry a lack-of-efficacy sign, for example a dog that tested positive for heartworm while on a preventive. For Ivermectin the share is 69%.
What does the data show for Librela (bedinvetmab)?
19,619 dog reports name bedinvetmab, the active ingredient in Librela. 62% are marked serious and 20% record a death or euthanasia. Librela is an arthritis injection given mostly to older dogs, and the data holds no count of injections given, so these shares describe the reports and are not a risk estimate.
Can I use this to decide whether to give my pet a medicine?
No. The dataset cannot tell you how likely a side effect is. Use it to prepare questions, then talk to a veterinarian who knows your animal. If your pet has a reaction to a medicine, contact your veterinarian, and you can report it to the FDA.
Can I reuse the data?
Yes. The compilation is published under CC BY 4.0: reuse is free with attribution to Vincent Wesley Couey and a link to this page. The underlying reports are published by the FDA through openFDA.
Does it have a DOI?
Yes. Version 1 is https://doi.org/10.5281/zenodo.24144162. The DOI for all versions is https://doi.org/10.5281/zenodo.24144161.
Bottom line
The FDA's adverse event reports are the largest public record of what goes wrong, or is thought to go wrong, with pet medicines: 1,152,544 reports for dogs and cats. Read as counts, they show what owners and veterinarians report and how the reports for one medicine differ from the rest. Read as rates, they mislead. Bring the numbers to your veterinarian as questions.