Guide
How FDA CVM Tracks Animal Adverse Events
The reporting process, voluntary versus mandatory submissions, limitations of the database, and what report volume trends actually tell you about drug safety.
Key Takeaway
The FDA's animal adverse event system is a signal-detection tool, not a comprehensive registry of every medication reaction in the US. Manufacturer reporting is mandatory; veterinarian and owner reporting is voluntary. This means the database likely captures a fraction of actual events. Despite this, the 1M+ reports in PlainBreed's database represent the most comprehensive public record of animal medication reactions available in the US, and the patterns it reveals have led to real label changes and safety communications.
The FDA CVM's Role in Veterinary Drug Safety
The U.S. Food and Drug Administration's Center for Veterinary Medicine (CVM) regulates drugs, vaccines, and medical devices intended for animals. Like its human medicine counterpart, the FDA does not simply approve a drug and move on, post-market surveillance is an ongoing responsibility. The adverse event reporting program is the primary mechanism through which the CVM monitors drug safety after a product enters the market.
Once a veterinary drug receives FDA approval, it enters a population of animals that may differ significantly from the controlled trial populations used for approval. Real-world patients have comorbidities, receive multiple drugs simultaneously, come from diverse genetic backgrounds, and are administered medications by pet owners under variable conditions. Adverse events that were statistically unlikely to appear in a trial population may surface as the drug scales to millions of animals.
This is precisely why the post-market surveillance system exists. The rankings section of PlainBreed shows which breeds appear most in adverse event reports, data that would not have been apparent in any pre-approval trial focused on drug efficacy.
Mandatory vs. Voluntary Reporting
The reporting system operates on a two-tier structure. Manufacturers are required by law to report adverse events they learn about, whether through their own monitoring, direct calls from veterinarians or pet owners, or scientific literature. This is called mandatory reporting. Failure to report is a federal violation that can result in enforcement action.
The second tier is voluntary. Veterinarians, pet owners, and animal caretakers can submit reports directly to the FDA, but are not required to do so. In practice, many pet owners who experience a medication reaction contact the drug manufacturer rather than the FDA, which means the manufacturer receives the report and has the obligation to file it. This pathway works reasonably well when pet owners call the number on the packaging.
The gap in voluntary reporting is widest for mild, self-resolving reactions. Vomiting that passes in a day is rarely reported; a death following a medication is much more likely to result in a report. This creates a selection bias in the database toward serious outcomes, which actually makes the serious-outcome data somewhat more reliable than mild-reaction data as a reflection of real-world experience.
What Gets Captured in Each Report
A complete adverse event report contains structured information across several categories:
- Animal details: Species, breed, age, sex, reproductive status, weight
- Drug information: Product name, active ingredient, manufacturer, lot number, dose, route of administration, duration
- Concomitant medications: Other drugs the animal was receiving simultaneously
- Adverse event details: Reaction type(s) observed, onset time after dosing, duration
- Outcome: Recovery (with or without treatment), ongoing at report, death, unknown
- Reporter assessment: Whether the reporter believed the drug was likely responsible
In practice, many reports, especially those filed directly by pet owners rather than veterinarians, are incomplete. Breed may be listed as "mixed" or unknown. Lot numbers are frequently omitted. The onset time may be imprecise. The FDA processes reports with available data and codes missing fields accordingly. PlainBreed's database reflects what the openFDA source actually contains, including these gaps.
How the FDA Detects Safety Signals
The CVM uses quantitative methods to distinguish statistical noise from meaningful patterns in the incoming report stream. The core approach is disproportionality analysis, comparing the observed frequency of a drug-reaction combination to what would be expected if all drugs and reactions were randomly associated. When a drug-reaction pair appears significantly more often than expected, that is a signal warranting investigation.
Signal detection is not automatic action. A detected signal triggers an investigation: reviewing the full case series, examining manufacturing records, consulting with the drug manufacturer, and sometimes requesting additional epidemiological data. Only after this process does the FDA decide whether regulatory action is warranted.
The isoxazoline class is a clear example of this process in action. The CVM detected a neurological reaction signal across multiple isoxazoline products in the years following their market introduction. The 2018 safety communication and required label update were the outcome of that investigation, not an immediate reflex to the first cluster of neurological reports. You can review individual isoxazoline drug profiles in the drugs directory.
Reporting Trends Over Time
Annual report volumes reflect a combination of real-world adverse events, reporting infrastructure improvements, and market forces. The table below shows approximate report volumes by period across the PlainBreed database.
| Period | Est. Reports | Context |
|---|---|---|
| 2010–2013 | 230,674 | Electronic Safety Reporting Portal adoption; early database growth |
| 2014–2017 | 290,258 | Isoxazoline class market expansion; FDA monitoring ramp-up |
| 2018–2021 | 274,655 | Isoxazoline 2018 neurological warning issued; COVID disruptions (2020–2021) |
| 2022–2025 | 253,328 | Post-COVID recovery; most recent full periods (database current) |
Figures computed from FDA openFDA Animal & Veterinary Adverse Events data (openFDA extract 2010–2025). Periods prior to 2010 are not available in this extract.
The growth from 2004 to 2019 reflects multiple factors: the launch of the electronic Safety Reporting Portal, increased public awareness of the database, growth in the pet pharmaceutical market, and the introduction of high-volume drugs like isoxazolines. The post-2020 dip partially reflects reduced veterinary visits during pandemic-related disruptions, which would naturally reduce both prescribing volume and the opportunities to observe and report adverse events.
Limitations You Should Know
The adverse event database has structural limitations that affect how any analysis, including PlainBreed's, should be interpreted:
- Underreporting: The true rate of adverse events in the field is unknown. Estimates suggest that for any given reaction, only a fraction of actual events are reported. This is especially true for mild reactions.
- No denominator: The database records adverse events but not total prescriptions. Without knowing how many animals received a drug, there is no way to calculate an adverse event rate, only a raw count.
- Reporter bias: Reports filed by owners versus veterinarians differ in completeness and clinical accuracy. The clinical assessment in a veterinarian-filed report is generally more reliable.
- Confounding medications: Many animals that experience adverse events are on multiple drugs. Attributing a reaction to one drug when the animal was receiving three is a significant analytical challenge that the database cannot resolve from report data alone.
- Association vs. causation: The database records that an event occurred while or after a drug was used. It does not, and cannot, confirm that the drug caused the event.
Using PlainBreed's Rankings to Spot Patterns
Despite its limitations, the adverse event database is a powerful tool when used to identify patterns rather than individual case claims. PlainBreed's rankings pages surface the breeds with the highest report volumes, highest death rates, and most frequently documented drug associations, giving you a population-level view that a single case report cannot provide.
For example, if a particular breed shows a death rate significantly above the median across all breeds in the database, and that pattern concentrates around a specific drug class, that is a meaningful signal. It may reflect a genetic predisposition, a dosing convention issue specific to that breed's typical weight range, or a reporting pattern where owners of that breed are more likely to report serious events. Each of these is worth investigating further with your veterinarian.
The breed directory and reaction directory on PlainBreed allow you to drill into these patterns at the level most relevant to your specific pet.
Frequently Asked Questions
What is the FDA CVM and what does it do with adverse event reports?
The FDA Center for Veterinary Medicine (CVM) is the division responsible for regulating veterinary drugs, devices, and food additives. It operates the adverse event reporting system for animal medications and uses incoming reports to conduct post-market safety surveillance. When the CVM identifies a safety signal, a pattern suggesting a drug may carry previously unknown risks, it can require label changes, issue public safety communications, or pursue further investigations with the manufacturer.
Is adverse event reporting mandatory for veterinarians and pet owners?
No. Reporting by veterinarians and pet owners is voluntary in the United States. Only drug manufacturers are legally required to report adverse events that come to their attention, including secondhand reports forwarded by veterinarians or owners who contacted the company. Because voluntary reporting is the dominant pathway, the FDA consistently acknowledges that the database underrepresents the true incidence of adverse events in the field.
How do I submit an adverse event report for my pet?
You have two main options. First, you can contact the drug manufacturer directly, the hotline number is on the product packaging. Manufacturers are required to file the report with the FDA on your behalf. Second, you can submit a report yourself through the FDA's Safety Reporting Portal at fda.gov. For complex or serious cases, your veterinarian can also submit a report. Providing as much detail as possible, drug name and lot number, dose, timeline, and specific signs observed, improves the quality of the report.
How long does the FDA take to act on adverse event signals?
The timeline varies significantly based on the severity and volume of incoming reports and the complexity of the analysis required. Minor pattern updates may result in label revisions within months. Large-scale investigations requiring manufacturer studies can take years. The FDA typically prioritizes signals involving serious outcomes (death, permanent disability) and situations where a pattern is strong and consistent. The database is reviewed on a rolling basis, not only at quarterly update intervals.
Can an adverse event report cause a drug to be pulled from the market?
It can contribute to a market withdrawal, but voluntary market withdrawal by the manufacturer or an FDA-directed recall is relatively rare. More commonly, adverse event signals lead to label updates, new contraindications, or restricted use indications. In the history of the CVM adverse event program, only a handful of veterinary drugs have been fully withdrawn specifically due to adverse event accumulation. Most interventions are corrective rather than prohibitive.
How is the FDA animal adverse event database different from VAERS or the human MedWatch system?
All three share the same fundamental limitation: voluntary reporting means underreporting is endemic, and reports represent association rather than causation. The animal system (openFDA Animal & Veterinary) is smaller in absolute terms than the human MedWatch database, partly because the animal health market is smaller and partly because veterinary reporting infrastructure is less developed. The animal database also includes species-specific variables (breed, weight class, species) that have no equivalent in human reporting systems.
Sources
- U.S. FDA openFDA, Animal & Veterinary Adverse Events API (api.fda.gov/animalandveterinary)
- FDA Center for Veterinary Medicine, Animal Drug Safety & Availability (fda.gov/animal-veterinary)
- FDA, Safety Reporting Portal (fda.gov/safety/reporting-serious-problems-fda)
- FDA, Adverse Event Reporting System (AERS): Background and Context (CVM guidance document)
- PlainBreed database, 1M+ reports, 506 breeds, 2,558 drugs (openFDA source, 2004–2025)
This content is for informational purposes only and does not constitute veterinary advice. Consult your veterinarian about your pet's medications and any reactions you observe.
The live report counts and ratios linked from this guide are rendered directly from FDA openFDA adverse event data. Worked examples and illustrative numbers cited in the guide text are for explaining the method, not drawn from this portal's live database. This page's data-linked figures draw directly on FDA openFDA data. See our editorial standards & corrections policy, the methodology behind these numbers, or report a data error.
Frequently asked questions
Where does this data come from?
Every figure derives from the U.S. FDA Center for Veterinary Medicine (CVM) animal drug adverse event reporting program, accessed through the openFDA animal & veterinary endpoint. We use no proprietary aggregators and no estimated data, see the methodology page for the exact source and extract date.
Do these reports prove a medication is dangerous?
No. Adverse event reports are voluntary and unverified. A report means only that someone observed a health event after a product was used, it does not establish that the product caused it. Always consult your veterinarian before starting, changing, or stopping any medication.
Why does one breed have far more reports than another?
Report volume is driven mostly by how common a breed is and how actively its owners and veterinarians file reports. A popular breed such as the Labrador Retriever sits near the top because the breed is enormous in number, not because it faces unusual risk.
How often is the data updated?
We refresh from the FDA CVM extract periodically. The current extract date is shown on the methodology page.