Filler Complications: Reading the FDA Reports, Not Just Counting Them
Reviewed for medical accuracy against our Medical Review Board · Last updated 23 August 2026
Part of Dermal Fillers
The FDA holds a public database of adverse events reported for medical devices, and fillers are devices. It contains 9,792 reports for Juvederm alone. We pulled the figures and then did the part nobody does: read the reports. That matters, because filtering for deaths returns eight results and reading them shows they do not support the claim that number implies. One of the eight involved a substance that was not Juvederm at all.
What follows is the real picture: which complications are genuinely documented, which counts fall apart on inspection, and the warning signs that are worth memorising.
1. Read this before any number
Everything below comes from MAUDE, the FDA's device adverse event database. It is genuinely useful and it is easy to misuse, so here is what it is not:
- These are report counts, not rates. There is no denominator anywhere in this data. We do not know how many treatments were performed, so no percentage can be calculated.
- More reports does not mean more dangerous. Juvederm leads every count in this article because it is the most widely used filler in the United States. That is the likeliest explanation, not a safety ranking.
- Reporting is voluntary and incomplete. Both under-reporting and duplicate reporting happen, and we found duplicates.
- A report does not establish causation. Anyone can file one. Section 2 shows exactly how much that matters.
Used properly, this data tells you which kinds of things go wrong and what the warning signs are. It cannot tell you your odds.
2. Why counting is not enough: the death reports
Filter the Juvederm reports by event type "Death" and you get eight results. That number, published on its own, would be alarming and wrong. We read all eight. Here is what they actually contain:
- One describes a patient injected in the cheeks with no issues who died the following day. The report states the event was "deemed not device related".
- One describes a doctor injecting "unknown substances from a non identified clear jar" while telling the patient it was Juvederm, followed by fatal anaphylaxis. Whatever was injected, the report itself says it was not identified as the product.
- One describes a patient who died in their sleep the night of a cheek injection, where further investigation with the injecting facility and the coroner's office found another explanation.
- Two are duplicate reports for the same product with no clinical detail. Information was requested and never supplied.
- One came from an anonymous caller who left no contact details, so it could not be followed up.
- One is a bare administrative response to an FDA inquiry letter, containing no event detail at all.
So the count is eight and the evidence supports none of them as a filler-caused death. That is not us defending the industry, it is what the reports say when you open them.
We are including this because it is the most useful thing on the page. Somebody will find that count and publish "8 deaths linked to Juvederm", and now you know what sits behind it. It is also the reason to treat every other number here as a starting point for reading rather than a conclusion.
3. Vascular occlusion, the one that matters most
This is the complication that justifies the whole subject being taken seriously, and unlike the death counts the data here is coherent and consistent with a recognised mechanism. Filler enters or compresses a blood vessel and the tissue downstream loses its blood supply.
| Coded problem | Reports |
|---|---|
| Obstruction or occlusion | 695 |
| Vascular system, circulation impaired | 322 |
| Necrosis | 261 |
Necrosis means tissue death. 261 reports describe it. The other filler brands show the same signature at smaller volumes: Radiesse has obstruction or occlusion as its second most reported problem at 267 of 1,641 reports, and Belotero shows necrosis in 37 of 321.
The reason this is survivable is that for a hyaluronic acid filler there is an antidote. Hyaluronidase breaks the gel down, and given promptly it can restore circulation. That is the single strongest argument for choosing a dissolvable product, covered in our filler guide.
4. Vision loss, and the case that recovered
Filler reaching an artery connected to the eye can cause blindness. This is the most feared complication and the data on it holds up:
| Coded problem | Reports |
|---|---|
| Blurred vision | 53 |
| Visual impairment | 46 |
| Visual disturbances | 40 |
| Loss of vision | 34 |
Unlike the death reports, these are internally consistent. The loss of vision reports are repeatedly co-coded with obstruction or occlusion and with impaired circulation, which is exactly the mechanism you would expect. They describe a recognised event rather than an unexplained one.
Two of them cite a peer-reviewed case report, and it is worth knowing about because of what it documents: Danks, Dalgliesh and Ayton, "Cosmetic Filler Blindness: Recovery After Repeated Hyaluronidase Injections", Aesthetic Surgery Journal, 2021. Recovery. Vision loss from filler is not automatically permanent if the response is fast and hyaluronidase is available.
Which turns an abstract fear into a practical checklist. Any vision change after filler is an emergency, the treatment exists, and whether your clinic has it on site is a question you can ask before anything is injected.
5. Nodules and granulomas
Less dangerous, far more common, and the reason many people end up unhappy rather than harmed.
| Coded problem | Reports |
|---|---|
| Nodule | 1,099 |
| Subcutaneous nodule | 306 |
| Granuloma | 272 |
| Abscess | 646 |
| Unspecified infection | 1,199 |
Worth separating two different things. A lump in the first few weeks is extremely common and usually just tissue settling: the FDA trial data on our filler recovery page found 81.1% of patients reported lumps or bumps, with 15.8% of those lasting beyond thirty days. A firm nodule appearing months later is a different event, sometimes triggered by an illness or vaccination, and is what the granuloma code describes.
The infection and abscess counts are a reminder that this is a sterile procedure being performed on skin, and that where it is performed matters.
6. The reports that blame the technique
MAUDE codes a device problem as well as a patient problem, and this is where it gets uncomfortable for the industry:
| Coded device problem | Reports |
|---|---|
| Adverse event without identified device or use problem | 6,968 |
| Improper or incorrect procedure or method | 952 |
| Gel leak | 272 |
| Off-label use | 304 |
| Migration or expulsion of device | 102 |
| Migration | 46 |
| Device remains implanted | 61 |
952 reports coded as improper or incorrect procedure or method. That is technique, not product. And 304 coded as off-label use, meaning the filler was placed somewhere its approval does not cover.
Set against 6,968 reports where no device or use problem was identified at all, the proportion is modest. But it is the clearest available evidence for something this site says constantly: who injects you matters more than which brand they use. Checking a licence is free and takes two minutes, and our guide to verifying one walks through it.
7. Patterns by product
Report totals track how much a product is used, so the totals themselves say little. What is informative is the shape of each product's reports, which tracks what the material is:
| Product | Reports | Signature problem |
|---|---|---|
| Juvederm | 9,792 | Oedema, skin irritation, inflammation |
| Versa | 2,461 | Largely no known impact, then pain |
| Sculptra | 2,009 | Granuloma, 258 |
| Radiesse | 1,641 | Obstruction or occlusion, 267 |
| Restylane | 1,217 | Swelling, pain, erythema |
| RHA | 588 | Foreign body embolism, 183 |
| Belotero | 321 | Necrosis, 37 |
| Revanesse | 209 | Swelling, nodule |
| Bellafill | 148 | Pain, swelling |
Sculptra works by provoking a collagen response, and granuloma is its second most reported problem. Radiesse is a firm particulate gel and shows vascular obstruction prominently. RHA's single largest reported problem is foreign body embolism, 183 of 588 reports. Those are coherent with the materials, which is why they are worth raising when choosing a product for a particular area rather than treating all fillers as interchangeable.
8. The warning signs
Learn these. They are the practical output of everything above:
- Severe or escalating pain, out of proportion to the procedure
- Blanching, meaning white or pale patches appearing on the skin
- Mottled, dusky, grey or bruised-looking skin in a pattern rather than a spot
- Any change in vision, including blurring or double vision
- Skin cold to the touch over the treated area
- Anything getting worse rather than better after day three
These can begin during the injection or hours afterwards. They are a same-day emergency, not a wait-and-see. For hyaluronic acid filler the treatment is hyaluronidase, and the published case above shows that even vision loss has recovered when it was given promptly and repeatedly.
9. Three questions before you book
- "Is this product hyaluronic acid?" If yes, it can be dissolved. Radiesse, Sculptra and Bellafill cannot.
- "Do you keep hyaluronidase on site?" Not "can you get it". On site.
- "What is your protocol if you suspect a vascular occlusion?" A good injector answers this immediately and without irritation.
Add a fourth if you are being treated in the nose, the temple or under the eye, which are the highest risk vascular territories: ask how often they treat that specific area.
10. Keeping this in proportion
A page that lists 261 necrosis reports and 34 vision losses needs an honest closing note, because the impression it leaves should be accurate rather than merely cautious.
Those reports accumulated across many years and many millions of syringes, and MAUDE has no denominator so no rate can be calculated from them. The overwhelming majority of the 9,792 Juvederm reports describe swelling, irritation, inflammation and lumps. The serious vascular events are a small minority of a database that itself captures a small and unknown fraction of treatments.
Fillers are, on the available evidence, a reasonably safe procedure performed badly quite often and catastrophically rarely. The honest summary is that the risk is concentrated in who injects you, where they place it, and whether they can reverse it. All three are things you can influence before anything happens, which is more than can be said for most medical risks.
Our dermal filler guide covers the products and the dissolvability split, filler recovery covers what normal looks like from FDA trial diaries, and our city guides cover the clinics we have researched.
Most of these figures are Juvederm’s. It has 9,792 reports, more than any other filler, almost certainly because it is the most used. The Juvederm guide splits them by product.
Frequently asked questions
Has anyone died from dermal filler?
Can filler cause blindness?
What is the most common filler complication?
What is vascular occlusion?
Do fillers ruin your face?
Which filler has the most complications?
What are the warning signs after filler?
Are filler lumps a complication or normal?
Can filler migrate?
Should I be worried about getting filler?
Where does this data come from?
Is hyaluronidase always available?
Know the count before the quote. The unit count is half the bill, and a low per-unit rate recovered by a high unit count is the commonest way a deal is not one. How to tell.
Deciding which area, not how many units? Botox by facial area covers what each zone actually achieves, which four are FDA approved, and the things Botox cannot fix at all.
