Dish assays are not a reason to swallow a farm syringe

A dish hit is not a swallowable dose

Cell-culture antiviral numbers sat far above human peaks.

Early in 2020 a lab reported that ivermectin slowed SARS-CoV-2 in Vero cells. The finding was real. The concentration was not a clinic dose. Levels that worked in the dish sat in a range you cannot safely hold in human plasma with oral 3 mg chips. Pharmacologists said so while the screenshot was still circulating.

Repurposing often starts in a dish. A hit is an invitation to measure whether a tolerable human level can do the same job. For this molecule and this virus, it could not. To chase the dish number you would push into the confusion, vomiting, and seizure range we already know from overdose.

Cheapness does not close a concentration gap. A Nobel for river blindness does not close it either. The honest first sentence was always: interesting signal, unusable as a home antiviral. The ivermectin monograph still lists worms and mites. That list did not grow a virus.

Large outpatient trials closed the virus-pill claim

TOGETHER and ACTIV-6 tested real people at ordinary doses.

Once you leave the dish, the question is simple. Does a counted human dose change hospitalization, recovery time, or death in people with COVID-19? Large randomized, placebo-controlled outpatient trials were built to answer that. TOGETHER and ACTIV-6 are the names this desk keeps on the blotter. Other groups ran their own stacks. The picture converged.

Those trials did not show a meaningful outpatient benefit. Not a shorter useful recovery that holds up. Not a drop in the hard events people actually fear. Different countries, different teams, same landing: the 3 mg chip stayed a parasite drug. That is how a repurposing idea is supposed to end when the effect is not there.

Moving the goalposts after each negative - wrong dose, wrong day, wrong variant - is not skepticism once the variations have been tried. At some point the stack of clean trials is the answer. This desk treats that stack as closed for outpatient COVID use.

Small early papers looked busy, then faded

Retracted preprints and tiny cohorts fed the first headlines.

A pile of small, poorly controlled studies can look like a chorus. Small samples swing. Groups that were never randomized carry hidden differences that impersonate a drug effect. At least one widely shared preprint was withdrawn over data problems, and meta-analyses that had swallowed it had to be rebuilt. The correction never traveled as far as the first graph.

Most of those authors were not cartoon villains. They were working in a frightened year. Enthusiasm still is not evidence. One large randomized trial outweighs a folder of uncontrolled series. When the large trials arrived, the apparent benefit thinned out. That arc - loud early signal, quiet later null - is ordinary in medicine. It is not a conspiracy.

If a social post offers 'dozens of studies' and none of them are the big RCTs, you are being shown the folder, not the verdict. Read the verdict.

Farm-store syringes showed up in poison logs

Poison centers logged a spike once livestock paste went viral.

A treatment that does not work still costs something if people swallow the wrong product. During the peak of the claim, poison centers reported more ivermectin exposures, many of them livestock paste measured by eye. Those syringes are built for horses and cattle. They are not a warehouse version of a 3 mg chip.

The clinical picture was the overdose we already knew: vomiting, confusion, tremor, unsteadiness, sometimes seizures. That is harm in exchange for a benefit the RCTs did not find. There is a quieter cost too. Time spent trusting a null drug is time not spent on measures that actually move risk.

The molecule itself got dragged into a culture fight. Some readers now distrust a superb antiparasitic for the jobs it does well. Others hoard paste. Neither reaction helps a person with Strongyloides. Keep the fights in the comments. Keep the chips on the parasite list.

A Nobel for worms is not a license for viruses

Avermectin prestige does not rewrite a negative trial stack.

Ivermectin is a great drug for the organisms that earned it. That prestige made the myth easier to sell. Cheap, available, already on shelves, medal on the wall - of course a frightened year wanted it to be more. A kernel of dish truth gave the story something to grow around. Official caution then read, to some ears, like suppression.

Agencies that said no to COVID ivermectin are the same agencies that spent decades moving this drug through river-blindness programs. They are not allergic to the molecule. They are allergic to a null outpatient claim. Cheap and off-patent also means there is no blockbuster machine that profits from hiding it. The boring explanation is the trial results.

Identity is sticky. Once a pill becomes a badge, contrary evidence bounces. The way out on this desk is still the same sentence we use for every other drug: show the large randomized trials. For COVID, they exist, and they are negative.

Keep the 3 mg chip on the parasite list

Nothing in the COVID record changes Strongyloides or onchocerciasis use.

If you need ivermectin for a named worm or mite, the COVID years did not cancel your script. Use human 3 mg tablets, count from kilograms, and follow the stomach clock on the slip. That practical file is the kilogram chart. The organism file is the parasite map. The product page is ivermectin.

If you do not have a parasite, do not swallow a farm syringe because a dish assay once looked pretty. Do not run a standing 'just in case' chip against a virus the RCTs already tested. Dr. Ivy Chukwu answers the leftover mail below.

This page is teaching, not your chart. Change nothing without the clinician who knows your history. The legal floor is the desk disclaimer.

Consultation

Reader questions, answered

Answered by Dr. Ivy Chukwu, MD · Infectious disease & internal medicine

Mail that still treats a dish assay as a prescription.

Theo Marchand asksIf the lab killed the virus, why do you refuse the chip for COVID?

The lab used a concentration you cannot safely hold in a person with oral 3 mg tablets. That is the whole catch. A dish hit only matters if a tolerable human level can repeat it at the infection. Here it cannot. Push toward the dish number and you are in the overdose range - confusion, vomiting, low blood pressure, seizures. When researchers then gave ordinary human doses to outpatients in large randomized trials, the clinical benefit was not there. So we are not 'ignoring the lab.' We are reading the lab with a pharmacokinetic ruler, then reading TOGETHER and ACTIV-6. FDA public notes at the FDA said the same thing without the comment-section heat.

Sable Winters asksI can show you a folder of positive studies. How can that folder be wrong?

A folder of small, messy studies is not a chorus of proof. Tiny samples swing. Unrandomized groups hide differences that look like a drug effect. At least one influential preprint was withdrawn, and pooled papers that had trusted it had to be rebuilt. Quantity is not quality. The studies built to strip those biases - large, randomized, placebo-controlled outpatient trials - did not find a useful COVID benefit. In evidence ranking, those trials sit above the folder. It feels upside down if you discovered the folder first. It is how the hierarchy is supposed to work. NIH's library at NCBI is where I send readers who want the trial records rather than a screenshot collage.

Diego Paredes asksMy cousin took horse paste after a video and ended up in the ED. What did we just watch?

You watched a concentration mismatch. Livestock paste is built for a horse or a steer, not for a human 3 mg chip. Measuring a ribbon by eye is how people take a massive overdose that still looks like a little bit of paste. Confusion, vomiting, tremor, unsteadiness - classic toxicity, and exactly what poison centers logged when the claim went wide. Call poison control or emergency care and bring the tube so they can read the strength. People usually recover with prompt care. The RCTs never owed anyone that risk. CDC guidance at the CDC covered the livestock-product spike in the same years the trials came back negative.

Freya Lindholm asksI still take a chip now and then so I do not catch COVID. Harmless habit?

Not a habit I would keep. Prevention trials did not show protection, so the standing chip is side-effect exposure without a gain. If the chips are livestock paste, you add overdose risk on top. There is also a quieter problem: if you believe you are covered, you may skip the steps that actually cut risk. Ivermectin is a parasite drug. A virus does not grow a chloride channel because you swallowed a tablet on Sunday. If your goal is fewer bad COVID outcomes, we should talk about the tools that moved endpoints in trials, not about an antiparasitic souvenir. WHO's public COVID pages at WHO still keep ivermectin off the outpatient treatment list for this reason.

Kwame Boateng asksIt is cheap and off-patent. Why would agencies bury it unless something is off?

Follow the money the other way. A generic with no exclusive profit is also a generic with no exclusive profit in hiding it. The same agencies spent decades pushing ivermectin through onchocerciasis programs and celebrating the donation story. They are not allergic to the molecule. They said no to COVID use because large randomized outpatient trials did not show benefit. Cheapness made the myth easier to love, not truer. Suspicion is fine. The test of suspicion is the primary papers, not the loudest voice. If a hidden cure were sitting in a 3 mg blister, TOGETHER and ACTIV-6 were built to find it. They did not.

Rosa Villanueva asksDoes this flop mean we should stop testing old drugs for new viruses?

No. Repurposing is a fair strategy and it has real wins in other diseases. Ivermectin got a fairer, larger COVID test than most folk remedies ever receive. That is the process working. A plausible dish signal, a dose you can actually reach, then randomized trials in people - every candidate has to walk that hall. This one failed the last door for COVID. Others will pass. Judging each idea on its own trial stack is the instinct to keep. Do not throw out remdesivir-style programs because one antiparasitic did not become an antiviral. And do not drag a failed COVID claim back onto the parasite jobs the molecule still does well.

Ellis Park asksIs there any respiratory illness where you would still reach for Stromectol?

Not for a viral respiratory illness. Flu, COVID, a common cold - no chloride channel, no job for the chip. The only time ivermectin belongs near a sick visit is when there is a named parasite in the same person, and that is almost never why someone is coughing. If you are short of breath or worsening, that is a reason to be examined, not a reason to open a farm syringe. Taking a 3 mg stack 'so at least I did something' adds risk and delays useful care. MedlinePlus pages at MedlinePlus keep the approved uses in one public list if you want a third voice next to this desk. Keep the blister for worms and mites. Leave the virus drawer empty.

General education from a clinician, not personal medical advice. Bring your own history to your own prescriber.