I weigh 72 kg. Why does this shelf say four Stromectol 3 mg chips when a Strongyloides table I found says five?
Four chips on the field slip
Draw four 3 mg chips before you listen to a viral claim. That is the whole method of this folio. River-blindness programmes did not invent a meme. They invented a countable swallow that a village line can repeat without a laboratory at the elbow.
Onchocerciasis ink aims at about 150 mcg per kilogram. Merck's donated Mectizan 3 mg tablet, and the Stromectol oncho table that clinics still photocopy, put four chips in the mid-adult band. This shelf locks that count to 66-79 kg so the SERP and the slip agree. The FDA oncho column prints the same four-chip swallow as 65-84 kg. Height over 158 cm draws four when a scale is missing. Same idea: a band, not a vibe.
A viral thread treats any blister as interchangeable. Field ink does the opposite. Strongyloides at 200 mcg/kg can ask five chips in that same 66-79 kg window. Scabies asks for a return visit. Loa loa can stop the line. If those sentences feel fussy, good. Fussy is how mass treatment stayed standing.
I will not retell glutamate-gated chloride channels here. The compendium slip already holds the molecule. This page keeps the programmes, the stool trials, the mite clock, and the three platforms that closed a rumour.
The dish never reached a human peak
Caly bathed Vero/hSLAM cells at five micromolar. At 24 hours, supernatant viral RNA dropped about 93% versus DMSO. Cell-associated RNA dropped about 99.8%. By 48 hours the paper described a roughly 5,000-fold reduction. The IC50 sat near 2 uM. That is a real dish result. It is also the sentence that escaped the lab without its units.
A standard 150 mcg/kg oncho swallow produces a peak in the tens of nanomolar, not micromolar. Commentaries in 2020 put therapeutic Cmax near 40 nM and even a 600 mcg/kg modelled peak around 0.12-0.14 uM. Schultze and colleagues noted that protein binding (about 93%) makes the free concentration smaller still. You cannot close a 50-fold to 250-fold gap by chewing extra veterinary paste. You only buy toxicity.
So the viral myth did not fail because parasites are 'the wrong species.' It failed because the concentration that stopped a monkey-cell culture is not a concentration a human field dose ever occupies. TOGETHER and ACTIV-6 then tested doses already above the labelled parasite bands and still did not move hospitalisation. That is the shelf-check: dish, then dose, then outcome. Skip a step and you get a thread.
Mectizan bands still count in kilograms
| Band on the field slip | 3 mg chips | What the ink is aiming at |
|---|---|---|
| Oncho / Mectizan, 65-84 kg (site lock: 66-79 kg) | 4 chips (12 mg) | ~150 mcg/kg single swallow; height >158 cm also draws four |
| Oncho / Mectizan, 45-64 kg | 3 chips (9 mg) | Same 150 mcg/kg target, one chip less |
| Strongyloides label, 66-79 kg | 5 chips (15 mg) | ~200 mcg/kg - same kilos, a different count |
| Scabies clinic practice | 200 mcg/kg, then again | Second swallow at 7-14 days; poorly ovicidal |
Mectizan still doses river blindness by weight or height, not by rumour. The donation programme opened in 1987. The target stayed simple: one oral dose near 150 mcg/kg, then come back in 6 or 12 months because adult Onchocerca worms are not cleared by a single swallow. The microfilariae in skin and eye drop. The adult file is a longer war.
When a scale is present, the mid-adult oncho band is four 3 mg chips. When a scale is not, a height stick does the same job: over 158 cm, four chips. Community distributors can run that without a lecture on pharmacokinetics. That is why the programme scaled. It is also why a screenshot that says 'just take ivermectin' is not a field protocol.
Note the fork that people flatten. Lymphatic-filariasis campaigns often pair the same tablet with albendazole, and sometimes DEC, on a height or weight stick of their own. The chip count can look similar. The indication is not. Mixing those slips is how a reader ends up quoting a filariasis line as if it were a COVID recipe.
Shelf-check the kilos before you admire a 'high dose' post. Four chips at 70 kg is 12 mg, about 171 mcg/kg - already in oncho range. The COVID platforms that later failed used 300-400 mcg/kg for three days. That is a different experiment, not a secret village trick the WHO forgot to publish.
Loa loa is the hold the field still keeps
High Loa loa microfilarial density is the field hold that still stops a mass campaign. Ivermectin kills Loa microfilariae quickly. In a person carrying tens of thousands per millilitre, that kill has been tied to an encephalopathic picture: confusion, lethargy, coma, incontinence. The Mectizan Expert Committee has treated counts above about 30,000 mf/mL as the serious neurological zone, with marked reactions already discussed from about 8,000 mf/mL.
Community screening uses RAPLOA - a history of eye-worm passage. A 40% RAPLOA rate maps, roughly, to a 20% microfilaremia prevalence, the threshold where a district is treated as higher risk. In those belts the answer is not 'skip all oncho work.' The answer is enhanced surveillance, test-and-treat, and sometimes a different tool. A later PLOS NTDs re-analysis still found SAE probability rising with density (over 1% above 20,000 mf/mL, over 2.5% above 30,000), with extra caution in some middle-aged men.
This is the opposite of a COVID thread. The field did not hide a danger. It wrote the danger on the slip and built a procedure around it. If a reader wants 'ivermectin is always safe because Africa takes it,' they have not read the Loa file. If a reader wants 'ivermectin is poison because of Loa,' they have not read the oncho file. Bind both. Do not flatten either.
Three platforms that closed a rumour
TOGETHER asked a blunt outpatient question in Brazil. Early, symptomatic, PCR-positive adults, ivermectin 400 mcg/kg daily for three days versus placebo, 679 versus 679 in the published pair. Primary outcome: hospitalisation for COVID progression or an emergency-department stay longer than six hours within 28 days. Events: 14.7% versus 16.3%. Relative risk 0.90, 95% Bayesian credible interval 0.70 to 1.16. The trial did not cross its superiority threshold. Secondary viral-clearance and death analyses did not rescue the story. NEJM, 2022. NCT04727424.
ACTIV-6 asked a different outpatient clock in the United States. Decentralised, age 30 and over, mild-to-moderate COVID, ivermectin 400 mcg/kg daily for three days versus placebo, 817 versus 774 who received drug. Time to sustained recovery - three straight days without symptoms - had a hazard ratio of 1.07 (95% credible interval 0.96 to 1.17). The chance that any benefit lasted more than 24 hours sat under 1%. Hospitalisation or death: 10 versus 9. Safe. Not useful. JAMA, 2022.
PRINCIPLE, published 2024 in the Journal of Infection, ran as an open UK platform in a largely vaccinated community. Ivermectin target 300-400 mcg/kg daily for three days (2,157 people) versus usual care (3,256). Time to first self-reported recovery: hazard ratio 1.15 (1.07 to 1.23), median about two days faster from a usual-care median near 16 days. The pre-set bar for a meaningful effect was a hazard ratio of 1.2 or more. Probability of hitting that bar: 0.192. COVID-related hospitalisation or death: odds ratio 1.02 (0.63 to 1.62). Six-month recovery looked similar. The authors said further community trials in vaccinated people look unwarranted. I agree.
Bind those three and stop shopping for a fourth platform that will finally 'open' a field drug. Small early papers, retracted preprints, and meta-analyses that leaned on them are not a secret majority. The large randomised outpatient file is negative on the outcomes that matter: hospital, death, and a recovery clock worth a national protocol.
A second swallow after the eggs hatch
Usha's Kerala comparison is the slip I still cite when someone wants one swallow to finish scabies. Eighty-five people, household contacts treated, 200 mcg/kg versus a single overnight 5% permethrin cream. One ivermectin swallow cured 70%. A second swallow at two weeks pushed that to 95%. One permethrin night cured 97.8%. Permethrin also looked faster. The authors' reading still holds: ivermectin is a weak ovicide, so eggs hatch into a second generation the first swallow never met.
Clinic practice took that life-cycle note and made a habit. Two oral doses, 7 to 14 days apart, plus treat the household, plus wash the linen. CDC-style teaching and later cluster trials keep the same rhythm even when they argue about which agent wins. A 2025 French cluster RCT (day 0 and day 10, 200 mcg/kg versus 5% permethrin) did not show oral ivermectin non-inferior to cream. That is a comparative quarrel. It is not a licence to skip the second swallow.
Crusted scabies is a different desk. Densities are huge. One or two ordinary swallows are not a programme. Those patients need a specialist plan, often topical plus repeated oral doses, isolation, and nail work. I will not invent a schedule here. I will say the field habit that leaked onto social media - 'one pill, done' - is the opposite of how mite clinics actually work.
Note the indication gap. Oral ivermectin for classic scabies is common practice and guideline language in many countries. It is not the same card as US Stromectol's labelled oncho and Strongyloides uses. A folio that pretends otherwise is doing marketing, not shelf-check.
Stool trials that beat albendazole
Marti and colleagues randomised 301 children in rural Zanzibar to a single 200 mcg/kg swallow or to albendazole 400 mg daily for three days. Strongyloides stercoralis cure was 83% versus 45%. Both drugs handled Ascaris. Ivermectin did almost nothing useful to hookworm. Trichuris cure stayed poor on ivermectin (11%) and only middling on albendazole (43%). That is a field note, not a slogan: the tablet is a Strongyloides and Ascaris tool, not a universal worm broom.
Datry's earlier open randomised study enrolled 60 people with confirmed strongyloidiasis and evaluated 53. Ivermectin at 150-200 mcg/kg cured 24 of 29 (83%). Albendazole 400 mg for three days cured 9 of 24 (38%). Failures were offered ivermectin again. Most cleared. The sample is smaller than Marti's. The direction is the same.
A Cochrane review later pooled ivermectin versus albendazole across four trials (478 participants). Parasitological cure favoured ivermectin - risk ratio 1.79, 95% CI 1.55 to 2.08. Single 200 mcg/kg and a two-day 200 mcg/kg schedule both beat albendazole; the review did not find a clean winner between those two ivermectin schedules. Versus thiabendazole, cure looked similar and tolerability looked kinder on ivermectin. That is the Strongyloides shelf, not a viral footnote.
Bind this to the chip count. The US strongyloidiasis table uses 200 mcg/kg, so 66-79 kg is five 3 mg chips, not four. If a reader quotes this site's four-chip oncho lock as if it were a Strongyloides order, they have mixed the slips. Send them back to kilograms and to a clinician who has seen a stool, not a timeline.
Bind the ink, leave the viral claim off the shelf
Bind this folio as two separate cards. Card one: counted chips for named parasites. Four 3 mg on an oncho band (66-79 kg on this shelf, 65-84 kg on the FDA oncho column). Five chips in the same kilos when the Strongyloides ladder is the one in play. A second swallow when the target is a mite egg that has not hatched. A Loa check when the map says so. Marti, Datry, Usha, Mectizan, RAPLOA.
Card two: the antiviral claim. Caly's micromolar dish. TOGETHER's 14.7% versus 16.3%. ACTIV-6's recovery hazard ratio 1.07. PRINCIPLE's recovery hazard ratio 1.15 under a 1.2 bar, hospital odds ratio 1.02. Those numbers do not borrow glory from a Nobel or from a village line. They sit in their own column and they are negative.
Take the parasite card to a clinician who can weigh kilograms, pregnancy, Loa risk, and the rest of the chart. Leave the viral card in the archive. The ivermectin slip holds the molecule. The onset file on Prozac 40 mg is a different clock entirely - weeks, not a field swallow.
Last Updated
Reader mail
Reader questions on this review
Answered by Dr. Ingrid Halvorsen, MD · Internal medicine & clinical pharmacology
Readers send field-math questions, not mechanism essays. I answer the way I would at a Copenhagen desk: named trials, counted chips, holds that still matter. Teaching marks. Not your chart.
Because those tables are aiming at different microgram-per-kilo targets. Onchocerciasis programmes and the Stromectol oncho column aim near 150 mcg/kg. At mid-adult weight that is four 3 mg chips - this folio locks the sentence to 66-79 kg so it matches the site's other slips; the FDA oncho print uses 65-84 kg for the same four-chip swallow. Strongyloidiasis labelling aims near 200 mcg/kg, so the same 66-79 kg window becomes five chips. Height sticks in Mectizan lines (over 158 cm = four) are a scale-free way to hit the oncho target, not a Strongyloides shortcut. Do not mix the cards. Bring the indication and the kilograms to your own clinician.
Did the Mectizan programme ever prove ivermectin wipes out adult river-blindness worms in one visit?
No. The donation programme, from 1987 on, is built on repeated mass swallows because ivermectin is a strong microfilaricide and a weak adulticide. Skin and eye microfilariae fall after a dose. Adult Onchocerca volvulus live on, so districts come back every 6 or 12 months. That is why a 'one tablet forever' story is not field ink. It is also why oncho success is measured in community microfilarial load and blindness averted over years, not in a 48-hour viral-style endpoint. Your own programme clinician decides the interval from prevalence and logistics, not from a post.
Marti's 83% versus 45% - is that enough for me to treat a possible Strongyloides infection myself after travel?
Marti (Zanzibar, 301 children, single 200 mcg/kg versus albendazole 400 mg for three days) and Datry (83% versus 38% in 53 evaluable adults) are why ivermectin sits as the better Strongyloides drug in immunocompetent people. They are not a home kit. Diagnosis still wants a lab that knows Baermann or agar-plate methods, because ordinary smears miss larvae. Immunosuppression, HTLV-1, and steroids change the stakes toward hyperinfection, where a missed diagnosis is dangerous. Cochrane's pooled risk ratio of 1.79 versus albendazole is a comparative fact, not permission to swallow a blister from a suitcase. Get a stool and a clinician.
If Usha needed two ivermectin swallows for scabies, why do some clinics still offer one?
Because older trials and busy rooms cut corners, and because permethrin cream after one night already looked excellent in Usha (97.8%). Oral ivermectin as a single 200 mcg/kg swallow cured only 70% there; the second swallow at two weeks brought 95%. The mite life cycle is the reason, not a marketing upsell: the first dose is a poor ovicide, so larvae that hatch later need a second hit, usually at 7-14 days. Household contacts and linen still matter. A later cluster trial that dosed day 0 and day 10 still did not prove oral ivermectin non-inferior to 5% permethrin. I treat 'one pill for scabies' as incomplete teaching unless a specialist has a reason.
TOGETHER's relative risk was 0.90. Isn't that a hint of benefit the media buried?
A point estimate of 0.90 with a credible interval from 0.70 to 1.16 means the data are compatible with a modest benefit, with no effect, and with modest harm. The trial's own superiority rule was not met. Most primary events were hospitalisations, not long emergency stays. Secondary analyses (clearance, death, ventilator days) did not turn the result. ACTIV-6 then showed a recovery hazard ratio of 1.07 and 10 versus 9 hospitalisations or deaths. PRINCIPLE's hospitalisation odds ratio was 1.02. That is not a buried miracle. That is three large files that failed to move the outcomes we care about.
PRINCIPLE said people recovered two days faster. Why do you still call it negative?
Because the investigators pre-specified what 'meaningful' meant: a hazard ratio of 1.2 or higher for time to first self-reported recovery. They observed 1.15 (1.07 to 1.23). The probability of crossing their own bar was 0.192. Usual-care median recovery sat near 16 days; two days off that clock, in an open trial, in a vaccinated community, without a drop in hospitalisation or death, is not a national protocol. Six-month recovery was similar (74.3% versus 71.2% feeling fully recovered). I will not pretend a missed threshold is a secret win.
Could Caly's dish still matter if we just used a much higher human dose?
The dish used 5 uM and an IC50 near 2 uM. Human peaks after labelled parasite doses sit in the nanomolar range. Even modelled high doses stay well under a micromole, and most of that is protein-bound. TOGETHER and ACTIV-6 already tested 400 mcg/kg for three days - above oncho and Strongyloides bands - and did not move hospitalisation. Pushing further toward veterinary paste is how people end up with toxicity, not with an IC50. The FDA warning on animal formulations exists for that reason. A dish is a hypothesis generator. These platforms were the test.
How does a clinic actually screen for Loa loa risk before ivermectin in a co-endemic belt?
At community scale, RAPLOA asks adults about a worm crossing the eye. A 40% positive history is treated as a flag that maps to roughly 20% microfilaremia, the Mectizan Expert Committee's higher-risk community threshold. At individual scale, the dangerous zone for encephalopathy after ivermectin has long been counts above about 30,000 microfilariae per millilitre, with marked reactions discussed from about 8,000. Some programmes use test-and-treat rather than blanket MDA. This is not a Copenhagen waiting-room test. If travel or origin includes Central African Loa belts, the conversation belongs with a tropical-medicine desk before anyone counts chips for oncho or something off-label.
Where should I go next if I want the molecule, not the field argument?
The ivermectin folio slip holds empty-stomach timing, labelled US uses, and the pharmacology this review refused to retell. If you want a completely different clock - weeks to an SSRI effect, not a counted parasite swallow - the Prozac 40 mg onset file is the next shelf. Neither page is a cart. Bring kilograms, maps, and the rest of your medicines to your own clinician before anyone changes a dose.
These replies are teaching, not a treatment plan for the person who asked. Take your own history, labs and medicine list to a prescriber who can see all three.