Ivermectin Fails to Statistically Beat Placebo in Colombian Study & Mass Media Finally Takes Notice—An Interesting Observation

Ivermectin Fails to Statistically Beat Placebo in Columbian Study & Mass Media Finally Takes Notice—An Interesting Observation

Recently, the results of a new ivermectin study out of Columbia were reported in the medical journal JAMA. Led by Eduardo Lopez Medina, MD, a physician and researcher at the Center for Pediatric infectious Diseases in Cali, Colombia, the double-blind, placebo controlled randomized clinical trial was organized to test the efficacy of ivermectin in preventing progression of disease in adult patients with early states of COVID-19. Interestingly, as the results were not compelling, the New York Times decided to swoop in on the news. The study (Lopez 2021) team found the patient participants taking the 5-day course of ivermectin failed to significantly improve the time to resolution of the symptoms. Although the investigative team concluded that the study results failed to support the use of ivermectin as a treatment of mild COVID-19, they acknowledge that perhaps larger trials may be needed to better understand the effects of the drug on other clinically relevant outcomes. TrialSite suggests that the results of this study do little to change the comprehensive meta analyses currently generated from researchers in America and the UK.

The study results indicated the time to resolution of symptoms was shorter (10 vs. 12 days) but not statistically significant. Very few participants in this trial got seriously ill, but there were “trends” in favor of ivermectin. For example, 2% vs. 3.5% clinically deteriorated—this is 2 points on the clinical scale but not significant. Additionally, a 2% vs. 5% required escalation of care, not statistically significant (although perhaps not a totally fair comparison as 4 of the recipients of placebo went to the hospital within 12 hours of randomization). Overall, the study drug was deemed safe with no increase in serious adverse events but again the statistically significant results were not there for ivermectin either.

In this study reported in JAMA, the team reported zero deaths associated with ivermectin and one control death. An interesting quirk in the study centered on the unusually high cumulative dosage of 1,500 μg/kg over five days. At that dose, 7.5x the normal, some non-life threatening and transient side effects can occur, notably blurring vision, and it was therefore appropriate for this study to track them, as shown from the study, supplementary file 2. Interestingly, the control group had exactly the same rate of these characteristics ivermectin-related adverse events as the ivermectin group; Diarrhea (30.2% v. 32.8%); Nausea (24.0% v. 23.7%); and blurred vision (16.0% v. 14.1%).

So, this study showed that both the ivermectin group and the control group had adverse events characteristic of ivermectin, at the same rate, within the limit of tolerability. Note that ivermectin is freely and inexpensively available over the counter without prescription in Colombia, the study locale. 

These quirky results for IVM to treat mild cases of COVID-19 cases with no significant differences in deaths (0 for IVM, 1 for controls) stand in contrast to what happened in Peru as reported in a recent study when the army went in with mass IVM treatments through operation MOT, and state by state for those 10 states, excess deaths dropped 74% (mean reduction) over 30 days within 1-11 days after MOT start date. For the 25 states of Peru, the reduction in excess deaths correlated with the extent of IVM distribution and treatment with a p value of 0.002. These results are far more precise than the battlefield experience for penicillin in World War II, in which only general indications of good success were known, no RCTs at all, and then were followed with mass treatments of civilians. 

Other Comments

While the results of this study cannot be denied, the measurement of time to symptom resolution as an outcome measure can also represent a challenge. Symptoms can be somewhat to quite heterogenous—that is they can fluctuate, and can tend to linger on for a while even after the individual has recovered. There was a hint of difference between groups on some other outcomes that were more closely related to clinical deterioration, but for those outcomes they did not have enough statistical power to show whether there is a true difference.

Interestingly, the primary outcome was changed 6 weeks into the trial but the concern of many TrialSite readers, as picked up by their comments, was that medical journals and press appear to not mind when the randomized controlled trial reveals negative or neutral results associated with a repurposed drug.

Of course, this particular trial was conducted among young people with mild COVID so one would not expect a lot of deaths and other poor outcomes. However, one death in the placebo group is one death too many.

Media Bias?

Among other notable points was the New York Times‘ fascination with this particular ivermectin study results as none of the overwhelmingly positive study results, nor positive meta-analyses results, gained any attention from that nor other mainstream media.

 Many in the TrialSite Network found it ironic that the prestigious paper swooped in on a neutral- to negative-leaning ivermectin study. After all, TrialSite has reported on one study after another that evidenced positive results. Yes, not all of these studies were designed the same but note there was no interest in major press for anything positive. But in this case, there was certainly considerable interest with bold declarations upfront.   Interesting, no doubt. 

The Importance of Meta Analyses

There are a few compelling meta-analysis studies circulating around North America and the United Kingdom, and these studies compare dozens of ivermectin-based trials. The net takeaway from these studies is an overwhelming positive outcome associated with the generic drug in the context of ivermectin and the coronavirus.  Many challenge the underlying study quality of design, however, discounting the power of these broader analyses.  

Whether one particular study is positive or negative, this hammers home the importance of meta-analysis studies. A relevant example is the remdesivir story. The U.S. FDA approved remdesivir on what Dr. Anthony Fauci himself declared as not the most compelling data (see TrialSite article) yet the World Health Organization in the Solidarity study revealed remdesivir wasn’t beneficial at all. The reality is that remdesivir is beneficial and just because one study shows it wasn’t doesn’t stop Gilead from generating nearly $3 billion in revenues. But TrialSite did ask the question whether power and politics was also at play with the drug. It takes many studies to make up a meta analysis, which provides a more accurate picture of a drug’s true efficacy.

Call to Action: TrialSite will continue to monitor the ivermectin study results as they are available and thereafter provide as much of an objective, non-commercially interested and unbiased perspective as possible. For example, despite this media platform’s favorable view of the current ivermectin meta analyses currently circulating, this particular Colombian study falls under our “TrialWatch Challenge” category, meaning the results were neutral to negative.


  1. This article ny Trialsite iseems to me no more objective than the study it assaults. Trialsite needs to stay out of the political fray in order to remain a reputable source of scientific research news.

  2. this is one of those studies designed to fail , like it was done with HCQ , giving extremely high doses to prove it does not work… shame on this people

  3. You knew there were going to be studies like this. This one seems reminiscent of one of the early HCQ built-to-fail tests where they gave the patients nearly lethal doses. So of course the media will run with this one. MSM is allied with Big Gov and Big Pharm. No kidding, right?
    And let’s not be led astray by "statisitical significance". 2 days less in the hospital is a blessing to those in the beds. It’s significant as all get out to them. A 4/7 drop in deterioration is HUGE, even if the "number" goes from 3.5% to 2%. That’s more than 50%.
    Maybe this trial should be dismissed like all the other ones. It’s too small, wasn’t properly controlled, and was done by brown-ish people in the 2nd or 3rd world, therefore it’s worthless. Follow the money; who paid who to generate which results?

    1. Wasn’t that "2 days less in the hospital" the original justification for remdisivir? At $3200.00 a pop it was to have saved something like $16,000.00 in hospital costs. At least that’s how I remember it.

  4. The original protocol is not provided by the authors and there is no information given on the changes between the original protocol and amendment 1. It would be good to know those changes and their justification. One change may have been in the title. The first line on the first page of "Trial Protocol" attached to the JAMA article is: "Re: 20-36255 – Efficacy and Safety of Ivermectin for Adult Patients with Mild Covid-19" . After amendment 1 the title of the protocol is: "DOUBLE BLIND, RANDOMIZED, CONTROLLED CLINICAL TRIAL TO INVESTIGATE THE EFFECTIVENESS OF THE D11AX22 MOLECULE IN ADULT SUBJECTS FROM CALI WITH INITIAL STAGES OF INFECTION BY SARS COV2 / COVID-19".

  5. this is unbelievable.
    all said hcq doesnt work,
    who said remdisivr doesnt work
    these people say ivm doenst work – and shockingly – placebo work. dont know what medicine were given to placebo
    therefore china virus is a placebo controlled virus

  6. TrialSite and others have raised the question of whether the control arm may have had exposure to ivermectin. One thing sticks out on reading the protocol. In the title of the protocol, the medicine is referred to as "D11AX22 MOLECULE". The authors should clarify if that terminology was used with the study subjects. If so, it’s conceivable that some of the participants wouldn’t have realized they were in an ivermectin trial.

  7. I noticed that a few of the secondary endpoints are different between the study protocol and the publication. In particular, one secondary endpoint is: "The duration of supplemental oxygen in each arm of the study." Results for that endpoint are not reported or mentioned.

  8. Very interesting, and what I have been expecting (especially from a hard-right Latin American political establishment deeply connected to the US establishment.
    I’ll call this a precursor to The Gates Foundation "study" which I have believed from the get-go to be nothing but a treatise on discouraging Ivermectin upon its results being released.
    Now, please, spell ‘COLOMBIA’ correctly.