The potential use of ivermectin as a treatment for COVID-19 has been a topic of intense debate and discussion throughout the pandemic. As one of the largest and most rigorous trials to investigate ivermectin's effectiveness in community settings, the PRINCIPLE trial's findings regarding this repurposed drug are of interest to a lot of people.
In this section, we look into the specific details of the ivermectin arm of the PRINCIPLE trial, including the methodology, participant characteristics, and the comprehensive results obtained after following participants for up to 12 months post-treatment.
We aim to provide a balanced and transparent overview of the trial's conclusions, addressing key questions and concerns raised by both proponents and skeptics of ivermectin's use for COVID-19 and to contribute to an evidence-based understanding of its potential role in the management of COVID-19 in non-hospital settings.
Contents:
Click the links to move to that section:
- What did we find?
- Why wasn't getting better two days faster considered 'clinically meaningful'?
- Why has it taken so long to publish the results?
- Why didn't you use an ivermectin dose of 600 μg/kg?
- Doesn't having some drugs tested in mostly unvaccinated populations and some in mostly vaccinated populations invalidate the results?
- Why was there a pause in the trial? Could this have affected the results?
- Some figures suggest ivermectin was effective at preventing hospitalisations or death, why do you say it wasn't?
- Is enrolling people up to 14 days after symptoms appeared too late for ivermectin to work?
- The ongoing conversation about ivermectin and what our findings say
Ivermectin results
Ivermectin was the sixth and final medication to be added to PRINCIPLE, and recruitment to the ivermectin arm of the trail began on 23 Jun 2021.
In total, the PRINCIPLE trial involved over 11,700 participants. 8,811 of those with confirmed COVID-19 infections contributed to the primary analysis of the ivermectin evaluation. Of these, 2,157 were ‘randomised’ to be given ivermectin.
Participants were given a three-day course of ivermectin or no additional treatment, and their recovery time, hospitalisations, and deaths were monitored for 28 days. Participants were followed up at 3, 6, and 12 months afterwards.
What did we find?
The key findings from the study are:
No significant difference in COVID-19 related hospital admissions or deaths:
- Ivermectin had no effect on reducing hospital admissions or deaths compared to usual NHS care.
- These outcomes have the greatest impact on patients and healthcare systems.
Modest improvement in recovery time:
- Those taking ivermectin reported recovering about two days faster on average than those who received usual care alone.
- However, this difference is not considered clinically meaningful – during a pandemic with symptoms lasting around 16 days, a reduction of one or two days is not significant enough to justify widespread use and funding by health authorities.
No long-term benefits:
- After 12 months, there were no clinically important differences between the ivermectin and usual care groups in terms of:
- Proportion of people reporting ongoing symptoms
- Quality of life
- Impact on work or study
These findings align with other large, well-conducted clinical trials and systematic reviews, which have also found no convincing evidence that ivermectin benefits patients with COVID-19.
While the results may seem disappointing to those who had high hopes for ivermectin as a COVID-19 treatment, it is essential to base healthcare decisions on robust scientific evidence. The PRINCIPLE trial provides valuable insights into the effectiveness of ivermectin and helps guide future research and treatment strategies.
Why wasn’t getting better two days faster considered ‘clinically meaningful’?
While a statistically significant difference – i.e. something we can be mathematically confident is a real difference between the two groups – of an average of two days might seem good on paper, the hazard ratio tells us whether the benefit is consistent for all patients and ‘clinically meaningful’ enough that it suggests we should change medical practice (usual care).
A reduction in symptom duration from 16 to 14 days, when the last two days of illness is mild, are not dramatic, and consistent with effects from a drug that has anti-inflammatory properties. There are safe anti-inflammatory drugs already in widespread use.
In the case of ivermectin the hazard ratio was 1.14, but we had prespecified a hazard ratio of 1.2 for all medications in PRINCIPLE as the threshold that needed to be crossed to be clinically meaningful.
For a more detailed explanation, please see ‘What’s a ‘hazard ratio’ and why did you use it?’
Why has it taken so long to publish the results?
The ivermectin arm of the trial randomised its last patient in July 2022. As such the final analysis used for this scientific publication couldn’t be begin until July 2023, after the 12 month follows had been completed.
In PRINCIPLE, we used routinely collected data from participants health service records as part of the follow up. It takes a minimum of three months after the last participant is followed up for these data to be pulled together by these data services and longer to be received by the trial. Additionally, the trial team had little control of the speed of the scientific papers submission and peer review process. The first version of the manuscript was submitted for publication in October of 2023.
Why didn’t you use an ivermectin dose of 600 μg/kg?
PRINCIPLE used dosages and timings recommended at the time the study was designed.
The ACTIV-6 Trial also tested a dose of 400 μg/kg of ivermectin daily for 3 days. They found a small, non-significant improvement in the time to recovery. Specifically, the median time to recovery was 12 days in the ivermectin group and 13 days in the placebo group. However, there was no impact on hospital admissions. These findings are similar to the PRINCIPLE trial results.
Another study in the ACTIV-6 trial tested a higher dose of 600 μg/kg of ivermectin daily for 6 days in outpatients with mild to moderate COVID-19. This study also found no improvement in the time to sustained recovery compared to placebo.
Doesn’t having some drugs tested in mostly unvaccinated populations and some in mostly vaccinated populations invalidate the results?
We explain elsewhere in the full trial article how we account for these differences in the section titled ‘If usual care changes over time, how do you account for that?’.
Briefly, to account for potential changes in recovery and hospitalisation rates over time due to factors like new COVID-19 variants or increasing vaccination rates, the researchers used statistical methods in their pre-specified analysis plan. This 'Bayesian analysis' approach, allowed them to use data from all participants enrolled in the study up to the point of analysis, regardless of when they joined the trial.
The comparison groups for ivermectin included participants who received usual care either at the same time as the ivermectin group (concurrently randomised controls) or before the ivermectin arm began (non-concurrently randomised controls).
The researchers also performed a 'sensitivity analysis' – a way of double-checking the results by slightly changing the way the data is analysed – comparing ivermectin to concurrently randomised controls among participants who tested positive for COVID-19, which showed similar results to the main analysis, providing additional confidence in the findings.
Why was there a pause in the trial? Could this have affected the results?
Unfortunately, and for reasons beyond the researchers control there was a delay in getting a batch of ivermectin from the USA through customs into the UK.
We did not include data from the control (usual care) group in our analysis from the period in which ivermectin was not in the trial. This meant that any changes in the nature or severity of the illness during the time ivermectin was not available to the study would have affected both arms equally.
As such, the researchers do not believe this has affected the results or their interpretation in any way.
Some figures suggest ivermectin was effective at preventing hospitalisations or death, why do you say it wasn’t?
The study found ivermectin had no effect on reducing these outcomes compared to usual NHS care.
When looking at the population who tested positive for COVID-19, 1.6% (34 out of 2157) of those who received ivermectin were admitted to the hospital or died, compared to 4.4% (144 out of 3256) in the overall study population.
At first glance, this looks like it suggests that ivermectin was effective at reducing hospitalisations. But this is misleading. The overall study population includes all control patients, even those from a time before the ivermectin arm of the study began, when hospitalisation was more common and no one was vaccinated, even without a positive COVID test.
This means that many of the usual care patients included in the ivermectin analysis were enrolled during a time when hospitalisation was more frequent, but ivermectin was not being studied during this period. This difference in timing can make the comparison between the ivermectin and control groups misleading.
When the researchers accounted for these changes over time in their pre-specified additional analysis, they found no difference in the proportion of people hospitalised between the ivermectin and usual care groups.
In the population that was enrolled at the same time as the ivermectin group (the concurrent and eligible population), 1.6% (34 out of 2157) in the ivermectin group and 1.5% (27 out of 1806) in the usual care group were admitted to the hospital or died.
These numbers are not significantly different from a clinical or statistical perspective.
Is enrolling people up to 14 days after symptoms appeared too late for ivermectin to work?
One common criticisms of the study is that it allowed people to be ill for up to 14 days before they were enrolled. As mentioned in the section “Why did PRINCIPLE still recruit people who had been ill for up to 14 days? Isn’t that too long?” above , the researchers included people with a range of illness durations to find out if ivermectin could help those who were still suffering from acute COVID-19.
Some people believe that ivermectin works best early in the illness because it might have antiviral properties, while others think it could also help people who have been sick for longer through its potential anti-inflammatory properties.
People who are ill with COVID-19 for long periods, might have ongoing viral infection (in which case the antiviral properties of ivermectin should have an impact), or they have ongoing inflammation triggered by the initial infection. Anti-inflammatory effects may be more relevant later on in the illness. The longer people have been ill and remail ill, the more pressing it is to find out if there is an effective treatment foir them. The researchers wanted to answer the question of whether someone should take ivermectin if they have prolonged symptoms of COVID-19.
The median (average) duration of symptoms before people joined the study was four days in the treatment groups.
To address this question, the study had to include people who remained sick for longer periods. The researchers used statistical models to estimate the time to recovery, taking into account the duration of symptoms before joining the study. They found no difference between the ivermectin and control groups, even when comparing the effect of ivermectin in people with symptom duration of less than seven days to those with symptom duration of more than seven days.
The ongoing conversation about ivermectin and what our findings say
Ivermectin has been a topic of much debate during the COVID-19 pandemic. Some have advocated for its use based on early studies that suggested it might be effective against the virus. However, many of these studies were small or had limitations that made their findings less reliable.
The PRINCIPLE trial is one of the largest and most rigorous trials to date that has examined the use of ivermectin for treating COVID-19 in the community (non-hospital) settings. Its findings underscore the importance of rigorous clinical trials in evaluating potential treatments. While early laboratory studies and small trials may show promise, larger, well-designed studies are necessary to confirm effectiveness and safety in real-world settings.
For patients, healthcare providers, and policymakers, the PRINCIPLE trial's results suggest that ivermectin should not be prescribed or recommended as a treatment for COVID-19 in the community, particularly in largely vaccinated populations. Instead, efforts should focus on evidence-based treatments and interventions proven to be effective in reducing the burden of COVID-19 on individuals and healthcare systems.
These findings align with the current recommendations from major health organisations, such as the World Health Organization (WHO) and the US Centres for Disease Control and Prevention (CDC), which do not recommend ivermectin for the treatment of COVID-19 outside of clinical trials.
- Glossary – key terms breifly explained.
- The PRINCIPLE Trial: A Guide for Patients and the Public – back to the main section.