Ivermectin
A new study on cheap, repurposed Covid treatment ivermectin has concluded that its findings "do not support the use of ivermectin to treat mild to severe forms of COVID-19". However, this conclusion is at odds with its findings.

The study, "Non-effectiveness of Ivermectin on Inpatients and Outpatients With COVID-19; Results of Two Randomised, Double-Blinded, Placebo-Controlled Clinical Trials", is published in Frontiers in Medicine. It includes among its authors Dr. Andrew Hill, who last year appeared to suggest to Dr. Tess Lawrie that pressure had been applied to him not to find in support of ivermectin in an earlier paper. He told her, "I'm in a very sensitive position here", and "I don't really want to get into" revealing who from Gates-funded charity Unitaid, which funded the study, really wrote the conclusion of the paper downplaying the benefits of the treatment.

The new study gives a helpful introduction to the drug.
Ivermectin is a low-cost established drug with clinical benefits and minimal safety concerns, which has been shown to inhibit SARS-CoV-2 in vitro in studies. Ivermectin has rapid oral absorption, with high lipid solubility is widely circulated in the body, metabolised in the liver, and excreted in faeces. The adequate concentration of ivermectin inhibiting SARS-CoV-2 in the in vitro experiment is higher than the approved dose of ivermectin concentration in plasma and the lungs of humans. However, a meta-analysis demonstrated that the administration of a standard FDA-approved dose shows a positive clinical response in COVID-19 patients.
The study is a follow-up to an earlier, smaller study which showed promise. However, the promise has not, the authors say, been borne out.
Despite our previous more favourable results from a multicentre, randomised clinical trial in 69 COVID-19 patients at the beginning of the pandemic which noted the effectiveness of ivermectin in recovery and decreasing duration of hospital stay, the current results of this extensive study on 609 admitted patients with moderate to severe form of COVID-19 and 549 outpatients with a mild form of COVID-19, did not show adequate support for the effectiveness of this drug.
Despite this downbeat assessment, the new study did actually find a significant 32% improvement in ivermectin hospital patients achieving complete recovery, with 37% of ivermectin patients vs 28% of placebo patients achieving the outcome [95% CI, 1.04-1.66].

A number of the other key outcomes, including ICU admission and death, were also better in the ivermectin group, though the study was underpowered (not large enough) for these results to be statistically significant (i.e., we can't be sure they weren't coincidence). These were:
  • ICU admission: 28 ivermectin vs 32 placebo patients; 9% vs 11%; 16% improvement [95% CI, 0.52-1.36].
  • Invasive mechanical ventilation: 3% ivermectin vs 6% placebo; 50% improvement [95% CI, 0.24 -1.07].
  • Supplemental oxygen by non-invasive ventilation: 244 ivermectin vs 252 placebo; 78% vs 85%; 7% improvement [95% CI, 0.86-1.00].
  • Death: 13 ivermectin vs 18 placebo; 4% vs 6%; 33% improvement [95% CI, 0.35-1.39].
The fact that all these outcomes showed an improvement, and mechanical ventilation and death considerably so, is a signal that the benefit is unlikely to be solely due to chance. Thus the conclusion should really have been that a larger study is needed to see if the promising results can achieve statistical significance.

For outpatients, there were also some significant clinical benefits:
  • Fever duration: 2.02 (± 0.11) days ivermectin vs 2.41 (± 0.13) days placebo; 16% improvement.
  • On the day seventh of treatment, fever, cough and weakness were significantly higher in the placebo group compared to the ivermectin group.
A few results went the other way, though none of these were statistically significant. For inpatients:
  • Length of hospital stay: 7.98 (± 4.4) days ivermectin vs 7.16 (± 3.2) days placebo; 20% worse [95% CI, 0.15-1.45]. The study claims this finding is "significant", but the wide confidence interval going through 1.0 indicates not. The authors write that "delays in discharging patients to other facilities such as rehabilitation centres... might be the reason for more extended hospital stay other than treatment for COVID-19".
  • Mean oxygen saturation at day seven: 92.01 (Range: 72-99) ivermectin vs 93 (Range: 48-99) placebo; 1% worse [95% CI, -2.89 to 0.91].
  • Relative recovery (where some symptoms persist on discharge): 53% ivermectin vs 60% placebo; 13% worse [95% CI, 0.76-1.00].
  • Persistent dry cough (until seventh day): 5 ivermectin vs 10 placebo; 3% vs 9%; 36% worse [95% CI, 0.13-1.03].
For outpatients:
  • Hospitalisation: 7% ivermectin vs 5% placebo; 36% worse [95% CI, 0.65-2.84].
  • PCR negative on day five after treatment: 26% ivermectin vs 32% placebo; 19% worse [95% CI, 0.60-1.09].
The authors write that "no evidence was found to support the prescription of ivermectin on recovery, reduced hospitalisation and increased negative RT-PCR assay for SARS-CoV-2 five days after treatment in outpatients". However, it's important to note that this was for ivermectin given more than a week after symptoms began. Proponents of ivermectin often argue that treatment should be given within five days of exposure, i.e., as soon as possible.

The paper does mention this issue, though in a strange sentence with typographical errors perhaps indicative of a late addition:
"Ivermectin may be going to be effective if it is given at the earliest possible time that clinical symptoms appear whiles [sic] the mean duration of symptoms before randomisation was 7.36 ± 3.43 days in the ivermectin group and 6.98 ± 3.63 days in the placebo group."
Typographical errors aside, the point is correct; an outpatient study really needs to start the treatment sooner.

There may also be a dosage issue. While the trial gave a dose of 0.4 mg per kg per day over a duration of three days, some have suggested a higher dose is required. The paper nods at this where it says:
"Krolewiecki et al. assessed antiviral activity and safety of a five-day regimen of high dose ivermectin, comparing the control group in 45 patients with COVID-19. The findings support the hypothesis that ivermectin has a concentration-dependent antiviral activity against SARS-CoV-2."
A further potential problem with the study, which was conducted in Iran where ivermectin has been popular as a Covid treatment, is the question of how many of the placebo group were also secretly taking ivermectin anyway. In the limitations the authors note that "after the allocation of ivermectin or placebo, a significant number of patients declined to be participants", which may be because they realised they wanted to be sure they were taking the drug. Taking an antiviral medication was an exclusion criterion for outpatients - 18 admitted to it, but how many continued with the trial (for which they were presumably paid) but took such drugs anyway? Furthermore, previously taking an antiviral does not appear to have been an exclusion criterion for inpatients, so it is unknown how many placebo-arm inpatients had taken ivermectin or another medication prior to hospitalisation. Once in hospital, I imagine they would not have been able to continue taking any medication secretly, and perhaps that explains why nearly a third of the inpatient participants were lost to follow up, most due to voluntary withdrawal or "incomplete intervention" (31.6%, 282 of 891; 136 ivermectin and 146 placebo).

Overall, I find the conclusion baffling given the findings. There were statistically significant benefits of ivermectin for complete recovery, shorter duration of fever and quicker clearing up of cough and weakness. There were also large but not-statistically-significant benefits for mechanical ventilation and death. The negative findings were mostly small and none were statistically significant. This is for a study which didn't start the treatment until over a week into symptoms, and may have been confounded by people in the placebo arm also taking the drug.

Perhaps we will never get to the bottom of exactly how effective ivermectin is against COVID-19. But since it's a safe drug (to quote U.K. Chief Medical Officer Chris Whitty, "Ivermectin has proven to be safe. Doses up to 10 times the approved limit are well tolerated by healthy volunteers") and this study shows once again that it gives some benefit - other studies show much greater benefit - why not be honest about that, allow medics to include it in their treatment protocol, and stop making such a fuss about stopping them?