Rock, paper, scissors: Ivermectin beats Paxlovid

On my list of on-the-go projects is a note to take a close look at the Paxlovid data as soon as it is released. Unlike the media, I think it is extremely unwise to take a press release at face value. There are two recent examples of press-release-science that come to my mind: the declaration that Molnupiravir is an effective drug and the South African study claiming the vaccine would remain effective, albeit less effective, against omicron.

As it turned out, Molnupiravir’s effectiveness came from poor or fraudulent randomization in the control trial. All it took was a slightly closer look at the data to see that something was seriously amiss. After the interim analysis, when the most serious chronic conditions were not heavily skewed to the placebo group, the drug’s efficacy took a nose dive. Unsurprisingly, regulators in the United States failed to (or ignored) pick up on the critical error in the control trial, but I suspect regulators in other countries have taken notice as the drug appears to be getting little international pickup.

Likewise, the vaccine was hailed as moderately effective against omicron in South Africa, allowing vaccine enthusiasts to breathe a sigh of relief at the prospect of a new variant. I was skeptical at the time purely based on the fact that the study used a test-negative control design, and when the data was released it looked horrid for the vaccines. As it turned out, the raw data, not the adjustment with the test-negative control design, turned out to be closer to the truth as has been shown in provinces around Canada and countries around the world.

When Paxlovid is being hailed as 89% effective and when both the media and the FDA are citing a press release as scientific evidence, you can color me skeptical.

So, what do we know?

Well, for starters, we know the AIDs drug ritonavir, which Paxlovid uses to slow the breakdown of nirmatrelvir, has a long history of adverse events, many of them linked to drug-drug interactions. You know, the kind of drugs that high-risk individuals are likely to be taking. Considering the fear surrounding the virus and the fact that, at least in my country, Paxlovid will be given to people who call in to a health hotline and not by someone with a full list of people’s medical history, any safety data around the drug is likely moot. There are strong incentives for someone living in fear of the virus to simply lie about the drugs they are taking because they do not understand the opportunity for adverse events will likely outweigh the benefits of the drug even if it was effective.

Of course, drug-drug interactions are well known and could be mitigated with cautious public health policy, but considering health officials have been elephants in a China shop for the last two years, I do not expect that. In any case, the problem is merely one of the limitations of the very controlled environment that randomized control trials take place in. Whether there are safety issues beyond the known ones remains to be seen, though I would be surprised if there are not.

More important is the efficacy. Again, we do not know if there were issues with randomization as we saw with Molnupiravir or problems as we saw in the Pfizer vaccine trials. Given the company we are talking about, I would be surprised if this trial was run honestly. However, as we wait for more data, we must take their numbers at face value. Note, there were 2 trials.

The EPIC-HR trial and the EPIC-SR trial. HR stands for high risk and SR stands for standard risk.

The EPIC-HR trial found that only 5/697 patients that took the drug were hospitalized or died after compared to 44/682 that took a placebo. The absolute risk reduction was 5.7% and the number needed to treat to reduce one hospitalization or death is 17.4. There were 9 deaths in the placebo group. Keep in mind, this is a trial with the highest risk group and the case fatality rate was 9/682 for the placebo group or 1.32%. Those deaths were not necessarily even from COVID or even with COVID. They were deaths within 28 days of getting COVID.

In the EPIC-SR interim analysis, 2/333 patients that took the drug were hospitalized compared to 8/329. This is a absolute risk reduction of 1.83% against hospitalization and a number needed to treat of 54.6. No patients died in either group for a case fatality rate of ZERO. Note that, all participants in this trial were unvaccinated. I am shocked that they found 662 unvaccinated patients willing to participate in a Pfizer trial at all, but this is actually quite damning to public health officials that have been screaming about an extinction event for the last two years.

The EPIC-SR trial after 80% of patients participated was 3/428 hospitalizations for the drug group compared to 10/426 for the placebo group bringing the absolute risk reduction down to 1.65% and the number needed to treat up to 60.7. Still. No. Deaths.

But we are also in an interesting transition period where there is a new variant that has been billed as “mild” meaning the absolute risk reduction has surely gone down, the number needed to treat has surely gone up, and yet Paxlovid has still been approved by regulators with no information. Meaning, people taking Paxlovid will be doing so under an Emergency Use Authorization with absolutely no scientific basis behind it. Given the fear and paranoia surrounding the virus, creating by health officials, there will be a strong incentive to lie to get the drug, which may lead to the kind of adverse events that even the corrupt media was not willing to ignore.

There are, of course, models that show Paxlovid may continue to be effective. This computational analysis is an interesting one, for example. The analysis looks at the binding energy of drugs against omicron. Yes, Paxlovid was shown to be effective. It was, in fact, the 7th most effective drug that ran through the simulation. The most effective drug? A well known drug, with few side effects, that was never approved for use against COVID: Ivermectin.