The Federal government has announced a policy of providing Pfizer’s anti-viral drug Paxlovid for free, to anyone with a positive COVID-19 test. I believe that a key aspect of this plan is that you can bypass going to a physician for a prescription. Hence, in theory, it will be immediately available to anyone with evidence of a positive COVID-19 test.
This drug is an anti-viral, that is, it interferes with viral DNA/RNA replication. (This news reporting explains that point, along with the proposed policy.) The idea behind providing this drug is to prevent severe COVID-19. With luck, the anti-viral will slow down the COVID-19 enough that a person’s immune system can tackle it without the need for hospitalization.
Paxlovid is advertised as having reduced the risk of hospitalization from COVID-19 by 90%. But that’s not true.
Or, to be more precise, it’s not true for the average person who tested positive for COVID-19. If you actually read a summary of the clinical trial, Paxlovid resulted in a roughly 90% reduction in hospitalization for:
- Persons with a positive COVID-19 test
- whose symptoms began less than five days earlier,
- and who had not yet been hospitalized,
- who had at least one significant comorbidity,
- And were unvaccinated
- And had no prior COVID-19 infection
- And who were not expected to receive monoclonal antibodies.
You can read a summary of the details at this reference. But all you really need to read is the title, which is:
Source: Same reference as above.
In other words, this drug reduced hospitalizations by 90% when it was a) the only treatment offered to prevent hospitalization, b) in a population with a high likelihood of hospitalization.
In particular, note that this drug was only given to individuals with no prior immunity to COVID-19. That means no vaccine, and no history of infection. By themselves, those two factors would likely have eliminated 90 percent or more of the U.S. adult population as being candidates for use of this drug.
I’m just trying to point out the vast gap between the population that the drug was tested on (those at high risk for hospitalization) and the population that can now receive the drug for free (any adult with a positive COVID-19 test).
Recall that vaccine/booster already reduces risk of hospitalization by about 90 percent (immediately following the booster shot). And that hospitalizations are highly concentrated among the elderly, and those with multiple risk factors.
Given that, it’s a pretty fair guess that if you give this drug to everyone — vaccinated or not, elderly or not, high-risk or not — you aren’t going to achieve anything like a (further) 90% reduction in hospitalizations. Most of the people you’d be giving this to would be at only the slightest risk of hospitalization in any case.
A crude cost-benefit calculation
The Feds will be shelling out $530 per course of treatment with Paxlovid. Not that anybody ever worries about cost.
The last time I looked into it, the average COVID-19 hospitalization cost somewhere between $20K and $25K. That was based on a then-current analysis of the cases covered and paid by Medicare.
Just looking at the round numbers, in order for Paxlovid to be cost-effective, for every 100 courses of treatment provided (at a cost of about $50,000), it would have to eliminate two hospitalizations (at a cost of about $25,000 each).
Sure, that’s only focused on the hospital facility bill. There may be disability beyond that, there will certainly be physician and other bills beyond the facility bill. There’s always the issue of the value of life (the dollar value assigned to preventing avoidable deaths).
But in a crude sense, to be roughly cost-effective, this drug needs to avoid two hospitalizations for every 100 courses of treatment given.
That raises an obvious minimum question: For what segment of the population does the crude case hospitalization rate for Omicron exceed 2 percent? That is, assuming this drug would eliminate all hospitalizations, for what portion of the population would this likely be cost-effective?
I have to do this by approximation, because there’s no single data source that lets me get at the answer. But as I weed-whack my way though the available data, among readily-identifiable groups, I get the sense that only the unvaccinated age 40+ have an Omicron case hospitalization rate in excess of 2 percent. (To those individuals you could plausibly all individuals age 85+).
Best guess, doing a quick back-of-the-envelope, about 15 percent of the U.S. adult population falls into the category of age 40+ and unvaccinated. Even assuming this drug absolutely eliminates the possibility of hospitalization, it’s not cost-effective for perhaps 85% of the population that’s now going to be eligible for it.
Contrast to vaccines.
Here’s what I find so odd about this. It’s not just about the money.
It’s not as if this drug is plentiful. To the contrary, near as I can figure, the U.S. will only buy enough pills for roughly 100,000 courses of treatment in March, upping that to 200,000 courses of treatment in April. (That’s based on stated counts of pills, from this source, with a five-day course of treatment requiring ten pills).
Contrast the 100,000 courses of treatment per month to the 58,000 new cases per day that we’re seeing currently in the U.S. The U.S. is buying enough doses to treat a few percent of new cases. But they’re making those doses available to 100 percent of new cases.
The last time a COVID-19 treatment was scarce, and made freely available, the government restricted access to those most in need. And here, I’m talking about vaccines. Vaccines were first made available to the oldest old, then to higher-risk persons within the general population. And only after those high-risk populations were taken care of could the average low-risk individual receive a vaccine.
By contrast, here we have a drug that was only tested for the highest-risk population: Individuals with comorbidities indicating high risk, and with no prior immunity (from vaccine or from prior infection). But for this latest treatment, the stated policy is a straight-up first come, first served.
I can’t quite grasp what’s changed. Maybe the Feds think that nobody’s going to want this treatment. Maybe they figure that most cases are mild enough that people won’t demand a this free anti-viral. (But that never stopped people from demanding useless antibiotics to treat colds and flu.) Maybe this is just a confidence-building measure. No matter who you are, you can be assured that if you get COVID, you’ll have immediate access to this treatment? So the impact is intended to be more psychological than medical?
Beats me. All I can say for sure is that’s an oddity to test a drug on one highly-circumscribed population, and then make that drug available to all comers. I don’t believe I’ve ever seen that before.
This still has to get Congressional funding to be implemented. At $500 a throw, you’d think somebody in the Congress would use the lack of cost-effectiveness as a reason to impede the President’s initiative, regardless of its merits or lack of merits.
I guess we’ll see how this plays out. But I just can’t see this sailing through the Congress with nobody questioning the discrepancy between the target population for this drug, as tested, and the target population for this drug, as delivered.
