“An ounce of prevention is worth a pound of cure” — attributed to Benjamin Franklin, love him or hate him — rings especially true as we start to see rapid development of novel COVID-19 treatments. We are on the cusp of a number of new developments, each with promise of life-saving potential, and each carrying a risk of detracting from what should be the real goal: prevention.
As of 21 Feb 2022, over 10.4 billion vaccine doses had been administered worldwide. The vaccine remains 90% effective in reducing infections and hospitalizations. Yet, the biggest news story in the last month is the launch of Paxlovid (nirmatrelvir/ritonavir, or simply nir/rit), for the treatment of COVID-19. A reactionary measure.
The optimism is largely driven by the EPIC-HR randomized controlled trial. The study’s strengths include that it enrolled 2246 patients, and showed a reduction in hospitalization or death by 89 percent. It should be noted that this reduction is a relative result, which is classically used to inflate the purported benefits of interventions. Looking closer at the details, here’s another way to look at the outcomes:
- Control (i.e., placebo) group event rate = 6.5 percent. (In other words, for every 1000 people included, 65 were hospitalized or died.)
- Experimental (i.e., active drug) group event rate = 0.7 percent. (For every 1000 people, 7 were hospitalized or died.)
This calculates to an absolute risk reduction of 5.8 percent. Because the outcome is very important to patients, even a 5.8 percent reduction is remarkable, but is a far cry from the 89 percent presented above. Another way to slice and dice these numbers is to consider that we’ll see the outcome in one less patient for every 17* patients we treat with nir/rit.
The cost of the intervention has not yet been made public, but some peg it at several hundred dollars; one quote from an American source lists it at over $500 per treatment course. When costs do get released, we’ll be able to calculate how much it costs to get this benefit in one patient (approximately 17 courses times $cost/course). Governments and third-party payers often do these calculations when determining whether or not they’ll list drugs on formularies.
So a question we may want to be asking as a society is: What is the cost-effectiveness of the vaccine compared with the cost-effectiveness of COVID-19 treatments? The answer to this calculation has surely been proposed, but almost none will be able to account for what we don’t yet know about COVID-19 infections. There are many viruses that we’ve been plagued with over the course of human history that don’t fully manifest for years.
So a question we may want to ask as individuals is: Would you rather roll the dice with an active infection and chase it down with a new drug studied in less than 2500 patients? Or would you prefer to prevent the infection with a vaccine given over 10 billion times, which simply activates your body’s natural immune system?
A final critical consideration is data for treatments skewed by studies conducted in unvaccinated populations, and/or during periods of different COVID-19 variants. We don’t know if the mechanisms, and thereby the results, will hold up in vaccinated populations afflicted by the next variant of concern. That said, data for vaccines, masking, etc., appears to be holding for well over a year now.
Innovation is great, and truthfully, we need more treatments. However, until we find the magic bullet for COVID-19, we need more focus on prevention.
* Number needed to treat (NNT): A commonly used evidence-based medicine calculation to determine how many patients would have to be treated to see the effect in one patient. In short, the formula is 100/(absolute risk difference). When the outcome is a good thing, a low NNT is a good thing; a high NNT is less ideal and often very expensive (both financially and in unwanted side effects).