Dr. Jeremy, once again I'm wowed by your dedication to public health issues and particularly COVID while so many hide their heads in the sand. I have avoided COVID so far mostly due to vaccines and living a very isolated life as I'm highly immune compromised due to Progressive MS breathing issues, being in a wheelchair, "big gun" drugs for the MS that knock my B and T cells to zip and also having Common Variable Immune Deficiency, and heart/stroke issues related to MS. And this month I was diagnosed with breast cancer, again. I'm also on a satin drug and this is good practical info for people like myself. (How do you possibly find time for this? Hoping you relaxed a bit on vacation.) And since Med page just put out an article that people like myself with neurological issues being more susceptible to Long Covid issues. What we do to protect ourselves can so often help others more vulnerable. Then we can participate more in helping others. Prime example, the MA Wheelchair Repair Bill passed the state Senate yesterday largely because disabled chronically ill patients pushed themselves further to advocate for themselves and others. We worked on it right through the holidays, non-stop. We're all stronger together. Thank you for all your advocacy and research. It really does count. Happy New Year.
Very interesting article. I faced this dilemma when I had COVID in October. In the end, I did opt for stopping the Statin and taking Paxlovid. I didn't want a repeat event of the virus I had in Jan-Feb 2019. Which caused a Pericardial Effusion. The provider concurred with my thinking at the time that stopping the statin for a few days would not increase the risk too much.
This is so helpful: the drug interaction calculators vary so much--the Liverpool vs epocrates. Personally I am chronically on a medication that the Pfizer prescribing guidelines say to "consider lowering dose, clinical observation"--I have immune issues and would want Paxlovid, but anyone other than my PCP would hesitate.
The interaction calculators really are not up to the task, are they? Most are binary. But they certainly do not consider opportunity costs.
What I did not say in the piece is that for a lot of people, statin and Paxlovid use would probably be fine! (Please pharmacologists, do not pilfer me for saying that!) Like...yes there's that interaction, but what is the risk of muscle breakdown that "matters" medically? Is it higher than the benefits of the individual drugs? Probably not, in most cases. The risk of that interaction being meaningful is quite low.
But to say this without all the caveats? Anathema!
100%. I have them hold and restart. I feel that the habit of not prescribing is due to fatigue (bc we are trying meet the unrealistic demands of the system, laziness bc they use Lexicomp and do not actually read the advisory, are under-informed or misinformed (often bc of their political ideologies), don’t stop to think about the fact that statins are part of a long ball game plan, run into some pharmacists who make themselves an inappropriate blockade, and APPs with insufficient or no physician oversight are less likely be properly informed about how meds can be held and restarted without patient harm because they have less training, even if they are overall good at their role. It isn’t like this is top secret information either. COVID fatigue definitely plays a role. PPIs and anticoagulants are another group, though with warfarin, it’s a sub therapeutic problem and INRs need to be checked and Lovenox added if needed.
You are absolutely correct that what drives clinicians is "what can I do that will cause the lead administrative headache." You've pointed out that these blockades often nudge docs towards doing the wrong thing, just so they can get along to the next patient. And it might the right meta strategy. Better to see one more patient and fix a big problem for them than continue to tinker with smaller questions on the last one.
For myself, I am on 30mg of Prozac (Fluoxetine) per day, 1mg Clonozepam on Sundays and Wednesdays, and 0.5mg on Fridays, and 10mg generic Flexeril on Tuesday evenings. I could count on one hand the number of times I have forgotten to take my anti-depressant over the 30+ years I've been on it, but I remember well each time. By the end of the day after I forgot to take it the night before, I was feeling a little off. I imagine that I'd be feeling pretty bad with each subsequent day if I were to stop cold turkey, and I absolutely DON'T want to find out how bad. I have sometimes deliberately not taken my muscle relaxant and usually my muscles start to tighten up and I end up taking it later in the week. As for my Clonozepam, I sometimes skip that as well, and the biggest thing I notice is that I lose sleep. Not sure if I'm taking enough that I would have a severe anxiety reaction if I stopped entirely, but I've always tried to keep the frequency low enough to avoid such a dependence. I have an emergency 30 day supply of these medications just in case in my GO Bag so that I could control how fast I reduce my dose if all hell breaks loose.
See, this is so interesting and what I was hinting at towards the end of the article. There are some things that "less dangerous" but "more disturbing". Like...having back pain is not a medical emergency, but if you had to chose that versus hospitalization for Covid, you'd chose the back pain. On the flip side, if your risk of hospitalization for Covid is actually really low, like <1%, to begin with, it's not a bad gamble to keep taking a back pain medicine that makes Paxlovid impossible (hypothetical). The complexity ramps up when you introduce these kinds of considerations.
There are both "discontinuation syndromes" and "resuming syndromes" (from resuming a med that you stopped), depending on the medication. Certainly something to watch and consider. (I assume you're talking about the other drugs, not Paxlovid; as per the comment below, "paxlovid rebound" is a separate phenomenon.)
Thanks for bringing this up though I was under impression that "rebound" is a major issue with prescribing Paxlovid. I am touching on this in my next substack post. Also, my sisters were told that it is just a bug and do not need antiviral. I do think that stopping statins for a week isn't not enough of a concern considering that this is addressing a chronic condition that existed for years, if not decades, prior to taking statins.
One thing I do not see is looking at Remdesivir, though it is getting harder to get this Tx. Very close to Paxlovid in efficacy and less interaction. Of course, Molnupiravir is last resort for really high risk Px due to it low efficacy and mutagenic properties.
There are others in the pipeline that I am addressing as well. Seems at VV116 and Ensitrelvir Fumaric Acid seem promising. Seems that the FDA put the on a fast track? Where are we at in this process?
Paxlovid-associated rebound is still an open question. Some say it is overblown and patients are not more likely to have a rebound of viral levels or symptoms if they take Paxlovid. My read of the overall data is that Paxlovid-associated rebound is real, and exceeds "background rates." I consider rebound in the risk-benefit calculation of lower risk patients. In higher risk patients, the risk of rebound becomes a rounding error compared to the hospitalization reductions of giving it.
I'm less enthusiastic about remdes and molnupiravir, in all honesty. I think both were really nice options in the pre-vaccine era, especially nice for high risk patients. But I've not seen great data showing their utility for the general population in the vaccine era.
I think you are asking about the difference between relative reduction and absolute reduction.
To compare two conditions, we have to do absolute reduction.
Let me use some hypothetical #s to illustrate.
Drug 1: Without treatment, 10% of people get hospitalized. With treatment 7% do. That is a relative reduction of 30% (3/10=0.3) and an absolute reduction of 3% (10-7=3). That means for every 100 people treated, 33.3 fewer would be hospitalized.
Drug 2: Without treatment, 1% of people get hospitalized. With treatment 0.7% do. That is a relative reduction of 30% (0.03/0.1=0.3) [no change compared to above!] and an absolute reduction of 0.3% (0.1-0.07=0.3). That means for every 100 people treated, 3.3 fewer would be hospitalized.
For the Paxlovid analysis above, 0.39% to 0.08%—is a relative reduction of 80% (!), but still an absolute reduction of 0.31% (0.0039 - 0.0008). That means for every 322 people treated, 1 fewer would die. Note: these are vaccine-era data. Pre-vaccine, it was much, much, much more dangerous to not get Paxlovid because the death rate without it was way, way higher than 0.39% in folks over age 65.
I realize that does not sound like much....and it isn't. But that's another story.
Upshot: paxlovid still works...but it is making a smaller difference than in the past.
In the nonmedical world, an absolute increase or decrease is always stated as "percentage points", not "%". It would be helpful to always state absolute increase/decrease if that's what you mean. Two example from the post that I found confusing, because you mean percentage points (absolute) but just use the % symbol without saying it's absolute:
"we can conservatively assume that the post-heart attack mortality rate for the first year is around 0.66% every 10 days. … If we assume that stopping the statin immediately erases the entire benefit of taking the drug … the 10-day mortality rate would rise to 0.82%—that is, an increase of 0.16%."
"Paxlovid use was associated with a decrease in mortality from 0.39% to 0.08%—a decrease of 0.31%."
Dr. Jeremy, once again I'm wowed by your dedication to public health issues and particularly COVID while so many hide their heads in the sand. I have avoided COVID so far mostly due to vaccines and living a very isolated life as I'm highly immune compromised due to Progressive MS breathing issues, being in a wheelchair, "big gun" drugs for the MS that knock my B and T cells to zip and also having Common Variable Immune Deficiency, and heart/stroke issues related to MS. And this month I was diagnosed with breast cancer, again. I'm also on a satin drug and this is good practical info for people like myself. (How do you possibly find time for this? Hoping you relaxed a bit on vacation.) And since Med page just put out an article that people like myself with neurological issues being more susceptible to Long Covid issues. What we do to protect ourselves can so often help others more vulnerable. Then we can participate more in helping others. Prime example, the MA Wheelchair Repair Bill passed the state Senate yesterday largely because disabled chronically ill patients pushed themselves further to advocate for themselves and others. We worked on it right through the holidays, non-stop. We're all stronger together. Thank you for all your advocacy and research. It really does count. Happy New Year.
Very interesting article. I faced this dilemma when I had COVID in October. In the end, I did opt for stopping the Statin and taking Paxlovid. I didn't want a repeat event of the virus I had in Jan-Feb 2019. Which caused a Pericardial Effusion. The provider concurred with my thinking at the time that stopping the statin for a few days would not increase the risk too much.
Sounds like you had a conversation that led to the right outcome!
This is so helpful: the drug interaction calculators vary so much--the Liverpool vs epocrates. Personally I am chronically on a medication that the Pfizer prescribing guidelines say to "consider lowering dose, clinical observation"--I have immune issues and would want Paxlovid, but anyone other than my PCP would hesitate.
The interaction calculators really are not up to the task, are they? Most are binary. But they certainly do not consider opportunity costs.
What I did not say in the piece is that for a lot of people, statin and Paxlovid use would probably be fine! (Please pharmacologists, do not pilfer me for saying that!) Like...yes there's that interaction, but what is the risk of muscle breakdown that "matters" medically? Is it higher than the benefits of the individual drugs? Probably not, in most cases. The risk of that interaction being meaningful is quite low.
But to say this without all the caveats? Anathema!
100%. I have them hold and restart. I feel that the habit of not prescribing is due to fatigue (bc we are trying meet the unrealistic demands of the system, laziness bc they use Lexicomp and do not actually read the advisory, are under-informed or misinformed (often bc of their political ideologies), don’t stop to think about the fact that statins are part of a long ball game plan, run into some pharmacists who make themselves an inappropriate blockade, and APPs with insufficient or no physician oversight are less likely be properly informed about how meds can be held and restarted without patient harm because they have less training, even if they are overall good at their role. It isn’t like this is top secret information either. COVID fatigue definitely plays a role. PPIs and anticoagulants are another group, though with warfarin, it’s a sub therapeutic problem and INRs need to be checked and Lovenox added if needed.
You are absolutely correct that what drives clinicians is "what can I do that will cause the lead administrative headache." You've pointed out that these blockades often nudge docs towards doing the wrong thing, just so they can get along to the next patient. And it might the right meta strategy. Better to see one more patient and fix a big problem for them than continue to tinker with smaller questions on the last one.
For myself, I am on 30mg of Prozac (Fluoxetine) per day, 1mg Clonozepam on Sundays and Wednesdays, and 0.5mg on Fridays, and 10mg generic Flexeril on Tuesday evenings. I could count on one hand the number of times I have forgotten to take my anti-depressant over the 30+ years I've been on it, but I remember well each time. By the end of the day after I forgot to take it the night before, I was feeling a little off. I imagine that I'd be feeling pretty bad with each subsequent day if I were to stop cold turkey, and I absolutely DON'T want to find out how bad. I have sometimes deliberately not taken my muscle relaxant and usually my muscles start to tighten up and I end up taking it later in the week. As for my Clonozepam, I sometimes skip that as well, and the biggest thing I notice is that I lose sleep. Not sure if I'm taking enough that I would have a severe anxiety reaction if I stopped entirely, but I've always tried to keep the frequency low enough to avoid such a dependence. I have an emergency 30 day supply of these medications just in case in my GO Bag so that I could control how fast I reduce my dose if all hell breaks loose.
See, this is so interesting and what I was hinting at towards the end of the article. There are some things that "less dangerous" but "more disturbing". Like...having back pain is not a medical emergency, but if you had to chose that versus hospitalization for Covid, you'd chose the back pain. On the flip side, if your risk of hospitalization for Covid is actually really low, like <1%, to begin with, it's not a bad gamble to keep taking a back pain medicine that makes Paxlovid impossible (hypothetical). The complexity ramps up when you introduce these kinds of considerations.
Very interesting and helpful. Is there any chance of a rebound effect from suddenly stopping any of those meds?
There are both "discontinuation syndromes" and "resuming syndromes" (from resuming a med that you stopped), depending on the medication. Certainly something to watch and consider. (I assume you're talking about the other drugs, not Paxlovid; as per the comment below, "paxlovid rebound" is a separate phenomenon.)
What do you mean by rebound? With so much noise about "paxlovid rebound", just asking for clarification.
Thanks for bringing this up though I was under impression that "rebound" is a major issue with prescribing Paxlovid. I am touching on this in my next substack post. Also, my sisters were told that it is just a bug and do not need antiviral. I do think that stopping statins for a week isn't not enough of a concern considering that this is addressing a chronic condition that existed for years, if not decades, prior to taking statins.
One thing I do not see is looking at Remdesivir, though it is getting harder to get this Tx. Very close to Paxlovid in efficacy and less interaction. Of course, Molnupiravir is last resort for really high risk Px due to it low efficacy and mutagenic properties.
There are others in the pipeline that I am addressing as well. Seems at VV116 and Ensitrelvir Fumaric Acid seem promising. Seems that the FDA put the on a fast track? Where are we at in this process?
Paxlovid-associated rebound is still an open question. Some say it is overblown and patients are not more likely to have a rebound of viral levels or symptoms if they take Paxlovid. My read of the overall data is that Paxlovid-associated rebound is real, and exceeds "background rates." I consider rebound in the risk-benefit calculation of lower risk patients. In higher risk patients, the risk of rebound becomes a rounding error compared to the hospitalization reductions of giving it.
I'm less enthusiastic about remdes and molnupiravir, in all honesty. I think both were really nice options in the pre-vaccine era, especially nice for high risk patients. But I've not seen great data showing their utility for the general population in the vaccine era.
Welcome back Dr Faust,
Would you clarify the efficacy of Paxlovid in preventing covid hospitalization and/or death? .31%
Thanks!
I think you are asking about the difference between relative reduction and absolute reduction.
To compare two conditions, we have to do absolute reduction.
Let me use some hypothetical #s to illustrate.
Drug 1: Without treatment, 10% of people get hospitalized. With treatment 7% do. That is a relative reduction of 30% (3/10=0.3) and an absolute reduction of 3% (10-7=3). That means for every 100 people treated, 33.3 fewer would be hospitalized.
Drug 2: Without treatment, 1% of people get hospitalized. With treatment 0.7% do. That is a relative reduction of 30% (0.03/0.1=0.3) [no change compared to above!] and an absolute reduction of 0.3% (0.1-0.07=0.3). That means for every 100 people treated, 3.3 fewer would be hospitalized.
For the Paxlovid analysis above, 0.39% to 0.08%—is a relative reduction of 80% (!), but still an absolute reduction of 0.31% (0.0039 - 0.0008). That means for every 322 people treated, 1 fewer would die. Note: these are vaccine-era data. Pre-vaccine, it was much, much, much more dangerous to not get Paxlovid because the death rate without it was way, way higher than 0.39% in folks over age 65.
I realize that does not sound like much....and it isn't. But that's another story.
Upshot: paxlovid still works...but it is making a smaller difference than in the past.
In the nonmedical world, an absolute increase or decrease is always stated as "percentage points", not "%". It would be helpful to always state absolute increase/decrease if that's what you mean. Two example from the post that I found confusing, because you mean percentage points (absolute) but just use the % symbol without saying it's absolute:
"we can conservatively assume that the post-heart attack mortality rate for the first year is around 0.66% every 10 days. … If we assume that stopping the statin immediately erases the entire benefit of taking the drug … the 10-day mortality rate would rise to 0.82%—that is, an increase of 0.16%."
"Paxlovid use was associated with a decrease in mortality from 0.39% to 0.08%—a decrease of 0.31%."