What is Dr. Green’s rationale for Z-pak/azithromycin over other antibiotics? Seems dangerous!

As title. Does anyone happen to know? Z-pak is a P-gp/ABCB1 Inhibitor which doesn’t really make sense to me and elevated sirolimus concentration (apparently takes 20 mg sometimes, which may increase ~4.2 fold) with a need for the ED to hold nirmatrelvir/ritonavir in case of COVID-19 with superimposed bacterial pneumonia doesn’t make sense.

Closest human data I am seeing:
“Erythromycin: Erythromycin is a substrate and inhibitor of CYP3A4 and P-gp;
co-administration of sirolimus oral solution or tablets and erythromycin is not recommended
[see Warnings and Precautions (5.17), Drug Interactions (7.2)]. The simultaneous oral
administration of 2 mg daily of sirolimus oral solution and 800 mg q 8h of erythromycin as
erythromycin ethylsuccinate tablets at steady state to 24 healthy volunteers significantly
affected the bioavailability of sirolimus and erythromycin. Sirolimus Cmax and AUC were
increased 4.4- and 4.2-fold respectively and tmax was increased by 0.4 hr. Erythromycin Cmax
and AUC were increased 1.6- and 1.7-fold, respectively, and tmax was increased by 0.3 hr.”

Rapa and Z-pak both may QT prolong which may lead to deadly torsades

Personally, I use a potential anti-tumor Doxy cocktail of my own right now in case of bacterial infections, but so far haven’t used it yet for years. It’s also the empiric antibiotic of choice in my geographical area for respiratory pneumonia for my situation that doesn’t look viral, and also accounting for the experimental vaccines I took.

Note that empiric antibiotics require a physician to determine because they are often highly situation-specific (i.e. likely UTIs in canagliflozin in T2D require very different things than sepsis in different geographical areas) and antibiotic stewardship is necessary. This is not medical advice.


I agree. Doesn’t make alot of sense to me either especially for skin infections. I also prefer either doxycycline or cephalexin.


Wow, this seems quite the contraindication use.

Can you suggest something mainstream to use instead? Do we wait for swab lab results re bacteria type, and then consider choice?

You have to remember that a prolonged QT doesn’t mean that you’ll get a serious arrhythmia. It’s just a risk factor.

It’s usually not practical to wait for culture results, though sometimes it is, but there’s numerous options . Personally, I almost never take them.


I normally don’t either, but I’m asking in the context of high dose rapamycin and suppressed (theoretical) immuno? Especially maybe older folks?

That’s true.

I would also add it is possible people are taking things like grapefruit which further increases high drug doses and direct inhibition of IKr channels. Or because of a positive COVID test they may be taking HCQ/CQ. Hypokalemia and hypomagnesemia can be secondary to infection and dehydration. (COVID has hyper activated RAAS on top of that.)

As for some people here, they often have FHx of heart disease or already heart disease on diuretics. Older folks also have higher susceptibility generally.

1 Like

I have a few guesses.

Dr. Green may be old school.

As a pathologist he understands things way beyond my lane of medicine - but pathologist do not usually practice in a clinical setting.

Z-pack in general are a decent choice for travel and have broad coverage when a possible atypical infection is involved, but have become an overused go-to antibiotic for years.

All of Dr Tongs points are valid, but most medications have some type of interactions - so clarifying the degree of interaction is helpful otherwise we would never Rx anything:)

1 Like

Yes, agree. Skin is usually staph or strep and both of your suggestions or PCN’s should cover and not interact. Z-pack does seem an odd choice. @tongMD


I use topical mupirocin bid 5 days personally in skin things like limited impetigo/MRSA nares based on my geography and individual situation. I also nose breathe, mask, and some other maneuvers to keep any possible bacteria infection at bay without a bias toward interventionism. I’m quite surprised there is so little quality information on immunology/infectious disease out there despite a literal COVID-19 pandemic for 2 years.

No need to go systemic in many scenarios if you include experimental vaccines (mostly viral stuff but also somewhat uncommon stuff like GBS bacteremia that already show potential efficacy) and take care of possible sites of entry and identify any possibilities quickly.

I’d also clarify that it’s not just a minor drug-drug interaction as per the FDA label - it warrants therapy modification in most scenarios not just monitoring therapy, especially with Dr. G’s older age patient group prone to QTC prolongation. I want to give Dr. G the benefit of the doubt with proper due diligence first, but I couldn’t find his reasoning. I found it rather alarming that nobody mentioned these risks ever by sitting down and breaking down the entire claim and intervention risks, despite Dr. G stating bacterial infection is the biggest risk. Any rational risk manager would start with that evaluating that risk first.

Heart transplant patient here. I take Sirolimus to protect my donor heart.

Any time I’m sick I get immediate transfusions of Z-Pak and azithromycin. If it’s a virus, I’m usually given some plasma as well. They do this at the hospital I go to to protect the heart from further damage or to prevent infection that can be caused by other diseases or viruses.


@ashkir If you don’t mind, can you ask your transplant doc and/or doc managing meds why z-pak is ok? (Some places have a message option on patient portals) Especially in the case where older adults or other risk factors (ie drug interactions) where QTc prolongation increasingly becomes an issue. Relatively young patients without other risk factors might not be as much of a concern since there is some buffer on the QTc.

Especially my specific concern is rapa levels might get elevated. I’m assuming trough levels done?

I have an appointment with my heart transplant infectious disease doctor in a week, I can ask her :slight_smile:


The contrary opinion from Josh Farkas, MD.

About Josh Farkas
I am an assistant professor of Pulmonary and Critical Care Medicine at the University of Vermont (Burlington Vermont, USA). My training consists of medical school at Cornell University followed by Internal Medicine residency at Dartmouth University and a Pulmonary and Critical Care fellowship at Albany Medical Center. Although trained and board certified in both Pulmonology and Critical Care, I work exclusively as a clinician educator in the combined medical, cardiac, and neurological ICU. Interests include point-of-care ultrasonography, coffee, FOAMed, and all things related to the ICU.

1 Like

Azithromycin does have senolytic properties. So that’s one advantage.

Since his old article, there are new FDA warnings on cardiac death from new postmarketing surveillance reports. Animal models can be helpful, but new quality empirical data must be incorporated in humans when there’s enough pointing in the other direction (all macrolides have roughly the same amount of torsades cases) and pathological understanding has changed with new mechanisms.

We know the classically linked pathways are not just KCNH2 (LQT2 or previous old name in the article HERG) potassium channels as he claims. The bulk is from that, KCNQ1 (LQT1) and SCN5A (LQT3) combined - with other pathways from basic science adding up besides those 3 for LQTS. It seems his pathophysiology understanding is either vastly oversimplifying or largely incomplete at the time of writing. Not surprising since it’s a relatively old article.

Torsades is not a done deal pathophysiological process where it is completely understood or even close to it, as he implies - thus we should be prudent when strong predictors such as QTc numbers give big warning signs. As he mentions, “azithromycin prolongs the action potential itself”. Action potential prolongation at the ventricular cardiac cell level sets up increased vulnerability for early afterdepolarizations and triggers activity via re-entrant mechanisms, which is torsadogenic to initiate ventricular ectopy and torsades which is associated with long QT intervals. Keep in mind, there are multiple proposed mechanisms as well as a novel mechanism later found with azithromycin.

Not only that, I mentioned azithromycin has highly potent p-gp drug interactions that affect rapa a lot among many many other drugs on top of weak CYP3A4 inhibition. So his claim - minimal drug interactions is not true. Here’s an obvious example - increased risk of statin-associated rhabdomyolysis.

See 2021 FDA warning on the increased risk of acute cardiovascular death among all patients taking azithromycin - this shows potentially increased risk of acute cardiovascular mortality with azithromycin compared with amoxicillin exposure (HR 1.71, 95% CI 1.06-2.76). This includes reports of torsades.

Keep in mind, that transplant patients might face different risks than “healthy” users of weekly rapa hence if one is targeting specific bugs - z-pak is chosen over others because the mortality from dying of infection is far worse than risking prolongation of QT relatively speaking, but we simply don’t know exactly how much it matters for “healthy”.

Also, since his 2015 opinion - there are other links with azithromycin and other drugs interacting for much longer QTc and torsades than the article is implying:


It’s also no surprise the ACC/AHA has kept to the same guidelines with the last review - there is risk associated with azithromycin similar to other macrolides in humans. The association hasn’t changed since then to really be less concerned. If it did, I’d be more than happy because azithromycin has a decent range of bugs although there are some question marks regarding possibly higher likelihood of developing resistance. The generally cited reason for using azithromycin is more tolerability (particularly lower gastrointestinal toxicity) among patients than erythromycin.

Looking for contrary opinions might be useful in some cases, but one better have exceptional arguments with plenty of contexts (as the author himself warns) to go against the official recommendations in medicine because the base rate of contrarians being wrong is very high - while he has some points to consider, but there are clearly multiple inaccuracies in the article - so what he claims isn’t really a myth bust if I have to bust some of his myths.

Hence, I ask why exactly Dr. Green is using it first preferably straight from the source. Maybe he has something great to add that I haven’t thought of - or maybe he is just sticking to some old stuff as one suggested - I rather clarify.

Especially the newest extremely large datasets suggesting very high numbers of cardiac deaths relative to other antibiotics with azithromycin - if you just care about all-cause mortality (which is also far off the opposite against the author’s assertions of supposed low mortality) and don’t want to look into torsades in depth, as the causality is still not completely definitive, nor is the understanding extremely complete.


Fisetin is a supposed very hot “senolytic” that was a failure in the ITP. They even got one of the biggest proponents to get the dosing and experiment design right. Until azithromycin can have true replication even in preclinical, does it really matter when compared to empirically shown side effects in humans?


I asked and got an answer. I was told in transplant patients it reduces the risk of composite death. It’s a safer route for treating respiratory infections and preventing them in transplant patients. She said protocol and studies have shown it is the better alternative for us transplant patients.

I forgot the name of the other medicine, she said it’s very similar and it’s why I had to drink that yellow paint-like liquid for a few months to protect the lungs.


Did Dr Green say how long to take Azithromycin for? In the interview it sounded like one day if I remember correctly

He doesn’t say (or at least not to me). But it sounds like he recommends as short a time period as possible. So one day is likely correct if you are feeling any better.

1 Like