Sepsis. Bad stuff. Sepsis is when an infection, and occasionally other medical conditions, set off a cascade of out of control inflammation, damaging or destroying any number of organs and often leading to death.

As I said, bad stuff. Sepsis, in all its manifestations, has been one of the mainstays of my practice. Sepsis can be transient and respond rapidly to antibiotics and modern ICU supportive care. It can also be a relentless crash and burn, a rapid and uncontrolled spiral into death.

I have seen a few young meningococcal patients who were fine at breakfast and dead by dinner. Bad stuff.

For whatever reason, patients have recently been made aware of sepsis. For the past few years, it has not been unusual to have been asked by families if the patient was septic.

We have improved the management of sepsis over the years, mostly as we are better and more aggressive at supporting failing organs. But the basic approach has been the same for the 36 years I have been in infectious diseases: antibiotics, supportive care, and a gobbet of good luck.

Why some die and some survive is often a mystery. There is always variation in both the infecting organism and the host. All E. coli are not the same, for example, where different strains infecting identical mice will have different mortality rates. Conversely, the same E. coli strain will have different mortality rates in different strains of mice.

There are innumerable variations in the immune system that increase or decrease the risk of acquiring or dying from a variety of infections. My personal favorite is how variations in snot can increase the risk of meningococcal disease. As the Bard said, “The Sepsis, dear Brutus, is not in our stars / But in ourselves”. I will be long retired (3 months, 9 days, 2 hours, 11 minutes as I write this, but who is counting?) before we can scan the genome of a patient with a tricorder and point to the polymorphism that increased, or decreased, the chance of infection and death.

Vitamin C

Every once in a while a patient does survive sepsis who really should not have. It always fries my bacon when the family credits God with the save. They have no idea of the work and training the ICU team had accomplished to produce the ‘miracle’. Give credit where credit is due.

I have been around long enough to see many an adjunctive therapy come and go for sepsis. Xigris, steroids, and others. It is common in the ICU, perhaps more often for processes in which our interventions often fail and death is common, for doctors to glom onto therapies that appear to work based on suboptimal trials.

I have referred to it in the past as medical cold fusion: an effect that makes little sense, if any, when compared to known reality. What follows is a flurry of explanations about how the effect could be the real deal, some enthusiastic embracing of the intervention, slow discouragement, finally ending in better clinical trials that show, nope, the intervention did not do anything at all.

Sherman, set the Wayback machine to March 31, 2017, when you will find my last entry of the first series of my SBM bloggin’, “Vitamin C and Sepsis. All Sound and Fury? Much Ado About Nothing?

At the time I was, for lack of a better word, skeptical (me? Hardly.) that intravenous vitamin C and thiamine were actually effective in decreasing mortality in sepsis. But I noted at the time:

Never go in against vitamin C when death is on the line!

So the study was grudgingly embraced by our intensivists. For a while. You could not tell at the time if the intervention made any impact on outcomes. It did drive up the price of both IV vitamin C and thiamine. A lot. And then there were more clinical trials.

What had been impressive was the decrease in mortality in the vitamin C/thiamin study I evaluated last decade. From 40% to 8.5%. I mean, wow. That seemed too good to be true. And you know what follows. Was that result replicated?

Of course not. Based on eleven studies:

In this meta-analysis, the use of IV high-dose vitamin C in patients with sepsis was not associated with lower short-term mortality.


A total of 17 studies including 3133 patients fulfilled the predefined criteria and were analyzed. Pooled analysis indicated no mortality reduction in patients treated with vitamin C when compared to reference.

So another one bites the dust.

There were a few minor potential beneficial effects, which often show up in these studies. And they are likely meaningless, the usual background noise seen in complex medical interventions that looks promising in subgroup analysis or secondary endpoints. It’s usually crap:

…although it was associated with significantly shorter duration of vasopressor use and greater decline in the Sequential Organ Failure Assessment score at 72–96 hours.


…subgroup analyses revealed an improved survival if vitamin C treatment was applied for 3–4 days

Subgroup analysis is a Latin term used by statisticians for ‘likely bullshit’. Subgroup analysis invariably leads to nothing in later studies. I long ago learned to ignore subgroup analysis.

It was noted that:

…interestingly, six studies reported no adverse events related to the intervention, while three studies documented more frequent adverse events in patients treated with intravenous vitamin C (hypernatremia n= 24, hospital-acquired infections n= 14, hyperglycemia n= 13, gastrointestinal bleeding n= 3, and fluid overload n= 1).

Interesting? Naw. The usual noise in clinical trials.

Perhaps more interesting was the NEJM trial, published after the above meta-analyses:

In adults with sepsis receiving vasopressor therapy in the ICU, those who received intravenous vitamin C had a higher risk of death or persistent organ dysfunction at 28 days than those who received a placebo.

Oops. The numbers:

At day 28, 191 of 429 patients (44.5%) in the vitamin C group had died or had persistent organ dysfunction, as compared with 167 of 434 patients (38.5%) in the placebo group (risk ratio, 1.21; 95% confidence interval [CI], 1.04 to 1.40; P=0.01).

I have been convinced that a p needs to be at least .005 to perhaps represent a ‘real’ effect. So really? Vitamin C was no different than placebo. All the effects, good and bad, noted in these trials are, I would bet, just noise. More sound and fury, signifying nothing.

I mean, they didn’t even bother to collect “information regarding specific pathogens and the appropriateness of antimicrobial therapy”.

And that is a finding across all the studies that I could hunt down. No one cared about the microbiology, the source of infection, the host (for example, while transplant patients have more bacteremia, they may die less often) or the appropriateness of antibiotics. All of which make a difference in outcomes.

From the perspective of an ID doc, it renders all the studies of marginal applicability. Not all sepsis is the same, a concept apparently only an ID doc cares about. Perhaps why every study has demonstrated that ID involvement with infections of all kinds, including sepsis, improves outcomes: less morbidity and mortality. Odd how having someone who actually knows that they are doing improves care. Of course, since most of these studies were done by self-serving ID docs, well, go to Midas, get a muffler.

But it probably doesn’t matter. Vitamin C has a long history of no practical benefit for any process outside of scurvy, because there is no reason it should. Add sepsis to that list.


As mentioned above, having healthcare providers who know what they are doing improves care. What a concept. It makes me wonder how the opposite end of the spectrum impacts healthcare. Not just knowing nothing about diseases but having your medical approach is based on the nonsensical ravings of a lunatic mind (NRLM) cannot be good.

Great Britain at one time had 5 homeopathic hospitals, one as recently as 2018. I cannot tell if these institutions operated like real hospitals or not.

The Bristol Homeopathic Hospital continued to provide a full range of services.

Whatever those services that might cover. I can’t imagine presenting at a homeopathic hospital with sepsis where all they had to offer is nothing.

No one would rely solely on homeopathy for sepsis, or so one would hope. Homeopathy has been used as an adjunctive therapy, with no benefit. Much to my surprise, I discussed homeopathy and sepsis in 2009. The things I have done for which I have no memory. Sigh.

Despite being an “excellent piece of work on the role of homeopathy in critically ill patients“, at least according to a homeopath, there has been little research since I wrote that blog entry.

Part of the problem is the inability of septic patients to provide the bizarrely detailed and pointless history that guides the selection of homeopathic nostrums. Fear not. There is now, in caps:

…the Protocol for Objective Homeopathic Semiology for patients with an altered state of consciousness.

The Protocol is available online, but I will confess after reading the paper upon which it is based, I have no clue how one would apply it. The author’s website, in Portuguese, has no explanation I can find. “[S]eeking the actual characteristic symptomatic totality, with the hierarchization of homeopathic symptoms” is, I fear, beyond me.

The author goes on to demonstrate how the protocol was applied to a series (three) of septic patients:

Choice of patients for homeopathic intervention was based upon recognition that they were not responding satisfactorily to conventional treatment as judged by the attending ICU medical team.

This was followed by the application of 5 drops of a 30 C or 200 C homeopathic nostrum. A perforated gastric ulcer, a meningococcemia, and a pneumonia, all of whom survived. The conclusion?

The practice of homeopathy may effectively be extended beyond treating chronic cases toward the less familiar but highly important frontiers on the verge of life and death.

They actually thought homeopathy added to care of the patients. I keep wondering why the practice of homeopathy is not considered a DSM-IV delusional illness. NRLM indeed.


One meta-analysis notes:

Acupuncture has been used for various diseases, including severe infection, in China for more than 2,000 years.

That is true. There are multiple reports of infection and sepsis caused by acupunc… oh wait. This is acupuncture used to TREAT infection and sepsis. That’s different. Nevermind.


Previous studies reported that acupuncture at Zusanli (ST36) might be effective in treating sepsis

The point is just below the knee and lateral.

It is quite the point, stimulation of which has wide-ranging effects besides treating sepsis. In animal models,

…acupuncture at ST36 could be useful in reducing sepsis-induced injuries in heat, lung, kidney, liver, gastrointestinal tract, and immune system. Moreover, its potential mechanisms for antisepsis might include decreasing oxidative stress and inflammation, improving microcirculatory disturbance, and maintaining immune balance during sepsis.

I mean wow, truly the wonder drug that works wonders despite:

Poor methodological quality and publication bias exist.

How about in humans? And here we have a problem. I can’t access most of the original data from the review “Efficacy and safety of acupuncture as a complementary therapy for sepsis: a systematic review and meta-analysis” that purports to show a mortality benefit to adding acupuncture to standard care.

All the studies are from China, and as a result, are unfortunately suspect as has been mentioned before. As is always the case, a were included in the review, so there was no one acupuncture intervention that ‘worked’ – it is the concept that is effective.

The same issue, by the way, was in the meta-analysis that demonstrated decreased mortality in sepsis with Traditional Chinese Pseudo-Medicine: each trial in the series had a different treatment intervention but the collected data showed:

Addition of TCM has better effects in participants with sepsis.

It just does not matter the specific form of TCPM. I find that so…weird that these meta-analyses include very different interventions as if they were all the same. Meta-analyses are supposed to compare apple and apples, not dog shit and cow pies.

For the acupuncture trials:

Most included studies had an unclear risk or high risk of bias in allocation concealment and blinding of participants. In addition, included studies were mostly small sample single-center RCTs. Therefore, the degree of certainty around our findings is low to very low.

My certainty is high that acupuncture does nothing for any process so why would it alter the course of sepsis?

I went through the bibliography seeing if I could glean much from the titles and abstracts. Nope. More than a few dead links. None of the articles appeared to be evaluating mortality as a primary outcome, mostly looking at lab parameters or effects on the GI tract, such as “Electroacupuncture Improves Intestinal Dysfunction in Septic Patients: A Randomized Controlled Trial“, which found:

However, there were no significant differences in the days on MV, length of stay in ICU, and 28d mortality between two groups ( P>0.05).

And Yang et al., noted in the references, demonstrated that:

There was no significant difference of mortality in the 28th day between the two groups, with 5 deaths of 29 patients in the intervention group (17.2%) and 9 of 29 in the control group (31.0%).

But not included in the mortality section of the meta-analysis. Odd.

So in the end I can’t make detailed analysis of the meta-analysis, but that always begs the question: why bother to apply the concepts of reality to fiction?

Because it is what we do at SBM.


  • Mark Crislip, MD has been a practicing Infectious Disease specialist in Portland, Oregon, since 1990. He is a founder and  the President of the Society for Science-Based Medicine where he blogs under the name sbmsdictator. He has been voted a US News and World Report best US doctor, best ID doctor in Portland Magazine multiple times, has multiple teaching awards and, most importantly,  the ‘Attending Most Likely To Tell It Like It Is’ by the medical residents at his hospital. His growing multi-media empire can be found at

Posted by Mark Crislip

Mark Crislip, MD has been a practicing Infectious Disease specialist in Portland, Oregon, since 1990. He is a founder and  the President of the Society for Science-Based Medicine where he blogs under the name sbmsdictator. He has been voted a US News and World Report best US doctor, best ID doctor in Portland Magazine multiple times, has multiple teaching awards and, most importantly,  the ‘Attending Most Likely To Tell It Like It Is’ by the medical residents at his hospital. His growing multi-media empire can be found at