IV Thiamine in Septic Shock

Many think that the interest in IV thiamine in septic shock is a relatively novel idea after Paul Marik’s paper was published in CHEST back in 2017. It actually goes a lot further than that. We’ve actually known that 20% of critically ill patients have a thiamine deficiency upon presentation to the ICU. More recent data shows that this could be up to 70%. Over the life of this page, I have covered the renal benefits of thiamine on an older post, including a YouTube video on the matter. In addition, I have covered data reflecting on how you need thiamine to clear lactate and how this all decreases mortality on an older post. This paper I am going to be discussing today was published earlier this week in Critical Care Medicine. A link to the paper is down below. Unfortunately, it is not open access.

In this Japanese study, which is a retrospective observational cohort over the course of 7 years, they wanted to look at the effect of thiamine administration on mortality in patients with septic shock who were on vasopressors. In this case, norepinephrine. I really like that the initiation of thiamine was early in the sepsis course. Sounds great! I hope they leverage their data collection to also look at other proposed benefits listed above and currently in the process of being studied such as lactate clearance and renal function. That would be very helpful as those parameters could more easily be powered by this sample size than mortality.

Something is bothering me about this paper right from the start. In the introduction, they cited a paper written by Donnino where, as the authors of this paper state “found no differences in shock reversal or mortality”. While that is technically true for the entire study, in the subgroup analysis of the patients who were actually thiamine deficient, 35% of the participants, it reduced mortality from 46% to 13%. That’s a number needed to treat of just 3. Small sample size makes NNT less reliable, but you understand where I’m heading with that.

Defining Study Outcomes

When defining the study outcomes, they only looked at mortality. They did not check thiamine levels on these patients which is completely okay with me. Ultimately, the cost of the lab is around $90 in the US per data I’ve found. The cost of 4 days of 200mg IV twice a day for 4 days is only a little more than that. Every institution is different on cost, of course. Countries may vary even more.

IV Thiamine Dosing

The other trials looking at IV thiamine used larger doses than what was used here. Marik and Donnino, for example, used 200mg IV BID. Another study showing benefit written by Woolum, et al. used 500mg IV TID as their most common dose.
The dosing here is either 100 or 200mg IV daily.


Ultimately, there was no difference in outcomes. They ONLY looked at mortality. They did a bunch of statistical jumping jacks to no avail. No difference. It leaves us with an answer but more questions.
Could the dose not been high enough?
Could the prevalence of thiamine deficiency not be as high in Japan as it is in other parts of the country where the incidence of such is between 20-70%?
They harnessed the power of their national database for the mortality data, but it was unable to give them lactate levels. I wonder if renal function data is stored in there as they used ICD-10 codes. Could the authors leverage this data into two studies?

What does this paper offer regarding IV Thiamine?

It gives us a dose, 100mg or 200mg VI daily, that may be too low in Japan, and perhaps the rest of the world. It tells us that thiamine on its own will likely not change mortality. But the questions I present still remain. Hopefully the many studies on clinicaltrials.gov regarding thiamine will show something, including a much larger Donnino trial that will be complete in 2021 or 2022.


Miyamoto Y, Aso S, Iwagami M, et al. Association Between IV Thiamine and Mortality in Patients With Septic Shock: A Nationwide Observational Study [published online ahead of print, 2020 May 11]. Crit Care Med. 2020;10.1097/CCM.0000000000004394. doi:10.1097/CCM.0000000000004394

Link to Abstract

Although great care has been taken to ensure that the information in this post is accurate, eddyjoe, LLC shall not be held responsible or in any way liable for the continued accuracy of the information, or for any errors, omissions or inaccuracies, or for any consequences arising therefrom.

How to support my work: my efforts are at no cost to you and I would like to keep it that way. You have to look at ads on this website, listen to them on my podcast and YouTube content. Thanks for bearing with me. But if you want to help out a little more, also at no cost to you, consider a free trial with Audible where you will get a free book (and two books if you are an Amazon Prime member. If you CLICK HERE and sign up for Audible, they will provide me with a commission in exchange for you joining. They will remind you to potentially discontinue your membership so you don’t get charged. Thanks for your support!