Monthly Archives: November 2017

Hyponatremia in SSTI

I recently cared for a patient who was an injection drug user who presented with fever, tachycardia, and redness/swelling of the forearm near an injection site. Her initial workup was remarkable for an elevated white blood cell count, hyponatremia to the high 120s, and hyperlactatemia. The patient was not eager to be hospitalized (as I have found is often the case with patients who inject drugs), and during our conversation about the risks:benefits of hospitalization and aggressive treatment with IV antibiotics and supportive care for what I felt was likely a very serious soft tissue infection with sepsis, I mentioned that the patient’s blood level of sodium was low and that I was concerned this was a serious and grave sign that their infection might be quite severe. When I went home after the shift, I wondered what that was founded on other than knowing that sodium was included in the LRINEC score and having had some experience in the past seeing patients with serious / necrotizing skin/soft tissue infections (SSTIs) who presented with hyponatremia.

I did a brief literature search and turned up a few articles, including one publication from one of our program’s recent graduates looking at our county hospital’s population of IVDU-associated STTIs.

Briefly, what I found in these articles was:

a.) Hyponatremia is a commonly-reported finding in patients with skin and soft tissue infections, and it seems to be found more commonly in patients with IVDU-associated infections. It was found in 41% of the patients from the Detroit group, 38% of the San Francisco group (compared to 27% of non-IVDU patients), and in the study looking specifically at necrotizing infections (from UCLA Harbor, with 31% of the 124 patients reporting IVDU) 56% of the patients had hyponatremia.

b.) When present, hyponatremia seems to be related to worse outcomes and is prognostically generally a bad sign. The group from SFGH did not report the association between low sodium and outcomes, but the group from Detroit and UCLA both showed significantly higher mortality in the patients with low serum sodiums at presentation.

Another question I asked myself was why do these patients get hyponatremia? I don’t think anyone knows for sure, but the authors from UCLA Harbor postulated the following in their paper:

Sepsis leads to increased muscle glucose uptake, increased lactate production and decreased utilization, an increase in the calculated ratio of muscle membrane permeabilities to Na+ and K+, and an increased intracellular Na+ concentration. These effects may be mediated by complement activation. In addition, sepsis has been linked to an increase in antidiuretic hormone level as well as adrenocortical insufficiency, both of which may lead to hyponatremia. Finally, severe NSTIs lead to marked third spacing of fluids, which may be replaced by free water, leading to hypovolemic hyponatremia.

I don’t know for sure how all this translates into clinical practice, but I do pay attention when I see low sodium in patients with infection, particularly if I’m concerned about how aggressive of a soft tissue infection this might be. But I also know that hyponatremia can be caused by heart failure, dietary deficiency, water intake, drugs, all sorts of things. It is a laboratory finding associated with so many different underlying actual problems, making it rarely useful in isolation. I’ve seen this often used as the reason for justifying admissions for patients with vague symptoms and many comorbidities but I think it is rarely the isolated cause of all of what the patient is experiencing. It’s also something that’s intimidating to be tasked with correcting in people, given the apparent stakes we learn about in medical school! It would be interesting to learn more about hyponatremia acts as a prognostic indicator in all those other conditions — I’m sure it isn’t good.


Kievlan DR1, Gukasyan M1, Gesch J1, Rodriguez RM2. Clinical profile of injection drug users presenting to the ED. Am J Emerg Med. 2015 May;33(5):674-6. PMID: 25744147. [PubMed] [Read by QxMD]
Yaghoubian A1, de Virgilio C, Dauphine C, Lewis RJ, Lin M. Use of admission serum lactate and sodium levels to predict mortality in necrotizing soft-tissue infections. Arch Surg. 2007 Sep;142(9):840-6; discussion 844-6. PMID: 17875838. [PubMed] [Read by QxMD]

Scooped yet again! Teaching Point on a Non-Traditional IO Application

One of my senior residents, now graduated and a fancy informatics fellow and general badass at Stanford, Christian Rose and I used to discuss how we wondered if in extenuating circumstances one could use an intraosseous drill / needle to trephine a skull and drain a rapidly-expanding intraaxial hematoma in a crashing patients. Like most emergency medicine doctors, I find life-saving procedures interesting and cool and even though I would never want the lifestyle that accompanies neurosurgery, being able to do these interventions in a resource-limited setting is something I aspire to if it had to be done. (I’ve seen a burr hole and clot evacuation done once by an ED physician who had been trained in the military and done them abroad in forward operating theaters — he borrowed a drill from the ortho OR. It was successful, appeared fairly easy, and saved a life.)

This was all just theoretical until I read this post from the EMCore conference blog by Peter Kas. He relates a story recently published as a case report in Injury of using an EZ-IO drill and a 25 mm 15 ga intraosseous needle to drain a rapidly-expanding epidural hematoma in a pedestrian who was struck by an auto who presented and had a rapid decline in GCS and clinical signs of herniation. This procedure took about 8 minutes — shaving and sterilizing the scalp, and was done in the anesthesiology prep area while the neurosurgeons prepped for a craniotomy. They were able to aspirate ~ 30 mL of blood, with resulting improvement in the pupillary exam. A formal craniotomy was completed, and the patient had a good outcome. 

The authors note that this is unlikely to provide definitive management, given that most significant hematoma requiring craniotomy are of a more solid consistency that would be difficult to drain entirely via this route, but that it might be utilized by ED providers needing to transfer a patient over a long distance to definitive care. They specifically describe how they envision the procedure:

We propose the site of insertion should be the point of maximal clot depth and therefore localisation of the insertion point would be on a case by case basis. Most extradural haematomas requiring emergency craniotomy are of sufficient size that we postulate that failure to place the needle into the haematoma is unlikely but a potential complication.

Anyway, fun fact of the day, and perhaps someday (hopefully this will never come up) a last-ditch hail mary move that is a little easier and possibly safer/faster than borrowing a drill from your local orthopod and hoping you don’t go too deep.


Bulstrode H1, Kabwama S2, Durnford A3, Hempenstall J4, Chakraborty A5. Temporising extradural haematoma by craniostomy using an intraosseous needle. Injury. 2017 May;48(5):1098-1100. PMID: 28238447. [PubMed] [Read by QxMD]