TOX: Sweet talk & slurred speech



About halfway through your relatively quiet shift in the Critical Care & Trauma area of your ED you hear an overhead notification, “stroke code to triage!”

You and Padma the intern set down your tasty cappuccinos and speed walk to the triage area where you see a middle aged man slumped to one side and drooling on the non-disposable blood pressure cuff.

“What’s your name?” You ask.

“Eeeeuuuggggghhhhh…” he replies, in a manner which provides valuable information that his airway is indeed patent.

“Alright. Padma, please continue the assessment and verbalize your thought process,” you state, simultaneously noting tachycardia to 124 with otherwise normal vital signs on the machine.

Padma instructs the technician to get a finger stick and ascertain whether the CT machine is ready. “Airway intact and protected . . . bilateral symmetric breath sounds. . . “

“Sugar is 16 mg/dL,” mumbles the technician.

“. . . Intact carotid pulses . . . no facial asymmetry . . . moving all four extremities . . . WHAT?!” Padma continues in a loud clear voice, “Ok! Place an IV and push 50 grams of dextrose now! Draw labs right after! And cancel that stroke code.”

Within moments of receiving the intravenous dextrose the patient starts to perk up. Now, he appears awake and alert, but his affect is blunt and he speaks very little.


“What happened?” you inquire

“I just want to go home. Leave me alone, please,” he says to no one in particular, and shakes his head “no” in response to all questions.

Padma is trying to coax him into eating a PB&J sandwich, but the patient refuses to eat. You get the sense that there’s a lot more to this story but need to leave the patient’s bedside for 45 minutes when two patients with gunshots to the abdomen are brought in.

Moments after dispositioning both new patients to the operating room:


You receive a call from chemistry reporting the following:

Sodium               148 mEq/L
Potassium           2.6 mEq/L        CRITICAL
Chloride              110 mEq/L
Bicarbonate         24 mEq/L
BUN                      18 mg/dL
Creatinine          1.12 mg/dL
Glucose              150 mg/dL
Phosphate           1.0 mEq/L       CRITICAL
Magnesium          1.7 mEq/L


You immediately reassess the patient.

You note that the patient is diaphoretic and tremulous. “Padma, what do you want to do?” you inquire.

“Let’s recheck his blood glucose and give 40mEq of potassium phosphate. I had an advisor in medical school tell me once, ‘I don’t give a flying **** about phosphate except in DKA.’ I can’t remember why DKA though. Anyways, back to our patient!”

She asks technician to please repeat his finger stick, and then continues, “We still don’t know why he’s hypoglycemic with low potassium and low phosphate. He may be on a sulfonylurea … could have alcoholism with depleted glycogen stores … maybe he’s septic … I guess renal or liver failure can do this too… ”

As Padma continues to list the possibilities, the glucometer beeps and displays 38 mg/dL. You push another 50 grams of dextrose into the patient’s IV and appreciate how badly you need more information. As you inject the fourth ampule of viscous fluid, you conclude this cannot be the most efficient path to euglycemia.


The most efficient path continues...

“Place a central line,” you state with the understanding that profound hypoglycemia over the course of 45 minutes is rather abnormal. “We’re going to start an infusion of D25, and that infusion can’t be given peripherally because of the degree of hypertonicity. A typical dose for refractory hypoglycemia would be glucose at 400-1,000 mg/kg/hour which in a 70 kg man would be 28 – 70 grams of dextrose per hour. That seems low for this man who became hypoglycemic after 45 minutes – so let’s increase it! We’ll give him 75 grams of dextrose per hour, so the rate will start at 300 mL/hour and we will titrate down based on hourly finger stick readings.”

As the last suture stabilizes the central line, and the infusion is initiated, a woman comes into the ED looking for your patient. She is visibly concerned and when asked if she knows what happened to the patient, she pulls out a grocery bag.


“What happened?” You ask

“I found these in the trash!” she utters, handing you 3 empty vials and a banana peel.


Your heart sinks as you do the math and think, “There are 3,000 units of insulin in his body right now. Holy crap. Now what?”


Now what indeed.

You grab your ultrasound machine and start scanning common injection sites: thighs, deltoids, buttocks and abdomen. You quickly find this sitting right beneath his umbilicus:


“This is it!” you think. “This is the insulin depot! I calculate this fluid pocket is about 12mL. It’s only as deep as a cutaneous abscess, and if I can get it out, his insulin burden will decrease by 1,200 units!”

With that, you say to Padma, “Go get a scalpel!”


Real life
This is a fictionalized patient encounter based on multiple case reports of long-acting insulin overdose. Even with overdoses of short-acting insulin patients are often admitted to ICU for several days with recurring episodes of hypoglycemia for up to 5 days after initial presentation.

Glucagon should be considered if there is a delay to IV access. Glucagon is ineffective if hepatic glycogen stores are depleted.

One hour finger stick monitoring, concentrated glucose infusion via central line, and electrolyte repletion are the cornerstones of insulin overdose.

Source control via incision and drainage should be considered if a collection of subcutaneous insulin is identified. If there are multiple injection sites, consider surgical consultation for adipectomy. (12)

Enteral glucose is preferred if a patient has a normal mental status, and hypoglycemic patients should be encouraged to eat. Enteric tube feeds should be considered if clinically indicated.


Science (pharmacodynamics of glucose)
In one study 25 grams of dextrose (one “amp”) had a mean elevation in blood glucose of 166 mg/dL with a range of 37-370 mg/dL. (1)

In a study of hypoglycemia secondary to insulin, 20 grams of dextrose taken orally increased serum glucose from 60 mg/dL to 120 mg/dL in one hour. Ten grams of dextrose taken orally increased serum glucose from 60 mg/dL to 100 mg/dL in one hour. (1)

One can of soda has 25 – 50 grams of sugar.


This website has the sugar quantities of common candy treats.


Science (electrolytes)
As all clinicians must know from the management of hyperkalemia, insulin activates the Na/K ATPase pump and thereby lowers serum potassium via intracellular shift.

Insulin also drives glucose intracellularly, where it is immediately phosphorylated by the enzyme hexose kinase, thus sequestering both glucose and phosphate (through phosphorylation) within the cell. Insulin insfuions and insulin overdoses are rare situations in which an emergency department physician must give a flying **** about phosphate because the level may drop to dangerously low levels.

Clinically significant hypophosphatemia may lead to impaired cellular metabolism due to inability to convert ADP to ATP. Effects of hypophosphatemia may include dysrhythmias, heart failure, rhabdomyolysis, seizures, CNS dysfunction, and diaphragm weakness.


Take home points
– Lethality from insulin overdose is primarily from hypoglycemia and hypokalemia.
– Consider central line and D20 infusion for refractory hypoglycemia.
– Get a point of care blood glucose on stroke code patients prior to head CT.


Questions for discussion
  1. What are the management options for hypoglycemia from sulfonylureas?
  2. What other toxic substances can cause acute hypoglycemia?
  3. What medications may allow for diabetic ketoacidosis in the setting of euglycemia?


Bonus question


This is Mycroft Dorian-Gray Blumenberg. He is a criminal.
A physician providing reassurance is classically referred to as giving “medicine through the ear.” What famous poisoning in renaissance fiction takes place through the ear?


Written by Adam Blumenberg MD

Special thanks to Drs. Jackie Shibata MD and Sage Weiner MD


  1. Goldfrank’s Manual of Toxicologic Emergencies 10th edition, Chapter 53 Antidiabetics and Hypoglycemics
  2. Rosens Emergency Medicine 7th edition, Section eleven, Metabolism and Endocrinology
  3. Emedicine: hypoglycemia
  4. Emedicine: hypophosphatemia
  5. Fuller ET, Miller MA, Kaylor DW, Janke C. Lantus overdose: case presentation and management options. The Journal of emergency medicine. 36(1):26-9. 2009.
  6. Lu M, Inboriboon PC. Lantus insulin overdose: a case report. The Journal of emergency medicine. 41(4):374-7. 2011.
  7. Doğan FS, Onur ÖE, Altınok AD, Güneysel Ö. Insulin glargine overdose.Journal of Pharmacology & Pharmacotherapeutics. 2012;3(4):333-335. doi:10.4103/0976-500X.103694.
  8. Karatas F, Sahin S, Karatas HG, Karsli PB, Emre C, Kivrakoglu F. The highest (3600 IU) reported overdose of insulin glargine ever and management. Indian Journal of Critical Care Medicine : Peer-reviewed, Official Publication of Indian Society of Critical Care Medicine. 2015;19(12):750-751. doi:10.4103/0972-5229.171418.
  9. Gundgurthi A, Kharb S, Dutta MK, Pakhetra R, Garg MK. Insulin poisoning with suicidal intent. Indian Journal of Endocrinology and Metabolism. 2012;16(Suppl1):S120-S122. doi:10.4103/2230-8210.94254.
  10. Eldred AE, Mustafa OG, Hunt KF, Whitelaw BC. Problem based review: the patient who has taken an overdose of long-acting insulin analogue. Acute medicine. 12(3):167-72. 2013.
  11. Fromont I, Benhaim D, Ottomani A, Valéro R, Molines L, Vialettes B. Prolonged glucose requirements after intentional glargine and aspart overdose. Diabetes & metabolism. 33(5):390-2. 2007.
  12. Droste, J., Hundia, V., Pettit, A., Narayan, N. and Nejim, A. (2012), Excision of injection site substantially reduced serum insulin concentration in a potentially life-threatening insulin analogue overdose. Practical Diabetes, 29: 243–245. doi:10.1002/pdi.1700
  13. I. W. Campbell and J. G. Ratcliffe, “Suicidal insulin overdose managed by excision of insulin injection site,”British Medical Journal, vol. 285, no. 6339, pp. 408–409, 1982.
  14. McIntyre AS, Woolf VJ, Burnham WR. Local excision of subcutaneous fat in the management of insulin overdose. The British journal of surgery. 73(7):538. 1986.
  15. von Mach MA, Meyer S, Omogbehin B, Kann PH, Weilemann LS. Epidemiological assessment of 160 cases of insulin overdose recorded in a regional poisons unit. Int J Clin Pharmacol Ther. 2004 May;42(5) 277-280. PMID: 15176650.


The following two tabs change content below.
Interests: Emergency medicine, toxicology, humanism, cooking & eating, science fiction, movies, travel, cats, photography, film making.

Latest posts by ablumenberg (see all)

1 comment for “TOX: Sweet talk & slurred speech

  1. Lee
    September 13, 2016 at 6:52 pm

    I must say I often have a twinge of anxiety when I prescribe the standard long and short acting insulin to patients in clinic given the fact they are potentially lethal medications with a mere slip up! I give a lot of credit to IDDM patients.

    1. In experience I’ve just had them on D5 1/2NS at maintenance with frequent FS, occasionally with some octreotide if we think the patient took too much of there meds (it inhibits endogeous insulin release which is exactly what sulfonylureas stimulate). The octreotide can be given as bolus doses so I have never bothered with a ggt. Also if they can eat, the complex carbohydrates are probably helpful. Glucagon apparently can be a temporizing measure if the patient’s hepatic function is intact. I never used it.

    2. Other than insulin (endogenous or exogenous) and sulfonylureas we have the meglitinides. There are a whole host of drugs less strongly associated with hypoglycemia including quinolones, antimalarials, and pentamidine.
    Also although this is not toxicology-related, I have certainly seen more than one patient with inaccurate finger stick readings 2/2 vasculopathies (e.g. scleroderma, SLE). In the absence of symptoms, compare a finger stick to a venous blood sample. I had a woman admitted to KCHC for pseudohypoglycemia who was furious and frustrated. I read up on cases of pseudohypoglycemia in patients with SLE and keep it in the back of my mind. Just recently at UHB we saw a patient with a similar issue.

    3. As for ketoacidosis and euglycemia, I had a hunch it must be the SGLT2 inhibitors (the flozins) because I could imagine that mechanism (increased gucosuria without increased insulin). FDA has a warning out. These new drugs do wonders to lower serum glucose, but may accelerate ketoacidosis and decrease ketone clearance. I imagine when patients are started on flozins the regular dosing of insulin is decreased given the lower serum glucose readings. Unlike insulin, flozins likely do little to inhibit lipolysis and ketone production.

    PS Hamlet.

Leave a Reply

Your email address will not be published.