A gentleman presents to the emergency department as a notification for altered mental status. The first thing EMS tells you is that the patient’s blood sugar on scene was 28 mg/dL, so they gave him an amp of D50W with mild improvement in mental status. On your exam, however, the patient is still confused, opens eyes only to stimulation, and is not responding to your questions. You notice his blood pressure is 80/60 mmHg, and he is tachycardic to 160/min with atrial fibrillation with RVR. His extremities are cool. His repeat fingerstick in the resuscitation room is 41 mg/dL, but you’re more preoccupied with those vitals, so you slam him with two more amps of D50W.
Everything happens simultaneously: the patient is placed on a monitor, labs are drawn and sent, a fluid bolus is hanging. Time to do your RUSH exam. But wait, the repeat fingerstick after the most recent two amps of D50W is 32 mg/dL. What is going on?? As you’re finishing up your RUSH exam, the attending tells you the glucose on the VBG is 350 mg/dL.
What’s in a number?
Technically, hypoglycemia is defined as a blood glucose of less than 70 mg/dL; however, this number is fairly arbitrary as patients can experience symptoms of hypoglycemia at higher blood glucose levels and, paradoxically, have no symptoms with lower blood glucose levels. You will frequently see mention of Whipple’s triad as a clinical tool for the diagnosis of symptomatic hypoglycemia.
1. Fingerstick < 55 mg/dL
2. Symptoms (eg. nausea/vomiting, diaphoresis, altered mental status)
3. Symptomatic improvement with the administration of glucose-containing products
Without the appropriate clinical context and/or accompanying symptoms, a low fingerstick is meaningless. If a patient is alert, comfortable, and interactive, are you going to immediately act on a fingerstick of 20 mg/dL? It’s more likely that you would assume it was an erroneous reading.
Glucometers can be affected by any number of factors and falsely elevate or lower a reading.
– Equipment dysfunction (poor calibration, expired test strips)
– Temperature[1]
– Blood sample (dilution from upstream IV fluids, non-capillary sample, surface contamination)
– Peripheral vasoconstriction (hypothermia, shock, vasculitic disease)
Blood Samples
Glucose strips are sensitive to oxygen, so glucometers are calibrated to the expected oxygen content of a capillary sample. Blood samples with a higher than expected oxygen content (e.g. arterial samples or patients on supplemental oxygen) tend to underestimate blood glucose levels. Often, we may be tempted to use blood left over from a venous stick, but due to the lower oxygen content. A venous sample is likely to overestimate the blood glucose level. Hematocrit level, increased serum triglycerides, and increased uric acid levels can also affect glucometer accuracy.[2]
Hypoperfusion
States of hypoperfusion (such as shock, hypothermia, Raynaud’s, and peripheral vascular disease) tend to cause pseudohypoglycemia in peripheral tissues due to the increased uptake of glucose. Although the reading will accurately reflect the capillary glucose level, this may not be reflective of the blood glucose level circulating to the vital organs. In critically ill patients, don’t hesitate to treat a low fingerstick; however, if the fingersticks don’t seem to be responding to therapy, follow it up with a blood gas, otherwise, you may end up chasing your tail as with the patient introduced in the beginning.[3]
Normal fingerstick, but something’s off…
To complicate this subject further, a patient can have neuroglycopenia (neural glucose deprivation) without hypoglycemia. In salicylate toxicity, the uncoupling of oxidative phosphorylation leads to an increase in cellular metabolic activity and thus, an increase in glucose utilization. This may lead to low levels of CSF glucose (and thus, symptoms of neuroglycopenia) before there is a noticeable drop in the blood glucose level. Read more at Rebel EM.
Another example of neuroglycopenia without hypoglycemia can be seen in diabetic patients with poorly controlled blood sugars. Chronically high blood glucose levels can lead to altered glycemic thresholds such that poorly controlled diabetics can experience neuroglycopenic symptoms at a seemingly normal blood glucose level. By the same token, patients with frequent hypoglycemic episodes (such as in patients with frequent insulin overdoses, or less likely, patients with insulinomas) may have altered glycemic thresholds in the opposite direction and may reach a dangerously low blood glucose level without any neuroglycopenic symptoms, termed “hypoglycemic unawareness.” The patient can experience a sudden deterioration to seizures, coma, or death without any of the classic preceding symptoms (such as confusion or diaphoresis).[4]
It’s also worth mentioning that patients on beta-blockers may experience “hypoglycemic unawareness” as beta-blockers tend to mask the tachycardia, nervousness, and tremor that result from the initial sympathetic response to hypoglycemia.
PEARLS
– If a patient is critically ill, do not hesitate to give dextrose for a low fingerstick
– If a patient does not seem to be responding to dextrose, correlate with a VBG
– In asymptomatic patients with a low fingerstick, consider re-taking a fingerstick or correlating with VBG
– Avoid using blood from a blood draw on a glucose test strip. (Test strips are calibrated to capillary blood.)
References
1. Haupt A, Berg B, Paschen P, et al. The effects of skin temperature and testing site on blood glucose measurements taken by a modern blood glucose monitoring device. Diabetes Technol Ther. 2005;7(4):597-601. doi:10.1089/dia.2005.7.597
2. Ginsberg BH. Factors affecting blood glucose monitoring: sources of errors in measurement.J Diabetes Sci Technol. 2009;3(4):903-913. Published 2009 Jul 1. doi:10.1177/193229680900300438
3. Garingarao CJ, Buenaluz-Sedurante M, Jimeno CA. Accuracy of point-of-care blood glucose measurements in critically ill patients in shock. J Diabetes Sci Technol. 2014;8(5):937-944. doi:10.1177/1932296814538608
4. Cryer PE. Symptoms of hypoglycemia, thresholds for their occurrence, and hypoglycemia unawareness. Endocrinol Metab Clin North Am. 1999;28(3):495-vi. doi:10.1016/s0889-8529(05)70084-0
nicanthony
Latest posts by nicanthony (see all)
- Abortion Care in the ED - May 23, 2023
- DNR Orders in the Suicidal Patient - September 22, 2022
- Sono of the Week: Neck Swelling - August 2, 2022
2 Comments
Robby · January 27, 2022 at 7:21 am
Awesome post! Something I haven’t really thought of before. Are capillary FSG always “falsely” low in shock states? Is this an early or late finding of shock? Are there any populations, young/old who are more or less likely to have this phenomenon? In cardiac arrest cases where the initial glucose is “normal”, will this always be lower than the value on vbg? Alternatively, can you use the presence or absence of peripheral hypoglycemia to predict ROSC?
nicanthony · March 4, 2022 at 9:16 pm
According to this systematic review (https://pubmed.ncbi.nlm.nih.gov/23506841/), glucometer readings seem to be more inaccurate in critically ill patients, especially those that are hypotensive or on pressors. They do not comment on the frequency of glucose underestimation in the actual review, although they do in the individual studies themselves. This prospective, non-randomized study (https://pubmed.ncbi.nlm.nih.gov/2029097/) showed a tendency for peripheral fingersticks to underestimate glucose levels in shock states, but definitely not in all or even most patients; same with this cross-sectional study (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4455371/).
Not enough literature on fingersticks in CPR or during cardiac arrest, other than one tiny study, which seemed to support the above. And no mention at all of whether the absence or presence of peripheral hypoglycemia could be predictor of ROSC.