Four Glucometer Errors That May Surprise You
A 2023 paper describes the following case: A 55-year-old male with a known history of insulin-dependent diabetes was found down on the side of the road.1 His hand was in a bag of candy. Though he was able to protect his own airway, he was unable to follow commands or answer questions. His vitals were otherwise within normal limits: 135/77 blood pressure, heart rate of 93, breathing 18 times a minute, and a room air oxygen saturation of 94%. His blood sugar was, naturally, taken in the field and found to be “hi.” An end-tidal carbon dioxide reading wasn’t provided. It isn’t known whether the providers assessing him had access to this assessment tool.
Most providers at this point would assume the patient had diabetic ketoacidosis or hyperosmolar nonketotic coma, which have very similar treatments in the field. They would give fluid, transport, and not interfere with the patient’s breathing as long as that interference was avoidable.
However, this situation was far from typical: When his blood sugar was re-checked by hospital staff, the reading was 25.
How could this be possible? Every provider involved had acted both with integrity and to the best of their ability. It was noted on the patient’s hospital chart that his hands were covered with chocolate.
The field crew had, in a single and understandable moment of inattention or panic, failed to clean the site before taking the blood sugar measurement. As a result, the chocolate on his hand had contaminated the blood with sugar, causing the inaccurate and abnormal reading. The hospital staff, conversely, had thoroughly cleaned the chocolate off the finger, yielding an accurate reading.
Hypoglycemia can cause brain injury as reliably as hypoxia. This mishap, though borne of such a small error, could have led to catastrophe. Luckily, it didn’t.
Here are several common errors when taking blood sugar measurements in a prehospital setting:
1. Not choosing to check the blood sugar in the first place
The best practice is, generally, to check a blood sugar on most patients. The beauty in emergencies caused by an abnormal glucose level is that we never have to miss it. Anyone who is in the field for long enough will inevitably misdiagnose a head bleed or a pulmonary embolism or heart attack. These are subtle conditions that can hide from even the most experienced and dedicated in-hospital clinicians, even with the back-up of a whole lab and suite of imaging tools.
Emergencies driven by blood sugar are exactly the opposite. This 65-year-old receptionist may insist that she simply slipped and fell; this 44-year-old who is slurring his speech in a bar could easily be presumed to simply be drunk. Patients who have unusual behavior of any type or children with isolated abdominal pain may in fact be experiencing a life- and brain-function threatening emergency. It’s unlikely, statistically, that the code you’re working was caused by hypoglycemia. But it’s possible, particularly in children.2 And if you aren’t checking BGLs on code once you have your compressions, shocks, epinephrine, and airway well underway you could miss this opportunity.
Furthermore, EMS is all about pursuing the unlikely explanation, decompressing trauma codes, and doing stroke assessments on syncope patients, for the rare but life-affirming moments when you find precisely the patient you were looking for. When your thoroughness, patience and attention to detail yield a life saved that could have easily slipped through the fingers of someone less dedicated and fastidious.
2. Failure to prevent contamination of the blood sample
The glucometer requires a tiny amount of blood, which is excellent for EMS.1 It can be checked by BLS providers because it doesn’t require a whole vial of blood, which would require venipuncture. It also allows it to be performed with little pain: a small lancet puncture to the side of a finger is bothersome at worst.
There is a trade-off here; however, any amount of contaminant can ruin your reading. This especially applies if the contaminant is sugary itself, such as a syrup or candy residue. It’s for this reason crucial to clean the site thoroughly. Many patients coming into contact with EMS don’t have consistent access to hygiene facilities. If you are checking a BGL on a patient and note that the alcohol wipe is looks soiled, use a second one to ensure decontamination. Furthermore, failing to allow the alcohol to dry completely, even though this process typically takes seconds, can dilute the sample and cause a falsely low reading.
In terms of mistakes that could easily be made at 3 am, ensure that you are taking a sample from the same digit that you cleaned. It would be relatively simple to disinfect one finger, turn around to get the bandage ready, turn back to the patient and take a sample from a totally different one.
The number you will get is precisely as accurate as the drop of blood you provided is pure.
3. Some uncommon patient histories can cause inaccurate readings
It’s not particularly likely that an average provider would encounter these situations but nor is it impossible. In another case study from medical literature, a crew encountered a 25-year-old comatose female. Being thorough and canny clinicians, they attempted to take a blood sugar. Even though the blood filled the strip completely, they received an error message. They worked the problem, cycling through four more glucometers with the same result persistently: error. Lab tests confirmed that her blood glucose was 180.
It turned out that this young lady had deliberately overdosed on Tylenol and Benadryl. Tylenol causes glucometer strips to rust instantly, making obtaining a reading impossible in the field.
Altitude also impacts the usefulness of the glucometer. For every 300 meters in elevation gain, the reading will be falsely lowered by 1% to 2%. That being said, glucometry should never be relied on for backcountry triage anyway. Anyone who is confused or had a behavior change or a mysterious syncope on a mountaintop should be presumed to be having an altitude-related emergency. They should also be hastily evacuated to both a lower elevation and an actual doctor to do a definitive evaluation.
Even if you for some reason are doing medicine on the top of Denali, that would only cause a blood sugar reading to be 40% too low. For example, a patient with a blood sugar of 100 would read as having one of 60. The worst-case scenario there is that a hiker gets an extra granola bar they didn’t truly need.
4. Investigate readings that don’t make any sense
There is unfortunately no tool that can substitute for good judgement. This is one of the first strategies we learn in the field once we are done with school. You got a blood pressure reading of 280? Double check the placement of the bladder inside of the fabric and the cuff size. Oxygen saturation of 67? Unless the patient is looking quite sick, check for a good pleth wave and see if their fingers are cold or their nails artificial.
The glucometer, of course, is no different. If you got a blood sugar reading of 20 on an alert and oriented patient who called 9-1-1 for shortness of breath, investigate. Change something, the finger, the provider taking it, the machine, the pack of strips, and then try again. If the next couple of readings with a different machine are reasonable, do the next shift a solid and remember to take the faulty one out of service with a detailed note explaining the entire incident for repair or replacement.
If you change a few things and are still getting the bizarre reading, accept it and respond accordingly. Patients don’t read our textbooks. We have all seen the patient who is still conscious with a room air saturation that is 52 or a patient whose coma was caused by his blood sugar reading of 78.
As always, remember that we are never truly alone. Consult your fellow providers or a higher level of care. The skill of identifying when a situation is now beyond what you can control and address yourself, when you need help from a more knowledgeable power, is perhaps the most crucial one to have, yet the most difficult to teach.
Works Cited
- Majewski, J., Risler, Z., & Gupta, K. (2023). Erroneous Causes of Point-of-Care Glucose Readings. Cureus, 15(3), e36356. https://doi.org/10.7759/cureus.36356
- Reno, C. M., Skinner, A., Bayles, J., Chen, Y. S., Daphna-Iken, D., & Fisher, S. J. (2018). Severe hypoglycemia-induced sudden death is mediated by both cardiac arrhythmias and seizures. American journal of physiology. Endocrinology and metabolism, 315(2), E240–E249. https://doi.org/10.1152/ajpendo.00442.2017
