Glucometer Errors

Study shows venous blood tests artificially high on capillary glucometers.

Did you know that it matters where you get the blood sample from when you’re using a glucometer? If you did, congratulations. You understand more than I did about glucometers before I read this study that showed venous blood tests as much as 17.42 mg/Dl higher on capillary glucometers than capillary blood.

Why do you care? If your agency is using standard, over-the-counter capillary glucometers, these are specifically calibrated to give accurate readings on capillary blood. The kind of blood your finger oozes when you poke it with a lancet. Not the direct venous blood that you might get off your IV needle or directly from the end of the IV catheter.

Does this mean that you should stop using the IV site, or needle, or little drops of blood on your bench seat to test the patients glucose level? Not necessarily. Venous blood will still give you an accurate ballpark estimate of the patients glucose level. Just be aware that if you’re looking for a dead-on accurate blood glucose level on your diabetic or altered mentation patient, you need to do a finger stick. And know that a venous blood sample reading will most likely be an artificially high number.

Now you know.


  1. For us Canadian folk, 17.42mgdl is 0.96mmol

    Great mini-article.

  2. This is a great point to illuminate to EMS folks, however, we must also keep in mind that sometimes the venous draw can in fact provide us necessary measurements. Capillary tests may be severely affected at times due to hypotension which in turn will provide poor perfusion to the accessed site. This may prove vital in a medical / trauma situation (i.e. diabetic becomes hypoglycemic, loses conciousness, and severely wrecks vehicle or falls off ladder, etc.)

    One other point which is noted many times with newer providers is that in taking post-D50 readings. It is recommended to take the reading from the other arm, not the administered arm. This will allow a full system check to make sure that your levels have in fact gone systemic and just aren’t backlogging in your med arm.

    Great article Steve!

  3. Michael Endres says:

    Using alcohol to clean the site can also result in wrong results if the alcohol is not completely dried off and mixed with blood. I prefer to use a cotton ball and saline to give it a good wash (to clean off any sugar etc. from that last chocolate bar…..) and always discard the first drop because it contains too much tissue fluid (same when you squeeze too hard) which can result in wrong readings as well. So the right order would be:

    – keep finger warm (rubbing, warm towel etc.)
    – if necessary use saline or alcohol (only if you wait long enough to dry) to clean
    – cotton ball to wipe it clean and dry
    – stick it and wipe off first drop with clean cotton ball
    – wait for second drop, do not squeeze too hard and et voilà perfect result!

  4. I have 3 type 1 kids and just wanted to remind all ems out there to look past the flash and beads on bracelets even temporary tattoos for medic alerts and always check tennis shoes for “toe tags ” identifying diabetic kids. Also never use alternative site testing on a kid with a suspected hypo. Always fingers! But you knew that!

  5. Just going through this site. Great stuff. My tip for blood glucose is dont have it upside down. I had a hypoglycemic coma during my ride time. got 62, well that shouldnt be causing this.. preceptor went through history and couldnt find my number. try 29. we had an iv and pushed d50 laughing our asses off, we had a firefighter on board bagging who just looked at us like we are crazy. good times.

  6. Just like during pregnancy, breastfeeding requires much blood sugar level monitoring from the mother.
    This happens due to two factors either the pancreas is not producing
    enough insulin or the cells are not responding normally to insulin.
    If someone repeatedly undermines your diet, you may have to limit your time with them.