Two devices, two different numbers
You check your CGM, Mmine is a Dexcom, and see 5.2 mmol/L. A finger-prick test says 6.1 mmol/L. Check again a little later and the numbers may be closer—or further apart.
It is easy to assume one device must be wrong, but a CGM and a blood glucose meter are not measuring glucose in exactly the same place.
Blood and interstitial fluid
A finger-prick meter measures glucose in a small sample of blood. My Dexcom, like other CGMs measures glucose in interstitial fluid, the fluid surrounding cells beneath the skin.
Glucose moves from the bloodstream into interstitial fluid, which takes time. When levels are stable, the two readings may be quite close. When glucose is rising or falling rapidly, the difference may be more noticeable because the CGM can appear to follow behind the blood reading.
This physiological lag does not mean CGM is useless during changes. The current reading, trend arrow and graph together show how glucose has been behaving. It simply explains why a meter comparison is not always identical at that exact moment.
Rapid changes create bigger differences
Differences are particularly common:
-
Soon after eating
-
After taking insulin
-
During or after exercise
-
While treating a hypo
-
When glucose is rising or falling quickly
If I want to compare the devices, a period when glucose is relatively steady gives a more meaningful comparison than testing during a dramatic rise or fall.
The meter is not perfect either
It is tempting to treat a meter as an infallible referee, but home glucose meters also have an accepted accuracy range.
Results can be affected by:
-
Food or sugar residue on fingers
-
Wet hands that dilute the sample
-
Expired or poorly stored strips
-
Too small a blood sample
-
Extreme temperatures
-
The meter and strip accuracy limits
I wash and dry my hands thoroughly before using a meter to investigate an unexpected CGM reading. Using an alcohol wipe is a good idea if washing your hands isn't an option.
Sensor-related reasons for a mismatch
A new sensor may take time to settle. Pressure on the sensor can cause an apparent compression low. A site with irritation, bleeding or poor adhesion may also behave unexpectedly.
Dehydration, sensor errors and a sensor nearing the end of its session can sometimes contribute to readings that seem less reliable, although a mismatch should be assessed rather than automatically blamed on one cause.
Which reading should guide a treatment decision?
Dexcom’s general safety advice is clear: if sensor readings do not match symptoms or expectations, use a blood glucose meter to make diabetes treatment decisions.
For example, if Dexcom says 3.2 mmol/L but I feel completely normal and the graph has dropped suddenly while I have been lying on the sensor, I check with a meter if I can do so safely. If I have obvious hypo symptoms or cannot check safely, I follow my personal hypo plan rather than delaying necessary treatment.
The reverse matters too. If I feel low but Dexcom displays an in-range result, I do not ignore my body simply because the screen looks reassuring.
Should Dexcom be calibrated?
Calibration instructions differ between Dexcom models and circumstances. Some systems are factory calibrated and do not routinely require finger-prick calibration.
Entering an unsuitable calibration—particularly while glucose is changing quickly or when the meter result itself is unreliable—can make performance worse. I follow the instructions for my exact Dexcom model or contact Dexcom support rather than repeatedly calibrating in frustration, although, the temptation to do so is often strong!
When a difference needs investigating
Contact your CGM support or speak with a healthcare professional if:
-
Readings repeatedly disagree with a correctly performed meter check
-
The sensor does not match symptoms
-
Unexplained errors or data gaps continue
-
The sensor site is painful, swollen or inflamed
-
You are uncertain whether readings are safe to use
-
You experience recurrent or severe hypo symptoms
The aim is not to make your CGM and a meter display matching numbers every minute. It is to understand why reasonable differences happen, recognise when a reading does not make sense and know which safety check to use next.
BLOG DISCLAIMER
If you or any other person has a medical concern, you should consult with your health care provider or seek other professional medical treatment. Never disregard professional medical advice or delay in seeking it because of something that have read on this blog or in any linked materials. If you think you may have a medical emergency, call your doctor or emergency services immediately.
This blog provides general information and discussions about health and related subjects. The information and other content provided in this blog, or in any linked materials, are not intended and should not be construed as medical advice, nor is the information a substitute for professional medical expertise or treatment.
This blog is for education and support only, not medical advice. Always follow your diabetes care team's guidance.
These articles are written for a UK audience using mmol/L. They provide general information and lived-experience context, not individual medical advice. Readers should follow the training, user guides and treatment plan for their exact devices and speak to their diabetes team before changing insulin settings.
The opinions and views expressed on this blog and website have no relation to those of any academic, hospital, health practice or other institution.
We always aim Information used in our blogs to be accurate and up to date. We accept no responsibility for any inaccuracies. Some blogs are written by Cameron and based on his independent research, if you spot an error please reach out so he can be made aware of this.