Diabetes, glucose meter error, and harm

December 10, 2012


Say a diabetic patient has a hypoglycemic episode and exhibits the typical symptoms of low glucose. After treatment, the patient appears normal. What might be said about this? The patient has experienced temporary harm, temporary because they appeared to have completely recovered. The fact that there was harm was the appearance of symptoms. This could be called temporary harm with symptoms.

If a diabetic patient has retinopathy, they could be said to have permanent harm. But this type of permanent harm is likely to arise from a series of events where glucose was not in control. The resulting retinopathy could be called permanent harm with symptoms.

Each time that glucose is not in control, even when there are no symptoms, one could ask is there harm at a molecular level. This could be called (molecular) harm without symptoms. Unfortunately, I am unaware of any data about harm without symptoms.

Some authors (1) have suggested that if glucose is controlled to within 70-130 mg/dL, there is no harm. But isn’t it possible that the patient who is controlled within ± 5 mg/dL will be better off than a patient who is controlled within ± 30 mg/dL.

So the point is that if glucose meter error causes glucose to be out of control, the safest amount of error is no error, but since this is impractical, a reasonable error goal is the lowest possible amount of error, which is state of the art performance. And the problem with asking physicians about the level of error that is important is that they are likely to set levels based on symptoms.


  1. Breton MD and Kovatchev BP Impact of Blood Glucose Self-Monitoring Errors on Glucose Variability, Risk for Hypoglycemia, and Average Glucose Control in Type 1 Diabetes: An In Silico Study J Diabetes Sci Techol 2010;4:562-570.