Portable blood glucose meters gave readings averaging as much as 16 percent in error in a study of 102 women with gestational diabetes conducted by Australian researcher Dr. Nimalie Perera, of the Royal Prince Alfred Hospital in Sydney. The study appeared in January 2011 in the journal Diabetes Care.
Perera compared test results from six different models with those from laboratory tests. The most precise model, Stat-Strip, erred by an average of 6 percent. The least precise, Optium Exceed, erred by 16 percent. Both are sold in the United States.
Although non-diabetics may consider these errors small, they can be significant because many diabetics use their meters to help them decide when and how much insulin to inject. Close monitoring and accurate dosage are critical to maintaining a diabetic's health and safety.
The typical allowable error in a blood glucose meter is 20 percent, but many experts argue for tighter standards -- especially in gestational diabetes, which can lead to miscarriage or an oversized baby and a difficult delivery.
Commenting on the study, Dr. David Sacks, director of the clinical chemistry lab at the U.S. National Institutes of Health, told Reuters Health that people for whom accurate readings are critical should be extra careful in performing the tests and, when their results seem atypical, take second readings.
One reason portable glucose meters are prone to error is that they work by a complex electrochemical process not like an electronic calorie counter. Within the test strip, a special chemical reacts with the glucose in the blood to produce an acid. Then another chemical turns the acid into a substance called ferrocyanide. Finally, the meter runs an electric current through the ferrocyanide and derives the glucose level from the change in the current. With so many steps depending on one another, small errors can add up to bigger ones
