Is your blood glucose meter giving you accurate readings? Testing under certain circumstances may be giving you misleading results. While most users expect accurate readings from their meters at all times, recent studies have found that many meters on the market today are inaccurate during hypoglycemia and when used at high altitude.

One study, reported in the December 1996 issue of Diabetes Care, evaluated the effect of hypoglycemia on six blood glucose meters-Lifescan's One Touch II, Medisense's Companion II, Bayer's Elite, Boehringer Mannheim's Reflolux and Accutrend models and the HemoCue meter from the company of the same name.

A wide variation among meter readings was discovered. In addition, none of the six meters tested met the American Diabetes Associations (ADA) 1994 criterion for accuracy when their performance was evaluated during hypoglycemia.

The ADA's current accuracy criteria are especially strict. In 1994, the ADA changed its recommended variability to 5 percent. This means that meters must test 5 percent above or below the true BG level. Previous to this 1994 criterion, 15 percent variability was considered acceptable.

It should be noted that, according to the ADA, any meter produced before this change can claim to meet the ADA recommendations if it tests within the old 15 percent variability standard.

While the meters did not test within the 5 percent variability, all measurements of the One Touch II, Reflolux, Accutrend and HemoCue meters were in what this study called the "accurate zone," while the Companion II and Elite meters fell into what it called the "not acceptable zone."

The study reviewed other recent tests of blood glucose meters (BGM) accuracy and found that, while a number of tests had been performed, there was insufficient data on the effect of hypoglycemia on meters' accuracy. This is an important variable because the study's literature review found that a 1993 study in Diabetes Research and Clinical Practice discovered meter accuracy varied across different BG ranges.

The importance of testing meters in different BG conditions is illustrated by comparing the results of this study with those of a meter test by Consumer Reports in October 1996. Consumer Reports found that all "eight of the leading blood glucose monitors" that they tested scored within the ADA's old 15 percent accuracy standards. The current study, however, found that all of the meters it tested had approximately 20 percent variability in their readings during hypoglycemia. The two tests used different meters. But, in light of the current findings, it seems unlikely that the ones in the Consumer Reports test would have faired as well under hypoglycemic conditions.

Another study published in the December Diabetes Care showed people who live at high altitudes or regularly participate in sports activities such as trekking, skiing and climbing may be receiving misleading readings from their BG meters as well. The study found that several meters suffer from accuracy problems at high altitudes due to low barometric pressure.

Altitude changes were simulated in the lab by decreasing the oxygen concentration (barometric pressure) in a test chamber. The meters were then tested at simulated altitudes from 1,000 to 4,000 meters. Actual altitude change would also include changes in temperature and humidity. These are both factors that could affect meter accuracy as well.

The study tested five meters: Accu-Chek Easy, Reflolux SF, Tracer (by Boehringer Mannheim), One Touch II and Glucometer 3.

All of the meters underestimated BG levels at high altitudes, except for the Accu-Chek Easy, which consistently overestimated BG levels in all measurement ranges. The One Touch II performed better than the others across different altitudes, and the Reflolux also "seemed more accurate." Overall, however, people testing at altitudes over 1,000 meters should be careful when interpreting their BG test results no matter what meter they use.