lood glucose meters
that let you test on places where you have few nerve endings are much less painful than traditional fingertip meters. These so-called alternative site meters are the best thing for people with diabetes since the first meters became available more than 30 years ago.
Or are they?
It seems that all is not perfect in alternativeland. In fact, they may be downright dangerous. Under certain circumstances they might not warn you quickly enough that your blood glucose level is going too low.
There are already five different brands of alternative site meters on the market. First to be introduced, Amira Medical's AtLast became available in December 1999. Since then LifeScan's OneTouch FastTake, which was previously available, incorporated new test strips that require less blood, thereby permitting forearm testing. The OneTouch Ultra is a similar monitor, but works even faster, taking only 5 seconds to give a reading. The TheraSense FreeStyle takes the least blood of all available meters. The newest alternative site meter, the MediSense Sof-Tact, hides the test strip, lancet, and blood drop inside the meter.
Two German scientists, Karsten Jungheim and Theodor Koschinsky, stirred up a hornet's nest when they presented their report on lags in alternative site tests at the ADA Scientific Sessions on June 25. They followed up with a letter published in the July 2001 issue of Diabetes Care.
Their preliminary studies indicate that tests with TheraSense's FreeStyle meter on the arm may lag about half an hour behind fingertip results. This could have serious consequences preventing timely detection of hypoglycemia if blood glucose levels were falling quickly.
They based their report on tests of just six people with type 1 diabetes on intensified insulin treatment. They did not rub the forearm skin before taking the blood glucose sample.
Note the rather unusual protocol of their test: After an overnight fast, they omitted the usual prebreafast insulin. Instead of breakfast, they each took 75 grams of glucose so that their blood glucose readings would be between 300 and 400 mg/dl. Then they got their usual (6 to 15 units) short-acting dose of insulin. The researchers watched their blood gluocose drop every 5 to 15 minutes until they reached either a steady state or hypoglycemia (below 60 mg/dl). Oral glucose compensated for the hypoglycemia. Some experts note that this large glucose dose followed by intravenous insulin is hardly a typical situation.
A TheraSense rejoinder reported in the same issue of Diabetes Care showed little difference between fingertip and arm among 120 people the company studied—if they rubbed the arm first.
In these initial TheraSense studies using single points in time with many subjects, they saw close correlation between forearm, upper arm, back of the hand, thigh, and calf. But the correlation of all of these sites to the finger was not as good.
"This is in keeping with the circulatory physiology of the skin," according to e-mail from Mr. McGarraugh, TheraSense's director of chemistry. "The blood flow on the palms of the hands (including finger tips) and the soles of the feet is considerably
faster than the blood flow in the skin of the arms, legs and abdomen. We believe the faster blood flow allows faster communication of the blood glucose changes. We currently are conducting the time course studies comparing thigh and finger. We will be reporting these results as soon as we have enough data to make sound conclusions."
Based on its studies to date, TheraSense has reached five conclusions, he says:
- Changes in glucose can be detected first on the finger, and the changes can lag on the forearm.
- Rubbing the test site prior to testing stimulates blood flow and minimizes the difference between the alternate site and the finger.
- The faster the changes the greater the difference between forearm and finger.
- There are significant differences from person to person. In some people the lag time is very short, and the differences between arm and finger are negligible. In other people the lag time can be significant—15-20 minutes.
- The lag is most significant in the detection of hypoglycemia. For some people the detection of hypoglycemia will be much more effective if the finger is used as the test site.
TheraSense continues to study the question, Mr. McGarraugh writes. "We are expanding to additional alternate sites and to specific patient populations (children, women with gestational diabetes, and people with peripheral vascular diseases). There will be an article in the Fall edition of Diabetes Technology and Therapeutics on this subject."
LifeScan has also studied the issue, although it has not yet published its conclusions. A LifeScan study of 42 people using their One Touch Ultra meter seemed to support Jungheim's and Koschinsky's finding. Finger tests tended to produce higher average blood glucose readings than forearm and thigh tests, especially an hour and an hour and one-half after meals. Not all patients, however, experienced this lag.
LifeScan is just now sending new recommendations to medical personnel. Alternative site testing, they say, should be done only two or more hours after the last meal, two or more hours after taking insulin, or two or more hours after exercise. Otherwise, LifeScan recommends fingertip testing.
MediSense and Amira Medical haven't stated their positions. However, because they use a different method to draw blood—suction—it is possible that tests now in progress will not show the same lags as experienced while using the TheraSense FreeStyle and the two LifeScan meters on alternative sites. Meanwhile, anyone who uses any of the alternative site meters and is concerned that his or her blood glucose is falling rapidly is well advised to rub or heat the test site or complement the alternative site test with a finger prick.
by David Mendosa
Last modified: August 20, 2001
Reproduced with permission.
