Usefulness or Uselessness of GlucoWatch in Monitoring Hypoglycemia in Children and Adolescents
Dario Iafusco, MDMaria Katrin Errico, MD
Colomba Gemma, MD
Francesco Prisco, PhD, MD
Department of Pediatrics Second University of Naples 80100 Napoli, Italy
To the Editor.—
The article of Chase et al1 was the first report of the routine use of the GlucoWatch biographer in children with diabetes mellitus.
Since May 2002 in our pediatric department, 14 children and adolescents (7–17.8 years old) and 8 young adults (18–23 years old) with type 1 diabetes mellitus wore the biographer each one on an average of 12 different occasions for a mean period of 30 days.
We agree with Chase et al1,2 about the good correlation between noninvasive readings of glucose levels by GlucoWatch and finger-stick glycemia. Our patients too showed a good compliance with extemporaneous changes of therapy, diet, and physical activity in response to alarms for hypo- or hyperglycemia confirmed by finger-stick determinations ("behavioral therapy").
But in our opinion the usefulness of this device for investigating the incidence of unrecognized hypoglycemias is not absolute.
In our patients, during the night, the GlucoWatch biographer recorded 7 episodes of hypoglycemia (blood glucose from 30 to 70 mg/dL) lasting from 20 to 280 minutes and confirmed by monitoring blood glucose in 4 of the 6 verified cases (85%). However, the alarm failed to alert the patient in the majority of cases, probably because the reduced awareness consequent to hypoglycemia itself.
In particular, two adolescents who were sleeping alone in the room discovered the severe nocturnal hypoglycemia (values repeatedly <30 mg/dL) only in the morning when they reviewed the recorded glucose readings. In one patient the device failed to measure glycemia due to sweating resulting from hypoglycemia. During athletic performances, repetitive readings of glucose could be very useful to detect early hypoglycemia and to decide additional glucose ingestion. In our experience, the interference of the sweat on the readings was very high during physical exercise in children and adolescents, and in most cases the readings were lost in 100% of the cases.
The GlucoWatch biographer was instead very useful before and after exercise to assume the correct amount of carbohydrates and to diagnose late-onset postexercise hypoglycemia.
In our opinion, the GlucoWatch biographer is very useful in home routine monitoring of children with type 1 diabetes, but we hope that the future versions of this device will be perfected in detecting hypoglycemias during physical activity and in making alarms more effective.
REFERENCES
- Chase HP, Roberts Md, Wightman C, et al. Use of the GlucoWatch biographer in children with type 1 diabetes. Pediatrics.2003; 111
:790
–794
[Abstract/Free Full Text] - Garg SK, Potis RO, Ackerman NR, Fermi SJ, Tamada JA, Chase HP. Correlation of fingerstick blood glucose measurements with GlucoWatch biographer glucose results in young subjects with type 1 diabetes. Diabetes Care.1999; 22
:1708
–1714
[Abstract/Free Full Text]
Richard C. Eastman, MD
Cygnus, Inc. Redwood City, CA 94063, USA
In Reply.—
The letter by Iafusco et al reports the early use of the first-generation GlucoWatch biographer in children with type 1 diabetes in Italy. The authors found good correlation between biographer readings and blood glucose levels and felt that the Biographer is very useful in the care of their patients.
The major concerns expressed by the authors are that the biographer did not detect all episodes of hypoglycemia and skipped some readings, particularly because of sweating. We certainly agree that the biographer’s ability to detect hypoglycemia is not absolute. However, high sensitivity for detecting most events can be obtained by setting the alert level at a sufficiently high level. For example, if the low alert level is set at 6.1 mmol/L, the sensitivity of the first-generation device to detect hypoglycemia is 86%, with 16% false positives. It should be noted that, although the device may not always awaken patients with hypoglycemia, the values are recorded in the biographer’s memory and are available to the user and provider for subsequent adjustment of the treatment regimen to avoid future hypoglycemic events. There is also the opportunity for the alarm to alert a sibling, camp counselor, or parent to the possibility of hypoglycemia even though it may not awaken the subject. Importantly, most of the episodes of hypoglycemia detected by glucose sensors such as the biographer are mostly asymptomatic and are not detected at all because patients do not perform frequent finger sticks and blood glucose monitoring during the night. Thus, most hypoglycemia events are not detected by blood glucose testing. The biographer offers the opportunity to capture data for retrospective analysis and to alert the patient in most instances when glucose levels are low. This is clearly shown in the article by Chase et al, where biographer users were able to detect significantly more hypoglycemia when wearing the biographer than in those not wearing the biographer. The relative benefit was especially notable when the patients wore the biographer at night.
The second-generation GlucoWatch G2 Biographer has an improved alert, called the down alert, that projects forward from the data over the last several readings and alerts the patient to the possibility of hypoglycemia in the next 10 to 20 minutes. Given the delay between blood and biographer readings of 
15 minutes, the currently available G2 Biographer detects hypoglycemia in real time. The latest version also provides twice as many readings, as often as every 10 minutes, which also somewhat reduces lag time. This feature may provide additional value to users, allowing them to detect hypoglycemia earlier.
The biographers measure glucose at micromolar concentrations in nanoliters of fluid, compared to conventional meters that measure millimolar glucose concentrations in microliters of fluid. Measuring glucose in the micromolar range in such small fluid volumes is subject to environmental factors such as motion, temperature, and sweating, which can interfere with and corrupt the measurement process. Similar factors affect conventional glucose meters, but these measurements are done at a 1000-fold-higher glucose concentration with a hand-held device that is not closely applied to the user. To ensure that only accurate data are presented to the user by the biographer, numerous data integrity checks are performed during each measurement cycle. If the data do not pass any of these checks, the value is not reported to the user. This increases the overall accuracy of the device, because inaccurate data are omitted. Thus, the patient gets 55 readings per use with the G2 Biographer in our studies and skips 20% of the readings. This is many, many more readings than patients get with conventional glucose monitoring because of the low sampling frequency.
Recent research and development efforts at Cygnus, Inc have focused on improving the data-integrity checks, essentially making them smarter, allowing more data to be presented to the user. These results have been successful. Results from a recent study conducted in the home environment using the improved integrity checks show that the number of skipped readings is reduced by >60%.1 Future versions of the device will incorporate these improved data screens.
REFERENCE
- Harper W, Wei C, Eastman R, Lesho M. Further enhancements to the utility of the GlucoWatch G2 Biographer [abstract]. Diabetes.2003; 52(suppl 1) :A95
PEDIATRICS (ISSN 1098-4275). ©2004 by the American Academy of Pediatrics
CiteULike 
Connotea 
Del.icio.us 
Digg 
Facebook 
Reddit 
Technorati 
Twitter What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
