Use of the Cygnus GlucoWatch Biographer at a Diabetes Camp
* Stanford University School of Medicine, Stanford, California
Cygnus, Inc., Redwood City, California
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Background. Detection and prevention of nocturnal hypoglycemia is a major medical concern at diabetes camps.
Objective. We conducted an open-label trial of the Cygnus GlucoWatch biographer to detect nocturnal hypoglycemia in a diabetes camp, a nonclinical environment with multiple activities.
Methods. Forty-five campers (7–17 years old) wore a biographer. The biographer was placed on the arm at 6:00 PM, with the low alarm set to 85 mg/dL (4.7 mmol/L). Overnight glucose monitoring occurred per usual camp protocol. Counselors were to check and record blood glucose values if the biographer alarmed.
Results. Biographers were worn for 154 nights by 45 campers. After a 3-hour warm-up period, 67% of biographers were calibrated, of which 28% were worn the entire night (12 hours). Thirty-four percent of readings were skipped because of: "data errors" (65%), sweat (20%), and temperature change (16%). Reported biographer values correlated with meter glucose values measured 11 to 20 minutes later (r = 0.90). Of 20 low-glucose alarms with corresponding meter values measured within 20 minutes, there were 10 true-positive alarms, 10 false-positive alarms, and no false-negative alarms. Campers reported sleep disruption 32% of the nights, and 74% found the biographer helpful. Campers reported they would wear the biographer 4 to 5 nights each week.
Conclusions. Half of the biographer low-glucose alarms that had corresponding blood meter values were true-positive alarms, and the remaining were false-positive alarms. There was close correlation between the biographer and meter glucose values. The majority of campers found the biographer helpful and would use it at home.
Key Words: Cygnus GlucoWatch biographer • diabetes camp • hypoglycemia
Abbreviations: CITs, counselors in training
A major medical concern at diabetes camps is detection and prevention of nocturnal hypoglycemia. Because of this risk, most diabetes camps have a protocol for nighttime testing, occurring typically between midnight and 3:00 AM. The Cygnus GlucoWatch biographer, which provides noninvasive transcutaneous monitoring of interstitial glucose concentrations, has been found to be well-tolerated as well as accurate when used by children.1 The biographer is placed on the child’s forearm and calibrated after a 3-hour period. Following calibration, it provides a glucose reading up to every 20 minutes for up to 12 hours. It is not waterproof; it cannot be worn while swimming or showering. Biographer readings are also affected by excessive perspiration.
We sought to evaluate the biographer in a camp setting to detect nocturnal hypoglycemia. We assessed the acceptability and tolerance of the biographer as well as its functionality and accuracy.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
The study protocol was approved by the Western Institutional Review Board. All children with diabetes attending a summer (2001) camp for children (7–17 years old) with diabetes were invited to participate in the study. For children who consented, use of the biographer was voluntary for each night of the 6-night camp. Parental consent was required as well. The study was designed to have
3 study participants wearing a biographer in each cabin of 10 to 12 campers. In cabins where >3 campers agreed to participate in the study, use of the biographer was assigned randomly. On subsequent nights, if a child assigned to wear a biographer elected not to, one of the other children who had consented to the study was allowed to wear a biographer. Those campers who did not wear a biographer on any night served as the control group. The biographer was placed on the child’s forearm at dinner (
6:00 PM). Calibration of the biographer occurred at the bedtime snack (9:00 PM). Calibration attempts occurred every 20 minutes. The biographer was removed by some campers during the warm-up period, and the biographer was also removed from campers after 3 unsuccessful calibration attempts.
The biographer’s low-glucose alarm was set at 85 mg/dL (4.7 mmol/L), and the high-glucose alarm was set at 240 mg/dL (13.3 mmol/L). Counselors were asked to perform and record fingerstick blood glucose values if the biographer alarmed "GL LO," "DOWN," "GL HI," or "PRSP" (perspiration) during the night. If the biographer alarmed on 3 occasions when the fingerstick blood glucose value was within normal range, the counselors were allowed to turn off or remove the biographer. All biographers remaining on the campers were removed before or at breakfast (6:00–8:00 AM). Routine nocturnal monitoring was performed per camp protocol; 2:00 AM testing was required for children with a fingerstick blood glucose <80 mg/dL (4.4 mmol/L) at bedtime snack, for those with a history of nocturnal hypoglycemia, and for those in whom a change in insulin regimen was made that might result in nocturnal hypoglycemia. Children with a low blood glucose value were treated per camp protocol, with 15-g glucose tablets or juice followed by 10-g of complex carbohydrates. Blood glucose was rechecked 15 minutes later.
A gold-standard measurement of serum glucose (YSI) was not available at the diabetes camp. Interstitial glucose values provided by the biographer were compared with blood glucose values determined by using the Lifescan (Milpitas, CA) One Touch Ultra or the Bayer (Tarrytown, NY) Glucometer Elite glucose meters. The majority of overnight fingerstick blood glucose levels were obtained by using the Bayer Elite meter. For the Elite meter, at reference glucose values of 41 to 70 mg/dL (2.28–3.89 mmol/L), the correlation coefficient with reference lab values was 0.78, the percentage of measurements within 20% was 75.6%, and the percentage of measurements outside 40% of the reference values was 4.2%.2 Another study has shown that the estimated bias from YSI plasma glucose values of 59 mg/dL (3.3 mmol/L) was 1.8%. Of 1377 meter glucose values, 96% fell within ±15% of the YSI glucose values.3 For glucose values between 36.4 and 434 mg/dL (2.0 and 24.1 mmol/L), the correlation coefficient when One Touch Ultra glucose values were compared with YSI Model 2300 Glucose Analyzer (Yellow Springs Instrument Company, Inc, Yellow Springs, OH) was 0.984 (package insert).
Height, weight, and hemoglobin A1c (DCA 2000) were measured on all consented campers once during the camp. Consented children completed a questionnaire before the last night of camp. The questionnaire asked about sleep disruption, skin irritation, and helpfulness of the device.
Demographic data are presented as the mean and standard deviation. The Pearson product moment correlation coefficient for paired glucose values was calculated by using Microsoft Excel.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Of 115 campers and counselors in training (CITs, 15–17 years old), 57 consented to participate in the study. Of these campers, 45 (22 male and 23 female) wore the biographer at least 1 of 6 nights. Because only 3 campers in each cabin could wear a biographer, 12 (5 male and 7 female) campers who consented to the study were assigned randomly not to wear the biographer (controls). All participants had type 1 diabetes. There were no significant differences in the age, total daily insulin dose, hemoglobin A1c, or duration of diabetes among the biographer wearers and the controls (Table 1).
|
Of the 45 campers who chose to wear the biographer at least 1 night, 15 (33%) wore the biographer on all nights that a biographer was available for wear. Campers chose to wear the biographer 61% of the nights a biographer was available for their use. Of the biographers placed on the campers at dinner, 18% were removed by the camper before calibration. The biographer removal rate before calibration was 19% for females and 17% for males. Calibration of the remaining biographers was successful in 75% of females and 60% of males. Removal and calibration varied by age group. None of the biographers worn by CITs were removed before calibration, and all biographers were calibrated successfully. Nine percent of biographers worn by 7- to 9-year-olds were removed, 22% were removed by 10- to 11-year-olds, and 44% were removed by 12- to 14-year-olds. Calibration rates of the remaining biographers were similar: 67% for 7- to 9-year-olds, 57% for 10- to 11-year-olds, and 70% for 12- to 14-year-olds.
Of the biographers that were calibrated successfully, 28% were worn the entire night (considered 9 hours after calibration because the biographers were calibrated between 9:00 and 10:00 PM and removed between 6:00 and 8:00 AM). Thirty-seven percent of females and 19% of males wore successfully calibrated biographers for the entire night. Per age group, none of the campers 12 to 14 years old wore a biographer the entire night as compared with 40% of CITs, 38% of 7- to 9-year-olds, and 21% of 10- to 11-year-olds. Mean length of wear after successful calibration was 5.2 hours.
The GlucoWatch biographer provides a reading every 20 minutes after successful calibration. During the study, a total of 1263 readings were obtained. Of these, 838 (66%) were glucose readings, and 425 (34%) were "skipped" readings, where the biographer reports a "SKIP" message along with the reason for the skip (ie, SKIP/PRSP, where the reason for the skipped measurement is caused by excessive perspiration). Sixty-five percent of the skipped readings were SKIP/DATA, which can be caused by inconsistencies in results between the 2 sensors. The inconsistencies are sometimes the result of excessive movement or bumping of the biographer. Twenty percent of errors were caused by sweat (SKIP/PRSP), and 16% were caused by temperature change (SKIP/TEMP). Limited data were collected in regards to blood glucose values at the time of perspiration alarms. On one occasion when the blood glucose was 47 mg/dL (2.6 mmol/L), there were 2 serial perspiration alarms, 1 which preceded and 1 which followed the fingerstick blood glucose. For the remaining 82 perspiration alarms, there were 2 blood glucose levels recorded (248 mg/dL [13.8 mmol/L] and 287 mg/dL [15.9 mmol/L]).
The correlation between meter and biographer glucose values was the highest when comparing meter glucose values with biographer values obtained 10 to 20 minutes later (r = 0.90, n = 28). When analyzing the paired glucose values on a Consensus error analysis grid, the number of points falling in the clinically acceptable regions (A and B) was 100% (Fig. 1).4 Routine overnight blood glucose testing allowed for the collection of data before biographer alarms. Correlation values at other times ranged between 0.66 (20–30 minutes following a meter glucose) and 0.85 (0–10 minutes before a meter glucose value).
|
The biographer low-glucose alarm was set to 85 mg/dL (4.7 mmol/L). When data from the biographers were downloaded to a computer, there were 188 low-glucose alarms. Written records and downloads of camp glucose meters revealed 20 glucose values obtained within 20 minutes after the biographer low-glucose alarm (11%). On 10 of these 20 occasions, both the meter and the biographer glucose values were <85 mg/dL. On 10 occasions the biographer value was <85 mg/dL, with a meter value >85 mg/dL. There were no false-negative alarms (meter value <85 mg/dL and biographer value >85 mg/dL). As mentioned, there was a meter low of 47 mg/dL (2.6 mmol/L), at which time there was no biographer glucose reading, but a perspiration alarm preceded and followed the low value.
There were a total of 80 high-glucose (GL HI) alarms. The response to high alarms was not studied formally. Counselors were instructed to treat the child with fluid and check urine ketones, and, on occasion, additional insulin was administered after a confirmatory capillary glucose level was obtained. No pump malfunctions or set failures were detected by a hyperglycemic alarm.
A nocturnal hypoglycemic seizure occurred in one child who was not wearing a biographer. Another child was found unarousable with a meter glucose value of 20 mg/dL (1.1 mmol/L) while the biographer was alarming. She was treated successfully.
When questioned about their experience with the biographer, 74% of campers who wore the biographer found that the biographer was helpful overnight, and 67% felt more comfortable when wearing the biographer. Campers stated they would wear the biographer 4 to 5 nights a week and 4 to 5 days a week. Campers wearing the biographer reported that their sleep was disrupted by alarms 32% of nights.
Minor pruritis occurred when the biographers were initially placed. When formally questioned about their experience, skin irritation was perceived by the campers as not a problem in 43%, a minor problem in 43%, and a major problem in 14%. No severe or significant skin reactions such as blistering were noted at any time during the study. Two children requested cream or topical hydrocortisone for persistent pruritis.
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
The biographer was well-tolerated by all campers, and it was worn on nearly two thirds of the nights when a biographer was available for a camper to use. In terms of use by age group, calibration of the biographer was successful in all 15- to 17-year-olds. The lowest rate of calibration was in 12- to 14-year-olds. This may represent their increased activity and perspiration following placement of the biographer or increased preference to remove the biographer.
The rate of error or skipped readings of 34% is higher than the 10% to 20% skipped readings observed in other studies.5,6 This is likely because of increased skipped readings caused by perspiration. We observed 20% of errors caused by perspiration compared with 3.1% noted previously.5 The increased perspiration was possibly secondary to the intense physical activities that children were involved in each evening.
Correlation of biographer values and meter values was the highest with biographer values 10 to 20 minutes after the meter value, with a correlation of 0.90 (of 28 paired values). This is consistent with reports that the biographer readings are 17 to 18 minutes delayed from blood glucose values.5,7 This lag is believed to be because of the fact that changes in interstitial glucose concentration lag behind changes in blood glucose concentration and because a reading is reported every 20 minutes after a time-averaged measurement. Although paired biographer and meter values were limited, it is encouraging that the accuracy of the biographer in a camp setting, with a variety of activities and environments, was similar to the correlation values obtained in accuracy studies performed in controlled clinical or home settings.1,6–8
Although half of the low alarms were false-positive alarms, this false-positive rate could possibly be decreased by lowering the low-alarm limit. Of a total of 65 paired blood glucose and biographer values collected and recorded overnight, there were no false-negative alarms or instances of a recorded low-meter glucose value when the biographer did not alarm. There was one low-meter glucose when the biographer perspiration alarm sounded. A detailed analysis of the accuracy of the biographer was not the focus of this study. More information on the rates of false-positive and false-negative alarms would require dedicated personnel to collect overnight data, which was limited by the mandate to not disrupt usual camp routines.
Use of the biographer did detect one severe nocturnal hypoglycemic episode, which allowed for timely treatment. From anecdotal experience, the counselors in the cabins with younger campers reported feeling more comfortable when the campers were wearing the biographer.
A weakness of the study was that training of the counselors and campers was limited in attempts to avoid interference with camp activities. It is likely that data collection as well as the functionality of the biographer such as the calibration rate would improve with more extensive education of the counselors and campers. The relatively low threshold for the high blood glucose alarm (240 mg/dL, 13.3 mmol/L), also likely limited data collection because, on occasion, serial high-glucose alarms led counselors to remove the biographer prematurely.5 Another weakness was reliance on tired counselors who had many responsibilities in addition to that of checking blood glucose values overnight when a biographer alarmed. This is reflected by the fact that there were meter glucose values recorded for only 11% of the low-glucose alarms in the memory of the downloaded biographers. Many children, as well as counselors, either did not hear the alarm, chose not to check a meter glucose at the time of an alarm, checked a meter glucose and did not record the value, or tested campers’ blood glucose on a personal glucose meter, the data of which were inaccessible to our analysis. Our observations in other studies have also shown that the biographer may be turned off voluntarily while sleeping after the child is aroused by the alarm.
We believe that the GlucoWatch biographer was a helpful tool for detecting hypoglycemia at camp. It was well-tolerated, and its function was accurate and acceptable. Our experience highlights the practical issues of using the biographer at a busy diabetes camp. Children with experience using the biographer before camp should be allowed to use the biographer while attending camp, but we would make several recommendations based on our experience. 1) Counselors should receive training in how to turn on, calibrate, and turn off the biographer so they can assist campers wearing a biographer. If the biographer is persistently alarming despite meter blood glucose values that are normal, they should know how to turn off the biographer or recalibrate it. 2) Only one or two campers should wear a biographer in each cabin each night to avoid a high frequency of sleep disruptions. If multiple children are using biographers in a cabin, they can alternate nights of use, or the biographer can be assigned to the children with the greatest need. 3) Because of camp activities, we would recommend that the biographer only be worn at night with initiation of wear before dinner and calibration after campfire before the evening snack. 4) The high alarm should be set to >300 mg/dL (16.6 mmol/L) so that the majority of alarms will be for hypoglycemia. The hypoglycemic alarm should be set between 70 mg/dL (3.9 mmol/L) and 85 mg/dL (4.7 mmol/L) depending on the campers’ history of severe nocturnal hypoglycemia. Additional studies will help elucidate the optimal setting for the low alarm to maximize sensitivity and specificity.
This study was performed by using a GlucoWatch biographer. Since completion of this study, the G2 biographer has been approved. It has been our experience that there is a much higher initial calibration rate (>90%) using the GlucoWatch G2 biographer. The G2 biographer requires a 2-hour (instead of 3-hour) warm-up period before calibration and provides readings up to every 10 minutes (instead of every 20 minutes).
Diabetes camp represents an ideal setting where new technology can be used to help improve the health and safety of children with diabetes. The GlucoWatch biographer is one such new technology that we feel can be used successfully to improve detection of nocturnal hypoglycemia and even improve glycemic control in this setting.
|
ACKNOWLEDGMENTS |
|---|
We are grateful to the campers and counselors who participated in this study; the Diabetes Society of Santa Clara Valley; and Cygnus, Inc for providing the GlucoWatch biographers.
|
FOOTNOTES |
|---|
Received for publication Oct 28, 2003; Accepted Apr 9, 2003.
Address correspondence to Laura M. Gandrud, MD, Department of Pediatric Endocrinology and Diabetes, Stanford University Medical School, SUMC-S302, Stanford, CA 94305-5208. E-mail: lgandrud{at}stanford.edu
Financial disclosure: M. Michelle Van Wyhe, Amy Leptien, and Richard Eastman are Cygnus, Inc. shareholders; Richard Eastman is the Medical Director at Cygnus, Inc.; and M. Michelle Van Wyhe and Amy Leptien were employed by Cygnus, Inc. Bruce Buckingham was a principal investigator and Betsy Kunselman was a site coordinator in a study supported by Cygnus.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
- Eastman R, Hathout E, Fullerbyk L, et al. Accuracy of the GlucoWatch Automatic Glucose Biographer in subjects 7–17 years of age with diabetes [abstract]. Diabetes.2001; 50(suppl 2) :A378
- Trajanoski Z, Brunner GA, Gfrerer RJ, Wach P, Peiber TR. Accuracy of home blood glucose meters during hypoglycemia. Diabetes Care.1996; 19 :1412 –1415[Abstract]
- Harrison B, Markes R, Bradley P, Ismail I. A comparison of statistical techniques to evaluate the performance of the Glucometer Elite blood glucose meter. Clin Biochem.1996; 29 :521 –527[CrossRef][Web of Science][Medline]
- Parkes J, Slatin S, Pardo S, Ginsberg B. A new consensus error grid to evaluate the clinical significance of inaccuracies in the measurement of blood glucose.
Diabetes Care.2000; 23
:1143
–1148
[Abstract/Free Full Text] - Pitzer K, Desai S, Dunn T, et al. Detection of hypoglycemia with the GlucoWatch biographer.
Diabetes Care.2001; 24
:881
–885
[Abstract/Free Full Text] - Garg S, Potts R, Ackerman N, Fermi S, Tamada J, Chase P. Correlation of fingerstick blood glucose measurements with GlucoWatch biographer results in young subjects with type 1 diabetes.
Diabetes Care.1999; 22
:1708
–1714
[Abstract/Free Full Text] - Tamada J, Garg S, Jovanovic L, Pitzer K, Fermi S, Potts R. Noninvasive glucose monitoring: comprehensive clinical results. Cygnus Research Team.
JAMA.1999; 282
:1839
–1844
[Abstract/Free Full Text] - Tierney M, Tamada J, Potts R, et al. The GlucoWatch biographer: a frequent automatic and noninvasive glucose monitor. Ann Med.2000; 32 :632 –641[Web of Science][Medline]
PEDIATRICS (ISSN 1098-4275). ©2004 by the American Academy of Pediatrics
CiteULike 
Connotea 
Del.icio.us 
Digg 
Facebook 
Reddit 
Technorati 
Twitter What's this?
This article has been cited by other articles:
|
|
 |
|
  E. Hathout, N. Patel, C. Southern, J. Hill, R. Anderson, J. Sharkey, M. Hadley-Scofield, L. Tran, A. Leptien, M. Lopatin, et al. Home Use of the GlucoWatch G2 Biographer in Children With Diabetes Pediatrics, March 1, 2005; 115(3): 662 - 666. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. C. Klonoff The Need for Separate Performance Goals for Glucose Sensors in the Hypoglycemic, Normoglycemic, and Hyperglycemic Ranges Diabetes Care, March 1, 2004; 27(3): 834 - 836. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
