William L. Clarke

Evaluating Clinical Accuracy of Systems for Self-Monitoring of Blood Glucose

Although the scientific literature contains numerous reports of the statistical accuracy of systems for self-monitoring of blood glucose (SMBG), most of these studies determine accuracy in ways that may not be clinically useful. We have developed an error grid analysis (EGA), which describes the clinical accuracy of SMBG systems over the entire range of blood glucose values, taking into account 1) the absolute value of the system-generated glucose value, 2) the absolute value of the reference blood glucose value, 3) the relative difference between these two values, and 4) the clinical significance of this difference. The EGA of accuracy of five different reflectance meters (Eyetone, Dextrometer, Glucometer I, Glucometer II, Memory Glucometer II), a visually interpretable glucose reagent strip (Glucostix), and filter-paper spot glucose determinations is presented. In addition, reanalyses of a laboratory comparison of three reflectance meters (Accucheck II, Glucometer II, Glucoscan 9000) and of two previously published studies comparing the accuracy of five different reflectance meters with EGA is described. EGA provides the practitioner and the researcher with a clinically meaningful method for evaluating the accuracy of blood glucose values generated with various monitoring systems and for analyzing the clinical implications of previously published data.

Reduced Awareness of Hypoglycemia in Adults With IDDM: A prospective study of hypoglycemic frequency and associated symptoms

OBJECTIVE To prospectively evaluate the frequency and severity of hypoglycemic episodes in IDDM subjects who declare themselves to have reduced awareness of hypoglycemia, to validate their self-designations in their natural environment, and to determine objectively the presence or absence of autonomic and neuroglycopenic symptoms associated with their low blood glucose (BG) levels. RESEARCH DESIGN AND METHODS A total of 78 insulin-dependent diabetes mellitus (IDDM) subjects (mean age 38.3 ± 9.2 years; duration of diabetes 19.3 ± 10.4 years) completed two sets of assessments separated by 6 months. The assessments included reports of frequency and severity of low BG, symptoms associated with low BG, and a BG symptom/estimation trial using a hand-held computer (HHC). Diaries of hypoglycemic episodes were kept for the intervening 6 months. HbA1 levels were determined at each assessment. RESULTS Of the subjects, 39 declared themselves as having reduced awareness of hypoglycemia (reduced-awareness subjects). There were no differences between these reduced-awareness subjects and aware subjects with regard to age, sex, disease duration, insulin dose, or HbA1. During the HHC trials, reduced-awareness subjects were significantly less accurate in detecting BG < 3.9 mmol/l (33.2 ± 47 vs. 47.6 ± 50% detection, P = 0.001) and had significantly fewer autonomic (0.41 ± 0.82 vs. 1.08 ± 1.22, P = 0.006, reduced-awareness vs. aware) and neuroglycopenic (0.44 ± 0.85 vs. 1.18 ± 1.32, P = 0.004, reduced-awareness vs. aware) symptoms per subject. Prospective diary records revealed that reduced-awareness subjects experienced more moderate (351 vs. 238, P = 0.026) and severe (50 vs. 17, P = 0.0062) hypoglycemic events. The second assessment results were similar to the first and verified the reliability of the data. CONCLUSIONS IDDM subjects who believe they have reduced awareness of hypoglycemia are generally correct. They have a history of more moderate and severe hypoglycemia, are less accurate at detecting BG < 3.9 mmol/l, and prospectively experience more moderate and severe hypoglycemia than do aware subjects. Neither disease duration nor level of glucose control explains their reduced awareness of hypoglycemia. Reduced-awareness individuals may benefit from interventions designed to teach them to recognize all of their potential early warning symptoms.

Accuracy of Perceiving Blood Glucose in IDDM

Type I (insulin-dependent) diabetic individuals and health professionals often assume that the symptoms of extremely low or high blood glucose (BG) levels can be recognized and, consequently, appropriate treatment decisions can be based on symptom perception. Because no research has documented the validity of these assumptions, this study tested the ability to perceive BG concentration. Nineteen type I adults, experienced in self-monitoring of BG (SMBG), estimated their BG 40–54 times just before measurement of actual BG. This procedure was repeated under two conditions: (1) in the hospital (hospital condition) while connected to an insulin/glucose infusion system that artificially manipulated BG, leaving subjects only symptomatic, or internal, cues and (2) in the natural environment (home condition), where both internal and external cues, e.g., food and insulin consumption, were available. Estimates significantly correlated with actual BG for 7 of 16 subjects in the hospital condition and for 18 of 19 subjects in the home condition. Believed ability to estimate BG did not predict documented ability in either condition. An evaluation of the treatment significance of estimation errors showed that the majority of errors were relatively benign. The most common error affecting clinical outcome was estimated euglycemia when actual BG was hypoglycemic or hyperglycemic.

Multinational Study of Subcutaneous Model-Predictive Closed-Loop Control in Type 1 Diabetes Mellitus: Summary of the Results

Background: In 2008–2009, the first multinational study was completed comparing closed-loop control (artificial pancreas) to state-of-the-art open-loop therapy in adults with type 1 diabetes mellitus (T1DM). Methods: The design of the control algorithm was done entirely in silico, i.e., using computer simulation experiments with N = 300 synthetic “subjects” with T1DM instead of traditional animal trials. The clinical experiments recruited 20 adults with T1DM at the Universities of Virginia (11); Padova, Italy (6); and Montpellier, France (3). Open-loop and closed-loop admission was scheduled 3–4 weeks apart, continued for 22 h (14.5 h of which were in closed loop), and used a continuous glucose monitor and an insulin pump. The only difference between the two sessions was that insulin dosing was performed by the patient under a physicians supervision during open loop, whereas insulin dosing was performed by a control algorithm during closed loop. Results: In silico design resulted in rapid (less than 6 months compared to years of animal trials) and cost-effective system development, testing, and regulatory approvals in the United States, Italy, and France. In the clinic, compared to open-loop, closed-loop control reduced nocturnal hypoglycemia (blood glucose below 3.9 mmol/liter) from 23 to 5 episodes (p < .01) and increased the amount of time spent overnight within the target range (3.9 to 7.8 mmol/liter) from 64% to 78% (p = .03). Conclusions: In silico experiments can be used as viable alternatives to animal trials for the preclinical testing of insulin treatment strategies. Compared to open-loop treatment under identical conditions, closed-loop control improves the overnight regulation of diabetes.

Closed-loop artificial pancreas using subcutaneous glucose sensing and insulin delivery and a model predictive control algorithm: the Virginia experience.

Background: Recent progress in the development of clinically accurate continuous glucose monitors (CGMs), automated continuous insulin infusion pumps, and control algorithms for calculating insulin doses from CGM data have enabled the development of prototypes of subcutaneous closed-loop systems for controlling blood glucose (BG) levels in type 1 diabetes. The use of a new personalized model predictive control (MPC) algorithm to determine insulin doses to achieve and maintain BG levels between 70 and 140 mg/dl overnight and to control postprandial BG levels is presented. Methods: Eight adults with type 1 diabetes were studied twice, once using their personal open-loop systems to control BG overnight and for 4 h following a standardized meal and once using a closed-loop system that utilizes the MPC algorithm to control BG overnight and for 4 h following a standardized meal. Average BG levels, percentage of time within BG target of 70–140 mg/dl, number of hypoglycemia episodes, and postprandial BG excursions during both study periods were compared. Results: With closed-loop control, once BG levels achieved the target range (70–140 mg/dl), they remained within that range throughout the night in seven of the eight subjects. One subject developed a BG level of 65 mg/dl, which was signaled by the CGM trend analysis, and the MPC algorithm directed the discontinuance of the insulin infusion. The number of overnight hypoglycemic events was significantly reduced (p = .011) with closed-loop control. Postprandial BG excursions were similar during closed-loop and open-loop control Conclusion: Model predictive closed-loop control of BG levels can be achieved overnight and following a standardized breakfast meal. This “artificial pancreas” controls BG levels as effectively as patient-directed open-loop control following a morning meal but is significantly superior to open-loop control in preventing overnight hypoglycemia.

Assessment and management of hypoglycemia in children and adolescents with diabetes

Barbara Davis Center, University of Colorado at Denver,Aurora, CO, USACorresponding author:William Clarke, MDDepartment of PediatricsUniversity of VirginiaP.O. Box 800386CharlottesvilleVA 22908USA.Tel: 434 924 5897fax: 434 924 9181e-mail: wlc@virginia.eduEditors of the ISPAD Clinical Practice Consensus Guidelines2006–2007: William Clarke, Timothy Jones, Arleta Rewers,David Dunger, and Georgeanna J Klingensmith.

Organ-specific autoantibodies in children with common endocrine diseases

Sera from 438 children were examined for autoantibodies to thyroid microsomes, thyroglobulin, pancreatic islet cells, gastric parietal cells, and adrenocortical cells by indirect hemagglutination and immunofluorescence techniques. A modification of the indirect hemagglutination technique allowed specific detection of low titers of antithyroidal antibodies. The subjects included a control group (117) with no known autoimmune disease, and children with disorders of the thyroid (88), insulin-dependent diabetes mellitus (201), Turners Syndrome (24), and Addison disease (8). A subjects age at the time of disease onset and the race and sex were correlated with the prevalence of autoantibodies. The coincidence of autoantibodies to components of the thyroid with autoantibodies to gastric parietal cells was increased in children with disorders of the thyroid (94%, 18/19) over that observed in diabetes (29%, 4/14), Turner syndrome (0%), or Addison disease (0%), perhaps indicating different genetic propensities for the development of parietal cell antibodies in these groups. Islet cell antibodies were not found in subjects with Turner syndrome, nor were they more prevalent in white or black subjects with diabetes. The incidence of organ-specific autoantibodies in individuals without overt clinical disease may reflect an altered immunologic state that will lead eventually to autoimmune disease. Islet cell antibodies decline in prevalence in diabetes, whereas thyroid antibodies in disorders of the thyroid do not; this may reflect differences in the pathogenesis of these common autoimmune endocrine disorders in children.

Algorithmic Evaluation of Metabolic Control and Risk of Severe Hypoglycemia in Type 1 and Type 2 Diabetes Using Self-Monitoring Blood Glucose Data

The optimization of metabolic control in Type 1 and Type 2 diabetes mellitus (T1DM and T2DM, respectively) [i.e., the maintenance of near-normal hemoglobin A(1c) (HbA(1c)) without increasing the risk of hypoglycemia] could be enhanced by analysis of self-monitoring blood glucose (SMBG) data assessing complementary processes: exposure to hyperglycemia and hypoglycemia. We present algorithms that simultaneously estimate HbA(1)c and risk for significant hypoglycemia using 45-60 days of SMBG. The algorithms were developed using a primary data for 96 subjects with T1DM (n = 48) and T2DM, and were validated in an external data for 520 subjects with T1DM (n = 231) and T2DM. All subjects were on insulin. In the primary (external) data the estimation of HbA(1c) had absolute error of 0.5 (0.7) units of HbA(1c) and percent error of 6.8% (8.1%); 96% (96%) of all estimates were within 20% from reference HbA(1c). The SMBG-estimated value of HbA(1c) was closer to current reference HbA(1c) than a reference HbA(1c) value taken only 2-3 months ago. The results in T1DM and T2DM were similar. Linear model predicted future significant hypoglycemia (R(2) = 62%, p < 0.0001). The leading predictor was a previously introduced Low Blood Glucose Index, which alone had R(2) = 55%. Probability model assessed accurately the odds for future moderate/severe hypoglycemia (coefficients of determination 92%/94%). Four risk categories were identified; within moderate- and high-risk category, there was no difference between T1DM and T2DM in the occurrence of prospective significant hypoglycemia. SMBG data allow for accurate estimation of the two most important markers of metabolic control in T1DM and T2DM - HbA(1c) and risk for hypoglycemia.

Cognitive Function Is Disrupted by Both Hypo- and Hyperglycemia in School-Aged Children With Type 1 Diabetes: A Field Study

OBJECTIVE We developed a field procedure using personal digital assistant (PDA) technology to test the hypothesis that naturally occurring episodes of hypo- and hyperglycemia are associated with deterioration in cognitive function in children with type 1 diabetes. RESEARCH DESIGN AND METHODS A total of 61 children aged 6–11 years with type 1 diabetes received a PDA programmed with two brief cognitive tests (mental math and choice reaction time), which they completed just before home glucose readings. The computer recorded time to complete each test and number of correct responses. Children completed several trials per day over 4–6 weeks for a total of 70 trials. Performance variables were compared across glucose ranges. Individual impairment scores (IISs) were also computed for each child by calculating the SD between performance during euglycemia and that during glucose extremes. RESULTS Time to complete both mental math and reaction time was significantly longer during hypoglycemia. During hyperglycemia, time to complete math was significantly longer and reaction time was marginally significant (P = 0.053). There were no differences on task accuracy. Decline in mental math performance was equivalent at glucose levels <3.0 and >22.2 mmol/l. IISs varied greatly across children, with no age or sex differences. CONCLUSIONS A decrease in mental efficiency occurs with naturally occurring hypo- and hyperglycemic glucose fluctuations in children with type 1 diabetes, and this effect can be detected with a field procedure using PDA technology. With blood glucose levels >22.2 mmol/l, cognitive deterioration equals that associated with significant hypoglycemia.

Predictors of fear of hypoglycemia in adolescents with type 1 diabetes and their parents

Objectives:  This study tested the hypothesis that both trait anxiety and hypoglycemic history contribute to fear of hypoglycemia (FOH) both in adolescents with type 1 diabetes mellitus (T1DM) and in their parents, and relationships between FOH and other variables including metabolic control, symptom perception, and use of insulin pump therapy.