Multicenter evaluation of the Glucometer Elite XL meter, an instrument specifically designed for use with neonates.

OBJECTIVE: To evaluate the clinical performance of the Glucometer Elite XL Diabetes Care System in neonatal settings using a multicenter study RESEARCH DESIGN AND METHODS: A total of 388 blood specimens from 333 neonates were included in the study. A capillary or arterial sample was analyzed for determination of glucose with the Glucometer Elite XL system by an attending trained nurse. Through the same sampling site, a specimen was collected and sent to the laboratory for measurement of plasma glucose, bilirubin, and hematocrit. RESULTS: The regression analysis between the results of the Glucometer Elite XL system and comparative methods resulted in the following: Glucometer Elite XL meter = 1.01 x laboratory method + 0.02 mmol/l (n = 388). For the 1.1-4.0 mmol/l plasma glucose range, the regression was Glucometer Elite XL meter = 1.07 x laboratory method + 0.12 mmol/l (n = 150). A difference plot indicated a mean bias of 0.04 mmol/l (95% CI -0.01 to 0.10). No relationship was found between meter glucose biases and hematocrit levels (r = 0.10, P = 0.14). Although a statistically significant correlation existed between bilirubin levels and the glucose meter biases (r = 0.14, P = 0.005), the predicted mean biases were of little clinical significance. CONCLUSIONS: The Glucometer Elite XL system showed a good performance when used in neonatal settings.

Clinical evaluation of the impact of a patient education program.

A 5-day patient education program, taught on an outpatient basis, was evaluated to determine its effect on metabolic control as reflected by glycosylated hemoglobin test values. A quasi-experimental design was used, consisting of a pretest, a posttest, and a follow-up assessment made approximately 6 months after the posttest. The 72 experimental and 324 comparison subjects all had insulin-dependent diabetes mellitus (IDDM), were between 14 and 78 years of age, and had a duration of diabetes ranging from 1 to 20 years. The experimental group demonstrated a statistically significant improvement in Hb A1 values from pre- to posttest and sustained these posttest levels upon follow-up, although not at statistically significant levels. The comparison group showed no pre- to posttest difference, but demonstrated an improvement from posttest to follow-up assessment.

Psychometric study of patient knowledge test.

The results of a psychometric study of the Test of Patient Knowledge are reported. The 50-item, multiple-choice test, developed by Etzwiler and associates at the International Diabetes Center, consists of a total score and seven subscores based on seven nonoverlapping content categories: nutrition, insulin, general knowledge, methods of control, pattern control, exercise, and complications. The results described herein provide evidence for the validity of the test (content, concurrent, and discriminant validity), a high level of reliability (Cronbach's alpha = .88), readability for the layperson at the 7th- to 8th-grade level, and sensitivity to instructional gains. The literature on psychometric research with other tests of patient knowledge of diabetes is reviewed and compared with the results of this study.