Clinical evaluation of computer-assisted self-monitoring of blood glucose system

The Glucometer M Diabetes Management System includes a glucose-reflectance meter with memory that can interface with a microcomputer for data manipulation and analysis. We evaluated the system in a short-term randomized control trial to determine its impact on metabolic control, self-monitoring of blood glucose (SMBG) testing behaviors, regimen self-adjustment, understanding of insulin-dependent diabetes mellitus (IDDM) treatment, attitudes about SMBG, and perceived quality of patient-physician interaction. Twenty-nine adolescent subjects (experimental) with IDDM were randomly assigned the Glucometer M system for 4 mo. Twenty-eight control subjects used meters without memory. All subjects returned twice to the clinic at 2-mo intervals during the study. At clinic visits, both groups reviewed their SMBG data with their physician. Reviews on experimental subjects were conducted with computer-generated data formats. Control subject reviews used traditional logbooks. Both groups showed a significant drop in glycosylated hemoglobin during the study period (P less than .001); however, there were no between-group differences. There were also no differences in SMBG testing behavior or self-reported regimen self-adjustment between groups or within groups compared with baseline. Compared with control subjects, experimental subjects indicated a significant increase in self-reported understanding of IDDM treatment (P = .002), perceived importance of testing (P = .006), and the quality of interaction with their physician (P less than .001). These data suggest that use of computer-assisted SMBG systems in the outpatient setting does not improve metabolic control over 4 mo. It may, however, contribute to improving communication between the patient and health-care providers.