Home / Resources / Articles / Web-based Decision Support Program to Improve Diabetes Control

Web-based Decision Support Program to Improve Diabetes Control

Apr 10, 2015

Program helps aid primary care physicians in the management of patients with diabetes using individualized therapy….

A meta-analysis showed that the most efficacious strategies to improve glucose control were: case management, health care team changes, patient education, and facilitated transmission of patient data to clinicians. Using those 4 key strategies along with leading guidelines for the management of diabetes, the authors created a web-based program to help aid primary care physicians in the management of patients with diabetes using individualized therapy. This study reports the evaluation of the efficacy and safety of using this web-based program in diabetic patients from an outpatient VA medical center located in Atlanta.

The program was evaluated with a randomized, double blind, before-and-after designed study on 113 patients. The patients had an A1c >7% despite therapy with metformin ± sulfonylureas or insulin, and patients were given either exenatide or placebo. Patients were to record glucose levels daily (pre-breakfast, pre- and post prandial, and at bedtime) and the use of glucose-lowering medications once a week. Patients were asked to call or fax information to the study stuff and inform the staff of any hypoglycemia. The study staff would then enter data into the web-based program. A1c was also recorded at baseline, 3-, 6-, 9-, and 12-months.

The program used algorithms to evaluate glucose control and recommend changes (with “rationales” for the recommendation included) as needed, mimicking the decision making of an endocrinologist. The algorithm glucose goal was 75-100 mg/dL pre-breakfast, 80-120 mg/dL preprandial and 110-150 mg/dL at bedtime. Medication algorithms included: maximization of metformin at 2000 mg/day, glipizide as needed before meals and/or at bedtime (max: 40 mg/day or 20 mg at bedtime), and Lantus® at 0.15 mg/kg body weight given before evening meal. The program would adjust the recommended therapy based on hypoglycemia, variability of glucose levels, or if a low glucose level was predicted. The final management plan remains the responsibility of the clinician, but the program takes advantage of different health care members’ roles in the management of the individual and patient education was assigned to the staff.

The results showed a decrease in A1c levels from the average 8.18% to 7.16% at 6 months and a rise to 7.54% at 12 months; in the adherent group, the baseline was 8.12% and decreased to 6.98% at 6 months, and remained near this level at 12 months:


Although the study claims to have demonstrated both efficacy and safety, several limitations still exist. The design itself was a before-and-after design and not a randomized, controlled trial. The authors cannot be sure of the patients’ glycemic control outcome if they were to continue being managed with just the primary care. Some patients were given antidiabetic medications not listed in the algorithm, and hypoglycemia was shown to be less than 1 episode per patient per month, but safety was demonstrated using only 16 patients from the adherent group.

Practice Pearls:

  • Web-based decision support program used in the management of diabetes.
  • Algorithms designed to mimic the decision making of a endocrinologist.
  • Shown to be safe and effective, but with many limitations that requires further studies.

Phillips LS, Barb D, Yong C, Tomolo AM, Jackson SL, et al. Translating What Works: A New Approach to Improve Diabetes Management. Journal of Diabetes Science and Technology. 9 March 2015. Web. 31 March 2015.