A Computer Program, Directed by Clinicians for Outpatients, Which Re-adjusts Subcutaneous Multiple Daily Injections (MDI) of Insulin, Achieves an Improvement of 2.6 A1c% Points

A Computer Program, Directed by Clinicians for Outpatients, Which Re-adjusts Subcutaneous Multiple Daily Injections (MDI) of Insulin, Achieves an Improvement of 2.6 A1c% Points

Presentation

International Hospital Diabetes Meeting

Date

October 2012

Authors

PC Davidson,1 BW Bode,1 HR Hebblewhite,2 R Booth2

OBJECTIVE

This study examines the performance of a computerized algorithm for adjusting subcutaneous basal-bolus MDI insulin regimens for outpatients.

ALGORITHM

This algorithm determines the meal bolus for a meal by applying an adjustment to the previous day’s same-time meal bolus based upon the subsequent BG response. Basal doses are adjusted in a similar manner based upon the lower of pre-breakfast BG or any immediately preceding nighttime lower BG tested earlier on the same day. The patients reported their BGs by text-messaging, email, telephone or the TelcareTM real-time cell-based meter system. Adjusted dosing information was relayed to the patient.

STATISTICAL METHODS

Glycemic Control: All the subjects were started on the Glucommander SubQ algorithm at the beginning of the study. The statistics followed a paired before-and-after design.

  • Method 1: For patients treated for > 60 days, each patient’s pre-study A1c was compared to his/her post-study A1c.
  • Method 2: For all patients, the mean of the last 12 BGs of each patient was converted to an A1c value using the ADA-approved correlation. This value was compared to the pre-study A1c.

Hypoglycemia:

  • The BG readings below threshold were counted for thresholds of 40, 50, and 60 mg/dL.
  • The patients with at least one BG < threshold were counted for 40, 50, and 60 mg/dL.

RESULTS

Results for glycemic control were as follows.

  • Method 1: For 21 patients treated for >60 days, pre-study A1c was 10.1% and post-study A1c was 7.5%; P<0.0004.
  • Method 2: For all 30 patients, pre-study A1c was 10.2%, post-study mean BG was 134, and estimated post-study A1c was 6.3% (per ADA-endorsed linear correlation); P<0.00001.

Results for hypoglycemia were as follows.

  • Of a total of 5,811 BG tests, the % BGs < threshold was 0.0% <40 mg/dL, 0.4% <50 mg/dl, and 1.0% <60 mg/dL.
  • Of a total of 30 patients, the % patients with at least one BG < threshold was 0.0% <40 mg/dl, 17% <50 mg/dL, and 38% <60 mg/dL.

Affiliations

  1. Atlanta Diabetes Associates, Atlanta, Georgia.
  2. Glytec, Greenville, South Carolina.
Request Demo