Patient-driven dose adjustment led to slightly but significantly more non-major hypoglycemic events than physician adjustment (P<0.0001), but rates were still substantially lower than at baseline for both, they reported at the American Diabetes Association meeting.

The findings should be reassuring "for primary care physicians who may be uncomfortable putting the care in the hands of the patient," Dr. Meneghini said.  He and colleagues conducted the prospective, open-label PREDICTIVE 303 study to test the safety and efficacy of the simplified patient-driven algorithm.

In the study, 5,604 patients began insulin detemir as an add-on to oral antidiabetic agents or as a replacement for other basal insulin drugs. They were randomized by study site (1,083 in the United States) to the "303" algorithm or standard of care for 26 weeks.

The algorithm group was taught to adjust their dose of insulin detemir every three days using fasting plasma glucose values as a guide.

If glucose levels averaged less than 80 mg/dL, patients were to reduce their dose by three units. If levels averaged above 110 mg/dL, dose was to be increased by three units. Otherwise there was to be no change.

The control group had physicians adjust insulin detemir dose based on standard of care.  Dose adjustments were not enforced in either group.

Baseline characteristics were similar between groups. All patients were to have an HbA1c of no more than 12% and had to be judged "likely to benefit from initiation, addition, switching to, or continuation of insulin detemir therapy" to enter the trial. The majority (85%) of patients gave "improve glycemic control" as the reason for starting a new therapy.

After 26 weeks, overall HbA1c levels were improved slightly but significantly more by patient-driven dose adjustment than standard of care (P=0.0106), though both strategies improved levels compared with baseline (8.5% versus 7.9% and 8.5% versus 8.0%, respectively, both P<0.0001).

The advantage to the 303 algorithm was more substantial for fasting plasma glucose.

Again, both adjustment strategies improved glucose levels over baseline (175 versus 141 mg/dL and 174 versus 152 mg/dL, both P<0.0001) but the patient-driven algorithm achieved significantly better results (P<0.0001).

Likewise, the algorithm and control groups had significantly reduced hypoglycemia rates overall, during the daytime, and for major events compared with baseline (P<0.05 for all comparisons).

Body weight was essentially stable for both groups with no significant difference between them (P=0.0676). Nearly all of the insulin naïve patients were on once daily dosing at 26 weeks, although the average end dose was higher with the patient-driven algorithm (0.59 versus 0.40 U/kg).

Dr. Meneghini concluded that patient-driven dose adjustments "appear to be a safe and effective alternative to physician-directed dose adjustment in the primary care setting."

However, he noted that even the best algorithm cannot improve outcomes if it is not carried through.

American Diabetes Association meeting:  Meneghini L, et al "Efficacy and Safety of Insulin Detemir in a Large Cohort of Patients with Type 2 Diabetes Using a Simplified Self-adjusted Dosing Guideline: Results of the PREDICTIVE 303 Study" ADA meeting 2007; Abstract 197-OR.

==========================

Advertisement
Do you have patients in pain? Over 5,000 Anodyne Therapy Care Providers in 35 countries worldwide are helping thousands of patients per month with painful, circulatory problems get relief. To learn how you can help your patients get back to life, click here.