Send to

Choose Destination
Curr Med Res Opin. 2008 Jan;24(1):11-20.

Improving glycemic control with insulin detemir using the 303 Algorithm in insulin naïve patients with type 2 diabetes: a subgroup analysis of the US PREDICTIVE 303 study.

Author information

Diabetes Research Center, Tustin, CA 92780, USA.



PREDICTIVE 303 was a 26-week, prospective, randomized, open-label, multi-center study in patients with type 2 diabetes that investigated whether patient-driven adjustments of insulin detemir doses using the 303 Algorithm achieved similar glycemic control compared to standard-of-care, physician-driven adjustments in doses. This post hoc sub-analysis evaluates insulin naïve patients on oral anti-diabetic drugs (OADs) who were directed to start on once-daily insulin detemir as add-on therapy to any other glucose-lowering regimens.


Patients in the 303 Algorithm group were instructed to adjust their detemir dose every 3 days based on mean fasting plasma glucose (FPG) values using a simple algorithm: mean FPG < 80 mg/dL, reduce dose by 3 units; between 80-110mg/dL, no change; > 110mg/dL, increase by 3 units. Physicians adjusted the detemir dose for patients in the Standard-of-care group according to their usual practice. No control insulin was used for comparison to insulin detemir.


Reductions in glycosylated hemoglobin (HbA(1c)) from baseline were similar between those patients in the 303 Algorithm and Standard-of-care groups (-1.1 and -1.0%, respectively; between group p = 0.0933); patients in the 303 Algorithm group achieved a greater reduction in FPG. Patients in both groups experienced a similar, low rate of hypoglycemia. Over 95% and 92% of patients, respectively, used detemir once daily.


These data indicate that patients with type 2 diabetes naïve to insulin can effectively implement the 303 Algorithm to initiate and adjust a once-daily dose of insulin detemir to achieve improvements in glycemic control.

[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Taylor & Francis
Loading ...
Support Center