The effectiveness of insulin initiation regimens in patients with type 2 diabetes mellitus: a large national medical records review study comparing a basal insulin analogue to premixed insulin^*

ABSTRACT

Objectives: The objective of the study was to compare the level of HbA_1c reduction between a once-daily basal insulin analogue (glargine, GLG) and two twice-daily premixed insulin analogue formulations (premixed insulin lispro 75/25, PIL; premixed human insulin 70/30, PHI) in patients with type 2 diabetes mellitus (T2DM) initiating insulin therapy.

Research design and methods: Data were extracted from a US national medical records database for this retrospective, 18-month, observational, cohort study. Patients with T2DM were initiated on GLG (n = 3624), PIL (n = 895) or PHI (n = 3647). A combined premixed insulin group (CPI; n = 4542) was formed for data analyses. Propensity score methods were used to adjust for 19 baseline characteristics.

Main outcome measures: Adjusted and unadjusted reductions in HbA_1c at six time points post-insulin initiation for a period of 18 months.

Results: Reductions in mean HbA_1c relative to baseline were demonstrated by all cohorts for all treatment periods. After adjusting for baseline differences, the CPI cohort consistently demonstrated greater reductions in HbA_1c (0.04–0.14%; p < 0.05), compared to the GLG cohort. The PIL cohort consistently demonstrated the greatest reductions in HbA_1c (0.26–0.65%; p < 0.05), compared to the GLG cohort.

Limitations: Retrospective study design and vulnerabilities to patient drop-outs.

Conclusions: In clinical practice settings, greater reductions in HbA_1c were found in patients with premixed insulin than with a basal insulin analogue with the greatest reduction observed with premixed insulin lispro 75/25, confirming the observations of randomized, controlled trials.

Key words: :

• Insulin initiation regimens
• Outcomes research
• Premixed insulin
• Propensity score methods
• Treatment effectiveness
• Type 2 diabetes mellitus

* Some of these data were presented at the 2005 Annual Conference of the European Association of the Study of Diabetes (Podium Presentation, Athens, Greece, September, 2005); the 2006 Annual Conference of the International Society of Pharmacoeconomics and Outcomes Research (Poster, Philadelphia, PA, USA, May 2006); and the 2006 Annual Conference of the American Diabetes Association (Poster, Washington DC, USA, June 2006)

* Some of these data were presented at the 2005 Annual Conference of the European Association of the Study of Diabetes (Podium Presentation, Athens, Greece, September, 2005); the 2006 Annual Conference of the International Society of Pharmacoeconomics and Outcomes Research (Poster, Philadelphia, PA, USA, May 2006); and the 2006 Annual Conference of the American Diabetes Association (Poster, Washington DC, USA, June 2006)

Notes

* Some of these data were presented at the 2005 Annual Conference of the European Association of the Study of Diabetes (Podium Presentation, Athens, Greece, September, 2005); the 2006 Annual Conference of the International Society of Pharmacoeconomics and Outcomes Research (Poster, Philadelphia, PA, USA, May 2006); and the 2006 Annual Conference of the American Diabetes Association (Poster, Washington DC, USA, June 2006)