American Diabetes Association (ADA) 71st Scientific Sessions: Page 2 of 6

August 12, 2011

Comparison of Premixed Insulin Analogs and Long-Acting Insulin Analogs

A diabetes simulation model programmed from a US healthcare payer perspective found that patients with type 2 diabetes taking premixed insulin analogs were likely to have longer life expectancy and quality-adjusted life expectancy compared with those taking long-acting insulin analogs (LAIAs). The authors reported that the model associated premixed insulin analogs with higher lifetime and direct medical costs, calculating  incremental cost-effectiveness ratios per quality-adjusted life-year of $28,580 for the premixed insulin analog group as compared with $23,150 for the LAIA group.

Results were presented during a president’s poster session at the ADA meeting. The poster was titled Estimating the Long-Term Cost-Effectiveness of Premixed Insulin Analog Versus Long-Acting Insulin Analog in the United States.

The researchers noted that an earlier meta-analysis from the Agency for Healthcare Research and Quality had concluded premixed insulin analogs provide more glycemic control than LAIAs. The substantial cost for insulin, which a 2007 ADA report placed at approximately $3.7 billion annually, prompted the authors to compare the cost-effectiveness of insulin regimens. The drugs compared in the model were insulin lispro mix 75/25 (75% insulin lispro protamine suspension/25% insulin lispro), insulin lispro mix 50/50 (50% insulin lispro protamine suspension/50% insulin lispro), and an LAIA regimen.

The model, which was validated, was used to simulate mortality risk and incidence of diabetes-related complications and assumed each patient took an average of 40 insulin units per day. The authors obtained insulin costs and direct medical costs from published sources and diabetes complication costs from a 2008 study. They also used published sources to establish baseline cohort characteristics. The population was 57.5% male, and the mean patient age was 62.8 years. Patients were assumed to have had diabetes for a mean of 12.0 years and had a mean glycated hemoglobin level of 8.6%.

The model calculated a mean life expectancy of 7.82 years for the LAIA group, 7.89 for the insulin lispro mix 75/25 group, and 7.91 years for the insulin pro mix 50/50 group. The mean quality-adjusted life expectancy was 4.93 years in the LAIA group, 5.00 in the insulin pro mix 75/25 group, and 5.01 years in the insulin pro mix 50/50 group. Direct costs associated with the groups were $40,128 (LAIA), $41,852 (insulin pro mix 75/25), and $41,848 (insulin pro mix 50/50).

The authors also applied incremental cost-effectiveness scatter plots and acceptability curves to the data. They found that with a willingness to pay $50,000 per quality-adjusted life-year gained, there was a 58.8% probability that insulin lispro mix 75/25 would be more cost-effective compared with LAIA and a 64.5% probability that insulin pro mix 50/50 would be more cost-effective compared with LAIA.

The authors acknowledged limitations to the study, including their decision not to include hypoglycemia incidence in the base-case analysis. They also noted that using a meta-analysis of short-term clinical trials to make long-term predictions might have confounded the findings. Despite these concerns, the authors pointed out that using a validated and published model ranks among the best and most accurate ways to project long-term cost-effectiveness for a given study population.

This study was sponsored by Eli Lilly and Co.