OBJECTIVES: To project the long-term clinical and economic outcomes of intensive treatment with either biphasic insulin aspart 30 (BIAsp30) or insulin glargine among type 2 diabetes patients exhibiting particularly high HbA1c levels when taking oral antidiabetic therapy alone. METHODS: Baseline characteristics and end-of-study treatment effect data among a sub-population of insulin naïve type 2 subjects with baseline HbA1c levels ³ 8.5% (mean HbA1c: 10.2%) were derived from a multicenter, 28-week, head-to-head clinical trial (INITIATE). Significant improvements in HbA1c levels favoring randomization to twice-daily BIAsp 30 + metformin (met) ± thiazolidinedione (TZD) compared to bedtime insulin glargine + met ± TZD were demonstrated (-0.53% between arms; p<0.005). A peer-reviewed, validated Markov/Monte-Carlo model combining published literature for risk of long-term diabetic complications with quality-of-life utilities projected the incremental cost-effectiveness ratio (ICER) and cumulative incidences of diabetes-related complications over 35 years. Cost-effectiveness was measured as cost per life years gained (LYG) and cost per quality adjusted life years gained (QALY). Cardiovascular, neurological, renal, and retinal complication rates were assessed. Lifetime costs were calculated as the annual direct pharmacy costs plus complication costs (US Medicare perspective). Clinical outcomes and costs were discounted at 3% annually. Sensitivity analyses were performed. RESULTS: Improvements in glycemic control corresponded with incremental increases in LYG and QALY favoring BIAsp 30 versus glargine (0.28±0.21 and 0.27±0.15 years, respectively). Treatment with BIAsp 30 was associated with reductions in the cumulative incidence of diabetes-related complications, notably in renal (18% less end-stage renal disease) and retinal (12% less severe vision loss) co-morbidities. An ICER of $30,924 per QALY gained was deduced. Sensitivity analyses support the reliability of the results. CONCLUSIONS: Among a sub-population of poorly controlled insulin naïve type 2 patients, BIAsp 30 was estimated to reduce lifetime complication incidences and be cost-effective within commonly supported thresholds when compared to insulin glargine.