OBJECTIVES: To compare 1-year healthcare utilization and costs in patients initiating insulin glargine vs NPH. METHODS: Patients with T2DM (03/2001¨C03/2005) who failed oral agents and initiated insulin glargine or NPH were evaluated using the Integrated Health Care Information System, a US managed care health plan database. Patients were continuously enrolled with managed care health plans for °Ý6 months before and 12 months after insulin initiation. Propensity score matched NPH to glargine initiators by baseline demographics, HbA1c, co-morbidities, healthcare utilization, and pharmacy copayment. Conditional logistic regression, McNemar's test, and paired t-test were used to compare subsequent utilizations/ costs between two insulin groups. Costs were paid by health insurance, adjusting for inflation to the most current year value in database. RESULTS: Matched sample (n=1,468) was 46% female, mean age 54.6 yrs., A1C 9.2%, Charlson Comorbidity Index (CCI) 0.69, metformin-use 77.6%, sulfonylureas 77.6%, and thiazolidinedione 56%. Before matching, glargine initiators were more likely than NPH initiators to be female, had higher HbA1c, CCI , more use of TZD, sulfonylurea and statins, fewer visits to an endocrinologist, higher out-of-pocket drug copayment, lower total healthcare utilization and associated costs (except diabetes medications). After propensity score matching, no differences remained between matched pairs. During 12-month follow-up, glargine initiators showed a lower hospitalization rate (OR:0.73, 95%CI [0.57-0.94], P=0.0124) while outpatient and emergency service utilization was not statistically different between groups. Number needed to treat with glargine was 17 (95% CI: 9-59) to avoid hospitalization for a patient. For the same follow-up period, glargine use on average cost $532 vs $293 for NPH (P<0.0001) and $2097 vs $1820 for all antidiabetic medications (P<0.0001). CONCLUSION: Initiation of insulin glargine is associated with lower rate of hospitalization compared to NPH in individuals with T2DM. This clinical benefit is achieved with a modest increase in pharmacy expenditures for treating diabetes.