ObjectivesTo compare clinical outcomes and health care costs across three cohorts of uncontrolled diabetic patients who initiated treatment with one of the following: sulphonylureas (SU), thiazolidinediones (TZD) or sitagliptin (SITA).Materials and methodsWe performed a retrospective study based on a linkage between administrative and laboratory databases maintained by three Italian local health units. The index period ranged from July 2008–June 2010. Patients were treatment-naïve to either SU, TZD, or SITA, but they were already treated with other oral hypoglycemic agents. Demographics and clinical characteristics were assessed at baseline. Adherence was measured by the medication possession ratio and adherent was defined as a patient with a medication possession ratio of 80% or greater. We used a Poisson regression model to estimate the risk ratios for disease-related hospitalizations that occurred during the 18-month follow-up period. The total annual costs included all the pharmacological treatments and the direct costs due to hospitalizations and outpatient services.ResultsWe identified 928 patients treated with SU, 330 patients treated with TZD, and 83 patients treated with SITA. SITA patients were significantly younger and with fewer previous hospital discharges. The baseline mean glycated hemoglobin level was 8.1% for SU, 8.0% for TZD, and 8.3% for SITA patients. SITA-naïve patients were more adherent than the SU- and TZD-naïve patients (79.5% versus 53.2% and 62.8%, respectively; P<0.001). The SU and TZD group showed a significant increased risk of disease-related hospitalizations compared with the SITA group (the unadjusted rate was 10.42 and 7.16 per 100 person-years versus 1.64 per 100 person-years, P=0.003; compared with SU, the adjusted incidence rate ratio for SITA was 0.21, P=0.030). The total annual costs per patient were €972 for SITA, €706 for SU, and €908 for those treated with TZD.ConclusionUncontrolled diabetic patients who initiated – as a second-line therapy in addition to metformin – treatment with SITA, compared to those who initiated treatment with SU or TZD, showed a reduced risk of disease-related hospitalizations. The total annual costs per patient were not significantly different among the three groups.