OBJECTIVES: One important aspect for patients suffering from diabetes is an appropriate metabolic control to avoid micro- and macrovascular complications. There is a lack of information about adjustment of antidiabetic treatment  during hospitalization. METHODS: To identify a change in medication in patients suffering from diabetes, treatment regimens before and after hospitalization were investigated retrospectively. Information was taken from 2006 to 2010 German claims data. Data of about 25,000 hospitalizations from 180,000 diabetic patients were analyzed over a five year period. RESULTS: Study population of statutorily insured diabetic patients is representative as it is comparable to the overall German population by age, gender, mortality and morbidity. Treatment of most diabetic patients on oral anti-diabetic drugs remained unchanged before and after a hospital stay (55.8%). A smaller part of diabetic patients (37.6%) is admitted and discharged with the same insulin regimen. In 15% of the cases where patients are discharged with insulin, insulin treatment was initiated or changed during hospitalization. 53% of these patients received a first prescription of insulin. Basal supported oral therapy (BOT) and conventional insulin treatment (CT) regimens made up the largest fraction of all insulin initiations (62%). If a BOT was initiated, insulin glargine was favored (60%) compared to all other long acting insulin formulations such as human insulin (30%) and insulin detemir (10%). In 21% of the cases initiation results in a basal-bolus regimen (ICT) and in 17% in a prandial insulin therapy (SIT). CONCLUSIONS: Insulin treatment was first prescribed or switched during hospitalization in 15% of the cases where patients are discharged with insulin. Initiation of BOT and CT regimens turned out to be the overall preferred therapies in the underlying analysis. Given the fact that 26.7% of the German diabetic population has suboptimal HbA1c-values more of them should receive an adequate adjustment of antidiabetic therapy during hospitalization.