[Etiologic classification of cerebral infarct. Experience from a prospective data register].

Most classifications of stroke include clinically heterogeneous subgroups and therefore are of limited value for comparative studies or clinical protocols. The view is held that a classification according to stroke etiology is clinically more reasonable and more consistent for therapeutic strategies. In order to determine the frequency of various etiological subgroups in a series of stroke cases, the results of the Klosterneuburger Schlaganfall-Datenbank (KSDB) are reported. This stroke registry has prospectively recorded over 300 items on all stroke cases referred to one center since March 1988. Investigation rates include CT in almost 100% and the investigation of cerebral vessels in over 90% of all cases. By applying defined etiological categories (undetermined etiology, atherosclerosis of the large craniocervical vessels, cardiogenic embolism, lacunar, primary hemorrhage, and multiple and other causes) to the first 420 patients registered within the first two years it can be shown that even with CT and neurosonology in routine use, in 29% of all cases the cause of the stroke cannot be determined. To investigate this largest subgroup by means of additional new methods as well as by investigating the long-term natural course represents an important challenge for clinical stroke research.