[Standardizing the diagnosis of acute ischemic stroke--the concept of an institutional manual and its clinical impact].

Unlabelled: In the management of acute ischemic stroke, a diagnostic procedure is critically important. However, till date, no guideline or consensus regarding a standard diagnosis procedure has been established. We hereby report an institutional manual for the consistent diagnosis of acute ischemic stroke.

Methods: The institutional manual was based on the National Institute of Neurological Disorders and Stroke III classification; however, the criteria for each stroke subtype were clearly defined using a rating system that included the representative observations specific to each subtype. The present manual was prepared in order to clearly define the characteristics of the stroke subtypes, which are often ambiguous in the clinical settings, and to design a diagnostic procedure within the institute with more emphasis on standardization rather than achieving complete accuracy. Several characteristic points were considered while doing this: lacunar infarctions should be clearly differentiated from other small infarctions, subtypes of "suspicious cardioembolism," "atherothrombotic infarction (embolic mechanism)," and "infarction due to arterial dissection" should be determined separately in order to identify the causal mechanism to the extent possible. For a final diagnosis, the patients were examined 3 time points during: at admission, during the subacute stage (day 4-7), and at discharge or more than 2 weeks after the onset of stroke. The current version of the manual has been used since 2006 after a transitional phase from 2004 to 2005. Retrospective comparisons on stroke diagnoses and clinical outcomes were performed between the periods during which the present manual was used and those during which it was not.

Results: The present manual was retrospectively applied for diagnosing 311 consecutive ischemic stroke patients who were admitted to our institute within 7 days after the onset in 2003; in the case of 85 patients (29%), the diagnosis was different from that made in 2003. Of all the lacunar infections reported in 2003, 17 were diagnosed as other conditions when the manual was used. Similarly, 45% of the cases of atherothrombotic infarction (thrombotic mechanism) and 24% of those of cardiac embolism in 2003 were diagnosed differently when the present manual was applied. The mean modified Rankin scale at discharge was 2.63 +/- 0.07 (mean +/- standard error of the mean) in patients in 2002 and 2003 (n=491), which was significantly different from 2.32 +/- 0.06 in patients in 2006 through March of 2008 (n=903; p=0.01).

Conclusion: The present manual appears to be helpful to improve the uniformity of the diagnoses and the clinical results. It may also assist residents and their mentors in the educational field. The manual will require periodical evaluation and version upgrade in order to maintain its efficacy.