The term “stroke unit” first appeared in 1969 from medical literature [1], but the concept of dedicated in-hospital ward for stroke traces back to 1962 in New York, where the investigator evaluated rehabilitation ward for hemiplegic patient. At that time, most studies compared the dedicated rehabilitation ward for stroke to the other general ward. The first study to investigate modern concept of stroke unit [2], which is an organized inpatient care for acute stroke, would be the Edinburgh study in 1980 [3], which is a randomized controlled study with 311 acute stroke patients to compare stroke unit versus general medical ward. After then, a series of controlled trials followed to investigate the efficacy of a comprehensive stroke unit, mostly performed in the 1990s. Among them, the Trondheim study is most representative, which is a randomized controlled trial comparing stroke unit versus general medical ward in 220 acute stroke patients and showed big success in reducing death and dependency with an odds ratio of 0.36 (95% confidence interval (CI) = 0.21 ~ 0.61) [4]. As the evidences of stroke unit accumulate, stroke unit became one of the essential components of stroke care organization in the twenty-first century [5], and now most clinical practice guidelines in the world state that the acute stroke patient should be treated in the stroke unit with a high grade of recommendation.

According to the latest systematic Cochrane review (Stroke Unit Trialists’ Collaboration, 2013), a comprehensive stroke unit is estimated to reduce death and dependency by an odds ratio of 0.82 (95% CI 0.68 ~ 0.98) [6]. When we compare this result to other proven treatments of acute stroke [7], through indirect comparison by death and dependency, the efficacy of a stroke unit exceeds that of acute aspirin use and is almost comparable to that of intravenous tPA (Table 2.1). This rather unexpectedly big effect made stroke unit a mainstream of acute stroke care as important as thrombolytic treatment. Another study showed that this effect does not fade away as time goes by but lasts more than 10 years of follow-up [8, 9].

There can be many reasons for the abovementioned efficacy, but the biggest should be attributed to close monitoring and early intervention of patient status, such as vital signs, neurological status, and electrocardiogram, which leads to better management of blood pressure, prompt detection of early neurological deterioration or paroxysmal atrial fibrillation, and effective prevention of possible complications [10]. Optimized clinical pathway by written care protocol according to guideline and early rehabilitation can also accelerate in-hospital process and thus contribute to better outcomes. But all these effects are abolished in the general ward or other types of care, even though applying the same clinical pathway or protocol, probably because the caregiver is not dedicated to stroke care [11].

To be a functionally effective stroke unit, many components are needed, either as human resources, facilities, or operating protocols [12, 13]. These components can be different from hospital to hospital, and many regional variations also exist. In some countries there are criteria to be certified by academic or government authority. However, there are essential components which are almost uniform in most stroke units (Table 2.2).