Predicting functional outcome and symptomatic intracranial hemorrhage in patients with acute ischemic stroke: a glimpse into the crystal ball?

Abstract

When encountering a patient with ischemic stroke, a stroke physician needs to make an outcome prediction for several reasons. First, a prediction may help patients and families have more realistic expectations about the patient’s future and plan their long-term living arrangements.1 Second, outcome prediction may be useful in making treatment decisions: for example, avoidance or de-escalation of aggressive treatment measures in patients who would achieve an excellent outcome regardless of such therapy or in those who would experience a poor and unacceptable outcome despite that treatment. Third, outcome prediction could aid in stroke research to control for case-mix variation in nonrandomized studies and to refine selection criteria in randomized, controlled acute stroke trials. An ideal stroke prognostic score system should include a limited number of readily available and relevant parameters, be easily and quickly applicable in the clinical setting (eg, without the need of complex mathematical formulas or nonconventional time-consuming additional investigations), be accurate with low intra- and inter-rater disagreement, be validated externally in large independent and preferably multiple populations, and be proven reliably useful in guiding treatment decisions. Extensive research has been performed during the recent years on prognostic scores for the prediction of stroke outcome and risk of symptomatic intracranial hemorrhage (sICH).2–4 This review summarizes current prognostic scores, discusses the scientific background behind the scores’ main components, acknowledges their strengths and limitations, and highlights the areas which warrant further research. Because of the space limitations, scores which were initially developed for one outcome (eg, functional outcome) and were later validated also for other outcomes (eg, post-thrombolytic sICH) will be discussed only for the initially validated outcome. Furthermore, scores for strictly selected populations and scores for patients treated with endovascular procedures are not included in this review. Several prognostic scores (Table 1; Figure 1 …