Number needed to treat for stroke thrombectomy based on a systematic review and meta-analysis.

The positive results of recent clinical trials examining endovascular treatment of acute stroke were the culmination of nearly two decades of studies of endovascular stroke treatment. We systematically reviewed this body of work, evaluated the strength of evidence, and performed a meta-analysis to define the clinical impact of these investigations. Terms were entered into search engines in a systematic fashion. Articles were reviewed independently by study authors, graded for level of evidence, and combined in a meta-analysis. The overall body of evidence was evaluated using GRADE criteria. Our search yielded 948 articles. Twenty-five met predefined inclusion criteria. We identified 12 grade I, 1 grade II, 5 grade III, and 7 grade IV studies (κ=0.86). Meta-analysis for independence at 90days showed a benefit of endovascular treatment (grade I studies OR 1.58 [1.20-2.07]). When limiting the analysis to studies using stent retriever, the OR increased to 2.44 (1.77-3.36). The number needed to treat (NNT) was 8. Endovascular treatment was not associated with increased symptomatic intracranial hemorrhage, and forgoing endovascular treatment was associated with death at 90 days. The quality of evidence according to GRADE criteria was "moderate." In summary, we found impressive evidence for a benefit of endovascular treatment of acute stroke, particularly when using stent retriever devices. Our meta-analysis is unique in that it includes all studies related to this topic and defines the clinical impact of the data, providing NNT. We show that thrombectomy is among the most effective stroke treatments currently available.