Cerebrovascular disease is the most common cause of death in China, and the incidence of ischemic stroke (240 per 100,000 people) is higher than that of hemorrhagic stroke (82 per 100,000 people). More than 80 percent of strokes can be prevented by early control of risk factors. Therefore, identifying and managing high-risk groups is a top priority in preventing stroke. The CHA2DS2-VASc score is a key prediction tool for stratifying stroke risk in individuals with atrial fibrillation (AF) as follows: zero score is low risk; one is intermediate risk; and two is high risk. The present study was undertaken to evaluate the accuracy of the CHA2DS2-VASc scoring system for stratifying ischemic stroke risk in the non-AF population.

We searched PubMed, EMBASE, and the Cochrane Library in June 2018 for relevant diagnostic studies. Study selection, data extraction, and quality assessment (using the QUADAS-2 criteria) were performed independently by two authors. Methodological variation across the selected studies precluded meta-analysis, so the results were synthesized narratively.

Seven prospective studies involving 50,652 patients (6,760 with ischemic stroke) were included. The treatment threshold ranged from two to four across the studies. Three studies reported diagnostic accuracy at a threshold of two, with a sensitivity above 0.8 and a specificity ranging from 0.32 to 0.68. The diagnostic odds ratio was greater than two (seven studies). The two studies using a treatment threshold of four reported a sensitivity of 0.59 to 0.76 and a specificity of 0.43 to 0.69. One study used a threshold of three, with a sensitivity of 0.79 and a specificity of 0.39.

The CHA2DS2-VASc score may be used to predict ischemic stroke in the non-atrial fibrillation population. Treatment thresholds greater than two provide more optimal diagnostic accuracy, although the predictive performance of the CHA2DS2-VASc score may be better in patients with chronic obstructive pulmonary disease but not AF.