Characteristics of the Included Studies
Among the twelve included RCTs, eleven were published in English9-12,16,18-23 and one in Chinese. 17The mean ages of patients were similar in both groups (69.5 years vs. 69.6 years). The follow-up periods ranged from 90 to 180 days. Eight RCTs 10-12,16,18-20,22reported the NIHSS scores, ranging from five to 17. Since the data could not be obtained from the full text of three studies, we obtained these data from the published meta-analysis24 for mortality rate, the proportion of patients with mRS ≤ 2, and NIHSS score. Characteristics of the included studies were shown in Table 1.
Three studies were rated as low risk of bias, 10,11,22 two studies were at high risk of bias, 19,23 and the others were at unclear risk of bias. 9,12,16-18,20,21The randomization method of one study was unclear; 21randomization and allocation concealment were not performed for one study, 19 and the allocation concealment method was unclear for three studies; 9,20,21 participant blinding was unclear in one study; 17 outcome assessment was not blinded in three studies for partial outcomes;16,17,21 three studies did not mention the reasons for the loss to follow-up; 9,12,23 three studies had selective reporting. 18,21,23 The risk of bias assessments of the included studies were shown in Figure 2.
Meta-Analysis