There were 8 studies [5-12] encompassing 1649 patients who underwent surgery with MMD were selected for analysis. The detailed study selection progress was shown in Figure 1. The main characteristics of included studies are summarized in Table 1. The included studies were consisted of 3 prospective studies [6, 11, 12] and 5 retrospective studies [5, 7-10].
Ischemic event
Five studies [5, 8, 9, 11, 12] about preoperative ischemic event were included in the pooled analysis (Fig. 2). Among them, two studies [8, 12] reported results of TIAs and infarction, respectively. Heterogeneity across studies was detected (I2 = 74.4%; p= 0.001), so random-effect model was applied. In this analysis, preoperative ischemic event significantly increase the risk of postoperative stroke events (OR=1.40; 95%CI=1.02–1.92; P=0.039). The results of subgroup analyses were stratified by race, MMD patients in Caucasian with preoperative ischemic event had a significant increase in risk for post-stroke (OR=3.29, 95%CI=1.04–10.41; P=0.043), whereas the pooled OR revealed a non-significant increase in Asian (OR=1.27; 95%CI=0.93–1.72; P=0.130).
PCA Involvement
Five studies [5-8, 12] about PCA involvement were pooled in the analysis (Fig. 3). Heterogeneity across studies was detected (I2 = 83.4%; p= 0.000), so random-effect model was applied. Patients with PCA involvement had a strong significant increase in risk for postoperative stroke events (OR=2.64; 95%CI=1.17–5.95; p=0.019). In subgroup analysis, PCA involvement correlate with an increased risk of post-infarction (OR=4.60; 95%CI=2.61–8.11; P=0.000), whereas there is no association between PCA involvement and risk of in post-stroke (OR=1.31; 95%CI=0.80–2.13; P=0.280).
Surgery types
Among 4 studies [5, 6, 8, 11], 454 patients underwent DB, 295 patients underwent IB, and 265 patients underwent CB. No heterogeneity across studies was detected (I2 = 0.00%; p= 0.601), so fixed-effect model was applied. Compared to DB, patients who underwent IB or CB could significantly increase the risk of postoperative stroke events. (OR=1.17; 95%CI=1.03–1.33; p=0.017). The study by Wonhyoung et al. [8] contributed importantly to the pooled OR (weight 96.08%), when this study was omitted in the model, the results was statistically insignificant (OR=1.31; 95%CI=0.69-2.49).
Suzuki stage
For studies [5, 6, 10, 12] contributed to pooled outcome. Moderate heterogeneity across studies was detected (I2=32.8 %; p=0.215), fixed-effect model was applied. Preoperative Suzuki stage was not associated with the risk of postoperative stroke events (OR=1.20; 95%CI=0.97-1.49; p=0.101).The study by Xiangyang et al. [12] contributed importantly to the pooled OR (weight 71.65%), when this study was omitted in the model, the results was statistically significant (OR=1.70; 95%CI=1.13-2.57).
Age at onset
Five studies [5, 6, 10-12] contributed to analysis. Moderate heterogeneity across studies was detected (I2 = 38.2%; p= 0.167), fixed-effect model was applied. In MMD patients, age at onset were marginally associated with an increased risk of postoperative stroke events (OR=1.02; 95 % CI=1.00–1.04; p=0.090).
Male sex
Four studies [5, 6, 10, 12] contributed to pooled outcome. No heterogeneity across studies was detected (I2 =0.00%; p=0.726), so fixed-effect model was applied. In MMD patients, male sex was not associated with an increased risk of postoperative stroke events (OR=1.16, 95%CI=0.75–1.82; p=0.504).The study by Xiangyang et al. [12] contributed importantly to the pooled OR (weight 52.74%), but when this study was omitted in the model, it remained statistically insignificant.
Medical history
Three [5-7] and two [5, 6] studies respectively reported the association between MMD patients with diabetes and hypertension and postoperative stroke events. No heterogeneity across studies was detected (I2 = 5.90%, p= 0.345; I2 = 0.00%, p= 0.494, respectively), so fixed-effect model was applied. MMD patients with diabetes were associated with an increased risk of postoperative stroke events (OR=4.02, 95% CI=1.59-10.16; p=0.003). However, MMD patients with hypertension were not associated with an increased risk of postoperative stroke events (OR=0.70, 95% CI=0.31-1.58; p=0.392).
Publication bias
The Begg’s rank correlation test indicated no evidence of publication bias among the included studies regarding the risk of male sex, age at onset, preoperative ischemic events, PCA involvement, medical history, Suzuki stage and surgical type (P>0.05) (Table2).
There were 8 studies [5-12] encompassing 1649 patients who underwent surgery with MMD were selected for analysis. The detailed study selection progress was shown in Figure 1. The main characteristics of included studies are summarized in Table 1. The included studies were consisted of 3 prospective studies [6, 11, 12] and 5 retrospective studies [5, 7-10].
Ischemic event
Five studies [5, 8, 9, 11, 12] about preoperative ischemic event were included in the pooled analysis (Fig. 2). Among them, two studies [8, 12] reported results of TIAs and infarction, respectively. Heterogeneity across studies was detected (I2 = 74.4%; p= 0.001), so random-effect model was applied. In this analysis, preoperative ischemic event significantly increase the risk of postoperative stroke events (OR=1.40; 95%CI=1.02–1.92; P=0.039). The results of subgroup analyses were stratified by race, MMD patients in Caucasian with preoperative ischemic event had a significant increase in risk for post-stroke (OR=3.29, 95%CI=1.04–10.41; P=0.043), whereas the pooled OR revealed a non-significant increase in Asian (OR=1.27; 95%CI=0.93–1.72; P=0.130).
PCA Involvement
Five studies [5-8, 12] about PCA involvement were pooled in the analysis (Fig. 3). Heterogeneity across studies was detected (I2 = 83.4%; p= 0.000), so random-effect model was applied. Patients with PCA involvement had a strong significant increase in risk for postoperative stroke events (OR=2.64; 95%CI=1.17–5.95; p=0.019). In subgroup analysis, PCA involvement correlate with an increased risk of post-infarction (OR=4.60; 95%CI=2.61–8.11; P=0.000), whereas there is no association between PCA involvement and risk of in post-stroke (OR=1.31; 95%CI=0.80–2.13; P=0.280).
Surgery types
Among 4 studies [5, 6, 8, 11], 454 patients underwent DB, 295 patients underwent IB, and 265 patients underwent CB. No heterogeneity across studies was detected (I2 = 0.00%; p= 0.601), so fixed-effect model was applied. Compared to DB, patients who underwent IB or CB could significantly increase the risk of postoperative stroke events. (OR=1.17; 95%CI=1.03–1.33; p=0.017). The study by Wonhyoung et al. [8] contributed importantly to the pooled OR (weight 96.08%), when this study was omitted in the model, the results was statistically insignificant (OR=1.31; 95%CI=0.69-2.49).
Suzuki stage
For studies [5, 6, 10, 12] contributed to pooled outcome. Moderate heterogeneity across studies was detected (I2=32.8 %; p=0.215), fixed-effect model was applied. Preoperative Suzuki stage was not associated with the risk of postoperative stroke events (OR=1.20; 95%CI=0.97-1.49; p=0.101).The study by Xiangyang et al. [12] contributed importantly to the pooled OR (weight 71.65%), when this study was omitted in the model, the results was statistically significant (OR=1.70; 95%CI=1.13-2.57).
Age at onset
Five studies [5, 6, 10-12] contributed to analysis. Moderate heterogeneity across studies was detected (I2 = 38.2%; p= 0.167), fixed-effect model was applied. In MMD patients, age at onset were marginally associated with an increased risk of postoperative stroke events (OR=1.02; 95 % CI=1.00–1.04; p=0.090).
Male sex
Four studies [5, 6, 10, 12] contributed to pooled outcome. No heterogeneity across studies was detected (I2 =0.00%; p=0.726), so fixed-effect model was applied. In MMD patients, male sex was not associated with an increased risk of postoperative stroke events (OR=1.16, 95%CI=0.75–1.82; p=0.504).The study by Xiangyang et al. [12] contributed importantly to the pooled OR (weight 52.74%), but when this study was omitted in the model, it remained statistically insignificant.
Medical history
Three [5-7] and two [5, 6] studies respectively reported the association between MMD patients with diabetes and hypertension and postoperative stroke events. No heterogeneity across studies was detected (I2 = 5.90%, p= 0.345; I2 = 0.00%, p= 0.494, respectively), so fixed-effect model was applied. MMD patients with diabetes were associated with an increased risk of postoperative stroke events (OR=4.02, 95% CI=1.59-10.16; p=0.003). However, MMD patients with hypertension were not associated with an increased risk of postoperative stroke events (OR=0.70, 95% CI=0.31-1.58; p=0.392).
Publication bias
The Begg’s rank correlation test indicated no evidence of publication bias among the included studies regarding the risk of male sex, age at onset, preoperative ischemic events, PCA involvement, medical history, Suzuki stage and surgical type (P>0.05) (Table2).