A total of 870 records were retrieved from the databases, and only 116 records were screened by title and abstract after excluding duplicate records. We screened 303 full-text studies for potential eligibility in this study. Finally, 26 studies (2329 eyes) were involved in the meta-analysis (Fig. 1). The characteristics of the included studies are summarized in Table 1 [8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34].

Flowchart of literature search and screening

IOP at 1 month post-surgery was highly heterogeneous among 14 studies (CHi2 = 112.71, df = 13, I2 = 88%, P < 0.01) (Fig. 2) [10,11,12,13,14,15, 18,19,20, 25, 27,28,29,30]. However, the Egger’s test showed no risk of publication bias (P >| t |= 0.246, Fig. 3). The findings of Fea et al. [12] and Qureshi et al. [29] were significantly different from these 14 studies. However, considering the small number of subjects in this study (only 12 subjects), the risk of bias could be noteworthy. Baseline IOP was significantly higher in Qureshi et al.’s study [29] than that in other studies, which may explain significant difference in the results. After excluding the two studies, it was found that the change in postoperative IOP varied among studies with different baseline IOP. Therefore, subgroup analysis was conducted based on the baseline IOP (Fig. 4). The results indicated that there was significant heterogeneity among different baseline IOP groups (CHi2 = 68.13, df = 3, I2 = 95.6%, P < 0.01), rather than among studies within each group (group 1, IOP was 19.2–21.2 mmHg): MD = 7.60, 95% confidence interval (CI) [6.55, 8.66], I2 = 0%, P < 0.01); group 2, IOP was 21.6–22.8 mmHg: MD = 5.83, 95% CI [4.94, 6.71], I2 = 0%; group 3, IOP was 23.5–24.3 mmHg: MD = 9.22, 95% CI [8.33, 10.10], I2 = 0%; group 4, IOP was 24.4–26.2 mmHg: MD = 11.64, 95% CI [10.49, 12.80], I2 = 0%) [10, 11, 13,14,15, 18,19,20, 25, 27, 30, 31].

A forest map illustrating changes in IOP at 1month after XEN implantation

A funnel plot displaying changes in IOP at 1 month after XEN implantation

A forest map illustrating the results of subgroup analysis of IOP changes at 1 month after implantation

IOP at 3-month post-surgery was also significantly heterogeneous among 14 studies (CHi2 = 87.44, df = 13, I2 = 85%, P < 0.01) [10, 11, 14, 15, 18,19,20, 25, 29,30,31,32]. The Egger’s test revealed no risk of publication bias (P >| t |= 0.151, Fig. 5). The characteristics of each study were fully examined, and the heterogeneity could not be explained by clinical and methodological heterogeneity. Hence, a random-effects model was utilized, and it was found that IOP was significantly lower at 3 months after XEN implantation (MD = 8.31, 95% CI [2.54, 8.46], Z = 13.92, P < 0.00001, Fig. 6) [10,11,12, 14, 15, 18,19,20, 25, 29,30,31,32].

A funnel plot displaying changes in IOP at 3 months after XEN implantation

A forest map illustrating the results of IOP changes at 3 months after implantation

There was high heterogeneity in IOP at 6 months post-surgery among 15 studies (CHi2 = 69.19, df = 14, I2 = 80%, P < 0.01) [10,11,12, 14, 15, 18,19,20, 25, 27, 28, 31, 32]. The Egger’s test showed no risk of publication bias (P >| t |= 0.843, Fig. 7). Subgroup analysis indicated significant heterogeneity among different baseline IOP groups (CHi2 = 68.13, df = 3, I2 = 95.6%, P < 0.01), and no heterogeneity was noted in group 1 (IOP was 17.8–21.8 mmHg) (MD = 5.71, 95% CI [5.05, 6.36], I2 = 0%, P < 0.01) and group 2 (IOP was 22.1–23.9 mmHg) (MD = 7.92, 95% CI [7.15,8.70], I2 = 0%). Low heterogeneity was identified in group 3 (IOP was 24.3–26.2 mmHg) (MD = 9.32, 95% CI [8.66,9.97], I2 = 30%), (Fig. 8) [10,11,12, 14, 15, 18,19,20, 25, 27, 28, 31, 32].

A funnel plot displaying changes in IOP at 6 months after XEN implantation

A forest map illustrating the results of IOP changes at 6 months after implantation

There was significant heterogeneity in IOP at 12-month post-surgery among 25 studies (CHi2 = 279.63, df = 24, I2 = 91%, P < 0.01) [10, 13,14,15,16,17,18,19,20,21,22,23,24,25,26,