“…*Bivariate random-effects regression model. 7 8 8 (100.0, 63.1-100.0) 100 0 (0.00, 0.00-3.62) Liang (2013) 11 5 5 (100.0, 47.8-100.0) 401 1 (0.25, 0.01-1.38) Nicolaides (2013) 12 2 2 (100.0, 15.8-100.0) 227 0 (0.00, 0.00-1.61) Song (2013) 13 3 2 (66.7, 9.4-99.2) 1737 † 0 (0.00, 0.00-0.21) Comas (2014) 17 0 -315 1 (0.32, 0.01-1.76) Porreco (2014) 20 9 9 (100.0, 66.4-100.0) 3269 11 (0.34, 0.17-0.60) Shaw (2014) 21 3 3 (100.0, 29.2-100.0) 192 0 (0.00, 0.00-1.90) Song (2015) 27 0 -203 1 (0.49, 0.01-2.71) Persico (2016) 35 3 2 (66. 7, 9.4- Jiang (2012) 6 3 3 (100.0, 29.2-100.0) 900 0 (0.00, 0.00-0.41) Lau (2012) 7 1 1 (100.0, 2.5-100.0) 107 0 (0.00, 0.00-3.36) Liang (2013) 11 3 3 (100.0, 29.2-100.0) 403 1 (0.25, 0.01-1.37) Porreco (2014) 20 6 6 (100.0, 54.1-100.0) 3263 5 (0.15, 0.05-0.36) Shaw (2014) 21 1 1 (100.0, 2.5-100.0) 194 0 (0.00, 0.00-1.88) Song (2015) 27 1 0 (0.0, 0.0-97.5) 202 0 (0.00, 0.00-1.81) Persico (2016) 35 1 1 (100.0, 2.5-100.0) 248 0 (0.00, 0.00-1.48) Zhang (2016) 39 1 1 (100.0, 2.5-100.0) 86 0 (0.00, 0.00-4.20) Pooled analysis (% (95% CI))* 100 (83.6-100) 0.004 (0-0.08) I 2 (%) 0 0…”