Diagnostic accuracy of fetal MRI to detect cleft palate: a meta-analysis

This systematic review aims to determine the diagnostic accuracy of fetal MRI for detecting cleft palate in fetuses at risk for orofacial clefts. Pubmed, Embase, and CINAHL were searched systematically. A diagnostic study was included if it performed MRI (index test) and postnatal examination (reference test) in fetuses at risk for orofacial clefts. Methodological quality was assessed using the QUADAS-2. A meta-analysis was performed with a random-effects model, calculating the pooled sensitivity, specificity, and area under the curve. The search resulted in eight studies (334 fetuses) to be included: four prospective and four retrospective studies. The applicability concern was low. There was, however, a risk of selection and information bias. All studies showed that MRI well predicted the chance of cleft palate. The sensitivity results were homogeneous, but heterogeneity was assumed regarding the specificity estimate (Cochrane’s Q test: p = 0.00). The pooled sensitivity was 0.97 (95% CI 0.93–0.99); the pooled specificity was 0.94 (0.89–0.97). The area under the curve was 0.98 (95% CI 0.98–0.99). Conclusion: This meta-analysis shows that MRI has an excellent sensitivity and good to excellent specificity for diagnosing cleft palate in fetuses at risk for orofacial clefts. Future research should assess applicability for clinical care. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (10.1007/s00431-019-03500-x) contains supplementary material, which is available to authorized users.