Fig. 4: Forest plot for effect of maintenance treatment (MT) adherence and development of insidious onset movement disorder (MD) in patients identified by newborn screening (NBS) including and excluding noninformative studies.

In contrast to noninformative studies, informative studies compare two different interventions groups within one study. Patients not following MT recommendations show a trend for increased relative risk (log risk ratio [RR]) for development of an insidious MD compared with patients with MT adherence (p = 0.058; log RR 0.61). This effect becomes clearly significant if noninformative studies are excluded (p < 0.0001) increasing the RR to almost exp(2) = 7. CI confidence interval, RE random effect. Including noninformative studies: test of heterogeneity: QM(df14) = 29.1436; model results: p = 0.0585. I2 (residual heterogeneity/unaccounted variability): 51.28%. Excluding noninformative studies: test for heterogeneity: Q(df = 1) = 0.1765; model results: p < 0.0001. I2 (residual heterogeneity/unaccounted variability): 0%.