A causality algorithm to guide diagnosis and treatment of catatonia due to autoimmune conditions in children and adolescents

Objectives: Pediatric catatonia is a rare and life-threatening syndrome. Around 20% of juvenile catatonia is associated with organic condition (Consoli et al., 2012). Autoimmune conditions represent a diagnostic and therapeutic challenge since specific antibodies can be missed. To facilitate decision making, we recently formulated a causality assessment score (CAUS) using a stepwise approach and an immunosuppressive therapeutic challenge (Ferrafiat et al., 2016). Our objectives were to validate retrospectively CAUS and to define its threshold for an accurate distinction between organic catatonia and non-organic catatonia, and specifically between autoimmune catatonia and non-organic catatonia. Method: To obtain a sufficient number of cases with organic catatonia, we pooled two samples (N = 104) – one from a child psychiatry center, the other from neuro-pediatrics center – expert in catatonia and autoimmune conditions. Organic conditions were diagnosed using a multidisciplinary approach and numerous paraclinical investigations. Given the binary classification needs, we used receiver operating characteristic (ROC) analysis (Peacock and Peacock, 2010) to calculate the best classification threshold. Results: The cohort included 67 cases of non-organic catatonia and 37 cases of organic catatonia. ROC analysis showed that the CAUS performance in discriminating both organic catatonia vs. non-organic catatonia, and autoimmune catatonia vs. non-organic catatonia was excellent (Area Under the Curve = 0.99). In both analyses, for a CAUS threshold ≥ 5, accuracy equaled to 0.96. Conclusion: Regarding juvenile catatonia, the use of the CAUS score algorithm combining a therapeutic challenge and a threshold ≥ 5 may help to diagnose and treat autoimmune conditions even without formal identification of auto-antibodies.