José Vicente Arcos-Machancoses1, Cristina Molera Busoms2, Ecaterina Julio Tatis3, María Victoria Bovo4, Jesús Quintero Bernabeu5, Javier Juampérez Goñi6, Vanessa Crujeiras Martínez7, Javier Martín de Carpi8. 1. Sant Joan de Déu Hospital (HSJD), Department of Pediatric Gastroenterology, Hepatology and Nutrition, Barcelona, Spain. Electronic address: jvicentearcos@gmail.com. 2. Sant Joan de Déu Hospital (HSJD), Department of Pediatric Gastroenterology, Hepatology and Nutrition, Barcelona, Spain; HSJD-HVH Comprehensive Unit of Complex Hepatology and Pediatric Liver Transplantation, Barcelona, Spain. Electronic address: cristinamolera@gmail.com. 3. Sant Joan de Déu Hospital (HSJD), Department of Pediatric Gastroenterology, Hepatology and Nutrition, Barcelona, Spain. Electronic address: ecaterina02@hotmail.com. 4. Sant Joan de Déu Hospital (HSJD), Department of Pediatric Gastroenterology, Hepatology and Nutrition, Barcelona, Spain. Electronic address: mavibovo@gmail.com. 5. Vall d'Hebron Hospital (HVH), Department of Pediatric Gastroenterology, Hepatology and Nutrition, Barcelona, Spain; HSJD-HVH Comprehensive Unit of Complex Hepatology and Pediatric Liver Transplantation, Barcelona, Spain. Electronic address: jquintero@vhebron.net. 6. Vall d'Hebron Hospital (HVH), Department of Pediatric Gastroenterology, Hepatology and Nutrition, Barcelona, Spain; HSJD-HVH Comprehensive Unit of Complex Hepatology and Pediatric Liver Transplantation, Barcelona, Spain. Electronic address: jjuamperez@vhebron.net. 7. University Hospital Complex of Santiago de Compostela, Department of Pediatric Gastroenterology, Hepatology and Nutrition, Spain. Electronic address: vanecrujeiras@hotmail.com. 8. Sant Joan de Déu Hospital (HSJD), Department of Pediatric Gastroenterology, Hepatology and Nutrition, Barcelona, Spain; HSJD-HVH Comprehensive Unit of Complex Hepatology and Pediatric Liver Transplantation, Barcelona, Spain. Electronic address: javiermartin@sjdhospitalbarcelona.org.
Abstract
BACKGROUND: Children with autoimmune hepatitis (AIH) often exhibit particular features. Accordingly, seven pediatric-specific criteria have been proposed. AIM: To develop a prediction model based on them, transform it into a scoring system and study its accuracy. METHODS: A cohort of children under study for liver disease was consecutively selected. AIH diagnosis was based on classical criteria. Already proposed pediatric criteria were recorded. The best possible regression model was selected, and the beta coefficient of each criterion was translated into a whole number (points). Total scores were obtained following the points system and the best cut-off was calculated. Subsequently, accuracy of the diagnostic score was studied in the validation set. RESULTS: Among 212 included patients, 100 had AIH. The score included 5 criteria: autoantibodies (0-2 points), hypergammaglobulinemia, exclusion of viral hepatitis, exclusion of Wilson's disease (1 point each) and liver histology (3 points). In addition, a normal cholangiogram is mandatory. The validation set was formed of 70 patients (24 with AIH). In this subsample, a score of ≥6 renders a sensitivity/specificity of 95.8%/100%. The area under the receiver operating characteristic curve was 97.1%. CONCLUSION: Pediatric-specific criteria for the diagnosis of AIH can be reliably used as a scoring system.
BACKGROUND:Children with autoimmune hepatitis (AIH) often exhibit particular features. Accordingly, seven pediatric-specific criteria have been proposed. AIM: To develop a prediction model based on them, transform it into a scoring system and study its accuracy. METHODS: A cohort of children under study for liver disease was consecutively selected. AIH diagnosis was based on classical criteria. Already proposed pediatric criteria were recorded. The best possible regression model was selected, and the beta coefficient of each criterion was translated into a whole number (points). Total scores were obtained following the points system and the best cut-off was calculated. Subsequently, accuracy of the diagnostic score was studied in the validation set. RESULTS: Among 212 included patients, 100 had AIH. The score included 5 criteria: autoantibodies (0-2 points), hypergammaglobulinemia, exclusion of viral hepatitis, exclusion of Wilson's disease (1 point each) and liver histology (3 points). In addition, a normal cholangiogram is mandatory. The validation set was formed of 70 patients (24 with AIH). In this subsample, a score of ≥6 renders a sensitivity/specificity of 95.8%/100%. The area under the receiver operating characteristic curve was 97.1%. CONCLUSION: Pediatric-specific criteria for the diagnosis of AIH can be reliably used as a scoring system.