Predicción temprana del resultado neurológico a los 12 meses en neonatos de riesgo en Bariloche / Early prediction of the neurological result at 12 months in newborns at neurological risk

INTRODUCCIÓN: El objetivo de este estudio fue valorar el examen neurológico de Amiel-Tison (AT) y la ecografía cerebral a término para predecir el resultado neurológico a los 12 meses en recién nacidos con riesgo neurológico. Población y métodos: Ingresaron en el estudio 89 niños con riesgo neurológico internados en Cuidados Intensivos Neonatales del Hospital Zonal Bariloche. La evaluación comprendió AT y ecografía cerebral neonatal y a los 12 meses, AT y valoración del desarrollo. Se calculó sensibilidad, especificidad, valor predictivo positivo y negativo de las evaluaciones neonatales en relación al neurodesarrollo a los 12 meses. Se usó el modelo de regresión logística para evaluar la asociación entre factores perinatales y neurodesarrollo. RESULTADOS: Completaron el estudio 70 niños, de los cuales un 14% presentó examen neurológico anormal y un 17% desarrollo anormal a los 12 meses. AT y ecografía cerebral neonatal tuvieron baja sensibilidad para estimar neurodesarrollo a los 12 meses. Entre neonatos con AT normal un 93% presentó resultado neurológico normal y un 86% desarrollo normal a los 12 meses; en aquellos con ecografía normal los porcentajes fueron 90 y 81%, respectivamente. Los neonatos con tres o más factores de riesgo presentaron 5,4 veces más posibilidad de AT anormal a los 12 meses que aquellos con menos factores y 3,5 veces más posibilidad de prueba de desarrollo anormal. CONCLUSIONES: El examen neurológico y la ecografía cerebral neonatales tienen baja sensibilidad pero alto valor predictivo negativo para neurodesarrollo a los 12 meses. Tres o más factores de riesgo se asocian a neurodesarrollo anormal

INTRODUCTION: The aim of this study was to evaluate the Amiel-Tison neurological examination (AT) and cranial ultrasound at term for predicting the neurological result at 12 months in newborns with neurological risk. PATIENTS AND METHODS: The study included 89 newborns with high risk of neurological damage, who were discharged from the Neonatal Intensive Care of the Hospital Zonal Bariloche, Argentina. The assessment consisted of a neurological examination and cranial ultrasound at term, and neurological examination and evaluation of development at 12 months. The sensitivity, specificity, positive and negative predictor value was calculated. The relationship between perinatal factors and neurodevelopment at 12 month of age was also calculated using logistic regression models. RESULTS: Seventy children completed the follow-up. At 12 months of age, 14% had an abnormal neurological examination, and 17% abnormal development. The neurological examination and the cranial ultrasound at term had low sensitivity to predict abnormal neurodevelopment. At 12 months, 93% of newborns with normal AT showed normal neurological results, and 86% normal development. Among newborns with normal cranial ultrasound the percentages were 90 and 81%, respectively. Among children with three or more perinatal risk factors, the frequency of abnormalities in the neurological response was 5.4 times higher than among those with fewer risk factors, and abnormal development was 3.5 times more frequent. CONCLUSIONS: The neurological examination and cranial ultrasound at term had low sensitivity but high negative predictive value for the neurodevelopment at 12 months. Three or more perinatal risk factors were associated with neurodevelopment abnormalities at 12 months of age

[Early prediction of the neurological result at 12 months in newborns at neurological risk]. / Predicción temprana del resultado neurológico a los 12 meses en neonatos de riesgo en Bariloche.

INTRODUCTION: The aim of this study was to evaluate the Amiel-Tison neurological examination (AT) and cranial ultrasound at term for predicting the neurological result at 12 months in newborns with neurological risk. PATIENTS AND METHODS: The study included 89 newborns with high risk of neurological damage, who were discharged from the Neonatal Intensive Care of the Hospital Zonal Bariloche, Argentina. The assessment consisted of a neurological examination and cranial ultrasound at term, and neurological examination and evaluation of development at 12 months. The sensitivity, specificity, positive and negative predictor value was calculated. The relationship between perinatal factors and neurodevelopment at 12 month of age was also calculated using logistic regression models. RESULTS: Seventy children completed the follow-up. At 12 months of age, 14% had an abnormal neurological examination, and 17% abnormal development. The neurological examination and the cranial ultrasound at term had low sensitivity to predict abnormal neurodevelopment. At 12 months, 93% of newborns with normal AT showed normal neurological results, and 86% normal development. Among newborns with normal cranial ultrasound the percentages were 90 and 81%, respectively. Among children with three or more perinatal risk factors, the frequency of abnormalities in the neurological response was 5.4 times higher than among those with fewer risk factors, and abnormal development was 3.5 times more frequent. CONCLUSIONS: The neurological examination and cranial ultrasound at term had low sensitivity but high negative predictive value for the neurodevelopment at 12 months. Three or more perinatal risk factors were associated with neurodevelopment abnormalities at 12 months of age.

Profile information matrix for nonlinear transformation models.

For semiparametric models, interval estimation and hypothesis testing based on the information matrix for the full model is a challenge because of potentially unlimited dimension. Use of the profile information matrix for a small set of parameters of interest is an appealing alternative. Existing approaches for the estimation of the profile information matrix are either subject to the curse of dimensionality, or are ad-hoc and approximate and can be unstable and numerically inefficient. We propose a numerically stable and efficient algorithm that delivers an exact observed profile information matrix for regression coefficients for the class of Nonlinear Transformation Models [A. Tsodikov (2003) J R Statist Soc Ser B 65:759-774]. The algorithm deals with the curse of dimensionality and requires neither large matrix inverses nor explicit expressions for the profile surface.