Comparison of neurological outcome at one and two years age of very preterm infants: risk factors for a misdiagnosis of cerebral impairment at one year of age

Background. - The purpose of this population-based study was firstly to compare the neurodevelopmental outcome at one and two years of very preterm infants, and secondly to identify the risk factors for a misdiagnosis of cerebral impairment at the age of one year: Population and methods. - The preterm cohort included 203 infants born between 25 and 32 weeks of gestational age in the region of Franche-Comte (France) during a two-year period. The control group included 196 full-term infants born in the same maternity wards. Neuro-developmental assessments were performed by pediatricians or physicians, both at one and two years of age, on 94% (161/171) surviving preterms and 89% (173/195) full-terms. Results. - There is a fair correlation between the two neurological evaluations of the control group (170/173, 98% have the same classification at the age of one and two). There is a weak correlation (kappa = 0.37) between the two neurological evaluations of the preterm group. Sixteen preterms (10%) had been classified more abnormal at one year than they were at two years. The presence of a broncho-pulmonary dysplasia, linked to male sex and extreme prematurity was statistically linked to this first kind of misclassification. Seventeen preterms (10%) had been considered more normal at one year than they were at flue years. The presence of a diplegia, family precariousness and the examination at one year of age by a general practitioner were statistically linked to this second kind of misclassification. Conclusion. - This prospective population-based study identifies structural situations (bronchopulmonary dysplasia linked to extreme prematurity) and environmental situations (family precariousness, examiner's qualifications) linked to a misclassification of the neurological status of one-year-old former preterm infants. (C) 2000 Editions scientifiques et medicales Elsevier SAS.