EUROPEAN JOURNAL OF PAEDIATRIC NEUROLOGY I4(20I0)67-72 Patients and methods (for gestational age beyond 42 weeks of gestation). Since the subjects of the study included both term infants and preterm nfants(bom at a wide range of gestational ages), we used their Patients selected for this retrospective region-based survey of gestational age adjusted birth weights as appropriate indica children with CP included 197 children aged 0-12 years tors for intra-uterine growth rather than their actual birth (117 males, 80 females)with CP whose medical records were weights. These birth weights were classified as: small for the Haifa district in Israel (The Clalit Healthcare Services, The gestational age based on Usher McLean intra-uterine growth Bnai-Zion Medical Center, and the Maccabi Healthcare charts. 13 For all patients the mode of delivery was classified as Services). Children with CP bon after 1993 were included in normal delivery, instrumental(forceps or vacuum extraction the study. All patients entered the child development centers delivery and cesarean section. This group was further divided prior to age 5, most of them during the first two years of lif into children that were born by elective cesarean section and During the years 2005-2006 when the data on these patients those who were born by an urgent cesarean section due to was collected, the mean age of the patients was7.2+3.0 years, labor complication or acute fetal distress. Apgar score was Range 1.4-12.8 years It should be noted that the data collected defined as high if it was above 7 at one minute or equal to 8 or covered seizures that occurred from birth to 12.8 years. One above at five minutes: as moderate when it was between 4 and hundred and three patients(52.3% were of Jewish ancestry, 88 6 at one minute or between 4 and 7 at ive minutes and as low (44.7%were of Arab ancestry and 6(3.0%)were of other score when it was less than or equal to 3 at one and five evelopmental disorders in this area are referred to these guinity and the presence of neonatal seizures. The research enters, these patients represent the vast majority of the was approved by the Helsinki committee of all the three children with CP in this region. The remaining patients belong agencies collaborating in this study a medical insurance group that did not participate in the Statistics, in 2006 there were 145,000 children(age range 0-13 variables and odds ratio and 95% CI were calculated e test Data analysis was performed using SPSS. Chi-squar study. According to the Government Central Bureau of was used to assess the relationship between categorical years)in the Haifa region which means that the prevalence of In groups with a small number of patients the Fisher Exact CP in the region is at least 1.35: 1000, 95% CI(1.169, 1.548). test was employed. T-test was used to compare continuous Children variables between two independent groups. Multivariate ompared with children with CP without epilepsy logistic regression was used to asses the influence of different The clinical types of CP were divided into five variants sk factors for the development of epilepsy. Odds ratio and hemiplegia, quadriplegia, diplegia, dyskinesia, and ataxia- 95% confidence interval were calculated from the model. All p hypotonia. Whether treated or not, epilepsy was defined as the values were two-sided, and statistical significance was presence of two or more non-febrile seizures beyond the defined as p<0.05 neonatal period. We classified the seizures according to the modified criteria of the International League against Epilepsy into 7 clinical subtypes: infantile spasms, partial seizure (without or with secondary generalization), generalized tonic. 3. Results clonic seizures(GTCS), myoclonic seizures, atonic and/or tonic seizures with or without atypical absences, typical absenc 3.1. Epilepsy, neonatal seizures and seizures and non-classified seizures. In cases where seizures were mixed or changed with time, we referred to the seizure Sixty-five children(33.0%)with CP developed epilepsy.of type at the earliest presentation of the epileptic seizures (but these, thirty two children (49.2%) had their first seizures after the neonatal period). Imaging data such as computed (beyond the neonatal period) during the first 12 months and45 tomography (Cr) and magnetic resonance imaging(MRI) children(69. 2%)had their first seizure during the first two findings were classified as: non-specific atrophy, brain mal formations, primarily gray matter lesions (including infarcts) primanly white matter lesions (including periventricular leu normalimaging. Thinning of the corpus callosum due to white s komalacia), hydrocephalus, intracranial-hemorrhage, and atter atrophy was considered as white matter lesions rather a20 than being a cerebral malformation Cases with non-specific 8 atrophy associated with additional specific changes were classified according to the latter Head size(at time of admis sion to the child development center) was defined as: normo- cephalic, microcephalic (below 2nd percentile),or acrocephalic (above 98th percentile). The gestational age of the patients at birth was classified as: term infants(for gesta- tional age between 38 and 41 weeks), preterm infants(for age(years) a gestational age between 30 and 37 weeks), extreme preterm Fig. 1- The number of patients with epilepsy onset at each (for a gestational age below 30 weeks) and post-term infants 第64°2. Patients and methods Patients selected for this retrospective region-based survey of children with CP included 197 children aged 0–12 years (117 males, 80 females) with CP whose medical records were retrieved from the three main child development centers of the Haifa district in Israel (The Clalit Healthcare Services, The Bnai-Zion Medical Center, and the Maccabi Healthcare Services). Children with CP born after 1993 were included in the study. All patients entered the child development centers prior to age 5, most of them during the first two years of life. During the years 2005–2006 when the data on these patients was collected, the mean age of the patients was 7.2 3.0 years; Range 1.4–12.8 years. It should be noted that the data collected covered seizures that occurred from birth to 12.8 years. One hundred and three patients (52.3%) were of Jewish ancestry, 88 (44.7%) were of Arab ancestry and 6 (3.0%) were of other origins. Since more than 90% of all the children with neuro￾developmental disorders in this area are referred to these centers, these patients represent the vast majority of the children with CP in this region. The remaining patients belong to a medical insurance group that did not participate in the study. According to the Government Central Bureau of Statistics, in 2006 there were 145,000 children (age range 0–13 years) in the Haifa region which means that the prevalence of CP in the region is at least 1.35:1000, 95% CI (1.169, 1.548). Children who subsequently developed epilepsy were compared with children with CP without epilepsy. The clinical types of CP were divided into five variants: hemiplegia, quadriplegia, diplegia, dyskinesia, and ataxia￾hypotonia. Whether treated or not, epilepsy was defined as the presence of two or more non-febrile seizures beyond the neonatal period. We classified the seizures according to the modified criteria of the International League against Epilepsy12 into 7 clinical subtypes: infantile spasms, partial seizure (without or with secondary generalization), generalized tonic￾clonic seizures (GTCS), myoclonic seizures, atonic and/or tonic seizures with or without atypical absences, typical absence seizures and non-classified seizures. In cases where seizures were mixed or changed with time, we referred to the seizure type at the earliest presentation of the epileptic seizures (but after the neonatal period). Imaging data such as computed tomography (CT) and magnetic resonance imaging (MRI) findings were classified as: non-specific atrophy, brain mal￾formations, primarily gray matter lesions (including infarcts), primarily white matter lesions (including periventricular leu￾komalacia), hydrocephalus, intracranial-hemorrhage, and normal imaging. Thinning of the corpus callosum due to white matter atrophy was considered as white matter lesions rather than being a cerebral malformation. Cases with non-specific atrophy associated with additional specific changes were classified according to the latter. Head size (at time of admis￾sion to the child development center) was defined as: normo￾cephalic, microcephalic (below 2nd percentile), or macrocephalic (above 98th percentile). The gestational age of the patients at birth was classified as: term infants (for gesta￾tional age between 38 and 41 weeks), preterm infants (for a gestational age between 30 and 37 weeks), extreme preterm (for a gestational age below 30 weeks) and post-term infants (for gestational age beyond 42 weeks of gestation). Since the subjects of the study included both term infants and preterm infants (born at a wide range of gestational ages), we used their gestational age adjusted birth weights as appropriate indica￾tors for intra-uterine growth rather than their actual birth weights. These birth weights were classified as: small for gestational age, appropriate for gestational age, and large for gestational age based on Usher & McLean intra-uterine growth charts.13 For all patients the mode of delivery was classified as: normal delivery, instrumental (forceps or vacuum extraction) delivery and cesarean section. This group was further divided into children that were born by elective cesarean section and those who were born by an urgent cesarean section due to labor complication or acute fetal distress. Apgar score was defined as high if it was above 7 at one minute or equal to 8 or above at five minutes; as moderate when it was between 4 and 6 at one minute or between 4 and 7 at five minutes and as low score when it was less than or equal to 3 at one and five minutes. We also looked in all patients for parental consan￾guinity and the presence of neonatal seizures. The research was approved by the Helsinki committee of all the three agencies collaborating in this study. Data analysis was performed using SPSS. Chi-square test was used to assess the relationship between categorical variables and odds ratio and 95% CI were calculated. In groups with a small number of patients the Fisher Exact test was employed. T-test was used to compare continuous variables between two independent groups. Multivariate logistic regression was used to asses the influence of different risk factors for the development of epilepsy. Odds ratio and 95% confidence interval were calculated from the model. All p values were two-sided, and statistical significance was defined as p < 0.05. 3. Results 3.1. Epilepsy, neonatal seizures and CP Sixty-five children (33.0%) with CP developed epilepsy. Of these, thirty two children (49.2%) had their first seizures (beyond the neonatal period) during the first 12 months and 45 children (69.2%) had their first seizure during the first two 0 5 10 15 20 25 30 35 0-1 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8-9 9-12 age (years) number of patients Fig. 1 – The number of patients with epilepsy onset at each age (N [ 65). 68 european journal of paediatric neurology 14 (2010) 67–72 第 64 页