Outcome After Ischemic Stroke in Childhood
by V. Ganesan, A. Hogan, N. Shack, et al, Developmental Medicine and Child Neurology, 2000;42:455–461.
Kendra Tracy, Northwestern University
Characterization of outcomes following ischemic stroke in childhood may help those who plan and provide rehabilitation services. Because there are risks associated with acute interventions, such as anticoagulation or thrombolysis, for ischemic stroke in children, it is important to decide if the risks associated with these interventions are worth the benefits they provide. To determine appropriate interventions, the prevalence of physical, behavioral, and/or cognitive impairments following ischemic stroke must be understood. The authors of this study used a parental questionnaire along with assessment by a pediatric occupational therapist, a pediatric physiotherapist, and a neuropsychologist to evaluate outcome after ischemic stroke. Basic functional factors were included on the questionnaire. These were related to residual impairments the children experienced during daily activities. Impact of clinical factors, such as age at the time of stroke, time since stroke, lesion location, and incidence of seizure, were examined in this longitudinal study.
One hundred twenty-eight children who had ischemic stroke were included in this study. All of these children were treated at Great Ormond Street Hospital in London, England between 1990 and 1996. The study included children who had been seen for treatment of acute ischemic stroke, and children who had been referred to the pediatric stroke clinic. Children who had sustained ischemic stroke within the first month of life were excluded from the study. Outcome information was available for 105 of the children; 15 of the original children died, and the remaining 14 were lost to follow-up.
A questionnaire, which evaluated outcome, was developed and distributed to parents of children who had survived an ischemic stroke. All nonresponders were sent a second questionnaire. According to previous research, the use of simple questions to evaluate outcome after stroke in adults has a higher degree of correspondence with formal assessments than do complicated questions. Therefore, the questionnaire used in this study was simple in design. The majority of the questions required yes/no answers. Quantification was required for two questions, which focused on upper and lower limb motor impairments; rating guidelines were included for these questions. Some questions focused on whether or not the child required assistance with activities of daily living in the school or home setting. Other questions specifically addressed the child’s motor, speech, language, and behavioral functions. Parents were also asked if the child had epilepsy and was on anticonvulsant therapy. Parents of 90 children completed the questionnaire. To allow for comparison with parent ratings, a pediatric occupational therapist and a pediatric physiotherapist also completed the questionnaire. The composite scores for the children ranged from zero to 13, with a score of 13 indicating the greatest amount of residual impairments. Children with a score below four were placed in the “good” outcome group and were considered to have impairments that would not likely interfere with daily life. Children with scores above four were placed in the “poor” outcome group; they were considered to have impairments that resulted in significant disabilities.
Neuropsychological assessments were done on 22 of the children. The therapists had also examined 15 of these children. The Bayley Scales of Infant Development (BSID-II) were used to measure the overall cognitive function of the four children under the age of four years. Intelligence quotient (IQ) was measured in the other 18 children using the age-appropriate Wechsler Intelligence Scale. The Clinical Evaluation of Language Fundamentals (CELF) was used to measure receptive and expressive language abilities in 15 children. Magnetic resonance imaging (MRI) reports from the time of the stroke were reviewed to determine lesion location, arterial territories involved, and level of cortical involvement. Cohen’s κ statistic was calculated to analyze the agreement between the reports of the parents and therapists. Independent effects of age at time of stroke, time since stroke, risk factors for stroke, seizures during stroke, and location of infarct on outcome were examined using a logistic regression analysis. The results reported were obtained using a backward stepwise method with P < 0.1 as the criterion for withdrawal of variables from the model; the two-tailed significance value was 0.05.
The subjects ranged between three months and 16 years of age at the time of the stroke (median age: five years). The time between stroke and outcome data collection was between three months and 13 years (median: three years). Cortical involvement was found in 59% of the children. Bilateral lesions were found in 31%, left-hemisphere lesions in 34%, and right-hemisphere lesions in 34% of the children. The anterior cerebral circulation was affected in 73%, the posterior circulation was affected in 14%, both anterior and posterior distributions were affected in 7%, and the internal border zone was affected in 2% of the children.
According to parental report, 87% of the children had clearly developed handedness before the stroke, and 26% changed handedness following the stroke. Only 4% of the children were unable to walk at the time of follow-up. All of these children had experienced a second neurological insult after the stroke. Parents reported 42% of the children had speech and language difficulties. Sixteen of these children had bilateral lesions, 14 had left-hemisphere lesions, and eight had right-hemisphere lesions. Parents of 59% of the children reported that they needed more help than their peers in the school environment. A good outcome was reported in 41%, and a poor outcome was reported in 59% of the children. Overall, 14% of the children were reported by their parents to have no residual impairments.
The pediatric therapists completed the questionnaire for only 24% of the children. Agreement between parents and therapists was good or very good for the qualitative questions, with the exception of the question about speech and language difficulties for which agreement was moderate. Agreement as to whether there was a motor impairment in the upper or lower limb was very good; however agreement was only moderate for the quantification of the impairment.
Neuropsychological examinations were performed on 22 of the children. Two of the four children under the age of four years demonstrated developmental delay according to the BSID-II. The results of the Wechsler’s scale ratings indicated that more children fell in the bottom half of the IQ distribution than is expected in the normal population. Of the 18 children tested on the Wechsler scales, seven had significantly higher verbal than performance IQ scores and five had significantly higher performance than verbal IQ scores. Two of the four children tested with the BSID-II showed a language impairment of more than 12 months, the other two had age-appropriate language skills. Eight of the 15 children who were given the CELF had language scores that were average or above average, four demonstrated a mild impairment, three had a severe impairment. Fair agreement between parental report and neuropsychological evaluation about whether a child had speech or language impairments was reported. The only clinical factor found to be related to poor outcome following ischemic stroke was young age at the time of stroke.
These data suggest that most children who suffer ischemic stroke have residual difficulties that can include a wide range of problems. In the children studied, the area of slowest and most incomplete motor recovery was hand function. After a single ischemic stroke, children usually regain ambulation, as was observed in this study. These children have difficulties in both the home and school environment. There was overall agreement between parental and therapist perception of disability. However, parents tended to overestimate the severity of motor impairments and underestimate the severity of language impairments. The overall findings support previous studies that have reported that approximately 75% of children have residual impairments after ischemic stroke.
Limitations and Implications
Although the authors discussed the risk-to-benefit ratio of using acute interventions such as thrombolysis or anticoagulation as a purpose for this study, they neglected to report information regarding use of these interventions in the children studied. Another limitation of this study is that one fourth of the questions in the questionnaire were related to hemiplegia although nearly one third of the children studied had bilateral lesions. Based on the results of the questionnaire, the children were placed into either the good- or the poor-outcome group, a third subgroup may have helped to more specifically categorize the level of disability experienced by the children. Less than one third of the children were reviewed by the therapists and/or given a neuropsychological evaluation. Because parental report was not always consistent with the professional findings, professional evaluation of all children may have yielded more accurate results related to level of disability. Finally, the questionnaire used was not validated with a standardized outcome measure, as there is no formal measure to evaluate functional impairments after ischemic stroke in the pediatric population. Therefore, the results of this study should be confirmed; and the development of a standardized scale to evaluate outcome after stroke in the pediatric population is needed.