 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
* Department of Pediatrics, Medical University of South Carolina, Charleston, South Carolina
College of Health Professions, Medical University of South Carolina, Charleston, South Carolina
Center for Health Care Research, Medical University of South Carolina, Charleston, South Carolina
|| Office of Research and External Affairs, University of Georgia, Athens, Georgia
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Methods. Participants were 60 children who had sickle cell disease (HbSS) and had no documented history of stroke. Children were classified according to Stroke Prevention Trial in Sickle Cell Anemia criteria (normal, conditional, and abnormal), and their performance was compared on measures of intellectual abilities, academic achievement, sustained attention/concentration, executive function, and parent and teacher ratings of executive function.
Results. Children with abnormal TCD values performed more poorly than children with conditional TCD values on measures of verbal intelligence and executive function. Children with conditional TCD values performed more poorly than children with normal TCD values on measures of sustained attention/concentration and executive function. TCD values also were a significant predictor of auditory working memory in exploratory analyses.
Conclusions. Our findings support the hypothesis that neurocognitive functions subserved by the frontal systems (eg, sustained attention/concentration and executive function) seem to be the most useful indices of progressive cerebrovasculopathy in children with HbSS disease.
Key Words: transcranial Doppler ultrasonography • sickle cell anemia • neurocognitive functioning • pediatric
Abbreviations: SCD, sickle cell disease • CVA, cerebrovascular accident • MRI, magnetic resonance imaging • CNS, central nervous system • TCD, transcranial Doppler • STOP, Stroke Prevention Trial in Sickle Cell Anemia • Vmean, mean velocity • WASI, Wechsler Abbreviated Scale of Intelligence • VIQ, verbal IQ • BRIEF, Behavior Rating Inventory of Executive Function • MANCOVA, multivariate analyses of covariance • ANCOVA, analyses of covariance • SD, standard deviation
Perhaps some of the most devastating outcomes of pediatric sickle cell disease (SCD) are neurologic complications. Cerebrovascular accident (CVA) constitutes a leading cause of morbidity and mortality, occurring most frequently in children with the HbSS genotype at an estimated incidence of 4% to 8%.1,2 Neurocognitive sequelae of clinically apparent CVA in children with HbSS are characterized by pervasive impairments, including decrements in general intellectual function, language and verbal abilities, visual-motor and visual-spatial processing, memory, academic achievement, and sustained attention/concentration.3–7 Deficits in the processing of subtle prosodic information also have been documented.8
Children with SCD also may show evidence of neurologic pathology and associated neurocognitive deficits before signs and symptoms become apparent. As many as 11% to 17% of children who have HbSS disease and are otherwise asymptomatic have silent cerebral infarcts, lesions of infarction or ischemia identified on brain magnetic resonance imaging (MRI).9–11 Silent cerebral infarcts are most likely accounted for by chronic hypoxia in the microvasculature ensuing from progressive disease in the major cerebral arteries.10,12,13 Early investigators of the neuropsychological correlates of pediatric SCD provided data suggesting that children who have SCD with no known history of stroke were at increased risk for learning impairments.14–18 For example, children with SCD demonstrated poorer performance on measures of reading decoding and sustained attention/concentration than healthy sibling comparison control subjects. The results of these investigations, however, were variable, making comparisons across studies difficult.
More recent investigations have addressed the neurocognitive sequelae of silent infarcts in children with SCD by comparing the performance of MRI-defined groups.3,4,12,19–21 These studies demonstrated that the neuropsychological deficits associated with silent infarcts are typically less severe than those deficits associated with overt stroke. These results also lend support for the importance of measures of attention and executive function for the identification of patients with silent cerebral infarcts.19 In fact, Brown et al4 provided important evidence for group differences in the areas of attention and executive function; children with silent cerebral infarct demonstrated similar attentional impairments as children with a history of overt CVA. This finding was explained by the location of central nervous system (CNS) pathology. For the entire sample, the highest frequency of damage was overwhelmingly localized in the frontal lobes, present in 93% of children in the overt stroke and silent cerebral infarct groups.
The extant neurologic and neuropsychological literatures support a disease-specific model of neurocognitive decline in children with HbSS disease. Neurocognitive functions subserved by the frontal systems (eg, tasks of sustained attention/concentration and executive function) seem to be particularly vulnerable to the deleterious effects of progressive cerebrovasculopathy. In fact, these data also are consistent with the finding of pervasive frontal cortex blood flow dysfunction in adults with HbSS.22 The advent of transcranial Doppler (TCD) ultrasonography allowed for the reliable detection of vasculopathy, previously seen only on angiography or neuropathological examination.23,24 Furthermore, this technique led to the identification of asymptomatic children who have SCD and are at highest risk for cerebral infarction.25–27 In this context, it would be important to investigate the neuropsychological correlates of TCD ultrasonography in the pediatric SCD population.
Our study sought to extend the findings of the Stroke Prevention Trial in Sickle Cell Anemia (STOP).28,29 This study represents the first known investigation to examine the association between cerebral blood flow velocity, as measured by TCD ultrasonography, and neurocognitive functioning in children with sickle cell anemia (HbSS). Previous studies generally compared groups (eg, stroke versus nonstroke groups), which did not allow for examination of the progressive nature of cerebrovasculopathy in this population of children. In this investigation, we examined the continuum of disease severity using TCD data as indices of disease severity. We also examined the usefulness of TCD diagnostic groupings in terms of neurocognitive functioning. We hypothesized that relative to children who displayed TCD ultrasonography values that were within the normal range (<170 cm/sec), children with values in the conditional (170–200 cm/sec) and abnormal ranges (>200 cm/sec) would exhibit poorer neurocognitive performance. Finally, we hypothesized that measures of attention/concentration and executive function would be more sensitive to the identification of TCD ultrasonography in the pathologic range, relative to more global indices of intelligence and academic achievement.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
After approval from the institutional review boards of both participating institutions, informed consent was obtained from the participants’ caregivers and assent was obtained from children 12 years of age and older. Demographic information was obtained through a questionnaire, and medical information was obtained from computerized databases at both institutions. For each participating child, laboratory data (eg, hematocrit) obtained closest in time to the neuropsychological evaluation were recorded from the computerized database. Caregivers also provided written consent to obtain information from the teachers of participating children. Educational history, including grade retention, special education placement, and behavioral functioning, was obtained through a questionnaire and behavior rating scales completed by each participating child’s teacher. Caregivers were provided with a written report of the results of the neuropsychological testing for their child.
TCD Methods
TCD studies of blood flow velocity in the basal cerebral arteries were performed by a trained technician, as detailed by Adams et al.28 Velocities in the proximal middle cerebral arteries and distal internal carotid arteries and bifurcation of the internal carotid arteries were recorded. The highest time-averaged mean velocity (Vmean) among these recordings was used in the statistical analyses. Following the guidelines of the STOP trial, participants were assigned to the normal TCD group when the highest Vmean was <170 cm/sec (n = 25). Participants were assigned to the conditional group when the highest Vmean was between 170 and 200 cm/sec (n = 15). Participants were assigned to the abnormal group when the highest Vmean exceeded 200 cm/sec (n = 20). TCD studies for participants in the normal and conditional categories were obtained within 6 months of neuropsychological testing. Because of the relative infrequency of abnormal TCD findings, participants who had abnormal TCD studies within a 5-year period before neuropsychological testing were enrolled. As part of the STOP protocol, these patients also were receiving chronic blood transfusions for primary stroke prevention.
Neurocognitive Battery
On the basis of the literature on neurocognitive functioning in children with SCD,3,4,19,21 a neurocognitive battery was selected to assess 5 domains of functioning: general intellectual abilities, academic achievement, sustained attention/concentration, executive function (eg, working memory), and parent and teacher behavior ratings of executive function. Each child was individually administered the battery of measures described next.
Intellectual Functioning
Measures of intelligence were obtained using the Wechsler Abbreviated Scale of Intelligence (WASI).30 The WASI is composed of those subtests on traditional Wechsler scales that have demonstrated the highest loadings (eg, >0.70) on g or general intellectual functioning.31–33 The WASI allows for assessment of verbal IQ (VIQ), performance, and full-scale intellectual functioning. These age-referenced standard scores were used in the statistical analyses.
Academic Achievement
Measures of academic achievement were obtained using the subtests that compose the Broad Reading and Broad Math composites of the Woodcock-Johnson Psycho-Educational Battery: Tests of Achievement-Revised.34 Broad Reading consists of the Letter-Word Identification subtest, which assesses correct pronunciation of isolated letters and words of graded difficulty, and the Passage Comprehension subtest, which measures reading comprehension. Broad Mathematics consists of the Calculation subtest, which assesses written calculations, and the Applied Problems subtest, which assesses skill in analyzing and solving practical problems in mathematics. Broad Reading and Broad Math age-referenced standard scores were used in the statistical analyses.
Measures of academic attainment also were obtained. At study enrollment, participant caregivers provided information on each child’s academic history, including whether the child had a grade retention and/or special education placement. Grade retention was defined as a child repeating a grade in school as a result of either lack of educational progress or excessive school absenteeism secondary to SCD complications. Special education placement was defined as receipt of special education services either in the regular education classroom (eg, 504 plan) or in an alternative classroom placement (eg, resource classroom, self-contained classroom).
Sustained Attention/Concentration
The Conners’s Continuous Performance Test II35 is a standardized, well-normed computerized task designed to assess visual sustained attention/concentration. During this 14-minute task, the examinee is instructed to respond to target stimuli on a computer screen and inhibit responses to nontarget stimuli. A number of computer-generated parameters yield information about inattention and impulsivity. Age-referenced T scores for omission errors, commission errors, reaction time, and response time variability were used in the statistical analyses.
Executive Function
The Numbers Backward and Picture Locations subtests from the Children’s Memory Scale36 provided measures of auditory and visual working memory, respectively. Working memory, an executive function, enables an individual to hold information in mind for the purpose of multistep problem solving. Numbers Backward requires mental manipulation of a string of numbers that are repeated backward. Picture Locations requires correct location of a previously exposed series of picture stimuli. Both measures assess span of working memory as the number of stimuli become increasingly longer as the tests progress. The composite score for Attention/Concentration on the Children’s Memory Scale, which includes Numbers Backward and Picture Locations, demonstrates a high correlation (0.73) with the Freedom From Distractibility Index of the Wechsler Intelligence Scale for Children III, a purported measure of working memory.36,37 Numbers Backward and Picture Locations yield age-referenced scaled scores, which were used in the statistical analyses.
The Trail Making Test38 is a paper-pencil test designed to assess focused attention, psychomotor speed, sequencing, mental flexibility, and visual search. Scoring is expressed in terms of the time in seconds required to complete parts A and B of the task. Part B, which requires set shifting (ie, executive function), has effectively differentiated the performance of children with SCD without CNS pathology from the performance of children with silent cerebral infarction.4 Performance on part B, time to completion, was used in the statistical analyses.
Parent and Teacher Ratings of Executive Function
Participant parents and teachers were asked to complete a behavior rating scale designed to assess executive functions in the home and school environments. The Behavior Rating Inventory of Executive Function (BRIEF)39 is composed of 86 items designed to assess various aspects of executive function, including ability to inhibit emotional and behavioral responses, cognitive flexibility, working memory, planning and organization skills, and initiative. Data support the reliability, internal consistency, and construct validity of this instrument. The BRIEF yields age-referenced T scores. The Behavioral Regulation Index, Metacognition Index, and Global Executive Composite T scores from the parent and teacher questionnaires were used in the statistical analyses.
Data Analyses
A series of separate 1-way multivariate analyses of covariance (MANCOVAs), using the general linear model, was performed for each of the neurocognitive domains (measures of intellectual functioning, academic achievement, sustained attention/concentration, executive function, and parent and teacher ratings of executive function). Group assignment based on TCD findings served as the independent variable (normal: <170 cm/sec; conditional: 170–200 cm/sec; and abnormal: >200 cm/sec). Given the modest correlation between hematocrit and some of the dependent variables that were of interest, hematocrit served as a covariate.a Also, hematocrit served as a covariate to control for the effects of transfusion in the abnormal TCD group. These analyses, which treated TCD data as a categorical variable, were conducted to determine whether neurocognitive performance differed across the 3 categories of TCD diagnostic groupings, controlling for the differential effect of hematocrit. Thus, 5 MANCOVAs were performed. Significant multivariate tests were followed by univariate analyses of covariance (ANCOVAs), again using hematocrit as a covariate. Fischer’s least significant difference planned contrasts were performed on any dependent measure for which there was a significant ANCOVA. The a priori 
level was set at 0.05.
A series of exploratory multiple regression analyses, using the general linear model, were conducted to examine the relationship between cerebral blood flow velocity and neurocognitive functioning. Participants who were receiving transfusions were removed from the larger sample, to control for the potential confounding effect of this treatment condition (revised N = 44). In these analyses, the predictor variable of TCD was treated as a continuous variable. Age, gender, and hematocrit also were entered into each regression model to explore the extent to which these variables contributed to the linear relationship between cerebral blood flow velocity and neurocognitive functioning. Standardized test scores for each of the 5 neurocognitive domains served as the dependent variables.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
2 = 6.24, P < .05). In addition, the groups differed in terms of special education placement. As expected, significantly more children in the abnormal group were receiving special education services, compared with the normal and conditional groups (2 = 9.38, P < .01). Although the comparison did not reach statistical significance, a trend was found toward increasing grade retention as a function of TCD group membership. A greater frequency of grade retention was found with more impaired TCD values. No differences were found for the other demographic variables. A significant group difference was found for the highest time-averaged TCD Vmean (F(2,56) = 156.02, P < .001), consistent with the objectively defined TCD diagnostic groups. Post hoc analyses revealed that the highest time-averaged Vmean was higher in the conditional group than in the normal group (P < .001) and higher in the abnormal group than in the conditional group (P < .001). In contrast, group comparisons in terms of hematocrit did not reach significance. An examination of the means showed that the abnormal group, in fact, had a slightly higher mean value for hematocrit, compared with the normal and conditional groups. This difference is likely accounted for by the treatment that this group received; only children in the abnormal group were receiving chronic blood transfusions for primary stroke prevention.
Multivariate Analyses
The results of a 1-way MANCOVA on the intellectual functioning domain yielded a statistically significant main effect (Wilks’ 
F[2,54] = 2.73, P < .05). Separate ANCOVAs were performed on each of the dependent measures, and a statistically significant effect was revealed for the WASI VIQ composite score (F[2,55] = 3.75, P < .05). Post hoc analyses revealed that the conditional group obtained significantly higher VIQs compared with the abnormal group (P < .01; Table 2).
|
F[2,55] = 1.84, P = .056). Although this effect was not statistically significant at the a priori level, because of the exploratory nature of this investigation we elected to conduct ANCOVAs. A significant main effect for the Continuous Performance Test Hit Reaction Time (F[2,55] = 3.58, P < .05) and the Continuous Performance Test Variability score (F[2,55] = 3.16, P = .05) was revealed. Post hoc analyses revealed that the conditional group performed more poorly than the normal group for both reaction time (P < .05) and variability (P < .05).
A statistically significant main effect (Wilks’ F[2,43] = 2.57, P < .05) also was demonstrated in the 1-way MANCOVA on the executive function domain. Separate ANCOVAs revealed a significant main effect for Trails B, time to completion score (F[2,44] = 5.99, P < .01). Post hoc analyses indicated that both the normal and the abnormal groups completed this test more rapidly than the conditional group (P < .01).
The results of the 1-way MANCOVA on parent and teacher ratings of executive function (BRIEF) yielded a statistically significant main effect (Wilks’ F[2,53] = 2.17, P < .05). Separate ANCOVAs yielded a significant main effect for the parent BRIEF Behavioral Regulation Index (F[2,54] = 5.19, P < .01). Post hoc analyses indicated that parents rated the abnormal group as more impaired in behavioral regulation relative to the conditional group (P < .01). In addition, a significant effect was revealed for the teacher BRIEF Metacognition Index (F[2,55] = 3.48, P < .05). Post hoc analyses revealed that teachers rated children in the abnormal group as more impaired in metacognitive skills relative to children in either the normal (P < .05) or conditional (P < .05) groups.
Regression Analyses
A series of exploratory multiple regression analyses were computed for each of the dependent measures (neurocognitive measures) using the general linear model. Bivariate correlation coefficients are presented in Table 3. TCD, chronological age, hematocrit, and gender served as the predictor variables and were entered into the full model. Standardized ß weights for the predictor variables were examined to determine their relative predictive ability.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
The performance of children classified according to STOP study criteria of normal <170 cm/sec, conditional 170 to 200 cm/sec, and abnormal >200 cm/sec28,29 was compared on measures of intellectual abilities, academic achievement, sustained attention/concentration, executive function, and parent and teacher behavior ratings of executive function. Children with abnormal TCD values performed more poorly than those designated with conditional TCD values on measures of verbal intelligence and executive function. Children with conditional TCD values performed more poorly than participants with normal TCD values on measures of executive function. Children with conditional TCD values also performed more poorly on measures of sustained attention/concentration, although these data must be interpreted with caution given the exploratory nature of the analyses. Preliminary findings also revealed deterioration in neurocognitive performance as a function of chronological age. Specifically, a decline in intellectual abilities, sustained attention/concentration, and executive functions was evidenced with increasing chronological age. Finally, TCD values were a significant predictor of auditory working memory after controlling for age, gender, and hematocrit; again, these data must be interpreted judiciously given the exploratory nature of the analyses.
It should be noted that our findings yielded seemingly inconsistent results. For example, children with abnormal TCD values performed more poorly on measures of global functioning (verbal intelligence and behavioral ratings of executive functions), whereas children with conditional TCD values performed more poorly on measures of specific neurocognitive skills (attention/concentration and executive function). One possible explanation is that participants with conditional TCD values may represent a subgroup of the pediatric SCD population that is distinct from participants with abnormal TCD values. Children with abnormal TCD values may have both vascular and structural abnormalities, whereas children with conditional TCD values may evidence simply one of these abnormalities. This possibility is supported by evidence that TCD and MRI studies often are discordant for children with SCD, perhaps revealing different aspects of the pathophysiology of CNS injury.40,41 In addition, children with abnormal blood flow velocities were the most likely to have been in their respective TCD diagnostic group for the longest period of time. For this reason, they may have experienced the cumulative effects of SCD for a greater duration of time with concomitant global impairment of function. An alternative explanation may relate to the relatively small sample size of this investigation, which may have resulted in sampling error. Regardless of which explanation may be correct, the general finding that elevated TCD values are associated with some type of cognitive impairment is provocative and clearly warrants additional investigation.
The contribution of our investigation must be interpreted within the limitations of the study design. First, the lack of a consistent stepwise increase in neurocognitive and behavioral dysfunction with increasingly high blood flow velocity may be explained, in part, by the confounding influence of transfusion therapy. The performance of the abnormal TCD group in this study was necessarily confounded by the transfusion therapy that they received as part of the National Institutes of Health STOP trial.28,29 Future studies will need to use prospective designs and control for the possible rehabilitative effects of transfusion therapy to determine the influence of this treatment on neurocognitive functioning in the long term. Second, neuroimaging, which was not available for all participants at the time of this investigation, would have permitted identification of participants with structural abnormalities that are characteristic of silent infarcts. Future investigations will need to corroborate TCD and neurocognitive data with neuroimaging techniques, including MRI and magnetic resonance angiography. Finally, the small sample size may have mitigated power and diminished significant effects that otherwise may have been detected with a larger sample. Additional research will need to include multisite, collaborative clinical samples that ensure larger cohorts of children with SCD. Both longitudinal designs and experimental studies are needed to confirm the correlational data obtained in this investigation. Whether there are critical developmental periods for neurocognitive impairment among children with SCD remains to be tested.
Of interest in our investigation is the association between age and neurocognitive decline. Some experts have suggested that SCD may constitute a progressive multisystemic disease process that includes CNS involvement.14 Clearly, additional study of this phenomenon is needed to corroborate these preliminary findings, particularly longitudinal studies that permit the examination of the natural progress of CNS disease during childhood and adolescence.
Notwithstanding these limitations, our data lend support to the growing body of evidence suggesting deficits in sustained attention/concentration and executive function in children with HbSS disease with various stages of CNS pathology.4,7,19 Most important, these results support recent evidence for the sensitivity of measures of specific neurocognitive functions as indices of disease severity, compared with more global measures of intelligence and academic achievement.42 As such, our findings underscore the importance of using measures of sustained attention/concentration and executive function to assess children who have HbSS disease and are at risk for neurologic impairment. In fact, these findings were substantiated for both psychometric measures and cross-informant behavioral ratings; the multiple informants and multiple data sources used in our investigation support the robustness of our findings.
Because we found that children with the most severe cerebrovasculopathy (measured by TCD ultrasonography) evidenced the greatest impairment in neurocognitive functions subserved by the frontal systems (sustained attention/concentration and executive function), we posit that the frontal systems are vulnerable to the effects of progressive pathology in the major cerebral arteries. It remains unclear why this region of the brain seems most vulnerable to insult. However, our findings are consistent with evidence suggesting altered metabolism on positron emission tomography scans in the frontal lobe area of adult patients with SCD.43 Our data also are consistent with the finding of pervasive frontal cortex blood flow dysfunction in adults with HbSS disease, despite structurally normal tissue on MRI.22 For children, cerebrovasculopathy associated with HbSS disease may interfere with the dynamic nature of development in the frontal cortex over time. Clearly, future research efforts will need to investigate the mechanism by which the frontal systems are affected by the progressive nature of cerebrovascular disease in children with SCD.
|
ACKNOWLEDGMENTS |
|---|
We thank the following individuals for contributions to this investigation: Jerri Winfrey-Carter and Shannon Johnson, Division of Hematology/Oncology, Medical University of South Carolina; Dr Kathleen McKie, Dr Virgil McKie, Dr Robert Adams, Ivy Tillman, and Heidii Poplick, Department of Pediatrics, Medical College of Georgia; and the patients and their families.
|
FOOTNOTES |
|---|
Reprint requests to (M.C.K.) the Division of Genetics and Developmental Pediatrics, Medical University of South Carolina, 135 Rutledge Ave, Box 250561, Charleston, SC 29425. E-mail: kralm{at}musc.edu
a In a series of separate 1-way MANCOVAs in which hematocrit was not treated as a covariate, significant group differences were diminished. 
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  J. Schatz, C. B. McClellan, E. S. Puffer, K. Johnson, and C. W. Roberts Neurodevelopmental Screening in Toddlers and Early Preschoolers With Sickle Cell Disease J Child Neurol, January 1, 2008; 23(1): 44 - 50. [Abstract] [PDF]
|
|
|
|
 |
|
 C. M. Hill, A. M. Hogan, N. Onugha, D. Harrison, S. Cooper, V. J. McGrigor, A. Datta, and F. J. Kirkham Increased Cerebral Blood Flow Velocity in Children With Mild Sleep-Disordered Breathing: A Possible Association With Abnormal Neuropsychological Function Pediatrics, October 1, 2006; 118(4): e1100 - e1108. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. J. Strouse, C. S. Cox, E. R. Melhem, H. Lu, M. A. Kraut, A. Razumovsky, K. Yohay, P. C. van Zijl, and J. F. Casella Inverse correlation between cerebral blood flow measured by continuous arterial spin-labeling (CASL) MRI and neurocognitive function in children with sickle cell anemia (SCA) Blood, July 1, 2006; 108(1): 379 - 381. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. C. Kral, R. T. Brown, J. K. Cure, N. Besenski, S. M. Jackson, and M. R. Abboud Radiographic Predictors of Neurocognitive Functioning in Pediatric Sickle Cell Disease J Child Neurol, January 1, 2006; 21(1): 37 - 44. [Abstract] [PDF]
|
|
|
|
|
|
R. J. Adams, D. J. Brambilla, S. Granger, D. Gallagher, E. Vichinsky, M. R. Abboud, C. H. Pegelow, G. Woods, E. M. Rohde, F. T. Nichols, et al. Stroke and conversion to high risk in children screened with transcranial Doppler ultrasound during the STOP study Blood, May 15, 2004; 103(10): 3689 - 3694. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
||
|
||
Home | My Pediatrics | Journal Information | Current Issue | Past Issues | Subscriptions & Services | Contact Us Click here for faster international access © 2003 American Academy of Pediatrics. All rights reserved. |
||
|
||
