Late neuropsychologic status after childhood head trauma

Late Neuropsychologic Status after Childhood Head Trauma Hanan Costeff, MD, Esther Abraham, PhD, Tova Brenner, MA, Israel Horowitz, MD, Naomi Apter, MD, Nahum Sadan, MD and Theodore Najenson, MD

A neurologic and neuropsychologic test battery was administered to a sample of 35 children drawn from all those in a defined geographic area who had been hospitalized for head trauma before age 7 during the years 1970-1976. Examination was performed 3% to 10 years after injury, at age 6-15. Twelve subjects had been diagnosed at the time of injury as suffering moderate insult and had been referred to the metropolitan neurosurgical center, while twenty-three with only mild injury had been retained for observation in a local pediatric ward. The twelve with more severe insult were significantly inferior to the other subjects on the Block Design and Coding subtests of the revised Wechsler Intelligence Scale for Children. The Koppitz score of the Bender Test, the WISC-R scatter, the Benton Visual Retention Test, the GATB Motor Speed Test and the Bourdon- Wiersma Vigilance Test showed less diagnostic power and failed to distinguish between the group with more severe injury and that with less. A detailed and carefully scored neurologic examination also failed to distinguish between the two groups. The findings suggest that relatively common traumatic injury may be associated with detectable late cognitive deficit, and that some WISC-R sub tests may be among the best measures for detecting such deficit. Key words: Head trauma, follow-up, brain damage, childhood, neurologic examination, psychologic examination, Bender-Gestalt Test. Costeff H, Abraham E, Brenner T, Horowitz I, Apter N, Sadan N, Najenson T. Late neuropsychologic status after childhood head trauma. Brain Dev 1988;10:371-4

It is common practice in diagnosis of learning problems and neurodevelopmental disabilities to seek evidence of diffuse cerebral dysfunction or "brain damage." The Bender·Gestalt Copying Test, the WISC-R subtest scatter and a search for neurological "soft signs" are often taken as reliable indicators of such dysfunction, although they have rarely if ever been compared against totally independent measures of brain damage. In the course of a follow-up study on cerebral trauma in childhood, we have performed these examinations in two groups of children with apparently different degrees of brain insult as judged by independent criteria.

From the Neuropediatric Unit, Loewenstein Hospital, Ra'anana (HC, EA, TB, NA, TN); Department of Pediatrics, Meir Hospital, Kfar-Sava (IH, NS); Sackler School of Medicine, Tel-Aviv University, Tel-Aviv (HC, EA, IH, NA, TN). Received for publication: January 27, 1988. Accepted for publication: August 8, 1988. Correspondence address: Dr. H. Costeff, Loewenstein Hospital, PO Box 3, Ra'anana 43100, Israel.

MATERIALS AND METHODS In 1980 a chart review was performed of all children who were hospitalized for head trauma in the years 197076 from the emergency room of the Meir Hospital in Israel. During that period the catchment area of this regional hospital had a mean total population of 160,000 and a mean population of 31,000 children of ages 0-7. During the seven years studied, 368 children of ages 0-7 were hospitalized after head trauma with suspicion of concussion or worse. Of these, 313 were judged by the admitting physicians to have mild injury and were reo tained for observation in the pediatric ward. The remaining 55 were judged to have more severe injury, based on the following criteria: neurologic deficit, disturbance of consciousness at time of examination, or depressed skull fracture. These 55 children were forwarded for hospitalization in the neurosurgical ward of a metropolitan medical center. All 313 children hospitalized on the pediatric ward survived. By contrast, of the 55 children referred for neurosurgical hospitalization, seven (13%) died. The 361 survivors were invited for follow-up study in 1980. Of these, 178 responded and were examined.

There were no significant or even suggestive differences between the children who appeared for follow-up and those who did not in terms of age, gender, ethnic origin, proportion of skull fractures and state of consciousness and presence of localizing signs on initial examination. The 178 children who were followed up underwent pediatric examination in a related study [1]. These included 154 of the 313 pediatric hospitalizations and 24 of the 48 survivors of neurosurgical hospitalization. These two subgroups were compared in gender, age, cause of trauma, ethnic origin, parental education, number of children in family and home crowding index, and they showed no significant or even suggestive premorbid differences. However, follow-up examination revealed that significantly more of the survivors of neurosurgical hospitalization suffered from seizures, headaches, enuresis and tics [1]. Similar but low proportions - 63% of the pediatric group and 61 % of the neurosurgical survivors - were making good scholastic progress [1] . In the present study, more detailed neurologic and psychologic study was performed in unselected samples of 23 of the 154 pediatric hospitalizations and 12 of the 24 neurosurgical survivors who appeared for follow-up. These two samples showed similar mean ages (10 years), sex distributions (about 60% boys) and ethnic distributions. The 12 neurosurgical survivors who were studied in depth had shown the same frequency of lateralizing signs, prolonged unconsciousness (up to 6 hours) and skull fractures at the time of hospitalization as had the other 12 who appeared for follow-up but were not studied in detail. Two of the 12 examined had undergone surgery for depressed fractures during hospitalization; one had later required re-operation for a leptomeningeal cyst. The additional study performed in these two samples included the Bender-Gestalt Copying Test (scored according to Koppitz [2]), the Benton Visual Retention Test, the Similarities, Picture Completion, Block Design and Coding subtests of the WISC-R, and a detailed neurological examination. The psychologic tests were chosen after neuropsychologic consultation with Dr Samuel Melamed of the Occupational Health and Rehabilitation Institute at Loewenstein Hospital and were those thought most likely to show deficits in the presence of brain damage. Two other specialized neuropsychologic tests - the BourdonWiersma Vigilance Test [3] and the GATB Motor Speed Test [4] - were also administered, out of curiosity as to their sensitivity to brain damage. This study is based on the assumption that the late prognosis of cerebral trauma is influenced by the age at time of injury, the period elapsed after injury, and the educational and social milieu as well as the type, location, treatment mode and severity of the traumatic cerebral lesions. Since no difference was found between the two groups studied in age, follow-up time or social milieu, it assumes that any group differences in outcome must be

372 Brain & Development, VallO, No 6, 1988

causally associated with differences in type, location, treatment mode and/or severity of the cerebral trauma. All subjects in this study suffered from blunt trauma, and localizing signs were rare in both groups. Therefore the only relevant between-group prognostic factors remaining are assumed to be the severity of cerebral trauma and the treatment mode. We assume that treatment did more good than harm, leaving us with an assumed model of two groups of children with different degrees of cerebral insult which did not differ in any other important prognostic aspect. The basic hypothesis of this study is that an effective test for brain damage should discriminate between children with more serious head trauma as evidenced by referral to neurosurgery and those who had merely been retained for pediatric observation. RESULTS Results of the four WISC-R subtests and of the Benton test are shown in Table 1. Those of the Bender test are seen in Table 2. The survivors of neurosurgical hospitalization performed significantly worse than those hospitalized in the pediatric ward in the Coding and Block Design subtests of the WISC-R. In all other measures the Similarities and Picture Completion subtests of the WISC-R, the Bender-Gestalt Copying Test and the Benton Visual Retention Test - the neurosurgical survivors tended to score less well than the pediatric cases, but the difference did not approach statistical significance. The Bender test in particular showed surprisingly little difference between the two groups, with a composite probability of 0.29 that the small difference observed could have resulted from chance. Of the two specialized neuropsychologic tests, the GATB Motor Speed Test showed a nonsignificant trend toward slower execution among the survivors of neurosurgical hospitalization. The BourdonWiersma Vigilance Test gave identical mean results in the two groups. The neurological examination revealed one of the 12 neurosurgical survivors with epilepsy and hemiparesis, another with hyperkinetic attention deficit disorder, and four more with various "soft signs." The 23 pediatric hospitalizations revealed one child with postpoliomyelitic monoparesis and 10 with "soft signs." The neurological examination was completely normal in about half of each sample. Finger graphesthesia and speed of repeated finger-thumb oppositions were slightly but not significantly inferior among the neurosurgical survivors_ In each subject the WISC-R subtest variance was estimated by averaging the four scaled scores and summing the squared deviance of each from this mean_ Subtest variance was not significantly different in the two samples. Omission of the child with hemiparesis and epilepsy from analysis did not alter the findings relating to group differences on the various psychological tests.

Table I Comparison of children with two different degrees of brain insult on four WISC-R subtests and on the Benton Visual Retention Test Test

Similarities Picture Completion Block Design Coding Benton

12 Neurosurgical cases

23 Pediatric hospitalizations

t/DF

p***

9.6 ± 3.4

10.2 ± 2.3*

0.66/32

> 0.25

11.5 ± 2.1

12.1 ± 2.4

0.77/33

> 0.20

10.3 ± 3.8 10.9 ± 3.4**

1.83/33 2.01/31

<0.05 < 0.05

0.0 + 3.6

1.37/32

>0.05

7.9 ± 3.3 8.5 ± 3.1 -0.9 + 2.8

Scores are given as mean ± SD, * 22 subjects, ** 21 subjects, *** one-tailed distribution.

Table 2 Comparison of children with two different

degrees of brain insult on the developmental Bender Test score Age

Parameter

< 11 yrs

N Sigma score (mean ± SD)

;;. 11 yrs

Neurosurgical cases

Pediatric hospitalization

15 8 -1.1±2.3 -1.1 ± 2.3

N Sum of ranks (errors minus expected errors)

4

22

Composite prob ab ility

8 56

U* p**

0

- 0.5 12 0.29

0.29

* Mann-Whitney, ** one-tailed distribution.

DISCUSSION Our main finding is that the 12 neurosurgical survivors examined were significantly inferior on two subtests of the WISC-R to the 23 children with head trauma who were retained for observation in the pediatric ward. By contrast, the neurological examination, the Bender-Gestalt Copying Test and the Benton Visual Retention Test failed to distinguish between these two groups. We think it very likely that the two groups indeed suffered from different severities of cerebral trauma, in view of their different presentations in the emergency room, their different medical dispositions, and their different mortality rates. It seems likely that the neurosurgical survivors as a group also suffered from a small but perceptible degree of residual neurobehavioral dysfunction in view of their different frequencies of various signs and symptoms at follow-up [1]. The fact that seven of the eight psychological subtests reported here demonstrated a trend toward inferior performance of the neurosurgical survivors as a

group, while the eighth showed no trend in either direction (p < 0.01) would seem to support this supposition. It also seems very likely that there were no major social differences between the two groups which could have influenced the test results, in view of their similar distributions of sex, age, ethnic origin, level of parental education and crowding indices. The sample of neurosurgical survivors seems quite representative of those neurosurgical survivors who appeared for follow-up. The indices available show no differences between the subjects who appeared for follow-up examination and those who failed to respond. For these reasons, the findings seem to suggest that the Bender test, the Benton test and the neurological examination are poorer instruments for discriminating between groups with different degrees of old diffuse brain insult than are the Coding and the Block Design subtests of the WISC-R. Since the numbers are small and the questions are many, further study of this possibility is indicated. It should nonetheless be noted that various parameters of the WISC-R, especially the full-scale IQ and the Coding subtest have previously been found to be among the most sensitive indicators of old brain damage [5], while the clinical neurologic examination has not [6, 7]. Similarly, the diagnostic power in brain damage of the Bender test compared with that of the WISC-R has not been proven. In seven published studies of children with and without cerebral dysfunction, controlled for IQ [8-14], the Bender distinguished between the normal and malfunctioning groups only in three [8-10]. It may also be noted that WISC-R subtest scatter has previously been shown to be of limited diagnostic value in children with cerebral dysfunction[15]. Although our findings suggest some detectable group findings of central nervous system dysfunction after a relatively moderate degree of head trauma, it should be emphasized that these findings were of very limited degree in the neurosurgical survivors. Almost all their overall scaled scores were in the normal range, and their school placement and function were similar to that of the children with lesser insult who had been hospitalized on the pediatric ward. Indeed, the very possibility that this degree of brain insult may cause permanent cerebral dysfunction has been disputed [16]. For this reason also, our findings should be confirmed or refuted by independent study.

ACKNOWLEDGMENTS We thank Prof. M. Shalit, formerly Director of Neurosurgery at the Beilinson Hospital, Petah Tikva, and Professors B. Mann and R. Reiss, formerly Directors of Surgery at the Meir Hospital, KfarSava, for permission to study their patients. Tabulation and statistical analysis were kindly performed by the Computer Unit, Occupational Health and Rehabilitation Institute at Loewenstein Hospital. Miriam Gitter helped administer a number of the psy-

Costeff et al: Childhood head trauma 373

chologic examinations. Prof. Nancy Robinson provided valuable criticism.

REFERENCES 1. Horowitz I, Co steff H, Sadan N, Abraham E, Geyer S, Najenson T. Childhood head injuries in Israel. Epidemiology and outcome. 1nt Rehabial Med 1983;5:32-6. 2. Koppitz EM. The Bender-Gestalt test for young children. New York: Grune and Stratton, 1963. 3. Institute of Occupational Health. Behavioral test battery for toxicopsychologic studies. Helsinki: Institute of Occupational Health, 1976. 4. US Government Printing Office. Manual for the GATB screening service. Washington DC: US Government Printing Office, 1959. 5. McMahon SA, Greenberg LM. Serial neurological examination of hyperactive children. Pediatrics 1977;59: 5 84-7. 6. Adams RM, Kocis JJ, Estes RE. Soft neurological signs in learning disabled children and controls. Am J Dis Child 1974; 128:614-8. 7. Gillberg IC. Children with minor neurodevelopmental disorders. III. Neurological and neurodevelopmental problems at age 10. Dev Med Child Neurol 1985;27: 3-16. 8. McConnell 01. Bender-Gestalt scores in relation to organic and emotional problems in children. J Clin Psychol 1967;

374 Brain & Development, Vol 1 0, No 6, 1988

23:370-4. 9. Stavrianos BK. Can projective test measures aid in the detection and differential diagnosis of reading deficits? J Proj Tech Person Assess 1971 ;35 :80-91. 10. Hartlage LC, Lucas DG. Scaled score transformations of Bender Gestalt expectancy levels for young children. Psy· chology in the Schools 1971;8:76-8. 11. Ackerman PT, Peters JD, Dykman RA. Children with specific learning disabilities: II. Bender-Gestalt test findings and other signs. J Learn Disab 1971 ;4:437-46. 12. Friedman J, Strochak RD, Gitlin S, et al. Koppitz Bender scoring system and brain injury in children. J Clin Psychol 1967;23: 179-82. 13. Hayden BS, Talmadge M, Hall M, et aI.- Diagnosing minimal brain damage in children: a comparison of two scoring systems. Merrill-Palmer Quarterly of Behavior and Development 1970;16:278-85. 14. Routh DK, Roberts RD. Minimal brain dysfunction in children: failure to find evidence for a behavioral syndrome. Psychol Rep 1972;31:307-14. 15. Ackerman PT, Peters JD, Dykman RA. Children with specific learning disabilities: I. WISC profiles. J Learn Disab 1971;4:150-66. 16. Chadwick 0, Rutter M, Brown G, Schaffer D, Traub M. A prospective study of children with head injuries, II. Cognitive sequelae. PsycholMed 1981;11:49-61.