Jouirnal of Neurology, Neurosurgery, and Psychiatry, 1977, 40, 1211-1219
Head injuries in children: a prospective five year follow-up H. KLONOFF, M. D. LOW, AND C. CLARK From the Department ofPsychiatry and Medicine (Neurology), University of British Columbia, Vancouver, British Columbia, Canada
A five year follow-up study was conducted with two groups of head-injured children, younger than 9 years old at time of injury and 100 older than 9 years. The four aspects studied were neuropsychological function, neurological status, EEG status, and school progress. There was an extended recovery process over time, as well as evidence of a differential rate of recovery for the SUMMARY
131
four aspects measured. Our understanding of the natural history of head (Klonoff et al., 1969; Klonoff, 1971; Klonoff and injuries in children can best be increased by pro- Low, 1974; Klonoff and Paris, 1974); this publication spective, long-term, multidimensional studies. The deals with the long-term effects, with particular paucity of such clinical studies has been commented emphasis on predicting outcome. on (Black et al., 1971; Lishman, 1973; Klonoff and Paris, 1974). A study which would add to our knowl- Method edge of the head-injured child should be broad in scope. Firstly, it should include: antecedentfactors- PROJECT SAMPLE constitutional predisposition, premorbid personality, The clinical sample consisted of children aged age, sex, environmental hazards; circumstances at between 2.7 and 15.9 years at time of head injury. time of head injury-the nature and extent of injury, The children were examined during initial hospital resilience of the neural apparatus in childhood, the admission and each year for five consecutive years nature of intervention and management during the after head injury. The sample comprised consecutive acute and postacute phases; and consequence admissions with head injury to either the Health factors-the effect of brain damage on general Centre for Children or St Paul's Hospital in Vanadaptation and maturation, on the development of couver, British Columbia. The sample was divided post-traumatic epilepsy, effect on education, trans- into two groups: younger children (less than 9 years action with family, and the role of compensation and old)-initially 131 children, after five years 78; and litigation. Secondly, it should encompass a variety of older children (more than 9 years old)-initially 100, clinical examinations including neurological, electro- after five years 39. Initially for the younger group, encephalographic, and psychological. Thirdly, the there were 80 males and 51 females with a mean age of study should address itself to the interaction of 5.87 years (SD=-1.76) and a 52 % skull fracture rate. effects, namely, the differentiation of immediate from For the older group, there were 67 males and 33 short-term or long-term effects, and the nature and females with a mean age of 11.53 years (SD==1.81) and 32o% fractures. Within the two groups, no course of reconstitution. This study accordingly set out to investigate discernible differences were found on these variables prospectively a head-injured group of children from between the original and the followed groups. As the time of trauma (hospitalisation) to the fifth year previous studies (Klonoff et al., 1969; Klonoff and after trauma, within the context of the model noted Low, 1974) have indicated that with the examination above. Previous publications have reported our procedures outlined below, there are no sex differmethods, and the immediate and short-term effects ences, sex was not included in the analysis. Address for reprint requests: Dr H. Klonoff, Division of Psychology. Department of Psychiatry, University of British Columbia, 2255 Wesbrook Crescent, Vancouver, BC, Canada, V6T IW5. Accepted 23 June 1977
EXAMINATION PROCEDURE
Neuropsychological test battery The neuropsychological examination included the 1211
1212 Reitan-Indiana Neuropsychological Test Battery for Children, two of Benton's tests, a lateral dominance test, the Stanford Binet, form L-M for children under 5 years of age, or the Wechsler Intelligence Scale for Children (WISC) for those aged over 5 years, and for children over 9 years the Kl0ve Motor Steadiness Battery. Thus, for the younger group complete data for the six examinations were obtained on 18 tests or 32 variables, while for the older group on 23 tests or 48 variables. Test construction, scoring, and normative data have been reported by Benton (1959), Knights (1966), Klonoff et al. (1969), Klonoff and Low (1974), and Reitan (1974). For the neuropsychological test battery, each subject was matched on age (within three months) and sex with a normal control. These control subjects referred by six paediatricians in the Vancouver area were classed as normal by four criteria: no neurological deficit; no physical anomalies; no profound signs of emotional disturbance; and normal school progress for those children attending school. The two head-injured groups and their matched controls were compared using an analysis of variance and orthogonal contrasts for each variable at each yearly examination.
H. Klonoji M. D. Low, and C. Clark
scalp electrodes plus two ear electrodes in the International 10-20 positions. Both referential (ear) and bipolar montages were used, awake and asleep. All tracings were interpreted by the same electroencephalographer (MDL). Besides describing the qualitative features of the records, he assigned each a global rating of normal (1), borderline (1.5), minimally abnormal (2), moderately abnormal (3), or markedly abnormal (4). The criteria for this are fully described by Klonoff and Low (1974). These data were analysed using children who were seen on six occasions and then the followed and original groups were compared. PREDICTION FOR RESIDUAL EFFECTS
For each of the three examinations additional analyses were done with variables obtained during initial hospital admission, in order to ascertain if a model of prediction could be developed. Specifically, these variables were: Full Scale IQ, age (in years), unconsciousness (duration in minutes), retrograde amnesia (duration in minutes), anterograde amnesia (duration in minutes), gestation period (in weeks), EEG rating, and neurological rating. Results
Neurological examination During initial hospital admission, a neurological assessment of degree of head injury was assigned to each child based on the following schema: minor (1)-suspected but not proven loss of consciousness, no evidence of concussion; mild (2)-suspected but not proven loss of consciousness, concussion (lethargy, vomiting, drowsiness, contusion, nausea, dizziness); moderate (3)-loss of consciousness for less than 5 min, concussion; severe (4)-loss of consciousness for 5-30 min, concussion or skull fracture; and serious (5)-loss of consciousness for more than 30 min, concussion or skull fracture (depressed or compound or both), other sequelae (psychosis, aphasia, etc.). For the subsequent assessments, each child was examined for: complaints resulting from the head injury (for example, headaches); sequelae (that is, neurological, personality, subjective, or learning); and school progress. Again these data were analysed with respect to the two cohort groups (younger and older) and these groups were compared with the original samples.
NEUROPSYCHOLOGICAL FINDINGS
Tables 1 and 2 summarise the results of the analyses of variance on the neuropsychological variables for the younger and older head-injured groups and their respective matched controls. The 'between-groups' difference column indicates whether the two groups are different over all years combined. Of the possible 32 F values, 29 were significant for the younger group, while for the older group 43 of the 48 F values were significant. The 'between-year' differences show whether there is a change in performance with time regardless of membership in either the head-injured or control groups. For the younger group, 31 of these differences were significant, while for the older group 43 were significant. These findings indicated improved performance with repeated exposure to the test battery (learning) or increasing age (maturation), or both. The 'groups x year' interaction term indicates whether there are differential rates of learning or performance between the two groups with repeated exposure to the test battery. For the younger group, 18 of these tests were significant, while for the older EEG group 21 were significant. Initial and repeat EEG records were obtained from In the six remaining columns (that is, 'contrasts') each head-injured child during initial hospital the results for the yearly group contrasts are recorded. admission and then once a year for five consecutive These contrasts indicated whether the head-injured years. Hyperventilation for 3 min and sleep (induced were different at specific examinations compared with by chloral hydrate) were used routinely for activation. the normal groups. For the younger group, the total Recording techniques included the application of 19 numbers of significant differences were as follows:
1213 Head injuries in children: a prospective five year follow-uip11 Table 1 Comparison of younger head-injured patients with matched controls on neuropsychological variables-analysis of variance Contrasts
Groups Variables
1 Category test (err.) 2 Tapping-dom. (corr.)
Tapping-non-dom. (corr.) 3 Speech perception (err.) 4 Trail making-A (time) Trail making-B (time) Trail making-total (time) 5 Matching figures (time) 6 7 8 9 10 11 12
13
Matching figures (err.) Matching Vs (time) Matching Vs (err.) Star (corr.) Concentric squares (corr.) Progressive figures (time) Colour form (time) Colour form (err.) Target test (corr.) Marching test-dom. (time) Marching test-dom. (err.) Marching test-non-dom. (time) Marching test-non-dom. (err.) TPT-dom. (time) TPT-non-dom. (time) TPT-both (time) TPT-total (time) TPT-memory (corr.) TPT-location (corr.) Matching pictures (corr.) Sound recognition Right left orientation (corr.) Lateral dominance (corr.)
14 15 16 17 1 8 FullIIQ
Betweengroup
Between
x years
interdifference difference action n* n* n*
26.625,
3.39 5.56 §
31.04¶, 33.505T
23.785, 38.405 4.581:
4.931: 10.9411 19.435
13.0551 15.345~ 38.855, 15.165F1
24.405~ 19.97Th
11.7751 8.1611
13.4451
29.035~ 9.2511 2.59 13.611
13.875I 2.72 11.41¶T 11.0711 34.735,
19.125~
7.2611 194. 105J.
Total number of significant differences: 29
years
1.87
183.535,
158.3159_ 91.715 75.40¶ 51.17T
72.2051 69.65¶ 52.295
64.695I
81.145' 53.675 84.2 1¶
50.6551
67.125 40.955 129.97'T 97.3 1¶
38.045~ 107.205~
33.26T,
1 13.675T
116.185,
120.915J 146.195i 106.025,
82.365. 99. 105~
Time of
One
Two
Three
Four
Five
trauma
year
years
years
years
years
later n=113
later
later
later
later
n =97
n =86
n =78
n=7S
~~n=J31
5.141:
7.7611
3.9811
6.6311
3.871:
0.71 1.49
0.02 1.95
30.795
20.175-,
17.45¶
0.81 1.73 0.80 2.64
24.765,
4.11It
0.23 0.10 0.61 1.29 5.291: 5.80 § 0.00 0.06 0.12 0.08 0.17 0.17 0.32 0.14 0.02 0.75 0.34 0.21 0.45 1.58 0.05 0.00 0.05 0.03 0.19 0.01 1.37 1.22 2.42 1.09
0.95 0.01 0.01 1.07 1.22 0.56 1.29 0.00 0.05 0.26 1.12 0.28 0.04 0.11 0.04 0.04 0.93 0.33 0.01 0.38 0.65 0.11 0.02 0.06 0.00 0.00 0.12 0.89 1.44 0.01 0.31
23.23¶
19.0251
0.35 0.81 0.44
4.971: 7.0511
3.2011 1.30 3.5011 3.1611 1.56 1.79 2.431:
2.401: 3.9011 7.0 35T
8.53T5
3.3011 3.7511 1.01 1.70 3.43 1.41
4.30¶I 8.555T 2.301: 12.055q
14.855~ 0.39 1.02 0.79 3.8811 1.71
11.45
1
4.60t
7.6711 13.295
7.0011
3.991:
20.101:
2.86 3.00
9.7211 31.265~
20.93'lu
50.82¶~ 24.815T
2.78
7.3311
4.711:
15.235T
7.2411 5.34 §
24.475~ 22.67511
17.55
2.32 2.93 2.89 5.84 §
48.2951
53.425~ 15.855~
83.585,
95.1 1¶T 2.86 9.9511
10.5411 45.22¶~ 17.19Uv
35.835 0.40 0.66 1.15 5.38 § 0.09 6.35 §
66.955T
11.635~ 22.605~ 4.671:
26.865~ 16.885~
3.98t
4.801: 0.36 0.20
4.051: 0.01
4.471: 3.921: 13.46¶ 7.9911
1.68
3.931: 0.13 3.68 3.31 1.13 0.03 0.25 1.23 1.21 1.40 0.78 2.89
5.121:
0.00 0.00 3.25 0.07 0.11 0.03 0.26 0.07 0.51 1.57 0.13 0.01 0.42 1.11 0.07 0.00 0.19 0.01 1.02
3.811:
0.50
0.93
0.16 1.40 0.61 0.22
14.035i
0.84
67.595
52.905,
28.215T
25.805J
31
18
28
20
13
40.245~
35.095~ 22.875~
2.471:
19.665,
5.171:
5
4
1
*The total number used was 580. In the analysis of variance, comparisons between groups are based on the respective weights of the groups. tContrast deleted as a number of subjects were too young to complete the test. §P < 0.02 1:P < 0.05 IlI < 0.01 5,P < 0.001.
28 during initial hospital admission; 20 one year later; 13 two years later; five three years later; four four years later; and one five years later. For the older group, the number of significant differences for specific examinations were as follows: 42 during initial hospital admission; 31 one year later; 15 two years later; 12 three years later; eight four years later; and six five years later. To supplement the group data, a further analysis was done to determine if there existed a subgroup within the head injury groups which was different both from the normal controls and from the majority of the head-injured children. For both the younger and older groups, the within-group variances for each variable (IQ excluded) were compared in order to determine if there were any differences in the specific variable distributions between the normal children and head-injured groups on the last re-examination.
From this analysis, 11I variables had a ratio of head injured to normal variance greater than three. Each head-injured and normal child was assigned a score for these variables based on the scores of a
normal sample of 158 children who had been classed by age and had done the test battery six times. The scores were assigned in terms of standard deviations above and below the mean (that is, O==±1I SD, ±1I= ±1I to ± 2 SD, ± 2==greater than ±2 SD). Thus, the resulting scoring schema had a range from -22 to 22. If the head-injured and normal groups were truly equivalent, the resulting distributions of scores for each group should be more or less the same, with a mean near zero. These results (Table 3) clearly indicate that there was a subsample of head-injured children who were different both from the normals and from the majority of headinjured children. Empirically, 23.700 of the head-
H. Klonoft, M. D. Low, and C. Clark
1214
Table 2 Comiparison of older head-injured patients with matched conitrols on neuropsychological variables-analysis of variance Groups Variables
I Category test (err.) 2 Tapping-dom. (corr.)
Tapping-non-dom. (corr.) 3 Speech perception (err.) 4 Trail making-A (time) Trail making-B (time) Trail making-total (time) S Matching figures (time) 6 7 8 9 10 11 12
13
14 15 16 17 18 19
Matching figures (err.) Matching Vs (time) Matching Vs (err.) Star (corr.) Concentric squares (corr.) Progressive figures (time) Colour form (title) Colour form (err.) Target test (corr.) Marching test-dom. (time) Marching test-dom. (err.) Marching test-non-dom. (time) Marching test-non-dam. (err.) TPT-dom. (time) TPT-non-dom. (time) TPT-both (time) TPT-total (time) TPT-memory (corr.) TPT-location (corr.) Matching pictures (corr.) Sound recognition Right left orientation (corr.) Lateral dominance (corr.) Full IQ Maze coordination dam. (time) Maze coordination dom. (counter) Maze coordination non-dom. (time)
Between group
x years interdifference difference action n* n* n*
36.82:1
24.1011 30.53'11 10.51
22.671!1
10.84§ 16.48!! 9.49 §
3.00 15.85!! 4.171 1.15 1.14 24.88!! 18.40!! 3.74
20.48!! 34.29!! 13.16!! 29.33!! 1.99 10.12 § 13.32!! 19.62!! 178! 15.87 14.65q 7.05 § 27.89!!
5.011t
34.52!! 98.47!! 55.13!! 61.64!! 57.63!!
Contrasts
Between
years
73.991! 64.7811 70.181!1 31.33:! 74.2911 42.171!
60.75! 14.14!! 1.43 19.87:! 5.82!! 0.78 3.21 § 42.09!! 28.82!! 4.27!! 29.03!! 62.47!! 0.92 74.32!! 0.48 46.29!! 47.77!! 39.48!! 62.14!! 32.78!! 24.30!! 8.98!! 36.50!! 3.40 § 1.11 18.33!! 57.54!! 79.13!! 77.45!!
Four
Fire
One
year later years later years later years later years later n=39 n=SO n =S6 n=80 n=6S
13.421!
0.46 0.58 0.22 4.34,1 0.58
9.77 § 8.90 § 36.91!!
2.38t
24.2111
1.96 0.79 0.61 1.48 0.55 0.85 1.89 2.02 6.54!! 3.77 § 0.97 3.31 § 0.93 2.991; 0.52 1.44 1.87 4.86!! 2.251 3.36 §
Three
Time of trauma
n=JOO
11.51!! 24.74!! 10.07!!
4.92t
15.13!! 6.69 §
4.10t
11.58!! 31.99!! 40.43!! 14.98!! 14.79!! 45.23!!
6.281; 40.59!! 1.45 15.46!! 22.13!!
48.271! 27.10!!
30.26!!1
15.59!! 13.80!! 9.86 § 2.64 8.43 §
5.371; 7.62 § 2.93 2.04 12.31!! 0.85 0.03 0.00
Two
8.59 §
5.25t
8.52 § 0.77 8.35 § 0.47 1.67
3.79t
2.99 2.20 2.83 2.40 2.50 0.57 1.15 0.06 0.24 0.41 0.17 0.71 0.21 1.11 0.55 2.20 1.61 1.60 0.22 1.74 0.32 1.25 0.63 1.80 15 0.37 0.07 1.15 1.76 2.13 4.401 15.45!! 5.621: 8.85 §
2.59 1.44 4.941 0.02 1.50 0.50 0.94 0.40 0.07 1.66 0.48 0.61 0.15 1.48 0.06 0.00 4.06 0.90 3.881 1.95 0.13 0.13 0.35 0.17 .9 0.74 1.84 0.02 1.47 0.02 8.84 § 9.95 § 0.99 2.09 2.35
3.55 1.31 1.70 0.32 0.47 0.00 0.08 0.45 0.34 0.04 0.45 0.02 0.08 2.24 0.00 0.00 0.05 0.66 1.94 0.59 0.47 0.30 0.72 1.25 06 0.03 0.00 0.53 0.01 0.08
3.79t
1.96 4.501 5.701 1.12 1.35 1.13 1.08 1.36 1.44 1.75 1.35
8.77!! 12.81!! 14.71!! 1.70 13.95!! 2.06 10.63 20.42!! 4.46!! 4.231 7.27 §
0.04 2.28 0.05 1.22 0.38 2.87 3.48 0.21 2.81 8.17 § 2.99 2.34 0.91 0.22 0.62 0.11 0.97 1.29 0.04 1.68 0.44 2.73 2.67 § 16.88!! 10.22 § 16.73!! 8.41 §
5.671; 4.531 0.15 0.02
13.86!! 9.00 § 4.291 4.621
8.77 § 2.59
2.16
3.68 4.521; 1.13
0.66
1.24 4.68t 11.97;! 1.75 8.26 § 2.03 2.63 6.70 §
0.72 0.34 0.78 0.93 2.57 1.40 3.43 6.441;
6.27t
6.08t
12
8
5.411; 27.87!1 10.97 § 10.16!! 13.09! 5.011~ 12.87,! 2.96 5.941; 5.81
5.29t
3.861; 0.72 6.89!! 0.41 0.37 0.98 13.05!! 8.27!! 10.86!!
28.91!! 8.89 § 66.78!! 1.59 7.00 § 54.51 136.49!! 101.73!! 124.53!!
6.74!! 0.22 0.81 0.65
96.19! 8.92 §
4.78!! 6.93 84.39!! 83.98!!
2.271 0.52 1.12 1.90 0.62 0.08 0.52 0.65 0.57
16.24'! 9.45 § 2.66 28.30!! 3.04 1.90 12.17!! 21.62!! 21.49!!
0.47 3.17 0.62 2.45 1.33 11.67!! 9.44 §
0.58 7.75 § 1.33 1.98 1.70 3.53 1.80 2.13 3.32
43
16
42
31
15
5.651;
4.45t
7.39 § 0.43 0.90 0.44
Maze coordination non-dom.
(counter)
66.89!!
29.68!! 20 Grooved steadiness dom. (time) Grooved steadiness dom. (counter) 21.88!! Grooved steadiness non-dom. (time) 7.07 § Grooved steadiness non-dam. 7.33 § (counter) 20.44!! 21 Steadiness dam. (time) 12.74!! Steadiness dam. (counter) 19.49!! Steadiness non-dam. (time) 14.85!! Steadiness non-dam. (counter) 12.28!! 22 Grooved pegboard dam. (time) Grooved pegboard non-dom. (time) 16.90!! 42.43!! 23 Foot tapping dam. (score) 47.49!! Foot tapping non-dam. (score) Total number of significant differences: 43
98.16!! 25.53!! 27.07!!
14.76:!1 6.26!! 20.70!! 20.04!! 24.43!! 34.36!!
5.961; 0.21
4.79t
*The total number used was 390. In the analysis of variance, comparisons between groups are based on the !p < 0.001 §P 9yr 9yr
Head injuries in children: a prospective five year follow-up H. KLONOFF, M. D. LOW, AND C. CLARK From the Department ofPsychiatry and Medicine (Neurology), University of British Columbia, Vancouver, British Columbia, Canada
A five year follow-up study was conducted with two groups of head-injured children, younger than 9 years old at time of injury and 100 older than 9 years. The four aspects studied were neuropsychological function, neurological status, EEG status, and school progress. There was an extended recovery process over time, as well as evidence of a differential rate of recovery for the SUMMARY
131
four aspects measured. Our understanding of the natural history of head (Klonoff et al., 1969; Klonoff, 1971; Klonoff and injuries in children can best be increased by pro- Low, 1974; Klonoff and Paris, 1974); this publication spective, long-term, multidimensional studies. The deals with the long-term effects, with particular paucity of such clinical studies has been commented emphasis on predicting outcome. on (Black et al., 1971; Lishman, 1973; Klonoff and Paris, 1974). A study which would add to our knowl- Method edge of the head-injured child should be broad in scope. Firstly, it should include: antecedentfactors- PROJECT SAMPLE constitutional predisposition, premorbid personality, The clinical sample consisted of children aged age, sex, environmental hazards; circumstances at between 2.7 and 15.9 years at time of head injury. time of head injury-the nature and extent of injury, The children were examined during initial hospital resilience of the neural apparatus in childhood, the admission and each year for five consecutive years nature of intervention and management during the after head injury. The sample comprised consecutive acute and postacute phases; and consequence admissions with head injury to either the Health factors-the effect of brain damage on general Centre for Children or St Paul's Hospital in Vanadaptation and maturation, on the development of couver, British Columbia. The sample was divided post-traumatic epilepsy, effect on education, trans- into two groups: younger children (less than 9 years action with family, and the role of compensation and old)-initially 131 children, after five years 78; and litigation. Secondly, it should encompass a variety of older children (more than 9 years old)-initially 100, clinical examinations including neurological, electro- after five years 39. Initially for the younger group, encephalographic, and psychological. Thirdly, the there were 80 males and 51 females with a mean age of study should address itself to the interaction of 5.87 years (SD=-1.76) and a 52 % skull fracture rate. effects, namely, the differentiation of immediate from For the older group, there were 67 males and 33 short-term or long-term effects, and the nature and females with a mean age of 11.53 years (SD==1.81) and 32o% fractures. Within the two groups, no course of reconstitution. This study accordingly set out to investigate discernible differences were found on these variables prospectively a head-injured group of children from between the original and the followed groups. As the time of trauma (hospitalisation) to the fifth year previous studies (Klonoff et al., 1969; Klonoff and after trauma, within the context of the model noted Low, 1974) have indicated that with the examination above. Previous publications have reported our procedures outlined below, there are no sex differmethods, and the immediate and short-term effects ences, sex was not included in the analysis. Address for reprint requests: Dr H. Klonoff, Division of Psychology. Department of Psychiatry, University of British Columbia, 2255 Wesbrook Crescent, Vancouver, BC, Canada, V6T IW5. Accepted 23 June 1977
EXAMINATION PROCEDURE
Neuropsychological test battery The neuropsychological examination included the 1211
1212 Reitan-Indiana Neuropsychological Test Battery for Children, two of Benton's tests, a lateral dominance test, the Stanford Binet, form L-M for children under 5 years of age, or the Wechsler Intelligence Scale for Children (WISC) for those aged over 5 years, and for children over 9 years the Kl0ve Motor Steadiness Battery. Thus, for the younger group complete data for the six examinations were obtained on 18 tests or 32 variables, while for the older group on 23 tests or 48 variables. Test construction, scoring, and normative data have been reported by Benton (1959), Knights (1966), Klonoff et al. (1969), Klonoff and Low (1974), and Reitan (1974). For the neuropsychological test battery, each subject was matched on age (within three months) and sex with a normal control. These control subjects referred by six paediatricians in the Vancouver area were classed as normal by four criteria: no neurological deficit; no physical anomalies; no profound signs of emotional disturbance; and normal school progress for those children attending school. The two head-injured groups and their matched controls were compared using an analysis of variance and orthogonal contrasts for each variable at each yearly examination.
H. Klonoji M. D. Low, and C. Clark
scalp electrodes plus two ear electrodes in the International 10-20 positions. Both referential (ear) and bipolar montages were used, awake and asleep. All tracings were interpreted by the same electroencephalographer (MDL). Besides describing the qualitative features of the records, he assigned each a global rating of normal (1), borderline (1.5), minimally abnormal (2), moderately abnormal (3), or markedly abnormal (4). The criteria for this are fully described by Klonoff and Low (1974). These data were analysed using children who were seen on six occasions and then the followed and original groups were compared. PREDICTION FOR RESIDUAL EFFECTS
For each of the three examinations additional analyses were done with variables obtained during initial hospital admission, in order to ascertain if a model of prediction could be developed. Specifically, these variables were: Full Scale IQ, age (in years), unconsciousness (duration in minutes), retrograde amnesia (duration in minutes), anterograde amnesia (duration in minutes), gestation period (in weeks), EEG rating, and neurological rating. Results
Neurological examination During initial hospital admission, a neurological assessment of degree of head injury was assigned to each child based on the following schema: minor (1)-suspected but not proven loss of consciousness, no evidence of concussion; mild (2)-suspected but not proven loss of consciousness, concussion (lethargy, vomiting, drowsiness, contusion, nausea, dizziness); moderate (3)-loss of consciousness for less than 5 min, concussion; severe (4)-loss of consciousness for 5-30 min, concussion or skull fracture; and serious (5)-loss of consciousness for more than 30 min, concussion or skull fracture (depressed or compound or both), other sequelae (psychosis, aphasia, etc.). For the subsequent assessments, each child was examined for: complaints resulting from the head injury (for example, headaches); sequelae (that is, neurological, personality, subjective, or learning); and school progress. Again these data were analysed with respect to the two cohort groups (younger and older) and these groups were compared with the original samples.
NEUROPSYCHOLOGICAL FINDINGS
Tables 1 and 2 summarise the results of the analyses of variance on the neuropsychological variables for the younger and older head-injured groups and their respective matched controls. The 'between-groups' difference column indicates whether the two groups are different over all years combined. Of the possible 32 F values, 29 were significant for the younger group, while for the older group 43 of the 48 F values were significant. The 'between-year' differences show whether there is a change in performance with time regardless of membership in either the head-injured or control groups. For the younger group, 31 of these differences were significant, while for the older group 43 were significant. These findings indicated improved performance with repeated exposure to the test battery (learning) or increasing age (maturation), or both. The 'groups x year' interaction term indicates whether there are differential rates of learning or performance between the two groups with repeated exposure to the test battery. For the younger group, 18 of these tests were significant, while for the older EEG group 21 were significant. Initial and repeat EEG records were obtained from In the six remaining columns (that is, 'contrasts') each head-injured child during initial hospital the results for the yearly group contrasts are recorded. admission and then once a year for five consecutive These contrasts indicated whether the head-injured years. Hyperventilation for 3 min and sleep (induced were different at specific examinations compared with by chloral hydrate) were used routinely for activation. the normal groups. For the younger group, the total Recording techniques included the application of 19 numbers of significant differences were as follows:
1213 Head injuries in children: a prospective five year follow-uip11 Table 1 Comparison of younger head-injured patients with matched controls on neuropsychological variables-analysis of variance Contrasts
Groups Variables
1 Category test (err.) 2 Tapping-dom. (corr.)
Tapping-non-dom. (corr.) 3 Speech perception (err.) 4 Trail making-A (time) Trail making-B (time) Trail making-total (time) 5 Matching figures (time) 6 7 8 9 10 11 12
13
Matching figures (err.) Matching Vs (time) Matching Vs (err.) Star (corr.) Concentric squares (corr.) Progressive figures (time) Colour form (time) Colour form (err.) Target test (corr.) Marching test-dom. (time) Marching test-dom. (err.) Marching test-non-dom. (time) Marching test-non-dom. (err.) TPT-dom. (time) TPT-non-dom. (time) TPT-both (time) TPT-total (time) TPT-memory (corr.) TPT-location (corr.) Matching pictures (corr.) Sound recognition Right left orientation (corr.) Lateral dominance (corr.)
14 15 16 17 1 8 FullIIQ
Betweengroup
Between
x years
interdifference difference action n* n* n*
26.625,
3.39 5.56 §
31.04¶, 33.505T
23.785, 38.405 4.581:
4.931: 10.9411 19.435
13.0551 15.345~ 38.855, 15.165F1
24.405~ 19.97Th
11.7751 8.1611
13.4451
29.035~ 9.2511 2.59 13.611
13.875I 2.72 11.41¶T 11.0711 34.735,
19.125~
7.2611 194. 105J.
Total number of significant differences: 29
years
1.87
183.535,
158.3159_ 91.715 75.40¶ 51.17T
72.2051 69.65¶ 52.295
64.695I
81.145' 53.675 84.2 1¶
50.6551
67.125 40.955 129.97'T 97.3 1¶
38.045~ 107.205~
33.26T,
1 13.675T
116.185,
120.915J 146.195i 106.025,
82.365. 99. 105~
Time of
One
Two
Three
Four
Five
trauma
year
years
years
years
years
later n=113
later
later
later
later
n =97
n =86
n =78
n=7S
~~n=J31
5.141:
7.7611
3.9811
6.6311
3.871:
0.71 1.49
0.02 1.95
30.795
20.175-,
17.45¶
0.81 1.73 0.80 2.64
24.765,
4.11It
0.23 0.10 0.61 1.29 5.291: 5.80 § 0.00 0.06 0.12 0.08 0.17 0.17 0.32 0.14 0.02 0.75 0.34 0.21 0.45 1.58 0.05 0.00 0.05 0.03 0.19 0.01 1.37 1.22 2.42 1.09
0.95 0.01 0.01 1.07 1.22 0.56 1.29 0.00 0.05 0.26 1.12 0.28 0.04 0.11 0.04 0.04 0.93 0.33 0.01 0.38 0.65 0.11 0.02 0.06 0.00 0.00 0.12 0.89 1.44 0.01 0.31
23.23¶
19.0251
0.35 0.81 0.44
4.971: 7.0511
3.2011 1.30 3.5011 3.1611 1.56 1.79 2.431:
2.401: 3.9011 7.0 35T
8.53T5
3.3011 3.7511 1.01 1.70 3.43 1.41
4.30¶I 8.555T 2.301: 12.055q
14.855~ 0.39 1.02 0.79 3.8811 1.71
11.45
1
4.60t
7.6711 13.295
7.0011
3.991:
20.101:
2.86 3.00
9.7211 31.265~
20.93'lu
50.82¶~ 24.815T
2.78
7.3311
4.711:
15.235T
7.2411 5.34 §
24.475~ 22.67511
17.55
2.32 2.93 2.89 5.84 §
48.2951
53.425~ 15.855~
83.585,
95.1 1¶T 2.86 9.9511
10.5411 45.22¶~ 17.19Uv
35.835 0.40 0.66 1.15 5.38 § 0.09 6.35 §
66.955T
11.635~ 22.605~ 4.671:
26.865~ 16.885~
3.98t
4.801: 0.36 0.20
4.051: 0.01
4.471: 3.921: 13.46¶ 7.9911
1.68
3.931: 0.13 3.68 3.31 1.13 0.03 0.25 1.23 1.21 1.40 0.78 2.89
5.121:
0.00 0.00 3.25 0.07 0.11 0.03 0.26 0.07 0.51 1.57 0.13 0.01 0.42 1.11 0.07 0.00 0.19 0.01 1.02
3.811:
0.50
0.93
0.16 1.40 0.61 0.22
14.035i
0.84
67.595
52.905,
28.215T
25.805J
31
18
28
20
13
40.245~
35.095~ 22.875~
2.471:
19.665,
5.171:
5
4
1
*The total number used was 580. In the analysis of variance, comparisons between groups are based on the respective weights of the groups. tContrast deleted as a number of subjects were too young to complete the test. §P < 0.02 1:P < 0.05 IlI < 0.01 5,P < 0.001.
28 during initial hospital admission; 20 one year later; 13 two years later; five three years later; four four years later; and one five years later. For the older group, the number of significant differences for specific examinations were as follows: 42 during initial hospital admission; 31 one year later; 15 two years later; 12 three years later; eight four years later; and six five years later. To supplement the group data, a further analysis was done to determine if there existed a subgroup within the head injury groups which was different both from the normal controls and from the majority of the head-injured children. For both the younger and older groups, the within-group variances for each variable (IQ excluded) were compared in order to determine if there were any differences in the specific variable distributions between the normal children and head-injured groups on the last re-examination.
From this analysis, 11I variables had a ratio of head injured to normal variance greater than three. Each head-injured and normal child was assigned a score for these variables based on the scores of a
normal sample of 158 children who had been classed by age and had done the test battery six times. The scores were assigned in terms of standard deviations above and below the mean (that is, O==±1I SD, ±1I= ±1I to ± 2 SD, ± 2==greater than ±2 SD). Thus, the resulting scoring schema had a range from -22 to 22. If the head-injured and normal groups were truly equivalent, the resulting distributions of scores for each group should be more or less the same, with a mean near zero. These results (Table 3) clearly indicate that there was a subsample of head-injured children who were different both from the normals and from the majority of headinjured children. Empirically, 23.700 of the head-
H. Klonoft, M. D. Low, and C. Clark
1214
Table 2 Comiparison of older head-injured patients with matched conitrols on neuropsychological variables-analysis of variance Groups Variables
I Category test (err.) 2 Tapping-dom. (corr.)
Tapping-non-dom. (corr.) 3 Speech perception (err.) 4 Trail making-A (time) Trail making-B (time) Trail making-total (time) S Matching figures (time) 6 7 8 9 10 11 12
13
14 15 16 17 18 19
Matching figures (err.) Matching Vs (time) Matching Vs (err.) Star (corr.) Concentric squares (corr.) Progressive figures (time) Colour form (title) Colour form (err.) Target test (corr.) Marching test-dom. (time) Marching test-dom. (err.) Marching test-non-dom. (time) Marching test-non-dam. (err.) TPT-dom. (time) TPT-non-dom. (time) TPT-both (time) TPT-total (time) TPT-memory (corr.) TPT-location (corr.) Matching pictures (corr.) Sound recognition Right left orientation (corr.) Lateral dominance (corr.) Full IQ Maze coordination dam. (time) Maze coordination dom. (counter) Maze coordination non-dom. (time)
Between group
x years interdifference difference action n* n* n*
36.82:1
24.1011 30.53'11 10.51
22.671!1
10.84§ 16.48!! 9.49 §
3.00 15.85!! 4.171 1.15 1.14 24.88!! 18.40!! 3.74
20.48!! 34.29!! 13.16!! 29.33!! 1.99 10.12 § 13.32!! 19.62!! 178! 15.87 14.65q 7.05 § 27.89!!
5.011t
34.52!! 98.47!! 55.13!! 61.64!! 57.63!!
Contrasts
Between
years
73.991! 64.7811 70.181!1 31.33:! 74.2911 42.171!
60.75! 14.14!! 1.43 19.87:! 5.82!! 0.78 3.21 § 42.09!! 28.82!! 4.27!! 29.03!! 62.47!! 0.92 74.32!! 0.48 46.29!! 47.77!! 39.48!! 62.14!! 32.78!! 24.30!! 8.98!! 36.50!! 3.40 § 1.11 18.33!! 57.54!! 79.13!! 77.45!!
Four
Fire
One
year later years later years later years later years later n=39 n=SO n =S6 n=80 n=6S
13.421!
0.46 0.58 0.22 4.34,1 0.58
9.77 § 8.90 § 36.91!!
2.38t
24.2111
1.96 0.79 0.61 1.48 0.55 0.85 1.89 2.02 6.54!! 3.77 § 0.97 3.31 § 0.93 2.991; 0.52 1.44 1.87 4.86!! 2.251 3.36 §
Three
Time of trauma
n=JOO
11.51!! 24.74!! 10.07!!
4.92t
15.13!! 6.69 §
4.10t
11.58!! 31.99!! 40.43!! 14.98!! 14.79!! 45.23!!
6.281; 40.59!! 1.45 15.46!! 22.13!!
48.271! 27.10!!
30.26!!1
15.59!! 13.80!! 9.86 § 2.64 8.43 §
5.371; 7.62 § 2.93 2.04 12.31!! 0.85 0.03 0.00
Two
8.59 §
5.25t
8.52 § 0.77 8.35 § 0.47 1.67
3.79t
2.99 2.20 2.83 2.40 2.50 0.57 1.15 0.06 0.24 0.41 0.17 0.71 0.21 1.11 0.55 2.20 1.61 1.60 0.22 1.74 0.32 1.25 0.63 1.80 15 0.37 0.07 1.15 1.76 2.13 4.401 15.45!! 5.621: 8.85 §
2.59 1.44 4.941 0.02 1.50 0.50 0.94 0.40 0.07 1.66 0.48 0.61 0.15 1.48 0.06 0.00 4.06 0.90 3.881 1.95 0.13 0.13 0.35 0.17 .9 0.74 1.84 0.02 1.47 0.02 8.84 § 9.95 § 0.99 2.09 2.35
3.55 1.31 1.70 0.32 0.47 0.00 0.08 0.45 0.34 0.04 0.45 0.02 0.08 2.24 0.00 0.00 0.05 0.66 1.94 0.59 0.47 0.30 0.72 1.25 06 0.03 0.00 0.53 0.01 0.08
3.79t
1.96 4.501 5.701 1.12 1.35 1.13 1.08 1.36 1.44 1.75 1.35
8.77!! 12.81!! 14.71!! 1.70 13.95!! 2.06 10.63 20.42!! 4.46!! 4.231 7.27 §
0.04 2.28 0.05 1.22 0.38 2.87 3.48 0.21 2.81 8.17 § 2.99 2.34 0.91 0.22 0.62 0.11 0.97 1.29 0.04 1.68 0.44 2.73 2.67 § 16.88!! 10.22 § 16.73!! 8.41 §
5.671; 4.531 0.15 0.02
13.86!! 9.00 § 4.291 4.621
8.77 § 2.59
2.16
3.68 4.521; 1.13
0.66
1.24 4.68t 11.97;! 1.75 8.26 § 2.03 2.63 6.70 §
0.72 0.34 0.78 0.93 2.57 1.40 3.43 6.441;
6.27t
6.08t
12
8
5.411; 27.87!1 10.97 § 10.16!! 13.09! 5.011~ 12.87,! 2.96 5.941; 5.81
5.29t
3.861; 0.72 6.89!! 0.41 0.37 0.98 13.05!! 8.27!! 10.86!!
28.91!! 8.89 § 66.78!! 1.59 7.00 § 54.51 136.49!! 101.73!! 124.53!!
6.74!! 0.22 0.81 0.65
96.19! 8.92 §
4.78!! 6.93 84.39!! 83.98!!
2.271 0.52 1.12 1.90 0.62 0.08 0.52 0.65 0.57
16.24'! 9.45 § 2.66 28.30!! 3.04 1.90 12.17!! 21.62!! 21.49!!
0.47 3.17 0.62 2.45 1.33 11.67!! 9.44 §
0.58 7.75 § 1.33 1.98 1.70 3.53 1.80 2.13 3.32
43
16
42
31
15
5.651;
4.45t
7.39 § 0.43 0.90 0.44
Maze coordination non-dom.
(counter)
66.89!!
29.68!! 20 Grooved steadiness dom. (time) Grooved steadiness dom. (counter) 21.88!! Grooved steadiness non-dom. (time) 7.07 § Grooved steadiness non-dam. 7.33 § (counter) 20.44!! 21 Steadiness dam. (time) 12.74!! Steadiness dam. (counter) 19.49!! Steadiness non-dam. (time) 14.85!! Steadiness non-dam. (counter) 12.28!! 22 Grooved pegboard dam. (time) Grooved pegboard non-dom. (time) 16.90!! 42.43!! 23 Foot tapping dam. (score) 47.49!! Foot tapping non-dam. (score) Total number of significant differences: 43
98.16!! 25.53!! 27.07!!
14.76:!1 6.26!! 20.70!! 20.04!! 24.43!! 34.36!!
5.961; 0.21
4.79t
*The total number used was 390. In the analysis of variance, comparisons between groups are based on the !p < 0.001 §P 9yr 9yr
