National Academies Press: OpenBook
« Previous: Introduction
Suggested Citation:"Method and Materials." National Research Council. 1961. Follow-Up Study of Head Wounds in World War II, by a. Earl Walker and Seymour Jablon. Washington, DC: The National Academies Press. doi: 10.17226/18529.
×
Page 5
Suggested Citation:"Method and Materials." National Research Council. 1961. Follow-Up Study of Head Wounds in World War II, by a. Earl Walker and Seymour Jablon. Washington, DC: The National Academies Press. doi: 10.17226/18529.
×
Page 6
Suggested Citation:"Method and Materials." National Research Council. 1961. Follow-Up Study of Head Wounds in World War II, by a. Earl Walker and Seymour Jablon. Washington, DC: The National Academies Press. doi: 10.17226/18529.
×
Page 7
Suggested Citation:"Method and Materials." National Research Council. 1961. Follow-Up Study of Head Wounds in World War II, by a. Earl Walker and Seymour Jablon. Washington, DC: The National Academies Press. doi: 10.17226/18529.
×
Page 8
Suggested Citation:"Method and Materials." National Research Council. 1961. Follow-Up Study of Head Wounds in World War II, by a. Earl Walker and Seymour Jablon. Washington, DC: The National Academies Press. doi: 10.17226/18529.
×
Page 9
Suggested Citation:"Method and Materials." National Research Council. 1961. Follow-Up Study of Head Wounds in World War II, by a. Earl Walker and Seymour Jablon. Washington, DC: The National Academies Press. doi: 10.17226/18529.
×
Page 10
Suggested Citation:"Method and Materials." National Research Council. 1961. Follow-Up Study of Head Wounds in World War II, by a. Earl Walker and Seymour Jablon. Washington, DC: The National Academies Press. doi: 10.17226/18529.
×
Page 11
Suggested Citation:"Method and Materials." National Research Council. 1961. Follow-Up Study of Head Wounds in World War II, by a. Earl Walker and Seymour Jablon. Washington, DC: The National Academies Press. doi: 10.17226/18529.
×
Page 12
Suggested Citation:"Method and Materials." National Research Council. 1961. Follow-Up Study of Head Wounds in World War II, by a. Earl Walker and Seymour Jablon. Washington, DC: The National Academies Press. doi: 10.17226/18529.
×
Page 13
Suggested Citation:"Method and Materials." National Research Council. 1961. Follow-Up Study of Head Wounds in World War II, by a. Earl Walker and Seymour Jablon. Washington, DC: The National Academies Press. doi: 10.17226/18529.
×
Page 14
Suggested Citation:"Method and Materials." National Research Council. 1961. Follow-Up Study of Head Wounds in World War II, by a. Earl Walker and Seymour Jablon. Washington, DC: The National Academies Press. doi: 10.17226/18529.
×
Page 15
Suggested Citation:"Method and Materials." National Research Council. 1961. Follow-Up Study of Head Wounds in World War II, by a. Earl Walker and Seymour Jablon. Washington, DC: The National Academies Press. doi: 10.17226/18529.
×
Page 16
Suggested Citation:"Method and Materials." National Research Council. 1961. Follow-Up Study of Head Wounds in World War II, by a. Earl Walker and Seymour Jablon. Washington, DC: The National Academies Press. doi: 10.17226/18529.
×
Page 17
Suggested Citation:"Method and Materials." National Research Council. 1961. Follow-Up Study of Head Wounds in World War II, by a. Earl Walker and Seymour Jablon. Washington, DC: The National Academies Press. doi: 10.17226/18529.
×
Page 18
Suggested Citation:"Method and Materials." National Research Council. 1961. Follow-Up Study of Head Wounds in World War II, by a. Earl Walker and Seymour Jablon. Washington, DC: The National Academies Press. doi: 10.17226/18529.
×
Page 19
Suggested Citation:"Method and Materials." National Research Council. 1961. Follow-Up Study of Head Wounds in World War II, by a. Earl Walker and Seymour Jablon. Washington, DC: The National Academies Press. doi: 10.17226/18529.
×
Page 20
Suggested Citation:"Method and Materials." National Research Council. 1961. Follow-Up Study of Head Wounds in World War II, by a. Earl Walker and Seymour Jablon. Washington, DC: The National Academies Press. doi: 10.17226/18529.
×
Page 21
Suggested Citation:"Method and Materials." National Research Council. 1961. Follow-Up Study of Head Wounds in World War II, by a. Earl Walker and Seymour Jablon. Washington, DC: The National Academies Press. doi: 10.17226/18529.
×
Page 22
Suggested Citation:"Method and Materials." National Research Council. 1961. Follow-Up Study of Head Wounds in World War II, by a. Earl Walker and Seymour Jablon. Washington, DC: The National Academies Press. doi: 10.17226/18529.
×
Page 23
Suggested Citation:"Method and Materials." National Research Council. 1961. Follow-Up Study of Head Wounds in World War II, by a. Earl Walker and Seymour Jablon. Washington, DC: The National Academies Press. doi: 10.17226/18529.
×
Page 24
Suggested Citation:"Method and Materials." National Research Council. 1961. Follow-Up Study of Head Wounds in World War II, by a. Earl Walker and Seymour Jablon. Washington, DC: The National Academies Press. doi: 10.17226/18529.
×
Page 25
Suggested Citation:"Method and Materials." National Research Council. 1961. Follow-Up Study of Head Wounds in World War II, by a. Earl Walker and Seymour Jablon. Washington, DC: The National Academies Press. doi: 10.17226/18529.
×
Page 26
Suggested Citation:"Method and Materials." National Research Council. 1961. Follow-Up Study of Head Wounds in World War II, by a. Earl Walker and Seymour Jablon. Washington, DC: The National Academies Press. doi: 10.17226/18529.
×
Page 27
Suggested Citation:"Method and Materials." National Research Council. 1961. Follow-Up Study of Head Wounds in World War II, by a. Earl Walker and Seymour Jablon. Washington, DC: The National Academies Press. doi: 10.17226/18529.
×
Page 28
Suggested Citation:"Method and Materials." National Research Council. 1961. Follow-Up Study of Head Wounds in World War II, by a. Earl Walker and Seymour Jablon. Washington, DC: The National Academies Press. doi: 10.17226/18529.
×
Page 29
Suggested Citation:"Method and Materials." National Research Council. 1961. Follow-Up Study of Head Wounds in World War II, by a. Earl Walker and Seymour Jablon. Washington, DC: The National Academies Press. doi: 10.17226/18529.
×
Page 30
Suggested Citation:"Method and Materials." National Research Council. 1961. Follow-Up Study of Head Wounds in World War II, by a. Earl Walker and Seymour Jablon. Washington, DC: The National Academies Press. doi: 10.17226/18529.
×
Page 31
Suggested Citation:"Method and Materials." National Research Council. 1961. Follow-Up Study of Head Wounds in World War II, by a. Earl Walker and Seymour Jablon. Washington, DC: The National Academies Press. doi: 10.17226/18529.
×
Page 32
Suggested Citation:"Method and Materials." National Research Council. 1961. Follow-Up Study of Head Wounds in World War II, by a. Earl Walker and Seymour Jablon. Washington, DC: The National Academies Press. doi: 10.17226/18529.
×
Page 33
Suggested Citation:"Method and Materials." National Research Council. 1961. Follow-Up Study of Head Wounds in World War II, by a. Earl Walker and Seymour Jablon. Washington, DC: The National Academies Press. doi: 10.17226/18529.
×
Page 34

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Chapter II METHODS AND MATERIALS A. SAMPLING PLAN AND ORGANIZATION OF STUDY The Medical Statistics Division, Office of The Surgeon General, Depart- ment of the Army, made available duplicate punchcards representing the 31,178 battle casualty admissions to Army hospitals in 1944 with the diagnosis of a traumatism of the head. It had been hoped to use this file as the source of all cases for the study, but elimination of the thousands of death cases, scalp lacerations, ear wounds, etc., reduced the roster to less than a third of its original size. The records of the Veterans Administra- tion were then consulted to obtain current addresses. Since it was feared that a poor response would be encountered by an attempt to examine men living at great distances from the clinical centers, certain areas were defined for each of the 4 centers, and only men residing in the defined areas were allocated to the centers for examination. The choice of particular areas for each center was influenced by ease of transporation and the number of patients available. The areas chosen and the numbers of men allocated (hereafter called collectively the "Army roster") are shown below. Men Johns Hopkins Hospital (Baltimore, Washington, Philadelphia, and vicinity of each) 159 Montefiore Hospital (New York City) 193 Cushing VA Hospital (Boston and vicinity) 73 Long Beach VA Hospital (Los Angeles and vicinity) 38 Total 463 The method of selecting the Army roster was not such as to produce a sample representative of all head wounds. This was, however, no part of the purpose, which was to elucidate the course of the patient with a head wound in terms of the characteristics of the injury and, if possible, to assess the influence of certain factors associated with treatment. Since the Army files did not supply adequate numbers of cases for the clinical centers, the roster was supplemented by men drawn from lists of veterans receiving disability compensation from the Veterans Administra- tion. Each of the more than 60 regional offices of the Veterans Adminis- tration maintains lists of the veterans of the area who receive a disability compensation or pension. Men are classified by the nature of the disability; if two or more disabilities coexist, for the purpose of these listings, assign- ment is made to the most important single disability. Of course, in the individual veteran's record all causes of disability are shown. The lists of veterans receiving disability compensation seemed quite attractive as a source of case material since, as they are kept on a regional office basis, it was possible to obtain lists which were very largely made up of men whose place of residence made them suitable for study at one or another of the 4 centers.

A review of the records of 60 head-injured men drawn from the Army roster revealed that the majority was being compensated for 1 of 3 particular causes: a. Posttraumatic personality disorder. b. Posttraumatic encephalopathy. c. Loss of part of inner or outer tables of skull. Lists were obtained from the Veterans Administration which included all men who resided in the study cities and who were, in 1950, receiving dis- ability compensation for one of these disorders. The number of men allocated to the 4 centers from the "VA roster" were: Baltimore 157 New York 107 Boston 93 Los Angeles 112 Total 469 After review of the records of men drawn from the VA roster had begun, it became plain that this roster included many men who had suffered noncombat injuries—falls and the like. Thereafter, men were included only if they had suffered a bullet or shell fragment wound. However, 98 men for whom records had already been abstracted, and for whom the wounding agent was not a missile, were continued in the study. The other 371 men had missile wounds of the type specified. All except 4 of the 98 men referred to above were Baltimore cases, since the Baltimore center operated as a pilot study and Baltimore cases were the first for which rec- ords were abstracted. The New York area was the only one in which it was possible to find more cases than could be examined in the time available. It was thought that a total of 300 represented a maximum practicable load. Therefore, the men drawn from the VA roster for New York represented the first 107 eligible men on the listing. Since the listing was arranged in claim number order, this was equivalent to choosing those eligible men who earliest filed claims for compensation. A sample which is selected because of the manifestation of a residual of injury cannot ordinarily be expected to contribute usefully to a study of the probability of residuals, although it may be employed in a study of the course in time of specified residuals. In fact, however, in this instance the VA roster, or at least part of it, is capable of providing useful information with respect to the probability of specified residuals for severely injured men. The reason for this is that men who were badly wounded almost invariably received VA disability compensation for one reason or another; for lightly injured men, the VA roster is heavily biased in the direction of residuals, but for heavily injured men it is not. This subject is discussed more fully below. B. CHARACTERISTICS OF THE SAMPLE AS DETERMINED FROM ARMY RECORDS Age. Only 14, or 1.5 percent, of the 932 men were nonwhite. Since, in the main, these men were combat troops they were predominantly young

men, 81.0 percent being under 30 years of age at the time of the head in- jury (table 1). Agent. Three-quarters of the men (693, or 74.4 percent) had been injured by shell or grenade fragments, while 132, or 14.2 percent, had bullet wounds of the head. These percentages do not reflect the proportions of all injuries to the head caused by shell fragments and bullets, respectively, because men who were killed in action or who died of wounds are not in- cluded in the study, and there is evidence (14) that bullet wounds are more often lethal than shell fragment wounds. In 33 instances, or 3.5 percent, the head injury was caused by a blunt object or resulted from a fall; and in 68 cases, or 7.3 percent, the wound was due to such miscellaneous causes as blast injury, sharp-cutting objects, etc. For six men the agent could not be determined from the records. Table 1.—Age at Time of Injury Age Number Percent Age Number Percent 18-20 170 18.2 33-35 60 6.4 21-23 . . 234 25 1 36-38 25 2.7 24-26 208 22. 3 39-41 3 . 3 27 29 143 15 3 30-32 89 9. 5 Total 932 100.0 Type of Wound. The type of wound was most commonly a penetrating wound of the brain, i.e., a compound, comminuted fracture; such wounds accounted for almost 60 percent of the cases. The other large group, in contrast, consisted of men who had scalp lacerations, usually with concussion i (about 22 percent). A complete breakdown is in tables 2 and 3. Depth of Wound. In almost 30 percent of the cases, only the scalp was overtly involved in the trauma; in another 19.4 percent, there was cranial fracture without deeper penetration; and in 2.6 percent, although the dura mater was torn, there was no obvious evidence of damage to the underlying brain tissue. Thus, in a total of 482 cases, or not quite 52 percent, there was no apparent destruction of gray matter. In 410 cases (44.0 percent) the brain itself was penetrated either by metallic foreign bodies or by bony fragments of the cranium; and in 27 additional cases, 2.9 percent, fragments penetrated into the ventricles. Cranial Defect. There was no cranial defect in 347 instances (37.2 per- cent) ; in 22 cases (2.4 percent) the cranial defect was of the nature of a gutter wound, the size of which could not be easily defined. In 72 cases the oversea records were so vague that it was not possible to determine the size of the cranial defect which apparently was present, but in 491 cases the average diameter of the cranial defect could be determined. The most frequent size of defect (diameter) among men who survived was in the range 2 to 4 cm. (table 4). Additional Wounds of the Head. In 111 instances (11.9 percent) a second head wound, distinct from the primary injury, was present. Of these, 1 Concussion is rarely defined in the records; presumably it implies loss of conscious- ness for an appreciable period of time. 830802—62

Table 2.—Type of Injury Type of head wound Number Percent Type of head wound Number Percent Closed, without frac- ture 65 7.0 Compound commi- nuted fracture of Closed, with fracture outer table only. . . . 43 209 4.6 22.4 base 1 . 1 Other and unspeci- Closed, with de- ^ fied 6 .6 Perforating 28 3.0 Total 932 99.9 Linear fracture, com- pound 551 59 1 26 2.8 Table 3.—Deepest Structure Penetrated Depth of penetration Number Percent Depth of penetration Number Percent 277 29 7 27 2. 9 Cranium 181 19.4 Unspecified ... 13 1.4 24 7 fi Brain 410 44. 0 Total 932 100.0 Table 4.—Size of Cranial Defect Average diameter of cranial defect Number Percent Average diameter of cranial defect Number Percent No defect 347 37.2 >4 cm 100 10.7 22 2 4 72 7.7 <2 Cm 157 16 8 2-4 cm 234 25 1 Total 932 99.9 78 (70.3 percent) were wounds of the scalp only; 13 (11.7 percent) involved the cranium; and in 1 case the dura mater was penetrated but apparently not the brain itself. In only four cases was brain tissue damaged by the second wound as well as by the first. Other Wounds. Accompanying wounds of other parts were present in 497 cases, or 53.3 percent. Many of the men who had additional wounds were hit in several places. Some 497 men were coded for a total of 936 locations other than the head, so that these men suffered, on the average, at least 1.9 additional hits each. Most of these, of course, were caused by shell fragments rather than bullets. Most frequently involved in additional wounds were the upper extremities, closely followed by the lower extremities (table 5). It should be emphasized that the group of men being discussed did not die of their injuries; presumably one of the reasons that thoracic and abdominal wounds are not more frequent in this sample is the higher fatality rate associated with wounds of these regions.

Table 5.—Site of Other Wounds Location Number Percent of men Location Number Percent of men Ear 53 10 7 9 1 8 Face 144 29.0 tebrae. Neck 51 10 3 209 42 1 286 57 5 Thorax. . 122 24 5 497 Abdomen 27 5.4 men with addi- Pelvis . . 35 7 0 Traumatic Unconsciousness. Although the oversea records were often not explicit, apparently unconsciousness was usual after the head wound. For 165 men (17.7 percent) the records definitely indicated that there was no unconsciousness, and for 213 men (22.9 percent) it was impossible to decide whether or not there had been coma. However, the remaining 554 individuals (59.4 percent) were known to have been unconscious after the injury. Of these 554, almost half (202) were unconscious for 2 hours or less and 105 (19.0 percent) for more than 24 hours. However, in 144 cases (26.0 percent), although it was clear from the records that the soldier had been unconscious, it was not possible to determine how long. Confusion. Because amnesia related to the wounding is considered an index of the severity of brain injury, an attempt to examine this factor was made by noting the state of confusion or disorientation of the patient after recovery from coma (if present); it was even less well recorded than the facts regarding unconsciousness itself. In 444 cases, or 47.6 percent, it could not be determined whether there had been confusion. In 91 instances it seemed clear that there had been no disorientation, while in 397 cases there clearly had been. However, even in these cases it was impossible to deter- mine the duration of confusion in 142, or 35.8 percent. Of the remaining 255 men, i.e., those known to have been disoriented for a specific time period, the duration of disorientation was under 2 hours in 60 cases (23.5 percent); between 2 and 24 hours in 36 cases (14.1 percent); between 1 day and 1 week in 77 cases (30.2 percent); between 1 and 4 weeks in 60 cases (23.5 percent); and over 4 weeks in 22 cases (8.6 percent). Complications. In 42 cases the presence of complications of the head wound could not be determined, but in the remaining 890, for which this factor was known, in 707 cases, or 79.4 percent, there were no complications. The complications most frequently noted were intracranial hemorrhage and herniation of brain tissue, each present in 88 cases, or 48.1 percent of the 183 men known to have had complications. Other complications were relatively infrequent: seven instances of cerebrospinal fistula; five of cere- brospinal rhinorrhea; three each of cerebrospinal otorrhea, frank infection, and abscess; and two instances of meningitis. There were also seven cases with other unclassified complications. Neurological defects following the wounding were usually documented if present; in only seven cases was the neurological deficit unknown. Of the

925 for whom the records were clear, 538, or 58.2 percent, had no deficit. The most frequent defect was hemiplegia or hemiparesis (sensory or motor), which was present in 180 cases, or 19.5 percent. Hemianopsia was noted in 73 instances (7.9 percent) and aphasia in 118 (12.8 percent). The combination of hemiplegia and aphasia was quite common, being found in 63 cases, or about a third of the cases of hemiplegia and over half of the cases of aphasia. In addition to the categories mentioned above, there were 144 cases of a miscellaneous nature involving such conditions as nerve deafness, monoplegia, triplegia, quadriplegia, quadrantanopsia, facial paralysis, etc. Exactly 100 of the 144 men with miscellaneous deficits had none of those specifically categorized (hemiplegia, hemianopsia, aphasia). Debridement. As might be expected from the fact that many of the wounds seemed rather minor, 128 men (13.7 percent) had no debridement, while for 15 others the oversea hospital records were incomplete, and it could not be determined whether or not a debridement had been done. Of the 789 men who were debrided, in 590 (74.8 percent) the debridement was done within 24 hours of the wounding and in 126 (16.0 percent) be- tween 1 and 3 days posttrauma. However, 47 (6.0 percent) were delayed beyond 3 days, and in 26 instances the time could not be determined from the records. Second debridements were performed on 211 men. The most frequent causes were retained fragments (91 men); infection (20 men) ; intracranial bleeding (16 men); abscess (12 men); and draining sinus (9 men). In 63 instances, the reason for the second debridement was not specifically coded; a host of miscellaneous reasons, such as exploration for suspected (but not discovered) fragments, secondary closures to aid healing following the first debridement, and the like, were reported. Third debridements were done in 52 men and, indeed, 14 had 4 or more. Again, the most frequent reason for a third debridement was an attempt at removal of retained fragments (16 cases), while the second most common reason was infection (7 cases). Three men had draining sinuses, three had abscesses, and one had intracranial bleeding. The remaining 22 tertiary debridements were done for miscellaneous reasons which were not explicitly coded. Cranioplasty. Cranioplasties were performed on 371, or 39.8 percent, of the 932 men studied. These were distributed by time with respect to the original injury as shown in table 6. In 21 instances the cranioplasty was done at a debridement, 4 times at a first debridement, 13 times at a second debridement, and 4 times at a third debridement. Table 6.—Time From Injury to Cranioplasty Time from injury to cranioplasty Number Percent Time from injury to cranioplasty Number Percent < 3 months 88 23.7 2—3 years 1 .3 3—6 months 180 48. 5 3—4 years 2 . 5 6—9 months 63 17 0 1 .3 25 6 7 1—2 years 11 3.0 Total 371 100 0

The overwhelming majority of the plates (341, or 91.9 percent) were of tantalum. However, there were 20 methacrylate plates used, as well as 5 of bone and 1 of celluloid. In two instances cartilage grafts were employed. Intracranial Foreign Bodies. According to the records, no intracranial foreign bodies were present in 380 men; in 476 men the status of intra- cranial foreign bodies both before and after definitive debridement was explicitly given, but in 45 men the records are unclear in this regard. Bone fragments only were present in 237 cases; in 168 of these (70.9 percent) the fragments were entirely removed at the debridement, while in the re- mainder at least some of the fragments remained in situ. Only 31 men had metal fragments alone, and in only 8 of these (25.8 percent) were all frag- ments removed. Finally, 208 men had both bone and metal intracranial foreign bodies; all fragments were removed at debridement in 49 cases (23.6 percent); all bone fragments (but not all metal) were removed in 67 cases (32.2 percent); all metal foreign bodies (but not all bone frag- ments) were removed in 21 cases (10.1 percent); and foreign bodies of both kinds remained after the debridement in the remaining 71 cases (34.1 percent). Postdebridement Complications. Few men had complications following debridement: 655, or 83.0 percent, of the 789 men with debridements definitely had none. Most frequent of the complications reported was frank infection (30 cases), followed by abscess (19 cases); fungus cerebri (16 cases); hematoma (14 cases); cerebrospinal fistula (12 cases); menin- gitis (11 cases); and cerebrospinal rhinorrhea and cerebrospinal otorrhea (3 cases each). Cranial Roentgenograms. A report of the predebridement X-ray of the cranium was available for 448 of the 789 men who were debrided. Of these, 117 (26.1 percent) showed a normal skull. The most frequent find- ings were of retained bone fragments with or without accompanying metallic pieces: 227, or 50.7 percent, of the reports mentioned bone frag- ments. In 151 cases (33.7 percent) intracranial metallic foreign bodies were present, usually (104 cases) accompanied by bone fragments (table 7). For 744 men it was possible to obtain the report of the last X-ray taken in any Army hospital (table 8). Only 142 (19.1 percent) were thought to show a normal skull. Bone fragments had been entirely removed in many instances. Only 77, or 10.3 percent, still had such fragments, whether with Table 7.—Radio graphic Findings Before Debridement X-ray report Number of men X-ray report Number of men Normal skull 117 47 Fracture of outer table only. Fracture only 8 21 Retained bone fragments 104 Skull defect only . 13 1 14 Retained bone fragments .... 123 Total 448 11

Table 8.—Radiographic Findings at Last Army Examination X-ray report Number of men X-ray report Number of men Normal skull 142 103 Fracture of outer table only. . 17 14 Retained bone fragments 31 Skull defect only . 131 17 33 Defect plated 210 Total 744 Retained bone fragments .... 46 or without foreign bodies. However, 134 men (18.0 percent) had foreign bodies still present, accompanied only in 31 cases by bone fragments. Neurological Complaints at Discharge From Service. More than a third of the men (376, or 40.3 percent) reported no neurological symptoms at the time of discharge from service. Records were inadequate for 15 men, but the remaining 541 men reported a host of symptoms (table 9). The term "posttraumatic syndrome" as used here implies the presence of at least 3 of the specific complaints listed in the table; the specific complaints were coded for men who claimed not more than 2 symptoms. Hence, the total number of men having headache at the time of discharge probably approxi- mated 464. Neurological Abnormalities. More than two-fifths (405, or 43.5 per- cent) of the men were free from objective neurological abnormalities when discharged from service. Records were incomplete for 14 men. The 513 men for whom abnormalities were present were classified as shown in table 10. The very large number of men coded "Other" had a wide variety of abnormalities, ranging from nerve deafness and loss of sense of smell to psychoneurosis. Pneumoencephalograms. Pneumoencephalograms had been performed in Army hospitals on 129 of the men. These were usually made within 6 months of the head injury, although in 23 instances they were made after this time. Only 41 of the 129 men so examined were found to have normal air studies. The most frequent abnormal findings (table 11) were gen- eralized ventricular dilatation and unilateral ventricular dilatation on the side of the lesion, each of which was observed in 28 men (21.7 percent). Table 9.—Neurological Symptoms at Discharge From Service Symptom Number of men Symptom Number of men 292 7 Dizziness. . . . 117 Easy fatiguability 4 Impaired memory 32 Other 26 Tinnitus 29 Posttraumatic syndrome 172 of concentration 25 T«ta' 541 Irritability 14 12

Table 10.—Neurological Abnormalities at Discharge From Service ; . Finding Number of men Finding Number of men Hemiplegia or hemiparesis . . . 131 Personality changes 27 Aphasia 82 20 Hemianopsia or other field Paraparesis 6 defect 74 Other 314 40 Spasticitv .... . . . 29 Total 513 Table 11.—Pneumoencephalographic Findings in Military Hospitals Finding Number of men Finding Number of men Normal 41 3 Generalized ventricular dila- tation Local absence of air in sub- arachnoid space 28 10 Unilateral ventricular dilata- tion on side of lesion Inadequate filling of ventri- cle 28 13 Local ventricular dilatation Other abnormality 13 on side of lesion or over cortex 34 Electroencephalograms. Electroencephalograms (EEC's) were made more frequently than pneumoencephalograms. Tracings made during the first 6 months after injury were coded for 295 men, and during the second 6 months for 162 men. Beyond 1 year after injury the material became much thinner, but 103 examinations were made in the second year, 23 in the third year, 25 in the fourth year, and 43 were made more than 4 years posttrauma. About half of the 295 EEC's which were done within 6 months of injury were considered to show no definite abnormality: 103 (34.9 percent) were normal, while 45 (15.3 percent) were borderline. Thus, only 147 (49.8 percent) were interpreted as definitely abnormal. Focal abnormalities were most frequently found and were present in 102 tracings; i.e., in 34.6 percent of all tracings and in 69.4 percent of those considered abnormal. Gen- eralized abnormalities were noted in 85 instances, or 28.8 percent of the whole number, and 57.8 percent of those called abnormal. In 41 instances generalized and focal abnormalities were found in the same men, so that 40.2 percent of the 102 men with focal abnormalities had generalized ab- normalities also, while 48.2 percent of the 85 men with generalized ab- normalities had accompanying focal abnormalities. In 49 cases the focal abnormalities were right-sided, in 47 cases left-sided, and in 5 instances bilateral. In one case it was not possible to determine from the records on which side the focal abnormality was found. The nature of the generalized abnormalities was usually either dys- rhythmia or paroxysmal slow waves (table 12). It will be noted that 122 different abnormalities were coded for the 85 men having generalized abnormalities. 13

Table 12.—Generalized EEG Abnormalities Finding Number of men Finding Number of men Dysrhythmia 42 Paroxysmal slow waves 37 Slow waves .... 13 Fast waves 10 waves 20 Focal abnormalities were frequently paroxysmal slow waves. These were noted for 57 men, or 55.9 percent, of the 102 men with focal abnormalities. The only other abnormality which occurred frequently was asymmetry, which was noted 28 times, accounting for 27.5 percent of focal abnormalities (table 13). Focal abnormalities showed no remarkable tendency to cluster in particu- lar locations (table 14). Hyperventilation activation was employed in 219 of the 295 EEC's. Changes were not noteworthy (table 15). Disposition. Finally, of the 932 men studied, 691, or 74.1 percent, were discharged from the Army for disability (CDD). Routine discharges were given to 201 men (21.6 percent), while 26 (2.8 percent) remained in service. In 14 instances, the circumstances of separation from service were not explicitly shown in the records. C. THE FOLLOW-UP EXAMINATION After the identification of those veterans who resided in one of the 4 selected areas, their names and current addresses were supplied to the clini- cal centers, whose responsibility it was to persuade the men to report for examination. While practice at the four centers varied somewhat, ordinarily appoint- ments were made for two patients each morning. The examination was divided into two parts: One consisted of the history, neurological examina- tion, electroencephalogram, and roentgenogram of the head, while the other part consisted of the personality evaluation (Minnesota Multiphasic Per- sonality Inventory), intelligence test (Wechsler-Bellevue), and performance test (Goddard Tactual). The two parts were administered in alternate order to the two patients seen each day—one part in the morning and the other part in the afternoon. Thus the neurologist and clinical psychologist each examined one man in the morning and one in the afternoon. Table 13.—Focal EEG Abnormalities Finding Number of men Finding Number of men Irregularity 14 Paroxysmal fast waves . . . 13 Asymmetry 28 Spikes 7 Spindles 5 Spike and wave 1 Paroxysmal slow waves 57 Other 2 14

Table 14.—Site of Focal EEC Abnormalities Region Number of men Region Number of men Entire side 6 Frontoparietal 5 Frontal . 16 Frontotemporal . . . 6 Parietal 12 Parietotemporal 9 Temporal 19 5 Occipital . . . 16 Temporo-occipital . .... 7 Vertex 1 The examiner had available an abstract of the Army clinical history in the form of "code sheets" (app. 2). These sheets were prepared at the Follow-up Agency and represented information culled from the clinical records of the Army hospitalization, which were available usually from the VA claims folders but sometimes from the Army's Demobilized Personnel Records Branch, St. Louis, Mo. Results of examination were entered into code booklets (app. 3), which were returned to the Follow-up Agency for cardpunching and subsequent analysis. The practice of having the clinical centers enter examination results in coded form has both advantages and drawbacks. Reports in narrative style often lack the specificity which is needed for later coding. On the other hand, although space was provided for remarks and the examiners were encouraged to enter notes of conditions which were not adequately represented by the code positions, it did happen that certain cases were given unusual combinations of code positions which were not always immediately explicable. Weighing the advantages and disadvantages, it is considered that the coding of cases by the clinician worked well. The initial approach to the patients was made by the examiner, who wrote a letter to the veteran in an effort to gain his cooperation (app. 4). Re- peated correspondence and telephone calls sometimes failed to induce cooperation. American Red Cross social workers rendered invaluable assistance in reducing the number of nonparticipants by visiting the most reluctant men, explaining to them the purpose of the study, and arranging transportation. Not infrequently veterans who had made appointments under the pressure of repeated urging failed to appear, and further efforts Table 15.—Results of EEC Activation Finding Number of men Finding Number of men No change 130 Spike and wave complex .... 5 Generalized slowing, within 5 normal limits . . 36 Focal slow wave 16 Generalized fast, within nor- Other alterations 2 mal limits 5 Generalized subnormal slow- inar . . . ..... 18 15

Table 16.—Results of Solicitations for Examinations Center Number allocated Number examined Percent Baltimore 316 257 81.3 New York . . 300 221 73.7 Boston 166 152 91.6 150 109 72.7 Total 932 739 79.3 were necessary. Despite all efforts, however, a certain proportion of men could not be examined (table 16). There is clear cause for some concern over the fact that just over 20 percent of the veterans who had been selected for study could not be induced to cooperate. It seemed plain that a variety of motivations were at work in different men, but the possibility could not be excluded that the men who did not cooperate were quite different from those who were examined, and that the results of the study might be seriously affected. In order to obtain some measure of the magnitude of error so introduced, the VA claims folders of the men not examined were reviewed. Claims folders contain not only information concerning payments for disability compen- sation but also copies of the reports of physical examinations on which rates of payment are based. It was thought that comparison of rates of disability compensation of the nonexamined men with those for a sample of men who were examined, coupled with comparisons based on the original Army clinical records, would provide the basis for an estimate of the import- ance of bias attributable to nonresponse in the survey. D. DIFFERENCES BETWEEN MEN WHO REPORTED FOR FOLLOW-UP EXAMINATIONS AND THOSE WHO DID NOT 1. Characteristics of the Injury and Status at the Time of Separation From Service Analysis of differences between men who could be persuaded to report for the follow-up examination and those who declined revealed some inter- esting contrasts, which were not always immediately explicable. Table 17 presents an extract of these data. The men who were examined did not differ in age from those who were not examined, but they more often had penetrating or perforating wounds, as opposed to scalp lacerations; they more often had large cranial defects, definite penetration of brain tissue, and epileptic seizures before discharge from the Army. However, they also were more often free of neurological symptoms and objective abnormalities at discharge. 2. Information From VA Claims Folders Of the 616 men allocated to the Baltimore and New York centers, 473 had VA disability claims. Of these men, 359 were examined, while 114 were not. Since the VA claims folders contain the report of examination on which the disability rating is based, it was decided to compare the re-

spondents and nonrespondents with respect to these examination reports to learn how much importance might attach to neurological status at follow- up in influencing men to report for examination or to decline. A summary of the comparisons is presented in table 18. It was surprising to find that the 2 groups of men seemed to differ very little with respect to follow-up status. Only a single factor, of those examined, differed significantly be- tween the 2 groups of men: While 30 percent of examined men were para- lyzed or had impaired motor power, the corresponding figure for non- examined men was only 20 percent. For many other observations, differences, even if not statistically significant, were in the direction sug- gesting that the more severely impaired men were more likely to report for examination than those with little disability. Thus, less than 16 percent of nonrespondents had ever had a fit, while 22 percent of respondents had had at least 1 seizure. Average percentage disability was slightly higher for the examined men, who also had somewhat more aphasia and muscle atrophy. On the other hand, nonexamined men more often complained of impaired hearing, were more often reported to show apathy or anxiety on psychiatric examination, and were more often unemployed. Table 17.—Comparison of Respondents With Nonrespondents; Date From Army Clinical Records Characteristic Respondents Number Percent Nonrespondents Number Percent Total <20 years old at injury 30 + years old at injury With perforating wound With penetrating wound With scalp laceration No cranial defect Cranial defect >4 cm. diameter No penetration of brain tissue Definite penetration of brain No fits prior to separation At least 1 fit in service Posttraumatic syndrome at separation Other or unknown symptoms at separation. No neurological symptoms at separation... No neurological abnormalities at separa- tion Hemiplegia or hemiparesis Aphasia Field defect 739 100.0 193 100.0 133 142 24 448 158 264 85 366 363 578 161 127 303 309 325 110 71 54 18.0 19.2 3.2 60.6 21.4 35.7 11.5 i 49.5 i 49.1 78.2 21.8 17.2 41.0 41.8 44.0 14.9 9.6 7.3 37 35 4 103 51 83 15 116 74 162 31 45 81 67 80 21 11 20 19.2 18. 1 2. 1 53.4 26.4 43.0 7.8 i 60.1 i 38.3 83.9 16.1 23.3 42.0 34.7 41.5 10.9 5.7 10.4 i Percentage difference between respondents and nonrespondents significant at P > .05 level. 17

Table 18. — Comparison of Respondents With Nonrespondents; Data From VA Claims Ratings Examinations Baltimore and New York Cases With Claims Respondents Nonrespondents Information from claims folden Number Percent Number Percent Total 359 100.0 114 100.0 Clinical symptoms: 42 95 64 75 44 79 11.7 26.5 10 28 26 22 21 25 8.8 24.6 22.8 19.3 18.4 21.9 Visual complaints. ... Parasthesias 17.8 20.9 12.3 22.0 Impaired hearing or deafness Psychiatric status: Normal 143 14 120 39.8 3.9 33.4 39 8 34.2 7.0 36.8 Aoathv Anxiety 42 Neurological examination: Aphasia 38 19 34 109 44 48 12 14 30 74 10.6 5.3 9.5 '30.4 12.3 13.4 3.3 3.9 8.4 20.6 9 7 12 23 9 17 2 5 6 30 7.9 6.1 10.5 '20.2 7.9 14.9 1.8 4.4 5.3 26.3 Impaired facial movement Paralyzed or impaired motor power. . Muscle atrophy Impaired deep sensation . Romberg positive Work status: Unemployed VA disability rating: <60 percent 172 185 79 47.9 51.5 22.0 60 52 18 52.6 45.6 15.8 70-100 percent Epileptic seizures .... i Percentage difference between respondents and nonrespondents significant at P=.05 level. 3. Data From Telephone Interview For those patients who would not report for examination, attempt was made by telephone interview to obtain certain information in regard to status at the time of follow-up. The most significant findings are shown in table 19. The differences were small with respect to employment, but with respect to clinical symptomatology were quite large. Respondents were much more likely to present complaints characterized as a posttraumatic syndrome; the nonrespondents usually had but a single complaint, usually headache, and more nonrespondents were free of symptoms. Similarly, fits were less com- mon in the nonrespondents than in respondents, although for about 5 per- cent of nonrespondents it was not clear from the telephone interview whether the man had epilepsy or not. In particular, a significantly greater propor- tion of the respondents than the nonrespondents had had more than a single fit.

4. Conclusions It seems plain that the men who failed to report cannot be simply charac- terized; they appear to have, as a group, somewhat less objective evidence of neurological disability than those who were examined, but also to have somewhat more in the way of personality deviation. However, it seems unlikely that differences of the magnitude found here would seriously affect the conclusions of the study. E. FOLLOW-UP RESULTS AND VARIABILITY BETWEEN CENTERS As a preface to discussion of the results of the clinical examinations, it is useful to consider the difficulties which may arise from the fact that ob- servations were made by four different clinical groups. Table 19.—Comparison of Respondents With Nonrespondents; Data From Telephone Interview for Nonrespondents Respondents Nonrespondents for examination but with telephone interview Information from telephone interview Number Percent Number Percent Total 739 100.0 116 100.0 Employment status: 8 110 1.1 Unemployed or occasional jobs 14.9 78.1 20 96 17.2 82.8 Employed regularly .... 577 Clinical symptomatology: Posttraumatic syndrome 419 256 64 »56.7 134.6 8.7 22 79 15 119.0 168.1 12.9 No symptoms .... Epilepsy: No fits at any time . . . - 532 10 26 169 72.0 1.4 3.5 i22.9 88 3 1 18 75.9 2.6 .9 "15.5 Borderline attacks only One definite attack More than 1 definite attack i Percentage difference between respondents and nonrespondents significant at P=.05 level. 1 Percentage difference between respondents and nonrespondents significant at P=.01 level. The subject of what has been called "observer error" was brought to the fore in striking fashion, insofar as clinical observations are concerned, by the work of Birkelo et al. (18), with respect to the reproducibility of inter- pretations of chest X-rays. By now most medical scientists have come to realize that observations called by the same names by different observers may represent quite different measurements. In clinical medicine it is seldom that the details of the process of observation are specified to that degree of exactitude which would be needed for the reproducibility of the observation by different observers or even by the same observer at different

times. This is sometimes summed up by saying that there is a large "sub- jective element" in many observations. In a very real sense "observer error" is often not an apt designation for the phenomenon in question. Error, according to one definition at least, is: "The difference between an observed or calculated value and the true value." While this may represent a valid point of view for considering, say, the values for hemoglobin reported by different laboratories on aliquots from a single specimen, it seems not to be germane to the situation when presumably equally competent observers differ as to, say, whether a given patient's liver is or is not palpable. There are a host of clinical variables which share this characteristic, namely, that there is no "true" value, but each observer measures according to his own private scale, which may or may not be susceptible of calibration against the scales of others. This is especially true of those observations which depend wholly or in part upon an interaction between the observer and the patient; the personalities of different observers react differently with that of the patient. Despite all this, clinicians continue to make observations and to find them useful. This is natural enough. Each individual, in the course of time, creates a body of experience appropriate to his own measuring rods, and judges each new observation, whether explicitly or implicitly, by this yard- stick. Difficulties become of importance when he attempts to communicate to others, and the situation becomes critical when the two parties to the communication erroneously believe that because they employ the same name they are talking about the same thing. This difficulty was anticipated in this study and, in an attempt to minimize it, the authors visited each center at the beginning of the follow-up, reviewing the desired clinical interroga- tion and examination and the criteria for coding. In spite of this briefing, the centers seemed to have maintained their individuality. From this point of view, it was essential to examine the evidence regarding the degree to which the four participating clinical centers employed com- parable instrumentation (in the broadest sense) in making the follow-up observations. The ideal, of course, would have been to present a sample of patients to the four centers in turn, and to compare the results. That being impossible, the next best is to subject comparable samples of men to exami- nation at the four centers and then to compare the results. That has been done below in the course of the description of the results of the examination. However, data from centers which apparently did not differ have been combined. The Baltimore center began pilot work about 2 years prior to the entrance into the study of the other three centers. Thus, most men examined in Baltimore were 5 to 8 years after injury, while at the other centers the intervals were 7 to 9 years (table 20). More than one-fourth of the Los Angeles patients gave a history of hav- ing been knocked unconscious prior to the head injury which brought them into this study; only 6.8 percent of the men seen at the other centers related a history of having been knocked out. This gross difference is unexplained, but seems unlikely to be caused by a real difference of this magnitude in the previous histories of the men examined. We must conclude that the exam- iner at Los Angeles inquired into this question in a more searching way than did the others (table 21). M

Table 20.—Time From Injury to Follow-up Examination Number of men Number of men Time after injury Time after injury Baltimore Other centers Baltimore Other centers 3—4 years 1 8-9 years 5 148 4—5 years 6 9-10 years 12 129 1 6—7 years 66 11 Total 257 482 7—8 years 50 310 The information with respect to previous illnesses (table 22) is rather discouraging from the point of view of center variability. The proportion (percent) reported as "none" was, respectively, 87.2, 94.6, 75.5, and 79.8. This variation far exceeds what the blind workings of chance alone might be expected to produce. Again, 13.2 percent of the men examined at Bos- ton gave a history of enuresis, far higher than at any other center. Allergic conditions were noted for only a single man at the New York center, but were not uncommon at the other centers. It seems clear that the ordinary history, although taken with care, produces information of dubious validity. In Boston 68.4 percent denied nervous disorders in the family, while at the other centers the combined proportion was 85.7 percent—strikingly different. Some of die particular differences which were quite marked were: "fainting," 6.6 percent and 1.7 percent, "neurosis," 13.2 percent and 2.9 percent, "psychosis," 6.6 percent and 3.4 percent, and "psychopathic traits," 2.6 percent and 0.2 percent (table 23). Some evidence of difference between the centers with respect to report- ing of work status may be seen in the relatively large numbers coded at Los Angeles for school or rehabilitation training and at Baltimore for "occa- sional jobs" (table 24). Certain differences between the centers in "general work progress" are readily apparent (table 25); the Boston center classified many more men as "none" or "regression" (71.7 percent) than did any other center, while Table 21.—History of Previous Head Injuries Numbe r of men Numbei of men Type of injury Type of injury Los Angeles Other centers Los Angeles Other None significant Knocked out once or twice 77 19 580 42 Severe head injury without sequelae . . . centers 2 4 Knocked out more with sequelae 1 than twice 10 1 2 2 Total men 109 630 11

Table 22.—History of Previous Illness Condition reported Number of men Baltimore New York Boston Los Angeles None, other than uncomplicated childhood diseases 224 209 115 87 Childhood diseases with neuro- logical complications 2 1 3 2 4 1 1 Febrile convulsions 1 3 Allecgic conditions 16 1 12 15 2 6 20 2 1 1 Syphilis 6 1 3 Other 1 Unknown 2 1 2 2 Total men 257 221 152 109 New York so classified relatively few (30.3 percent). At New York many men were coded for "advancement" (61.1 percent), while only a few (15.9 percent) were so classified at Boston. It seems clear, retrospectively, that such words as "advancement" are much too imprecise for comparable use by different observers. The Boston center tended not to classify men at the high end of the home adjustment scale as frequently as the other centers: Only 55.3 percent were considered to be happy and content as against 74.4 percent at the other centers. On the other hand, the distinction between home adjustments which are "happy and content" and those which are merely "reasonably satisfactory" seems arbitrary (table 26). The characterization of social adjustment did not appear to be more consistent between centers than were the other judgment items (table 27). Almost 90 percent of men at Baltimore were said to "mix well," while at Boston this was said of less than two-thirds. As an isolated observation, this might be ascribed to the proverbial Boston reserve, but it more prob- ably represents yet another instance of lack of agreement between examiners Table 23.—Familial History of Nervous Disorders Other Than Epilepsy Numbe r of men Number of men Condition reported Condition reported Boston Other centers Boston Other centers None 104 503 Other nervous dis- Fainting 10 10 eases 10 15 Migraine 4 18 Unknown 3 8 20 17 10 20 Total men . 152 58' Psychopathic 4 1 22

as to the meaning of the words used in defining the classifications. Few men were considered at any center to be antisocial, but the proportions characterized as "feels out of place" or "asocial" vary markedly. Thus, at — Baltimore twice as many men were labeled asocial as were given the former a Arc designation, while at New York the ratio was more than reversed. Table 24.—Work Status at Time of Examination I Work status Vumber of mei I Baltimore Los Angeles Boston and New York 27 9 53 Army or VA hospital 1 4 3 Return to former job, or advanced in former job 15 13 13 56 20 72 20 9 21 New work 121 51 207 Occasional jobs 17 2 2 Unknown 1 2 Total 257 109 373 Table 25.—Changes in Work Status Trend Number of men HJJUJ- "rca# Baltimore New York Boston Los Angeles tol*: tvtit- None or regression 107 67 109 56 School and advancement 52 17 18 9 1 " Wf 97 132 24 41 1, w Unknown 1 5 1 3 oh*"- molt' Total 257 221 152 109 nfflf Table 26.—Assessment of Home Adjustment Numbe r of men Numbei of men Adjustment Adjustment Boston Other centers Boston Other centers 84 437 9 26 Reasonably satisfac- Unknown 3 12 45 75 2 9 Total 152 587 Restless 9 28 830802—82 23

Table 27.—Assessment of Social Adjustment Adjustment Numbei of men Baltimore New York Boston Los Angeles Mixes well and quite satisfactorily. Feels out of place 230 8 171 29 97 28 75 12 Asocial 16 13 20 17 Antisocial 1 2 2 1 2 6 5 4 Total 257 221 152 109 Assessment of economic adjustment was reasonably uniform at the low end of the scale, but apparently the distinction between "satisfactory" and "fair" adjustments was variously interpreted by the raters (table 28). The proportions called "fair" at the four centers were, respectively, 22.2 percent, 64.3 percent, 29.6 percent, and 41.3 percent. The clinical symptomatology was reported more uniformly by the four centers than were many of the clinical observations (table 29). The Los Angeles center much more frequently characterized men as having "im- paired mentation" or "impaired memory" than did the other centers; Boston and Los Angeles characterized few men as without symptoms. One feature complicating analysis of the clinical symptomatology was the practice of coding as "posttraumatic syndrome" any combination of three or more specific complaints, so that the fact that the Boston center reported a smaller proportion of men complaining of headache must be coupled with the fact that this center also reported the highest proportion with posttrau- matic syndrome. Complaints with respect to speech, vision, and hearing were not frequent, and specific tables are therefore omitted. It was apparent that the centers did not define comparably anomia, global aphasia, tinnitus, impaired hear- ing, blurred vision, and field defects. Complaints of paresthesias and motor weakness did not appear to vary remarkably between centers. The number of men who complained of sensory and motor disturbances was about the same (table 30). Table 28.—Assessment of Economic Adjustment Adjustment Number of men Baltimore New York Boston Los Angeles Satisfactory ... 151 55 88 49 Fair 57 142 45 45 Dissatisfied 40 23 17 12 9 1 2 3 Total 257 221 152 109 24

Table 29.—Complaints at Follow-up Examination Symptom Numbei of men Baltimore New York Boston Los Angeles 30 20 8 6 Headache 71 50 25 37 Dizziness 17 13 8 11 Impaired mentation or lack of 21 17 23 9 1 7 2 12 Impaired memory 4 7 1 8 Easy fatiguability 9 15 4 7 2 2 Posttraumatic syndrome 140 131 98 50 Unknown 1 1 2 1 Total men 257 221 152 109 Table 30.—Sensory and Motor Complaints Number of men Number of men Location to which complaints refer Somato- sensory parcs- thesias Location to which complaints refer Somato- sensory pares- thesias Motor weakness Motor weakness None 538 12 46 22 40 30 513 Other, unknown, and complaints not due to head Face 17 45 27 72 32 Arm Lee 52 39 Total men . . 739 739 The Los Angeles center was clearly less sensitive to complaints referable to other systems with the single exception of the peripheral nervous system. The Boston center, on the other hand, reported many complaints with respect to the alimentary and respiratory systems. Other discrepancies include the emphasis at Baltimore on the autonomic system and at New York with respect to the endocrine system (table 31). The four centers differed markedly in their reports of the mental status of the men (table 32). The Boston center concluded that two-thirds of their patients had abnormalities of mental state not related to the cerebral injury, whereas other centers had few such cases. The New York center reported 22 percent of their men as having impaired judgment or mentation, more than three times the number of the other centers except Los Angeles, which reported 15 percent. As might be expected, the tests of memory and the serial 7 subtraction test (table 33), being relatively objective, varied little by center in contrast with some of the more subjective items which have been considered.

Table 31.—Symptoms Referable to Other Systems System Numbei of men Baltimore New York Boston Los Angeles None . . .... 197 191 91 101 Alimentary 25 13 37 1 6 3 11 Cardiovascular 2 1 8 1 1 3 Genitourinary 12 1 6 1 Autonomic . . 12 4 2 7 3 6 4 Other and unknown 6 8 8 3 Total men . . 257 221 152 109 The Baltimore center noted an aphasic abnormality in about 6 percent of examinations, New York and Boston in about 10 percent, and Los Angeles in about 18 percent (table 34). The New York center most frequently found visual abnormalities (only 76 percent of men had normal vision); Los Angeles occupied middle ground (about 83 percent normal), while Baltimore and Boston found relatively little abnormality (about 88 percent normal) (table 35). At Los Angeles almost all defects found consisted of homonymous hemianopsia, while at the other centers the defects noted were more varied. The difference among centers in the proportions reported to have homonymous hemianopsia is statistically significant at the 5 percent level. The examination of variability between centers, demonstrating differ- ences between examiners, not only for subjective judgments but also for Table 32.—Examiner's Assessment of Mental Status Status Number of men Baltimore New York Boston Los Angeles Normal 228 147 35 78 1 8 4 1 2 Uncooperative 2 14 2 2 1 1 Depression 4 9 1 Wisecracking Impaired judgment or mentation, or lack of concentration 3 2 48 12 10 2 16 6 Irritability Abnormal finding not due to head wound 6 10 4 105 3 Other and unknown 1 Total men 257 221 152 109 26

Table 33.—Results of Memory and Serial 7 Test Number of men Number of men Number of errors Number of errors Memory Serial 7 test Memory Serial 7 test None 204 244 127 68 22 14 8 9 333 129 75 53 39 16 15 11 8 9 8 1 9 3 7 14 2 >10 17 3 Not done or un- satisfactory 4 14 39 5 6 Total 739 739 7 Table 34.—Assessment of Aphasia at Follow-up Number of men Type Baltimore New York and Boston Los Angeles None 242 337 89 Expressive 2 14 6 Nominal 5 1 Global 7 12 9 Combinations 3 2 Other and unknown 3 3 4 Total 257 373 109 Table 35.—Assessment of Vision at Follow-up Vision Baltimore and Boston New York Los Angeles 359 168 90 Blind 4 3 i Vision impaired 9 13 Homonymous hcmianopsia 28 24 14 Other, unknown, and defects not due to head wound 14 17 5 Total men . . . 409 221 109 Number of men

presumably objective evaluations, is not reassuring. To be sure, the in- ference of differences between examiners rests on the assumption that the men examined at the various centers were essentially alike, and this assump- tion is probably not wholly valid. This is illustrated by the consideration of the hemianoptic disturbances, the demonstration of which is essentially objective. It seems more reasonable to conclude that the differences demonstrated for this observation are attributable more to variability in the samples of men studied at the centers than to differences in interpreta- tion of examiners. On the other hand, the very great variation in the pro- portions of psychiatric diagnoses assigned by the centers hardly allows any conclusion but that standards varied between examiners. This is not un- expected, since the examiners in the various centers had different specialty backgrounds, representing psychiatry, neurology, and neurological surgery. The existence of this variability must be borne in mind in connection with the analyses to be reported below and, where especially pertinent, will be referred to explicitly. F. THE INJURY SCORE It seems clear, in an intuitive sense, that head wounds vary with respect to a parameter that may be called "severity," but just how to assess this is not immediately apparent. Evidently, from the standpoint of severity, head wounds are multidimensional; such aspects as which organs or kinds of tissue have been damaged, and how much tissue of a particular kind has been destroyed, are not simply additive. The necessity in this study for finding some relatively simple way to characterize the severity of each head wound arises from the fact that the extent of damage is correlated with every aspect of treatment and prognosis. Thus, if it be inquired whether a cranioplasty increases the probability of subsequent epilepsy, there has to be taken into consideration that (7) the more severe the wound, the greater the chance that a cranioplasty was done, (2) the more severe the wound, the greater the chance that the patient has had fits; and, therefore, (3) the incidence of epilepsy is higher in men with cranioplasty than in men without. Since what is desired is a measurement of the direct effect of cranioplasty on epilepsy, independent of correlations with wound severity, it seemed necessary to create a simple characterization of the latter. Tabulations might be controlled by such a measure, and groups of men whose wounds were of equivalent average severity might be defined. For these purposes an "Injury Score" has been devised. The general method used to create the scoring system was to select an end-result criterion, and by this to evaluate the apparent relative importance of the different aspects of severity, and so to assign a score to every partic- ular category in each aspect. The resulting scores were added together for each patient to form the "Injury Score." The specific end result employed was the presence or absence of epileptic seizures during the first 2 years after injury. This information was gen- erally available, even for men who were not examined. This criterion was assessed against each of 15 items descriptive of the injury. The different categories in each item were then combined into a smaller number of

groups rather arbitrarily, the choice depending not only on what seemed to be a natural way of grouping, but also on the number of cases available for each category and the variation in the proportion with epilepsy. The logarithm of the percent with epilepsy was then used as a raw score for the category. The final score for the category was obtained from the raw score by first multiplying by the ratio of the mean difference between men with and without epilepsy to the variance of the raw scores in all 932 men, and then subtracting a suitable constant so that the category corresponding to the least severe injury would have a score of zero. The procedure may be expressed algebraically more understandably than in prose : Let Ti be the percent with epilepsy in the ith category. Then zt is the raw score, where 2,=log rt. Let nt and mt be the numbers of men with and without epilepsy, respec- tively, in category i. If " 22= t then and 2_-1_ ° ~ Sfa+m,) The final score, x«, is given by v &— gp)(gi— 50 xt- —j— where 20 is an arbitrary constant. The categories that were established, and the final scores for them are shown in table 36, while the final result of all these arithmetic procedures is displayed in table 37. The Injury Score, which is the sum of the 15 individual item scores, varied between 0 and 750. Men whose scores fell into the low range (below 250) had an incidence of epilepsy of only 7.4 percent, while in the intermediate range (250-499) the incidence was 24.2 percent, and in the high range (500 and over) it was 50.9 percent. More- over, among men with low scores who did have fits, only about one-tenth had focal attacks; among men with high scores the majority (58 percent) had focal attacks. It is worth remarking that the particular scoring system used here is dependent on the choice of epilepsy as an end-result criterion. It is quite possible that if another criterion had been used the scores given in table 36 might have been quite different. Since the main interest of the present study is in posttraumatic epilepsy, the scoring system derived here seems appropriate to our purposes. It may or may not be suitable to other purposes. It should be noted that the data used in constructing the scores were those available in military records. The significance assigned to each factor by the scoring system depends not only on the intrinsic importance of the factor but also on the accuracy with which the records reflect the true con- dition of the subject.

Table 36.—Scoring System for Injury Scores Total number of men With epilepsy Characteristic of injury Number Percent Score Type of wound: Closed head injury without fracture, fracture of outer table only, scalp 323 27 8.4 25.9 Closed head injury with fracture of convexity or base, or compound noncomminuted (linear) 27 7 44 Closed head injury with depressed ikull fracture or compound commi- nuted fracture (penetrating wound) . . . 554 28 150 8 27.1 28.6 46 49 Agent: 74 858 8 184 10.8 21.4 0 26 Bullet, shell fragment, blunt object. . . . Regions wounded: Vertex, basal, sinuses, mastoid, orbit, sphenoid, other or combinations of these 94 155 10 20 10.6 12. <T 0 1 Temporal or occipital only Frontal alone, or temporal or occipital 143 74 213 253 22 14 15.4 18.9 25.4 28.5 21 25 37 46 Frontal and other, but not parietal. . . 54 72 Parietal and other Mean diameter of cranial defect: No defect 347 229 35 55 61 32 9 10.1 24.0 26.1 32.0 40.9 0 38 42 52 66 2 -{-through 4 cm 234 100 22 ^4 cm Depth: Scalp, cranium or dura mater (after 471 20 414 27 47 4 130 11 10.0 20.0 31.4 40.7 0 31 54 69 Cranium, before debridement Dura mater before debridement or brain . Ventricle Second head wound: None, or not involving frontal or temporal regions 882 50 174 18 19.7 36.0 0 41 Of frontal or temporal region Immediate complications: None 707 116 16.4 0 Hematoma, frank infection, abscess, 126 11 39 4 33 31.0 36.4 32 46 50 Meningitis or cerebrospinal fluid leak . . 88 37.5 Period of unconsciousness: None, or <C[2 hours, or unknown 578 204 88 55 15.2 27.0 0 35 2—24 hours, or unconscious for un- stated time 1-7 days, or initially conscious, with later coma 131 19 42 7 32.1 36.8 42 50 *^>\ week Period of confusion: None 91 540 10 91 11.0 16.9 0 21 <24 hours or unknown 1 —3 days or confused for unstated period 184 117 46 45 25.0 38.5 43 70 >3 days

Table 36.—Scoring System for Injury Scores—Continued Characteristic of injury Total number of men With epilepsy Number Percent Score Neurological deficit at time of wound: None 538 Hemianopsia or unspecified deficit alone 152 Aphasia, with or without hemianopsia. 55 Hemiplegia, without aphasia or he- mianopsia 107 Hemiplegia with aphasia or hemia- nopsia or both 80 Extent of first debridement: None, or involved scalp or skull only. . 441 Involved dura mater 59 Unstated 53 Involved brain 361 Involved ventricle 18 Intracranial foreign bodies: No foreign bodies, or no debridement. . 387 Unstated 46 Bone or metal or both, completely removed at debridement 236 Bone or metal or both, incompletely removed at debridement 263 Complications of debridement: No debridement 143 None, or cerebrospinal fluid leak only. 669 Hematoma, frank infection, abscess, meningitis, fungus cerebri or other... 120 Second debridement. Not done 716 Done for retained foreign bodies or unspecified reason 146 Done for infection, draining sinus, intracranial bleeding or abscess 70 Additional debridements beyond second: Not done 880 Three debridements done 38 >3 debridements 14 68 30 17 40 37 44 13 14 113 8 48 9 56 79 17 131 44 128 34 30 174 10 8 12.6 19.7 30.9 37.4 46.3 10.0 22.0 26.4 31.3 44.4 12.4 19.6 23.7 30.0 11.9 19.6 36.7 17.9 23.3 42.9 19.8 26.3 57. 1 35 51 61 76 0 35 44 53 74 0 21 30 43 0 25 66 0 16 61 0 39 G. ROSTER-INJURY GROUPS As stated earlier, there seemed reason, initially, to fear that rosters drawn from VA compensation lists might be biased in the direction of residual impairment. The Injury Score, previously described, was used in an at- tempt to determine to what extent cases drawn from the two rosters might differ as to the probability of residual impairment, given wounds of equal severity. With respect to the incidence of epilepsy, it was quite apparent that among men with low scores (relatively mild injury), those drawn from the VA rosters were much more likely to have fits than those who came from the Army admissions lists (table 38). For men whose scores were 250 or more, however, there seemed to be essentially no difference. Between 250 31

8-" r^ oa .«. I Numbe -« Q iu k O a '5* . 2 K n £ 9 .5 i Numbe in o *^ ^^ >o T c v 2 £ Numbe o § TI C4 CO CS N CM ?. •r I ^ Numbe r 2 o H 32

Table 38.—Comparison of VA and Army Rosters as to Incidence of Epilepsy Within 2 Years of Injury in Relation to Injury Score From VA rosters From Army rosters From VA rosters From Army rosters Total Total Injury score Number of men Percentage fits in 2 years Total 932 469 463 20.6 19.2 22.0 0-99 186 142 103 193 196 88 24 107 105 58 73 68 43 15 79 37 45 120 128 45 9 * 4. 3 12.7 f> 9.7 23.3 27.6 52.3 45.8 7.5 15.2 10.3 20.5 20.6 58.1 40.0 100-199 5.4 8.9 25.0 31.3 46.7 55.5 200-299 300-399 400-499 500-599 >600 and 500, the incidence of epilepsy seemed to vary but little, but rose rather sharply for scores of 500 or more. On the basis of these findings it seemed appropriate to divide the men into four groups (table 39), henceforth referred to as the Roster-Injury groups (or R-I groups), defined as follows: R-I group 1: From Army roster, injury score 0-249. R-I group 2: From VA roster, injury score 0-249. R-I group 3: Both rosters, injury score 250-499. R-I group 4: Both rosters, injury score 500-750. These four groups were subsequently used throughout the analysis in an at- tempt to free correlations from at least the grossest artifacts. Thus, as will be seen, the frequency of abnormal electroencephalographic tracings in- creases as one passes from R-I group 1 to group 4, as does the frequency of epilepsy. Hence, if the material be taken as a whole, it would be im- possible to know to what extent correlations between these two factors were fortuitous and to what extent the brain wave tracings were actually revealing inherent characteristics related to epilepsy. No post hoc survey, such as the present study, can ever hope really to free itself of such difficulties, but it is believed that here it has been possible to minimize these disturbances by means of controlling tabulations by the R-I groups. Table 39.—Composition of R-I Groups Number of men Incidence of epilepsy in 2 years Number of men Incidence of epilepsy in 2 years R-I group Percent 1.5 R-I group Percent 50.9 1 135 4 112 2 243 10 7 3 442 24.2 Total 932 20.6 33

Next: The Chronic Neurological Deficit »
Follow-Up Study of Head Wounds in World War II, by a. Earl Walker and Seymour Jablon Get This Book
×
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF
  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!