APPENDIX H
Comparison of World War II Missile Casualty Data
Allan Palmer, M.D.
Detailed reports of missile casualty data obtained duringWorld War II by special wound ballistics teams have already been presented inthis volume. (See chapters IV, V, VI, VII, VIII, andIX.) A compilation of these data are presented in the statistical material inthis appendix.1 In addition, casualty data from previous wars have also beenincluded as a matter of interest. However, no extensive comparisons have beenmade between casualties sustained during World War II and those sustained inprevious wars because of the difference in weapons employed and of thedifference in the medical and surgical eras during which the casualties weresustained.
The percentages in the tables that pertain to the regionaldistribution of wounds refer to the total number of wounds. In the case ofregional frequency, only the frequency with which the various regions of thebody are wounded is considered regardless of the number of wounds in each bodyregion. The percentages in the tables which pertain to regional frequency ofwounds refer to the number of casualties. From such tabulations, casualties whosustained wounds in more than one region of the body must be excluded or anadditional entry made for them.
Table 1 shows the regional distribution of wounds due to allmissiles in WIA (wounded-in-action) only, in three wars. The presentation ofdissimilar samples is unavoidable since the statistical data have not beencollected in a uniform manner. The outstanding difference in wound distributionin the various surveys is the relatively low incidence of chest and abdominalwounds in casualties sustained by the Eighth Air Force bomber crew memberswearing body armor.2
There are only two surveys available on regional distributionof wounds in KIA (killed-in-action) casualties where the exact locations of allentry wounds have been recorded. Table 2 shows the regional distribution of theentry wounds due to all missiles in these two studies. Except for the moderatelyhigh incidence of chest hits (22.9 percent) in the Fifth U.S. Army dead, thelocation of hits approaches a random distribution; that is, the percentages ofhits in the various regions are proportional to the mean projected areas of thevarious body regions. The protective effect of body armor in the trunk region,particularly the chest, is again demonstrable in the Eighth Air Force dead asshown by a wound incidence of 20.0 percent (chest and abdomen combined) ascompared with 29.4 percent for the Fifth U.S. Army dead.
Table 3 shows the regional frequency of wounds due to allmissiles in the killed in action of the Fifth U.S. Army and the Eighth Air Forcesurveys plus two additional surveys of U.S. battle deaths in the South PacificArea. The regional frequency is shown by single and multiple regions wounded,and the numbers and the percentages refer to the number of
1Some variations will be notedin the statistical data presented in this section as compared to the datapresented in the chapters referred to. This is due in part to the separation andreevaluation of the original survey findings in an attempt to compare theresults from the varied sources upon similar terms. In addition, certain minorchanges were made in the statistical data during the preparation of this volume.The statistical material pertaining to other than U.S. Army casualties waspersonally collected by the author during his Army service.-J. C. B.
2The presentation of datacomparing ground troop and aircrew casualties might seem unfeasible, but they do indicate regional incidence of missile-inflicted wounds with a commonimplication for the development of personnel armor.-J. C. B.
844
TABLE 1.-Percent regional distribution of wounds due to all missiles, from casualty samples of wounded in action only, in three wars
War and survey | Casualty sample | Body region | |||||
Number of casualties | Number of wounds | Head | Chest | Abdomen | Upper extremity | Lower extremity | |
American Civil War | --- | --- | 9.1 | 11.7 | 6.0 | 36.6 | 36.6 |
World War I: | |||||||
| United States | --- | --- | 11.4 | 3.6 | 3.4 | 36.2 | 45.4 |
| British | --- | --- | 16.8 | 7.8 | 4.7 | 30.4 | 40.3 |
World War II: | |||||||
| Canadians1 | --- | 469 | 11.0 | 11.0 | 4.0 | 31.0 | 43.0 |
| New Zealanders2 | 476 | 769 | 13.0 | 7.0 | 5.0 | 25.0 | 50.0 |
| British2 | 112 | 202 | 14.0 | 6.0 | 3.0 | 28.0 | 49.0 |
| German2 | 370 | 504 | 9.0 | 9.0 | 3.0 | 40.0 | 39.0 |
| Allies3 | 183 | 352 | 10.0 | 11.0 | 7.0 | 33.0 | 39.0 |
| United States: | |||||||
New Georgia-Burma | 4230 | --- | 18.2 | 13.5 | 6.1 | 27.0 | 35.2 |
Bougainville | 41,162 | --- | 20.7 | 12.4 | 5.7 | 27.4 | 33.8 |
Cassino | 100 | 133 | 20.0 | 11.0 | 7.0 | 22.0 | 40.0 |
Normandy | 33,000 | --- | 16.1 | 9.8 | 5.6 | 28.2 | 40.3 |
Eighth Air Force | 51,007 | 1,298 | 19.8 | 4.9 | 2.2 | 29.4 | 43.7 |
| U.S.S.R. | --- | --- | 9.1 | 11.4 | 6.2 | 28.0 | 45.3 |
1At Dieppe Raid.
2In Tunisia.
3In Sicily.
4Casualties with multiple regions wounded are omitted from sample,and percentages are based upon number of casualties, not number of wounds.
5Heavy bomber aircrew personnel wearing body armor.
TABLE 2.-Regional distribution of wounds due to all missiles in 1,000 Fifth U.S. Army and 164 Eighth Air Force KIA casualties only
Body region | Fifth U.S. Army, Italy | Eighth Air Force, Europe | ||
Number of wounds | Percent of wounds | Number of wounds | Percent of wounds | |
Head | 680 | 10.5 | 105 | 23.3 |
Chest | 1,484 | 22.9 | 58 | 12.9 |
Abdomen | 425 | 6.5 | 32 | 7.1 |
Extremities: | ||||
| Upper | 1,538 | 23.7 | 145 | 32.1 |
| Lower | 2,360 | 36.4 | 111 | 24.6 |
| Total | 6,487 | 100.0 | 451 | 100.0 |
casualties. Comparison of tables 2 and 3 shows the strikingdifferences between the regional distribution of wounds and their regionalfrequency when dealing with samples of killed in action and died of wounds only.It has been observed that the dead are more frequently hit in more than oneregion of the body than is the case with the wounded. It is of interest to notethat the regional frequency of hits in the Fifth U.S. Army casualties approachesmore closely that for the Eighth Air Force dead than it does that for the deadthat were studied in the Pacific theater. The effect of the wearing of bodyarmor is again apparent in the air force study. The similarity in these twosurveys is probably due to the fact that in both of
845
them the preponderance of missiles causing the casualtieswere high explosive shell fragments. The incidence of multiple regions hit wasat least twice as great in both samples as it was in either of the samples ofdead from the Pacific theater. The increased proportion of small arms or"aimed" fire characteristic of the warfare in the Pacific theateraccounts for the high incidence of head and trunk wounds in these samples. Thisis an extreme departure from the randomness of hits as well as from the highincidence of wounds in more than one region of the body characteristic ofcasualties exposed to shell fragments.
TABLE 3.-Regional frequency of wounds due to all missiles in four samples of battle deaths and KIA casualties
Body region | New Georgia-Burma1 | Bougainville2 | Fifth U.S. Army, Italy | Eighth Air Force, Europe | ||||
Number of casualties | Percent of casualties | Number of casualties | Percent of casualties | Number of casualties | Percent of casualties | Number of casualties | Percent of casualties | |
Single region wounded: | ||||||||
| Head | 32 | 31.7 | 144 | 36.4 | 171 | 17.4 | 50 | 30.5 |
| Chest | 32 | 31.7 | 87 | 22.0 | 138 | 14.0 | 16 | 9.8 |
| Abdomen | 11 | 10.9 | 48 | 12.2 | 30 | 3.0 | 3 | 1.8 |
| Upper extremity | 1 | 1.0 | 1 | .3 | 25 | 2.5 | --- | --- |
| Lower extremity | 2 | 2.0 | 14 | 3.5 | 62 | 6.3 | 11 | 6.7 |
Multiple regions wounded | 23 | 22.7 | 101 | 25.6 | 559 | 56.8 | 84 | 51.2 |
| Total | 101 | 100.0 | 395 | 100.0 | 985 | 100.0 | 164 | 100.0 |
1Includes both KIA casualties andcasualties who died of wounds.
2Primarily KIA casualties only.
In the consideration of causes of death, a distinction hasbeen made between the causes of death on the one hand and fatal wounds on theother. It was obvious that in many cases more than one wound could have been thecause of death. The following criteria were followed in order to determine thecause of death:3
1. Only the severest one of multiple fatal wounds wasregarded as the cause of death in any one casualty.
2. When the severity of a head and a chest or an abdominalwound appeared to be the same, the cause of death was arbitrarily attributed tothe head wound.
3. When the severity of a chest and an abdominal woundappeared to be the same, the cause of death was attributed to the chest wound.
4. Decapitations were regarded as causes of death due towounds in the head and neck region in cases where the head was missing as wellas in cases where a head wound was very extensive and associated with completeevulsion of the brain.
5. In the case of extensive mutilating wounds, the cause ofdeath was attributed to a wound of the region of the body nearest the center ofthe area of mutilation.
Table 4 shows the causes of death in six studies of bothmilitary and civilian casualties due to all missiles according to the region ofthe body in which the primary fatal wound occurred regardless of the regionfirst struck by the missile and regardless of the multiplicity of fatal wounds.Thus, the causes of death in these samples are not more numerous than the numberof casualties.
3During the conduct of a survey,it is frequently necessary to adopt arbitrary criteria for the determination ofthe cause of death. However, autopsy studies have revealed the shortcomings ofsuch a method. A thorough study of smaller group of casualties can be moreinformative than a superficial survey of a larger number. Therefore, casualtysurveys should be conducted with adequate personnel to permit complete externalexamination of all wounds and an adequate autopsy study for the determination ofthe cause of death.-J. C. B.
846
TABLE 4.-Percent distribution of cause of death in military and civilian casualties, by region in which the primary fatal wound occurred
Body region | American Civil War | British civilians in London1 | New Georgia-Burma2 (78 casualties) | Bougainville2 (294 casualties) | Fifth U.S. Army, Italy (981 casualties) | Eighth Air Force, Italy (164 casualties) |
Head | 41.5 | 37.0 | 41.0 | 49.0 | 43.7 | 45.1 |
Chest | (3) | 20.0 | 41.0 | 29.6 | 36.7 | 38.4 |
Abdomen | (3) | 33.0 | 14.0 | 16.3 | 8.3 | 7.4 |
Extremities: | ||||||
| Upper | 2.6 | 4.0 | 2.0 | .3 | 2.3 | --- |
| Lower | 4.5 | 6.0 | 2.0 | 4.8 | 9.0 | 9.1 |
| Total | 4100.0 | 100.0 | 100.0 | 100.0 | 100.0 | 100.0 |
1Killed by bomb splinters during the"blitz" in 1941.
2Casualties with secondary fatal wounds omittedfrom sample.
3Data are not available for the individual bodyregions; for the combined regions, the figure is 51.4 percent.
4Includes 51.4 percent for wounds of the chest andthe abdomen combined.
It may be seen in all of the casualty surveys that wounds of the head andneck region account for the greatest number of fatalities. The chest region issecond in all samples except in that of British civilians in London during the"blitz" of 1941. It is possible that the suddenness of wounding bybomb splinters in unarmed and unprotected civilians, in contrast with themilitary, might account for a greater number of deaths due to abdominal woundsin civilians.
Whether or not there is complete random distribution of wounds in missilecasualties can only be ascertained in complete samples of unselected casualties.The sample must include the slightly as well as the severely wounded and thekilled. Table 5 shows the relative mean projected surfaces of the various bodyregions which may be regarded as the relative regional distribution of woundsexpected in a sample of casualties exposed to random distribution of themissiles causing wounds. The variations from the expected wound distribution forfive samples of casualty data are also shown in table 5. A lower than theexpected number of wounds in the chest and the abdomen in the case of the airforce casualties was due primarily to the wearing of body armor by aircrewpersonnel. The higher incidence of head and trunk wounds due to aimed fire orsmall arms is apparent in the casualties sustained by the ground forces in thePacific theaters.
Just as in the case of wounded in action only casualty studies, there areonly slight and insignificant differences in regional distribution and regionalfrequency of wounds in complete casualty samples. Table 6 shows the relativeregional frequency of wounds due to all missiles in three of the completecasualty surveys previously discussed. The incidence of casualties wounded inmore than one region of the body in the three complete casualty samples isfairly constant-ranging as it does from 14.9 to 18.6 percent. By excludingcasualties wounded in multiple regions from the data in table 6, the greatestdifferences between the regional distribution and the regional frequency ofwounds would be found in the Eighth Air Force survey. Although 40.4 percent ofall wounds occurred in the lower extremities (table 5), if those wounded inmultiple regions were excluded from the sample instead of being tabulated intable 6 as "multiple regions," the value of 38.3 percent in table 6would become 45.0 percent, the difference between regional distribution andregional frequency then being 4.6 percent. Thus, it may be concluded that in ananalysis of the regional distribution or frequency of wounds in completecasualty studies the exclusion of those casualties wounded in more than oneregion of the body does not materially alter the apparent incidence of wounds inthe various body regions.
847
TABLE 5.-Percent regional distribution of wounds due to all missiles, from six surveys of WIA and KIA casualties, by body region
Body region | Body surface area1 | Dieppe Raid2 | North Africa3 | New Georgia-Burma4 | Bougainville5 | Eighth Air Force Europe6 |
Head | 12.0 | 16.0 | 21.5 | 20.1 | 26.4 | 21.1 |
Chest | 16.0 | 13.0 | 17.3 | 20.6 | 15.9 | 6.4 |
Abdomen | 11.0 | 10.0 | 9.2 | 10.4 | 7.8 | 3.3 |
Extremities: | ||||||
| Upper | 22.0 | 26.0 | (7) | 21.5 | 22.0 | 28.8 |
| Lower | 39.0 | 35.0 | (7) | 27.4 | 27.9 | 40.4 |
| Total | 100.0 | 100.0 | 8100.0 | 100.0 | 100.0 | 100.0 |
1Percent expected hits.
2Based on 496 WIA and an estimated 124 KIA casualties.
3Based on a total of 3,919 casualties.
4Based on a total of 369 casualties, with 268 WIA casualtiesand 101 battle deaths, including killed in action and died of wounds.
5Based on a total of 1,546 casualties, with 1,162 WIAcasualties and 294 battle deaths, including killed in action and died of wounds.
6Based on a total of 1,115 casualties, with 1,007 WIA and108 KIA casualties.
7Data are not available for the individual body regions; forthe combined extremities, the figure is 52.0 percent.
8Includes 52.0 percent for wounds of the upper and lowerextremities combined.
It has been observed that shell fragments hit the body moreat random than the aimed fire of bullets. While initial fragment velocity isoften high, the striking velocity is commonly less than that of bullets atbattle ranges, due to rapid air retardation. This effect is largely due tosectional density and form factor. It is this fact which makes body armor ofvalue in protecting against fragment injury, while it would appear impracticalto contemplate an armor which could materially prevent rifle bullets fromcausing wounds of the protected areas. Thus, it is proposed that the protectiveeffect of body armor be evaluated on the basis of observed hits on personnelstruck by shell fragments only. Ideally, a comparison of the anatomic locationof hits on unselected samples of armored and unarmored troops would best revealthe effectiveness of protection. The exact anatomic locations of all hits byhigh explosive shell fragments on the surface of the body have been accuratelyrecorded in one casualty survey comprised of both the wounded and the killed,that being the 961 Eighth
TABLE 6.-Percent regional frequencyof hits due to all missiles, from three surveys of WIA and KIA casualties, by body region
Body region | Body surface area1 | New Georgia-Burma | Bougainville | Eighth Air Force |
Single region: | ||||
| Head | 12.0 | 20.1 | 21.5 | 19.8 |
| Chest | 16.0 | 17.1 | 12.9 | 3.4 |
| Abdomen | 11.0 | 6.8 | 6.4 | 1.5 |
| Upper extremities | 22.0 | 17.1 | 17.9 | 22.1 |
| Lower extremities | 39.0 | 22.4 | 22.7 | 38.3 |
Multiple regions | --- | 16.5 | 18.6 | 14.9 |
| Total | 100.0 | 100.0 | 100.0 | 100.0 |
1Percent expected hits.
848
Air Force flak casualties sustained during June, July, and August 1944 (ch.IX).All of these casualties may be regarded as being "armored."Although the exact incidence of those casualties who were not actually wearingbody armor at the time they were wounded or killed is not known, it is knownthat at least 11 percent were unarmored.
A further evaluation of the protection afforded by body armor may be madefrom a study of the quantitative relationship (indices of vulnerability) betweenobserved hits and expected hits based upon projected body surface areas. In arelationship of this sort, the nearest approach to random distribution of hitswould be expected in a selected sample of casualties due to only fragments fromhigh explosive shells, and the least evidence of randomness would be expected ina selected sample of casualties due only to bullets; that is, "aimed"fire. Since body armor is the subject under discussion, it is felt that selectedsamples of casualties due to high explosive shell fragments are best suited forthis demonstration. Warfare in which bullets cause the majority of casualtieswould not be the type of warfare in which body armor would be of greatest value.A purely random distribution of hits on unprotected individuals would cause allthe indices to be 1.00.
The regional frequency of hits due only to shell fragments in a sample ofunarmored ground force troops may be compared with the regional frequency ofhits sustained by the armored Eighth Air Force casualties. Table 7 shows therelative regional frequency of hits in the various body regions of the unarmoredBougainville casualties as compared with that of the armored Eighth Air Forcecasualties.
TABLE 7.-Relative vulnerability of differentbody regions to shell fragments (multiple wounds excluded) from two surveys of WIA and KIA casualties
Body region | Body surface area or hits expected (percent) | Bougainville1 | Eighth Air Force2 | ||
Hits observed (percent) | Index3 | Hits observed (percent) | Index3 | ||
Head | 12.0 | 24.5 | 2.04 | 16.1 | 1.34 |
Chest | 16.0 | 15.3 | .96 | 4.6 | .29 |
Abdomen | 11.0 | 6.6 | .60 | 1.8 | .16 |
Extremities: | |||||
| Upper | 22.0 | 21.9 | 1.00 | 29.0 | 1.32 |
| Lower | 39.0 | 31.7 | .81 | 48.5 | 1.24 |
| Total | 100.0 | 100.0 | 100.0 | ||
1Based on 707 casualties.
2Based on 818 casualties.
3Index=(Percent of hits observed)/(Percent of hits expected)
The action in which the Fifth U.S. Army in Italy participated and in which attimes as many as 85 percent of the casualties were due to shell fragments wasthe sort of warfare which defensively would be ideally suited to the wearing ofbody armor by ground force troops. Casualty survey observations on the regionaldistribution of hits due only to shell fragments in this action, however, wererestricted to a sample of KIA only casualties. Table 8 shows for comparison theregional distribution of hits due only to shell fragments and the indices ofvulnerability in samples of 850 unarmored Fifth U.S. Army dead and 144 armoredEighth Air Force dead.
It is not fair to attempt to evaluate protection afforded by armor on thebasis of observations confined to killed in action only. The chest and abdominalregions are still relatively vital regions of the body even when armored, andthe fatalities resulting from fatal wounds in these regions were obviously dueto the relatively higher velocity perforating flak fragments which struck theseregions in armored aircrew personnel. These fragments approached and
849
actually may have had velocities which were comparable to thevelocities of bullets. A point which may be observed, however, in the twosurveys with reference to protection is the difference in the distribution ofwounds. The sample of air force dead may be regarded generally as having wornhelmets as well as body armor as opposed to the sample of ground force deadwhich may be regarded generally as having worn helmets but not body armor.Therefore, with greater vital body area coverage by body armor as compared toarea of coverage by helmet only, the incidence of head wounds due to shellfragments in air force dead was more than twice that in ground force dead. Thelow incidence of head wounds due to high explosive shell fragments in the deadof the Fifth U.S. Army was the only instance in all of both the complete and KIA-onlycasualty surveys studied where the incidence of wounds was less than theprojected surface area of that region; that is, less than the expected woundincidence. Figures 1 and 2 show the anatomic location of the hits given in table8 for the ground force and air force casualties, respectively.
FIGURE 1.-Anatomic location of 6,003 hits on 850 KIA due toshell fragments, Fifth U.S. Army, Italy.
850
FIGURE 2.-Anatomic location of 373 hits on 144 KIA due toflak fragments, Eighth Air Force, Europe.
TABLE 8.-Percent regionaldistribution of wounds and relative vulnerability of body regions toshell fragments, from two casualty surveys of KIA casualtiesonly
Body region | Body surface area or hits expected | Fifth U.S. Army1 | Eighth Air Force2 | ||
Hits observed | Index3 | Hits observed | Index3 | ||
Head | 12.0 | 10.4 | 0.87 | 24.7 | 2.06 |
Chest | 16.0 | 22.4 | 1.40 | 11.5 | .72 |
Abdomen | 11.0 | 6.6 | .60 | 5.9 | .54 |
Extremities: | |||||
| Upper | 22.0 | 23.5 | 1.07 | 36.5 | 1.66 |
| Lower | 39.0 | 37.1 | .95 | 21.4 | .55 |
Total | 100.0 | 100.0 | 100.0 | ||
1850 casualties with 6,003 wounds.
2144 casualties with 373 wounds.
3Index=(Percent of hits observed)/(Percent of hits expected)
