CHAPTER 1
Internal State of Severely Wounded Men on Entry to the Most Forward Hospital
The effects on the human body of the destructive forces of warfare have been described many times in terms of organic damage and tissue loss. Our concern was rather with the internal state of the severely wounded man. Gross tissue damage is obvious, or becomes obvious on surgical exploration, but our purpose during the first phase of this investigation was to describe the latent consequences of the wound as revealed in impairment of organic unction and in abnormalities of the blood and the urine. These initial studies were made shortly after the patient entered the most forward field or evacuation hospital, before either vigorous resuscitative measures or operation had yet been undertaken. The physiologic studies were continued, whenever possible, throughout the patient`s course. Other aspects of the investigation as a whole relate to diagnosis, treatment, and pathology of the severely wounded.
The very severely wounded ("nontransportable patients") were those selected for study. They were the most critically wounded or injured battle casualties to reach a forward hospital alive. With few exceptions, chiefly cases ofinjury,1 the casualties2 studied were from the "wounded in action"3 group. The cases are listed in Appendix D.
1AR 40-1025, Sec V, par 79a, 12 Dec 44, sub: Definition [of injury]. "The term `injury` is used here in its broad sense to include such conditions as fractures, wounds, sprains, strains, dislocations, concussions, and compressions, commonly thought of as `accidents` . . ."
2ASF Manual M 807, 25 Oct 44, Glossary. "Casualty (Personnel). A soldier who is rendered unavailable for service as a result of disease, injury, or enemy action . . ."
3AR 40-1025, Sec II, par 26, 12 Dec 44, sub: [Definition of] WIA (wounded in action) cases. "The term will include wounds or injuries incurred as a direct result of a hostile act of a military enemy. It will not include injuries accidentally incurred while in combat, or those incurred on purely training flights or missions."
22
TABLE 1.-TIME FROM WOUNDING TO SURGERY, MEDITERRANEAN THEATER OF OPERATIONS
In all, 186 casualties were examined in the most forward hospitals by members of the Board. From previous studies made in the Theater, it was estimated that of 10,073 battle casualties in the area to reach forward hospitals alive during the period of the study, between 201 and 252 were seriously wounded. Hence the 186 studied here may be considered an adequate sample of the severely wounded in the Theater. One hundred and eight of these 186 casualties were seen at the time of admission and were studied rather completely (including blood chemistry and urine analyses) at that time. Account was taken of the nature and type of the wound, and also of the evacuation time, including the distance to be covered and the character of the terrain, since delay along the evacuation trail, the reaction of the patient to his wound, and his response to subsequent management all influence the factors under study and increase the significance of the laboratory data.
In addition to the data obtained as background material, the initial studies included determination of blood loss, of plasma protein and hemoglobin levels, analysis of other biochemic changes encountered, initial kidney function studies, and a study of liver function in the newly wounded man.
It will be observed in the tables and charts of this and following chapters that different groups and varying numbers of patients have been drawn from the total for consideration in given instances. This has been done because it was often found in comparing two or more factors that records were incomplete for the specific comparison in question and had to be omitted. As a re-
23
TABLE 1.-TIME FROM WOUNDING TO SURGERY, MEDITERRANEAN THEATER OF OPERATIONS-Continued
sult comparatively small numbers of cases are presented in some instances. No attempt has been made to keep the number of cases uniform in any of the various phases of the study; rather we have presented all the data that were complete for any one phase. This method was considered desirable because of the nature of the study and the exigencies under which it was carried out. In the tables throughout the study the standard error of the mean is shown whenever the data were sufficient to warrant this method of statistical treatment.
Initial Studies
Time from Wounding to Hospital
Entry and Surgery
Although some of the casualties were wounded near the forward hospitals, the majority had to be transported some distance, often over mountainous terrain, by litter carry or motor transport. Since the time required to transport a patient from the place of wounding to the most forward hospital may greatly influence his condition on arrival, some indication of the length of this period in the Mediterranean Theater of Operations is given in Table 1, which shows the average progress of three groups of casualties along the evacuation route. The first group consists of 100 men selected at random from those in our study
24
25
who were wounded during a relatively quiet period in the fall, winter, and spring of 1944-45. The other two groups represent men who were severely wounded during offensives in the spring of 1945. In the third locale cited, it was contended by those concerned that, considering the circumstances, evacuation had been effected rapidly. The table also shows the average time from hospital admission to surgery and the total time from wounding to surgery in the three groups.
Type and Location of Wounds
For various correlations throughout the study wounds are grouped according to their type, or location, or both. Many patients incurred multiple wounds, some multiple major wounds. For certain purposes two broad classifications of type were utilized: peripheral and non peripheral, and this terminology will be used whenever pertinent. Nonperipheral wounds were defined as those involving the major body cavities (the abdomen, the thorax, and the interior of the skull); all others were considered as peripheral. Crush cases are excluded in some of the correlations because they were studied separately.
In the following classification the severe wounds only are considered, since they were pertinent to the study. Thus in the patients with multiple severe wounds some wounds were listed as the principal major ones; no attempt was made to record minor wounds, such as fracture of a phalanx, for example. In general, the types of wounds found in our patients were as follows:
Severe peripheral wounds were present in 116 patients, constituting a major injury in 81 instances. Nearly all were wounds of the extremities. Thirty-three patients had peripheral wounds without fracture, 16 of which were the patient`s major wound. Fifty-three of 70 patients had major peripheral wounds with fracture, and 13 had traumatic amputation of an extremity. In 10 of these 66, a major wound was also listed in another category. Three patients among those with peripheral wounds had injury to the spinal cord.
Of the severe nonperipheral wounds, 34 patients had thoracic wounds(a major wound in 30 instances) and 56 patients had intra-abdominal wounds, a major wound in 50 instances. An additional 21 patients had combined thoraco-abdominal wounds and 2 patients had separate wounds of the chest and abdomen. Of the total abdominal wounds, there were 25 wounds of the liver, 20 wounds of the kidney
26
27
(treated by nephrectomy in 11 instances), and in 1 case it was not known whether a kidney or liver wound had been present. Wounds of the urinary tract involving the bladder or structures above it occurred in 9 patients. Ten patients with nonperipheral wounds had multiple major wounds.
Crush injuries were found in nine patients, and there was only one case of head injury.
Clinical Condition of Patients on Arrival at the Most Forward Hospital
Pain
The frequency and severity of pain in different types of wounds had been extensively studied under similar conditions and on the same types of patients shortly before the Board was organized and the study was therefore not repeated on these 186 patients. Part of the data obtained in the earlystudy4 is shown in Table 2. The incidence of severe pain was surprisingly low. The data showed that severe pain was not to be accounted for on the basis of the patients` having received less morphine or having received it earlier than patients who reported little or no pain. It was also pointed out that three factors are chiefly important in the distress of the wounded: pain, mental distress, and thirst. In the severely wounded patient in good general condition, the first two factors are important. In the man in shock, thirst is the main and often the only cause of evident distress, but it may be extreme.
Shock
Grading of Shock.-The view sometimes has been taken that shock is either present or absent in a given case and that to try to distinguish between degrees of shock is futile. In this study, however, it was found instructive to separate the patients arbitrarily into four categories; namely, those with "no shock," "slight shock," "moderate shock," and "severe shock." This was done on the basis of the criteria listed in Table 3 which in turn were based on preliminary observation of large numbers of battle casualties by members of the Board. A patient was assigned to a particular category if he exhibited the ma-
4BEECHER, H. K.: Pain in men wounded in battle. Ann. Surg. 123: 96-105, January 1946; also Bull. U. S. Army M. Dept. 5: 445-454, April 1946.
28
29
jority of criteria for that category as opposed to another. These signs were inadequate, of course, for management of a case, for a comprehensive appraisal of the patient`s condition must include not only an accurate concept of his present state but also a shrewd estimate of his probable course in the immediate future.
On the basis of this arbitrary classification the 186 patients understudy were evaluated as to the degree of shock they had at the time of their admission to the hospital. In three of them the degree of shock could not be ascertained. Those 78 patients who were not seen on admission by any member of the Board were classified by the Board on the basis of the available clinical data and on discussion with medical officers who had seen them on admission. In the 108 patients who had been observed on admission by some member of the Board the degree of shock was probably more uniformly classified. Table 4 shows the clinical evaluation of shock and its distribution in these 108 patients as well as in the entire series. It is apparent that the distribution remained about the same when the 108 were separated from the entire group. Since the percentages were essentially unchanged when the magnitude of the cases was roughly doubled, it was assumed that the size of the sample 108 cases was adequate.
TABLE 4.-CLASSIFICATION AND DISTRIBUTION OF SHOCK
Relationship to Time from Wounding.-Examination of the records of 167 of the entire group of severely wounded men under study on whom these data
30
TABLE 6.-WOUND COMPOSITION IN EACH SHOCK CATEGORY-121 PATIENTS
31
were complete failed to show any correlation between the time from wounding until examination (clinical appraisal and blood analysis) and the presence or severity of shock (Table 5). The time elapsed in each of the four groups was approximately the same. There was, however, a striking correlation between severity of shock and blood loss, as will be shown later.
TABLE 5.-RELATIONSHIP OF DEGREE OF SHOCK ON HOSPITAL ENTRY TO TIME FROM WOUNDING-167CASES
Relationship to Wound.-Table 6 indicates the wound composition of 121 patients in each shock category. It merits some comment. If the two types of serious extremity wounds--traumatic amputation of extremities and compound fractures of long bones--are combined (the two are often very similar as to blood loss), it can readily be seen from Table 7 that the incidence of such wounds rose progressively in each category of increasing severity of shock. In the section on Blood Loss it will be shown that the greatest loss of hemoglobin occurred when the wound involved compound fracture of the long bones or
32
A SEVERELY WOUNDED MAN receives treatment in the forward area.
traumatic amputation of an extremity. Since most patients having such wounds were in severe shock, one can generalize with probability of accuracy and say that it is the wounds that are associated with great hemorrhage that cause severe shock. Reasons for laboring this rather obvious point will be discussed later.
In contrast to the rising incidence of severe extremity wounds in progressiveshock categories, the percentage of penetrated abdomens, although ratherhigh, shows no such consistent rise. In the severe-shock group, abdominalwounds are definitely less often a cause of the poor condition of the patientthan are the combined extremity wounds. (Incidentally, this evidence doesnot support the view that clostridial infection plays an important partin producing shock in general.)
The question might be raised as to whether the relative importance ofab-
33
dominal wounds as a cause of shock has heretofore been exaggerated.The poor prognosis often encountered in patients with abdominal woundsprobably has a great deal to do with the apprehension felt in the presenceof such lesions. The concealed hemorrhage or concealed contamination oftenpresent in these cases may lead to subsequent profound shock. So, whileon the average abdominal wounds were not as often a cause of severe shockon hospital entry as were serious extremity wounds, the impossibility ofaccurate preoperative appraisal of the abdominal wound makes it difficultto exaggerate its potentialities.
Shock will be discussed in each of the following sections of this chapterand an attempt will be made to correlate degree of shock with the dataunder discussion whenever possible.
Cardiovascular System
Electrocardiographic Observations.-Prior to organization of theBoard, 58 electrocardiographic records were made on 30 patients in severeshock and after recovery.5 Since the observations were madeon the same type of patients as those studied by the Board and under similarcircumstances, electrocardiograms were not made on the Board`s cases. Theresults of that study are summarized here. In 10 patients (one-third ofthat series) the blood pressure could not be measured on hospital entry.In the other two-thirds the degree of circulatory collapse was somewhatless severe, but even so the systolic blood pressures ranged from 60 to70 millimeters of mercury and the diastolic from 20 to 40 millimeters.
Definite abnormalities of the electrocardiograms were observed in 5 of the 30 patients. The most striking feature was the normal character of the findings in the remaining twenty-five. In 2 of the 5 patients with abnormal findings, the electrocardiograms showed striking but transient inversion of the T wave in lead 1. In a patient with an intrathoracic injury there was a shift from marked right-axis deviation back to normal following operation. The electrocardiogram in the fourth patient showed bizarre QRS complexes of low voltage, and in the fifth showed evidence of an unusual degree of temporary cardiac irritability with paroxysmal fibrillation and ventricular tachycardia.
5BURNETT, C. H.; BLAND, E. F., and BEECHER, H. K.: Electrocardiograms in traumatic shock in man. J. Clin. Investigation 24: 687-690, September 1945.
34
This electrocardiographic evidence of abnormality is of some interestbut difficult to explain. In no instance were there clinical signs of cardiacweakness, such as abnormal accentuation of the pulmonary second sound,basal râles, gallop rhythm, or congestion of the cervical veins orof the liver. As stated, the majority of the electrocardiographic findingswere within normal limits. Several patients in the series were in severeshock, having low blood pressure for a period of hours with no effect uponthe electrocardiogram. It may be significant that in both patients withtransient inversion of the T wave in lead 1, the wound involved the leftside of the chest, although so far as could be determined by roentgenographicexamination and clinical findings at the time of operation, the heart andpericardium escaped injury. Furthermore, the transient nature of the inversionwas more in accord with a temporary functional disturbance (possible hypoxia)than with lasting tissue injury.
Pulse Rate.-The pulse rates of the patients in the present studywere considered in relationship to shock and no significant differencewas found between the four categories (Table 8). The pulse rates consideredwere those taken as close as possible to the time the condition of thepatient was evaluated. When the pulse was imperceptible at the time ofinitial examination, the first recordable rate was used unless the recordshowed evidence that the patient was well on the way to resuscitation.
TABLE 8.-RELATIONSHIP OFDEGREEOF SHOCK TO PULSE RATE-106CASES
The finding that the average as well as the minimum and maximum pulserates were about the same in all degrees of shock was surprising. Thereare two possible explanations for this: 1. The tachycardia in the lesserdegrees of shock may have been due in part to excitement. 2. In some casesthe elevation of the pulse rate (and the vasoconstriction accompanyingit) may have been adequate to ward off the signs of shock. It is interestingthat even patients judged to be
35
in severe shock can have a pulse rate as low as 60 beats per minute.Of greater significance than the actual rate of the pulse is its volume,which often was decreased so much in severe shock that the pulse couldno longer be felt.
Blood Pressure.-Blood pressures were analyzed in only those 70cases out of the 186 in which they had been recorded at the time the patient`scondition was evaluated. The volume of the circulating blood was also determinedin these 70 patients. There was no significant fall in the average systolicblood pressure except in those in moderate or severe shock (Table 9). Itwill be shown that these patients with considerable shock had lost on theaverage 33.6 percent of their calculated normal blood volume and nearly50 percent of the total circulating hemoglobin. In those with severe shock,the systolic blood pressure fell rapidly, the average being 49 millimetersof mercury. This group had lost approximately half the normal blood volume(see Table 22). There was, however, a progressive drop in the average diastolicblood pressure with increasing degrees of shock (Table 9). The averagediastolic blood pressure of the patients in severe shock was half thatof the patients in moderate shock. As severity of shock increased, therewas a significant and progressive decline in the pulse pressure (Table9). This confirmed the clinical observation that the volume of the pulsewas closely correlated with the degree of shock.
TABLE 9.-RELATIONSHIP OF DEGREEOF SHOCK TO BLOODPRESSURE-70CASES
"Irreversible" Changes in the Cardiovascular System.-Everyonewho has treated many patients for shock has encountered some who fail torespond to the transfusion of blood deemed adequate under ordinary circumstances,and this is often attributed to "irreversible" changes that have presumablytaken place during prolonged hypotension, ischemia, and anoxia. This problemis
36
further discussed in the section on Blood Loss. In most instances, adequateexplanation can be found for the failure of patients in shock to respondto blood transfusion; some common examples are concealed and continuinghemorrhage, hemothorax, irritant contamination of the peritoneum, peritonitis,clostridial myositis, and fat emboli. Four cases from our series are illustrative.
CASE REPORTS
Case 77.-A patient with a severe thoraco-abdominalwound was received at a forward hospital in severe shock 8¼ hoursafter he was wounded. Resuscitative measures were continued for nearly9 hours. During that time he received only 1,500 cc. of whole blood. Hiscondition failed to improve and he was operated upon but did not survivethe operation.
Necropsy showed massive collapse of the right lung witha plug of mucus in the right main bronchus. The lower lobe of the leftlung was collapsed and about one-third of the left upper lobe was atelectatic.There was gross dilatation of the right ventricle of the heart. On histologicexamination minimal evidence of fat embolism in the pulmonary vessels wasfound but considered of no clinical significance.
Comment.-There was adequate cause for this patient`s failureto respond to resuscitation. More aggressive measures should have beentaken, including bronchoscopy and the use of more blood in less time. Thereshould have been more concern when no improvement occurred during the first3 hours after the patient`s admission to the hospital.
Case 45.-A patient with a severe abdominal woundwas admitted to a forward hospital in severe shock 8 hours after wounding.During the next 3 hours, 2 units (600 cc. total volume) of plasma and 1liter of whole blood were transfused. The blood pressure during that timechanged from imperceptible to 90 millimeters of mercury systolic and 70diastolic. An additional unit of plasma was administered and an infusionof 500 cc. of 2-percent solution of sodium bicarbonate was given intravenously.Although this was the optimum time for surgery, operation was delayed and5 hours later the blood pressure was again unmeasurable. It was restoredto 86 millimeters of mercury systolic and 60 diastolic after transfusionof 1 liter of whole blood, and operation was performed which lasted 4 hours.
At operation the abdominal cavity was found to be "fullof blood." The blood pressure and pulse rate were unmeasurable during muchof the operation. The patient never regained consciousness and died 3¾hours after the end of the operation. Necropsy showed perforation of theinferior vena cava. There was histologic evidence of minimal fat embolismin the pulmonary vessels, probably of no clinical significance.
Comment.-The recurrent hypotension in this patient was probablydue to continued extraperitoneal and intraperitoneal hemorrhage. Operationshould
37
have been performed while he was responding well to resuscitative measuresduring the first 3 hours after admission to the hospital.
Case 100.-This patient had multiple wounds involvingboth arms, the left thigh, and the face. There were compound fracturesof the left humerus, radius, and ulna, and of the right ulna. There wasalso a transection of the right femoral artery with vascular insufficiencyin the leg. He was admitted in severe shock to a forward hospital 3½hours after wounding, and within 90 minutes he received 300 cc. (totalvolume) of plasma and 2 liters of whole blood. He showed general improvementbut his blood pressure was still only 80 millimeters of mercury systolicand 50 diastolic. His pulse rate was 144 beats per minute. Three hoursafter admission his blood pressure was 90 millimeters of mercury systolicand 58 diastolic. Operation was delayed for 3 additional hours. At no timeduring operation did the recorded blood pressure fall below 85 millimetersof mercury systolic. The patient had received a total of 4,500 cc. of wholeblood before, during, and immediately after operation.
Ten hours after operation the patient`s blood pressurewas low and he looked pale and "anemic." A transfusion was started, butan hour later he suddenly died. Ten minutes earlier he had carried on anintelligent conversation. Pulmonary embolus was suspected but at necropsyno cause could be found for the sudden death. Microscopically, a moderatelysevere grade of fat embolism was found in the lungs.
Comment.-The question was raised whether the 5- or 6-hour periodof hypotension in this patient could have caused irreversible changes inthe cardiovascular system so that it simply "gave out" when it did. Thiscannot be answered with certainty. The fat embolism in retrospect appearsto be the more important consideration.
Case 120.-This patient had a simple penetratingwound of the thigh caused by a shell fragment. The femoral artery belowthe origin of the profunda femoris was severed. During evacuation the patienthad received 4 units (1,200 cc. total volume) of plasma and when he reachedthe evacuation hospital, about 9 hours after wounding, he must have appearedin good condition for no resuscitation was deemed necessary. At operation,performed 4½ hours after admission, the femoral artery, vein, andnerve were found to be completely transected. The vessels were ligatedand the foreign body was removed.
At the conclusion of the operation, the systolic bloodpressure was only 70 millimeters of mercury and remained between 70 and60 throughout the day. Despite this the patient appeared to have good colorand his skin was not cold. Transfusion of 1 liter of whole blood did notimprove the blood pressure. The right leg looked as though it would notsurvive. Anuria developed and the patient died 48 hours after operation.Necropsy revealed nothing to account for the postoperative hypotension.There was no concealed hemorrhage and no evidence of clostridial myositisin the involved extremity. There was no histologic evidence of fat embolism.
Comment.-This patient probably had lost more blood than was realized.
38
Resuscitation
Before admission of the patients in this study to a forward hospital,resuscitative efforts had been limited chiefly to control of pain and hemorrhageand to administration of blood plasma. Relatively little whole blood wasgiven. The 108 patients seen by us on admission had received, on the average,2 units6 of plasma before the first blood sample was taken.Plasma administration was distributed in this group as follows:
|
Number of Patients |
Units of Plasma |
|
32 |
None |
|
25 |
1 |
|
17 |
2 |
|
13 |
3 |
|
12 |
4 |
|
3 |
5 |
|
1 |
6 |
|
1 |
7 |
|
2 |
8 |
|
1 |
9 |
|
1 |
11 |
Thus 69 percent (74) of these patients had received two units (6oo cc.)or less of blood plasma before or shortly after arrival at the most forwardhospital. Twenty-seven of the 108 patients received transfusions of wholeblood prior to withdrawal of the first blood specimen for laboratory analysis.Three of these, or 3 percent, had received whole blood in an aid stationbefore admission to a forward hospital. The blood transfusions were distributedas follows:
|
Number of Patients |
Units of Blood |
|
13 |
1/5 to 1 |
|
10 |
1½ to 2 |
|
2 |
3 |
|
1 |
4 |
|
1 |
6 |
The following tabulation summarizes the average quantities of bloodand of blood plasma used in resuscitating 157 of the very seriously woundedpatients in our series (the 108 referred to above and 49 others on whomwe had clinical notes):
|
Blood plasma preoperatively (average of 122 cases) |
3.08 units |
|
Blood plasma during operation (average of 10 cases) |
1.68 units |
|
Whole blood preoperatively (average of 127 cases) |
1,450 cc. |
|
Whole blood during operation (average of 95 cases) |
1,160 cc. |
61 unit = 250 cc. of normal plasma diluted to 300 cubic centimeters.
39
In round numbers, our average patient in this series had just over 3units (total volume) of blood plasma and 5 blood transfusions (total ofabout 2,500 cc. of whole blood) to support him from the time of woundinguntil his operation was completed. It is interesting to observe that plasmawas used during surgery in only 10 of these 157 patients.
The information concerning the cases referred to here was drawn fromthe shock tents of most of the hospitals of the Fifth Army and representsa broad sample of current practice in Italy over the last year of the EuropeanWar. Essentially the same type of case had been studied by two of us earlierat Anzio.7 In that series the average patient received 1,537cc. of whole blood (three transfusions) to prepare him for and carry himthrough surgery. These three transfusions contrast with the five referredto above. A notable difference between the Anzio study and MediterraneanTheater practice in general was in the time elapsed from hospital entryto start of surgery. In the Anzio study this averaged 2 hours, 21 minutes.Reference to an earlier part of this section on Time from Wounding to HospitalEntry and Surgery will show that over the Fifth Army Area as a whole, theaverage time from hospital entry to surgery varied from 5 hours to 8 hours.Two differing views as to the correct preparation of wounded men for surgeryare represented in these figures: the extended, and the rapid. The extendedrequired five transfusions of whole blood; the rapid, three. This has beendiscussed in a previous publication.8
Plasma Protein Concentration and Hematocrit ValuesOn Admission to Forward Hospital
The concentration of protein in the plasma and the blood hematocritvalue (both calculated from specific gravities measured by the copper sulfatemethod9) give a clue to the shifts that have taken place betweenthe blood stream and the tissues as well as to blood loss from the body.When considered
7BEECHER, H. K., and BURNETT, C. H.: Field experience in use of blood and blood substitutes (plasma, albumin) in seriously wounded men. M. Bull. North African Theat. Op. (no. 1) 2: 2-7, July 1944.
8BEECHER, H. K.: Preparation of battle casualties for surgery. Ann. Surg. 121: 769-792, June 1945.
9PHILLIPS, R. A.; VAN SLYKE, D. D.; DOLE, V. P.; EMERSON, K., JR.; HAMILTON, P. B., and ARCHIBALD, R. M.: Copper sulfate method for measuring specific gravities of whole blood and plasma. BUMED News Letter, U. S. Navy, vol. 1, June 25, 1943.
40
with quantitative measurements of whole-blood loss, a fairly accuratepicture of one consequence of the wound can be obtained. The plasma proteinand hematocrit levels were determined in our patients shortly after theiradmission to the most forward hospital, before resuscitation, anesthetization,or operation had been undertaken.
Relationship to Type of Wound
In Table 10 the relationship of the average plasma protein concentrationand the average hematocrit value to the type of wound is shown for 50 patientswho, prior to study, had received only 1 unit (300 cc.) of blood plasmaor less than 1 unit (in some instances no plasma had been administered).The patients whose wounds were peripheral are grouped and compared withthose having nonperipheral wounds; crush cases are not included.
From the table it may be seen that there is no decided difference betweenthe plasma protein levels of patients with peripheral and those with nonperipheralwounds. On the other hand the hematocrit values were significantly higherin.the latter group which is consistent with the hemoconcentration sometimesfound in such patients. There was also less loss of hemoglobin (as willbe discussed in the section on Blood Loss) in those patients with intra-abdominaland thoraco-abdominal wounds than in those having severe wounds of theextremities. Maintenance of a more nearly normal blood volume in patientswith such nonperipheral wounds doubtless reduces the need and tendencyfor blood dilution, although this factor alone would not account for thehemoconcentration when it is found in such cases.
41
Relationship to Blood Loss
When average plasma protein concentrations and average hematocrit valuesin patients with peripheral and nonperipheral wounds are compared withloss of blood volume (39 cases), it may be seen from Table 11 that thehematocrit values were significantly lower in all patients who had lostmore than 30 percent of their calculated normal blood volume. The hematocritlevel of 36 is 23.4 percent below the normal of 47 (Wintrobe method).
In the case of the plasma proteins, however, even when there was a lossof 30 percent or more of the normal blood volume, the average concentrationwas 6.1 Gm. per 100 cc. (only 6.1 percent below the normal of 6.5). Inother words, the hematocrit level fell proportionately about four timesas much as that of the plasma proteins. The blood appears to have beendiluted by protein-rich fluid (6.1 Gm. per 100 cc. of blood). The evidenceis too meager to justify much speculation here. However, as pointed outby Evans,10 the axial stream of corpuscles is surrounded bya plasma envelope. This varies in thickness and total volume, dependingupon certain hydraulic principles. It might be possible that the alterationsin the circulation caused by the loss of 30 percent or more of the normalvolume of blood (slowing of the peripheral circulation, for example) resultedin dragging an appreciable volume of plasma with normal protein contentinto
10EVANS, ROBLEY: Personal communication.
42
the circulating blood. Or it might be possible that protein was brought into the circulation from the liver.
Influence of Plasma Therapy
The influence of previous administration of blood plasma upon the concentrationof plasma protein and the hematocrit value was considered, and the findingsin different types of wounds are shown in Table 12 and Charts 1 and 2.Only three of the patients had had blood transfusions; these will be ignored.It is clear from the table and charts that plasma therapy did not influencethe plasma protein level, but it did have an important effect on the hematocritlevel.
The plasma protein concentration and hematocrit value were also analyzedin regard to shock, the data being broken down into two categories in which"no shock" and "slight shock" were grouped together, as were "moderateshock" and "severe shock." No important differences were found betweenthe two categories.
43
CHART 1. Influence of plasma therapy on plasmaprotein and hematocrit levels in peripheral wounds
CHART 2. Influence of plasma therapy on plasmaprotein and hematocrit levels in abdominal wounds
Relationship to Shock
Plasma protein and hematocrit levels were also studied in relation tothe clinical condition of about 100 badly wounded patients (crush casesexcluded) on admission to the forward hospital. The fall in average plasmaprotein concentration is probably significant as the cases are groupedin Table 13 and Chart 3; the fall in hematocrit value is definitely significant.There was no evidence of hemoconcentration. When the patients were groupedaccording to
TABLE 13.-PLASMA PROTEINAND HEMATOCRIT LEVELS ON ADMISSIONIN RELATION TO SHOCK*
44
TABLE 14.-PLASMA PROTEINAND HEMATOCRIT LEVELS ONADMISSION IN RELATION TO SHOCKIN PATIENTS WITH ABDOMINAL WOUNDS
CHART 3. Plasma protein and hematocrit levelsin relation to shock--all types of wounds
location of the major wound, there was no significant fall in the concentrationof plasma protein in those with peripheral wounds in relation to the degreeof shock. The findings in patients with abdominal wounds showed up differently(Table 14). In the group with minimal shock, the plasma protein concentrationmay possibly be accounted for by weeping of the irritated peritoneal surfaces,fluid being released which contained less protein than the plasma. As shockbecame moderate or severe, probably due to greater blood loss, the plasmaprotein concentration fell to a figure like that for extremity wounds,with hemodilution overcoming the effects of exudation. These data indicatethat the plasma protein and hematocrit values can vary independently.
45
Blood Loss
Volume and Hemoglobin
The quantity of blood a wounded man can lose and yet recover has generallybeen underestimated. One indication that this is so was the fact, wellshown in the prolonged campaigns of the Mediterranean Theater, that robustyoung soldiers tolerated surgery well, long before the blood volume oreven the blood pressure had been restored to normal. Actually the conceptof restoration of the patient in shock to normal prior to surgery is basedupon a false premise. Full organic restoration probably requires days toachieve. A good response of a young wounded man to treatment is by no meansadmissible evidence that his circulatory system has been restored to normal;it is evidence of the existence of safety factors in human physiology.These points have been discussed elsewhere.11
In the belief that measurement of the blood loss that had been sustainedby these severely wounded men by the time of their arrival at a forwardhospital would clarify the matter of the importance of whole blood forthe wounded, such a study was carried out. The direct relationship betweenquantity of blood lost and degree of shock had long been recognized, butfurther evidence of this relationship was desirable in view of the ever-recurringsuggestions that the cause of shock is mysterious and to be explained bythe presence of toxins in the body or by the breakdown of some vague butvital force.
The blood volume and the hemoglobin concentration were determined in67 patients12 shortly after their arrival at the most forwardhospital (which in most instances was a field hospital). The blood volumeloss and the total hemoglobin loss, expressed as percentages of a calculatednormal for each patient, were then determined on the basis of these findings.Normal blood volume was considered to be 8.5 percent of the body weight,after Gregersen. (See Appendix C for the method used.) Peters13has commented on the loss of the dye T-1824 from the blood stream. Suchloss of dye would of course
11See footnote 8. Also see section on Resuscitation in volume on general surgery of the series: The Medical Department of the United States Army. To be published.
12The blood studies were actually made in 71 cases (see Table 15), but one case was discarded because of a probable technical error and three others are excluded from the present discussion since they represent types not common to the group; namely, two crush injuries (Cases 93 and 124) and a head injury (Case A-7).
13PETERS, J. P.: Role of sodium in production of edema. New England J. Med. 239: 353-362, Sept. 2, 1948.
46
TABLE 15.-INITIAL BLOODCHANGES IN 71 SEVERELY WOUNDEDMEN
47
TABLE 15.-INITIAL BLOODCHANGES IN 71 SEVERELY WOUNDEDMEN--
Continued
48
give a falsely high value for blood volume. Our blood-loss values wereestimated by difference, difference between an average normal blood volume(8.5 percent of body weight) and the value found. Therefore the losseswe report are lower than the fact; they are minimal rather than maximal.
To compensate for the effect on his blood volume of the blood and plasmareceived by the patient, certain corrections were applied to our bloodvolume determinations. The majority of patients had received some plasmaprior to hospital admission and a few had received whole blood. Althoughblood volume determinations were made as soon after hospital admissionas possible, resuscitative procedures had likewise been initiated. Actually,in most cases, determinations and resuscitation proceeded concurrently.Two methods of correcting blood volume findings were therefore utilized:Correction A and Correction B.
Correction A was the subtraction of the total quantity of blood and blood plasma received by the patient from the time of wounding until
49
completion of the test.
Correction B was the subtraction of only the quantity of blood and plasma received from the time of the patient`s hospital admission until completion of the test.
Correction A was applied for the purpose of correlating the patient`s clinical condition on arrival at the hospital with the amount of blood (the percentage of his estimated normal blood volume) that he had actually lost due to his wounds, regardless of the amount that may have been replaced. Correction B was for correlation between the patient`s clinical condition on arrival and the blood deficit existing at that time. The same corrections were applied in calculating the total hemoglobin loss. However, in the case of hemoglobin
50
there was essential agreement between calculation A and calculation B because, with rare exceptions, whole blood had not been administered before the patient reached the most forward hospital.
Tables 15 through 22 show blood and hemoglobin loss in relation to other findings for patients individually and by groups. The loss is given as the percentage difference between a calculated normal and the observed value, or between the calculated normal and the observed value modified by Corrections A and B. For blood all three determinations are shown. For hemoglobin only the estimated loss derived by Correction B is shown. It will be noted that a few entries (shown as plus values in the tables) seem to indicate that the blood volume was greater than normal despite loss of blood. Several factors, indi-
51
vidually or collectively, might have brought about this apparent discrepancy:1. In the case of uncorrected entries, blood and plasma administered may have been in excess of blood loss. 2. Entries are from calculations based upon an average normal which may have been too low for some patients.3. There may have been errors in technique in blood volume determination.
Relationship of Blood Loss to Type of Wound
Fifty-nine of the patients listed in Tables 15 and 16 suffered primarily from a single major wound (abdominal, chest, or peripheral) and could therefore be more readily studied as to the relationship of blood loss to type of wound.
52
CHART 4. BLOOD VOLUME LOSS IN RELATION TO LOCATION OF WOUND
The findings in these 59 patients, and in 6 others who had combined thoraco-abdominal wounds, are presented in Tables 17 and 18, and in Charts 4 and 5. The huge standard errors present in several instances indicate the wide variations in the results found. From the table it would appear that loss of blood volume was greatest in the patients with peripheral wounds (Table 17). However the number of cases was small, the variability in the data was wide, and the true state of affairs possibly was masked by dilution of the blood volume after wounding by movement of fluid from the tissues to the blood stream.
The data on hemoglobin loss are probably more revealing. There is a significantly greater loss of hemoglobin in men with peripheral wounds than in those with abdominal wounds (Table 18) and in this case the situation is not
53
CHART 5. HEMOGLOBINLOSS IN RELATION TO LOCATIONOF WOUND
obscured by the factor of hemodilution. This difference is in agreement with the previous finding that more patients with compound fractures of long bones
TABLE 17.-RELATIONSHIP OFBLOOD VOLUME LOSSTO TYPE OF WOUND
54
TABLE 18.-RELATIONSHIP OF HEMOGLOBIN LOSS TO TYPE OF WOUND
or traumatic amputations were in severe shock than were those with abdominal wounds (Tables 6 and 7). Data on the comparison of blood loss in the two types of serious extremity wounds are scanty but are presented for what they are worth, Table 19. The blood loss in patients with compound fractures was greater than it was in those with traumatic amputation, probably because of the greater tissue damage usually found in the former.
TABLE 19.-LOSS OF BLOODVOLUME AND TOTAL HEMOGLOBININ 31 PATIENTS WITH SEVERE EXTREMITY WOUNDS
Table 20 shows the average blood volume and hemoglobin losses in 40patients with all types of wounds who had received either none or small amounts of plasma but in no instance more than 1 unit before the determinationswere made. The losses here are less than those shown in Tables 17 and 18because the group includes fewer of the patients in severe shock. Again a greater average loss of hemoglobin than of blood volume is shown. This is explained by
55
hemodilution which normally takes place after blood loss. Red blood cells are not replaced appreciably during the interval from wounding toarrival at a forward hospital except by transfusion.
Relationship of Blood Loss to Time from Wounding
There was no important average increase in blood volume loss or in hemoglobin loss with increased time elapsing between wounding and hospital entry (Table 21). However, those men who were suffering from continuing blood loss were probably given priority of evacuation. Somewhat greater blood losses were usually found in those who arrived at the hospital soon after wounding than in those who were brought in later.
TABLE 21.-RELATIONSHIP OFBLOODLOSS TO TIME FROM WOUNDINGIN 67 CASES
(ALL TYPES OF WOUNDS)
56
Relationship of Blood Loss to Degree of Shock
It is well known that individuals do not respond alike to a given blood oss; even previously healthy, normal young soldiers vary greatly in their response. This fact, together with the inexactness inherent in any clinical appraisal of the degree of shock plus the errors of the experimental method used, might have tended to obscure a real relationship between shock and blood loss. However this relationship was so striking that, even with the relatively small number of cases, the positive correlation of blood loss to the severity of shock was statistically significant (Table 22 and Chart6).
TABLE 22.-RELATIONSHIP OFBLOOD LOSS TO DEGREEOF SHOCK IN 67 CASES
(ALL TYPES OF WOUNDS)
The degree of wound shock as we saw it in men injured in battle precisely paralleled the quantity of blood actually lost. Conversely, recovery from shock resulted promptly from administration of whole blood. Although we made intensive search at the bedside of thousands of wounded men throughout the shock tents in Italy, we never found a clear case of "irreversible shock," mentioned so frequently in the literature. It is true that there were wounded men in whom the loss of blood was so rapid and so great that it was impossible to transfuse them with blood fast enough to save their lives. (For example, we were unable, in the case of the soldier who had both thighs blown off by a shell burst just outside our door at Anzio to get blood into him fast enough to save his life; he died in a very few minutes.) Nor were we able to resuscitate patients who had had inadequate circulation in the central nervous system
57
CHART 6. RELATIONSHIP OF BLOOD LOSS TO DEGREE OF SHOCK
so long that nearly all centers except the respiratory appeared to be dead; we were not able to overcome death of organs or of nervous tissue by resuscitative effort. But we believe that application of the term "irreversible shock" to either type of case is to use a definition that has no place at the bedside, however interesting it may be as a concept, and that may do real harm by providing an excuse for limiting resuscitative effort.
In short, if "irreversible shock" in the accepted sense was present, we missed it. If toxins caused any of the shock we saw, with the exception of that due to overwhelming and clinically apparent bacterial infections, we failed to recognize it. The shock we saw was caused by loss of blood(or of fractions of the blood). It was relieved by administration of whole blood.
From our data it may be said that, in general, when a third of the blood volume is lost, clinical shock of more than slight degree will result, and when half is lost, severe shock will result. It was also shown that in individual cases as much as 75 percent of the blood could be lost and the patient survive. Cases A-17, A-21, A-29, A-37, A-38, 127, and 139 are examples of severe blood loss and survival. The greatest blood volume losses found were 78.6 percent in Case 107 and 75.7 percent in Case 139 (Table16, C and D). Both de-
58
terminations were obtained by Correction A. Shock was moderate in the first case and severe in the second. The greatest loss of total hemoglobin ound was 83.8 percent in Case A-21 (Table 16, D). This patient had serious extremity wounds and was in severe shock.
Our data also indicated that, on arrival at the hospital, generally the deficiency in total hemoglobin was greater than the deficiency in blood volume. This is understandable in view of the well-known mechanism for replenishing blood volume at the expense of tissue fluid. On the other hand, once the readily available reserves of hemoglobin have been called into action, no mechanism is available to replace them rapidly; consequently it would be expected that greater deficiencies would be found in the total hemoglobin.
There was no correlation between passage of time and the degree of shock encountered, in that the elapsed time was the same in each of the four categories of no shock, slight, moderate, and severe shock. This does not say that continuing hemorrhage, for example, is not related to degree of shock; certainly it is, but the important factor in the development of shock is the character of the wound, particularly as it indicates the quantity of blood lost, not the passage of time per se.
We have tried various ways of handling the data on shock. Some investigators have attempted to separate the effects of blood loss alone from the effects supposedly accounted for by "sympathico-adrenal activity." Practically, this is impossible. Moreover, the validity of any such separation even if possible must be questioned. For example there is the matter of increased glycogenolysis in shock. Following all the recent advances in knowledge of carbohydrate metabolism, it seems to be too great a simplification to hang the explanation on the rather shaky peg of "sympathico-adrenal activity, "so we take refuge in our purpose of merely stating our findings, with little conjecture, and leaving to future study the search for their significance.
One great consequence of blood loss is the intense vasoconstriction, the shrinkage of the capacity of the vascular bed to accommodate the decreased blood volume. Contraction of the spleen probably plays a relatively small role in compensating for the blood lost in battle. Other body adjustments to blood loss do, however, take place, such as the entry of fluid into the blood vessels in an attempt at compensation. The greatest extravascular store of readily available fluid in the body is that in the extracellular space. Dehydration and oligemia may make quite early demands, not only on this extracellular but also on the intracellular supply.
59
Biochemic Changes Encountered
The clinical appearance of the newly wounded man, as well as his subsequent course, offers abundant evidence that profound changes have occurred in his internal state by the time he is admitted to a hospital. To be reported here are chiefly the changes found in the blood although certain urinary findings also are recorded. These changes are significant not only because they reveal the problem at hand, but also because they offer some basis for reasonable therapy. Several factors influence the presence or extent of the abnormalities found. These are discussed in the following sections.
Blood Chemistry Findings
Relationship to Location of Wound.-No significant relationship was found between the location or type of the wound and the plasma non protein nitrogen, creatinine, uric acid, phosphorus, or magnesium (See Table 24, page 60). Average values for serum sodium, plasma chlorides, and plasma carbon-dioxide combining power also were determined and likewise failed to show any significant relationship to the type of wound.
Relationship to Delay in Hospital Arrival.-With increased passage of time following wounding, the average plasma nonprotein nitrogen level rose (Table 23). This upward swing of the nonprotein nitrogen offers a basis for some inter-
60
TABLE 24.-BIOCHEMIC FINDINGS1IN RELATION TO TYPE ANDLOCATION OF WOUND
61
esting speculation, out of place here, but one might ask in passing:Is this rise a reflection of decreased renal blood flow? Does this risemean that renal impairment is initiated by the wound and continues, with accumulation of nonprotein nitrogenous waste products? Does this presumed malfunction set the stage for later trouble with the kidneys? The plasma creatinine level did not rise significantly with the passage of time preceding hospital entry, nor did that of uric acid, phosphorus, or magnesium (Table25). Also no correlation was found between passage of time and levels of plasma chlorides, serum sodium, or plasma carbon-dioxide combining power.
TABLE 25.-RELATIONSHIP OF SEVERALPLASMA CONSTITUENTS *TO TIME FROM WOUNDING
Relationship to Clinical Condition.-The interesting relationships of the plasma nonprotein nitrogen, creatinine, phosphorus, and magnesium to the degree of shock are shown in Table 26. Statistically significant rises occurred for nonprotein nitrogen, phosphorus, creatinine, and magnesium in progressive categories from "no shock" to "severe shock," with the chief difference occurring between the moderate- and severe-shock groups. The rise in uric acid was not statistically significant. When these substances were compared with blood loss, the same positive correlation was found for the nonprotein nitrogen, creatinine, phosphorus, and magnesium, as would be expected. Again the correlation in the case of uric acid was not significant. These comparisons are shown in Table 27; blood loss was divided into five categories from no loss to loss of more than 40 percent of the calculated normal blood volume.
Data relevant to acid-base balance are presented in Tables 28 and 29and Chart 7. An acidosis was present in those patients with severe shocka nd is
62
TABLE 26.-BIOCHEMIC FINDINGS *IN RELATION TO SHOCK
TABLE 27.-BIOCHEMIC FINDINGS *IN RELATION TO BLOOD LOSS
63
CHART 7. Plasma electrolytes (on hospital admission)in relation to shock
reflected in the carbon-dioxide combining power. A significant fall of carbon-dioxide combining power was correlated with shock (Table 28).Evidence that these low values are due to a metabolic acidosis will be presented in the discussion of azotemia in Chapter IV. Plasma chlorides were uniformly normal in all shock categories (Table 28); likewise, urinary chlorides were essentially normal although they showed wide variations in concentration (Table 29). There was, then, no evidence of salt deprivation on hospital entry. Phosphates, although significantly higher in terms of total anions in those in severe shock than in those with no clinical evidence of shock, could have had little effect on the acid-base balance. Plasma proteins are not included in Table 28, but they showed insufficient variation in terms of electrolyte concentration to have affected acid-base balance.
The plasma was examined for lactic acid in 5 patients shortly after they were wounded. The findings were compared with those in 7 normal, active
64
TABLE 28.-PLASMA ELECTROLYTESON ADMISSION IN RELATION TOSHOCK
65
TABLE 29.-PREOPERATIVE URINARYFINDINGS IN RELATION TO SHOCK
soldiers and in 10 bed patients convalescing from severe wounds. Theresults are recorded in Table 30. They show a twofold increase in the concentrationof lactic acid in the wounded when compared with normal, active soldiers and with convalescent bed patients.
TABLE 30.-CONCENTRATION OFLACTIC ACID IN PLASMAIN SEVERELY WOUNDED SUBJECTSAND CONTROLS
66
TABLE 31.-AVERAGE PLASMABILIRUBIN AND HEMOGLOBIN LEVELSIN PATIENTS WITH VARIOUSTYPES OF WOUNDS
TABLE 32.-AVERAGEPLASMABILIRUBIN AND HEMOGLOBINLEVELSIN RELATION TO TIME FROM WOUNDING
67
Plasma Bilirubin (van den Bergh Index)
and Plasma Hemoglobin Levels
Relationship to Type of Wound.-On comparison of the type of wound with the plasma bilirubin and hemoglobin, no impressive relationships were found (Table 31).
Relationship to Time from Wounding.-The van den Bergh index rose significantly with increases in time from wounding to examination (Table32). This may have been due to the absorption of breakdown products from hematomas, and to impaired liver function (see Chapter II). The situation was simpler at this time (when most of these patients had not yet been transfused with blood) than it would be later when large volumes of blood had been given which might tend to elevate the bilirubin level. The plasma hemoglobin level appeared to rise with the passage of time, but this was not significant so far as the data at hand are concerned.
Relationship to Clinical Condition.-There was no clear relationship between degree of shock and the plasma bilirubin or hemoglobin levels (Table33). However, when the bilirubin level was compared with the blood loss(Correction A), a significant relationship seemed to emerge, although the values were all at a rather low level (Table 34). Presumably the rise is to be accounted for by hemolysis of blood in damaged tissues, followed by absorption into the blood stream. There was no apparent correlation of blood loss with plasma hemoglobin levels.
TABLE 33.-AVERAGE PLASMABILIRUB IN AND HEMOGLOBIN LEVELS IN RELATION TO DEGREE OF SHOCK*
68
TABLE 34.-AVERAGE PLASMABILIRUB IN AND HEMOGLOBIN LEVELS IN RELATION TO BLOOD LOSS*
TABLE 35.-AVERAGE PLASMA GLUCOSE LEVELS IN SEVERELY WOUNDED MEN
69
CHART 8. PLASMAGLUCOSE LEVELS
Blood Sugar
The blood-sugar level was determined in 57 severely wounded men as they arrived at the forward hospital and was found to be above normal in allof them (Table 35 A and Chart 8). Normal by the method used was consideredto be from 80 to 90 mg. per 100 cc. of plasma. Some of the men had received
70
plasma but none had had a significant quantity of whole blood beforethe determination was made. The blood-sugar level appeared to rise significantly both with increased blood loss and with increased severity of shock (Table35 A and B). The glucose level may fall with the passage of time following wounding, but our data were not extensive enough to demonstrate this (Table35 C). Presumably the elevation in blood-sugar level was due to mobilization of liver glycogen following adrenal activity and probably reflected theemotional and physical stress the men had experienced.
TABLE 36.-URINE TESTS FOR CREATINE IN 32 PATIENTS
71
Urinary Findings
Acid-Base Relationship.-Sodium in the urine was measured in toofew cases for reliable averages. Most of the values found were within thenormal range. Magnesium, like phosphates on the acid side, was increasedbut not sufficiently so to affect total acid-base equilibrium. UrinarypH and specific gravity indicate that these men had essentially normalrenal function at the time they were wounded, in that they could make bothan acid and a concentrated urine. The acid-base relationship will be discussedfurther in Chapter III.
Creatinuria.-The appearance of creatine in the urine of adultmales is abnormal. On the assumption that there might be some abnormalityin the metabolism of creatine and creatinine in shock, the urine was examinedfor creatine in 32 patients (Table 36). In all, 69 urine specimens were tested; the results are summarized in Table 37. Creatine was found in theurines of 26 of the 32 patients. It was present in 6 out of 15 of the patients examined preoperatively, and in 20 out of 29 patients whose urines wereexamined postoperatively. In 13 of the 32 patients urine tests yielded only positive results, in 11 patients only negative, and in the remaining9, different specimens tested yielded both positive and negative results.
With one exception (Case 121) all patients who died showed creatinuria at one time or another. It was also present in approximately half the patients who survived. In relation to shock, 14 of 18 patients who had been in moderate or severe shock showed creatinuria. This tendency for creatine to be excreted in the urine in cases of moderate and severe shock and in fatal cases might indicate that it is the result of metabolic changes which accompany shock.
TABLE 37.-SUMMARY OF RESULTS IN 69 URINE SPECIMENS TESTEDFOR CREATINE
72
Liver Function in the Newly Wounded Man
As will be discussed in Chapter II, the only direct laboratory test of liver function carried out here was that of the Bromsulfale in excretion.The van den Bergh index and uric acid levels have also been considered as being in part at least related to liver function.
Bromsulfale in Retention
On Arrival at the Most Forward Hospital.-In 59 severely woundedpatients the average Bromsulfale in retention on forward hospital arrivalwas 12.4±1.2 percent (standard error of the mean) 45 minutes after5 mg. of dye per Kg. of body weight had been injected intravenously. Thisis well above the normal of 1.0±0.1 percent (as established in acontrol group of 45 subjects), and above our arbitrarily chosen upper limitof normal of 3 percent (see Chapter II).
Relationship to Time from Wounding.-There was no difference between the average bromsulfale in retention in 29 men examined within the first6 hours after wounding (14.4± 1.8 percent) and in 19 men examined after the first 6 hours (13.1 ±1.6 percent).
Relationship to Location of Wound.-In 22 patients with serious extremity wounds there was 13.3 ±2.3-percent average retention, and in 18 patients with abdominal wounds there was an average of 14.7±2.1-percent retention-no difference. In 11 men with penetrating chest wounds, however, the average retention was only 7.0±1.8 percent, which is significantly lower than that found in the other groups.
Relationship to Shock.-In 57 patients separated into the 4 shock categories (no shock, slight, moderate, and severe shock) no significant correlation with bromsulfale in excretion could be found; neither was there any correlation with blood volume or hemoglobin loss.
Relationship to Plasma Administration.-Curiously enough, there was a great increase in average bromsulfale in retention (14.0 ±2.0 percent) in 25 men who had had one or two units of plasma, over the average retention (8.0±2.0 percent) in 15 men who had had none. Three or more units of plasma did not increase the effect beyond that produced by one or two units. The effect was transient and disappeared between the first and second day following operation.
73
Plasma Bilirubin and Uric Acid
The bilirubin and uric acid concentrations in the blood have been discussed earlier in this chapter in relation to other factors studied.
SUMMARY
In the past, battle wounds have been described chiefly in terms of organic damage or tissue loss. The purpose of this chapter has been to describe, shortly after the soldier arrived at the forward hospital, the latent consequences of his wounds as they influenced organic function and produced changes in blood volume and chemistry, and abnormalities in the urine. These matters were studied, before vigorous resuscitative efforts had yet been made, in 108 patients. Altogether, 186 patients were studied in the course of the work carried out by the Board for the Study of the Severely Wounded.
In considering patients who had received not more than one unit of plasma, or none at all, it was observed that the hematocrit level was higher in those with abdominal wounds than it was in those with peripheral wounds, but even in the patients with abdominal wounds the average hematocrit values were somewhat below normal. While severe hemoconcentration can occur incases of burns, crush, and abdominal wounds, this was infrequent in our series and was by no means a general characteristic of shock as we saw it.
When the patients who had received little or no plasma or blood therapy at the time of first examination were divided into two groups, depending upon whether more or less than 30 percent of the blood volume had been lost, a puzzling situation was apparent: the average concentration of protein in the plasma in the more severely bled-out group was 6.1 grams per 100 cubic centimeters. This is 6.1 percent below normal. On the other hand, the average hematocrit value in this same group had fallen to 36 from the normal of 47, a reduction of 23.4 percent. The hematocrit level thus fell about four times as much as that of the plasma proteins. One implication of this is that the blood had been diluted with protein-rich fluid. Its possible source is discussed.
Evidence is presented that the plasma protein level was not influenced by plasma therapy, although the hematocrit level was. There was a sharp fall in hematocrit value with increasing severity of shock. Other examples of the independent variation of these two factors are given.
It is shown that men can lose about 75 percent of their blood and yet re-
74
cover-more than had been generally supposed. Blood loss in various types of wounds is discussed. Data presented show that there was a quantitative relationship between loss of blood volume or hemoglobin and the degree of shock met clinically. This supports the view that the major cause of the shock we encountered was hemorrhage.
No important differences in blood volume or hemoglobin loss were encountered with the passage of time from wounding to examination. This was possibly to be accounted for by the high priority and consequent rapid evacuation given to patients with bleeding wounds.
The clinical condition of the newly wounded man offers abundant evidence that his internal state has been profoundly altered by the time he enters a forward hospital. In addition to the matters already mentioned, this was studied in terms of nitrogenous waste products, electrolytes, bilirubin, and blood sugar. In general, these substances were found not to be influenced by the location of the wound. The plasma nonprotein nitrogen level rose rather strikingly with delay following wounding. The full significance of this is not clear, but it offers grounds for some interesting speculation. Examination of the four shock categories in sequence from "no shock" to "severe shock" shows significant rises in nonprotein nitrogen, creatinine, phosphorus, and magnesium.
Acidosis was present in the patients in severe shock. They showed a considerable fall in carbon-dioxide combining power as compared with that of patients with no clinical evidence of shock. The acidosis appeared to be of the "metabolic" type.
No evidence of salt deprivation was found on hospital entry. Examination of the admission urine specimen with regard to hydrogen ion concentration and specific gravity indicated that the men studied had essentially normal renal function at the time they were wounded.
The van den Bergh index rose significantly with increasing time from wounding to examination. This is discussed briefly. No clear relationship of shock to bilirubin or plasma hemoglobin levels was found. The blood-sugar level was found to be above normal. It was particularly high in the patients with severe shock.
Definite depression of liver function, as measured by bromsulfale in retention, was found on hospital arrival. We found no correlation between liver function and the degree of shock. The administration of one or two units of plasma appeared to impair liver function still further. This was a transitory effect and was not increased by giving three or four units of plasma.
