U.S. Army Medical Department, Office of Medical History
Skip Navigation, go to content







AMEDD MEDAL OF HONOR RECIPIENTS External Link, Opens in New Window






Chapter VI



Delayed Internal Fixation of Compound Battle Fractures – A Followup Study in the Zone of Interior

During the first 16 months' experience in the Mediterranean Theater of Operations, manipulative reduction followed by immobilization in a plaster cast or continuous skeletal traction proved satisfactory, or reasonably satisfactory, methods of management for a major portion of the combat-incurred fractures. Inadequate reduction, however, was by no means infrequent. It was particularly common in fractures involving the condylar and articulating portions of the large joints, fractures of both bones of the forearm and both bones of the leg, fractures in which loss of bone had caused segmental defects without contact of fragments, and fractures accompanied by massive loss of soft tissue. Even though the position obtained by repeated efforts at manipulative reduction or continuous skeletal traction had often not been satisfactory in these groups of injuries, inadequate reduction was accepted because of the fear that operative intervention would incite systemic infection or at least a severe wound infection with prolonged osteomyelitis.

For these reasons, delayed internal fixation of compound battle fractures was seldom performed during this period. On occasion, when adequate reduction had not been achieved by nonoperative methods, open reduction and internal fixation were performed, even though the wound was unhealed. The results were encouraging in a small series at one hospital, but in another series, at a different hospital, they were poor. A study of these cases and of sporadic cases treated in this way in other hospitals did not produce evidence that would warrant recommendation or approval of the procedure.

Even though it was realized that there would be a place for delayed internal fixation in the management of problem fractures once wound management was better understood, its use in the Mediterranean theater was forbidden by directive in November 1943. Section IV of Circular Letter No. 48, issued 18 November 1943 from the Office of the Surgeon, Headquarters, North African Theater of Operations, read as follows:


1. This procedure is still under trial with reference to indications, hazards, and incidence of serious complications. Its use is restricted to special groups authorized to assume the responsibility as a special study.


At this time, there were no groups in the theater authorized to make such studies. For one thing, further investigation of the problem and further experience in wound management as a whole were necessary before internal fixation could be considered an advisable technique. For another, it was thought desirable to wait to test the technique until adequate supplies of penicillin were available, as a safeguard against invasive infection.

The use of delayed internal fixation in the Mediterranean Theater of Operations as a part of the reparative surgery of compound fractures began in March and April 1944, during the hectic days of Anzio and Cassino. Its use was first permitted in the 300th, 23d, and 21st General Hospitals, in which penicillin therapy had been made available as an adjuvant to an aggressive surgical program. The patients treated by internal fixation fell into two groups, as follows: (1) Those with unreduced fractures and unhealed wounds, who had previously been treated unsuccessfully by other methods; and (2) those recently admitted from evacuation hospitals after initial wound surgery, with no previous attempts at reduction of the fractures. The progress of each of these patients was followed carefully and, when necessary, the theater holding period was abrogated to permit prolonged observation, in an effort to assure that later recommendations concerning the use of internal fixation would be soundly based.

The results of the initial test of delayed internal fixation as they were observed in the spring of 1944 were extremely satisfactory. Moreover, when the first 14 cases so treated were studied later in the Zone of Interior, it was found that in 10 instances the fractures had united and the wounds had healed without removal of the metal. The final results were equally satisfactory in the other four cases, although it had been necessary, in order to expedite healing of the wounds, to remove the metal and sequestra. The results were particularly impressive in seven in which established wound infection had been present when stabilization of the fracture was undertaken.

The immediate results as obtained in the 300th, 23d, and 21st General Hospitals by the use of delayed internal fixation warranted its recommendation, on specific indications, as part of the program of reparative surgery of compound fractures. It was pointed out, however, that the use of this technique was not, in itself, an objective of the program. The measures designed to obviate infection and achieve wound healing included (1) the excision of dead tissue as part of revision of wounds in general hospitals; (2) the closure of compounding wounds, especially to cover denuded bone; and (3) adequate drainage of residual dead space or of unexcisable bits of devitalized tissue. All these measures were of greater importance in the program than internal fixation of fractures. Internal fixation would have been doomed to failure if the other essential surgery had been ignored.

Internal fixation was neither advisable nor feasible in the majority of compound battle fractures because of severe comminution. The fractures which would permit its use were in the minority. Internal fixation was therefore reserved for specific indications, as will be pointed out shortly (p. 190).



Internal fixation was frequently used at the first operation of reparative surgery, but it was also employed later, after other methods had failed. It was often used at the first reparative operation (1) in fractures about the joints, to secure anatomic replacement of the articular surfaces; (2) in fractures of the long bones located deep in muscle tissue, particularly fractures of the femoral shaft or upper radius; (3) in fractures which experience had shown were difficult to hold in reduction by other means, such as fractures of the olecranon process associated with massive soft-tissue loss; and (4) in fractures in which there had been segmental loss of bone and in which contact of the fragments could not otherwise be secured.

Internal fixation was accomplished in three ways, as follows: (1) By standard plating, (2) by multiple screws, and (3) by wire sutures. Slotted plates and intramedullary pins were postwar developments. The only type of metal available in overseas hospitals was 18-8 chrome-nickel stainless steel. Fixation by multiple screws (2 or more) was often particularly useful in oblique fractures. This technique required little or no additional periosteal stripping and was therefore associated with minimal trauma. When the fracture could not be stabilized rigidly because of comminution, one or more wire sutures were used to hold major fragments in apposition. They could usually be placed without additional stripping of the periosteum.

Intimate contact of the bony fragments was essential for sound union and was regarded as of such importance that in occasional cases, in order to secure it, the extremity was deliberately shortened, to overcome segmental loss and permit approximation of the fragments. The same plan was sometimes adopted when there were deficits of nerve trunks or muscles without segmental loss of bone. Removal of bony fragments permitted approximation of nerve or muscle bundles in these injuries and was undertaken with the idea of accomplishing maximum functional restoration of the whole extremity rather than merely achieving a united fracture.

Although the operation was sometimes performed through the compounding wound, this technique had two disadvantages: (1) The metal was placed on boric which was usually devoid of periosteum. (2) It was also placed at the bottom of dead space created by the excision of devitalized muscle. It was therefore usually better, especially when a plate was applied, to utilize a separate surgical incision, which permitted coverage of the bone and metal by periosteum and soft parts.

The routine procedure for the management of these casualties in a fixed hospital must be clearly understood. Every patient was prepared for reparative surgery, usually 5 to 10 days after wounding. He was anesthetized in the operating room, where the plaster cast and dressing applied after initial surgery were removed. The extremity was cleansed, prepared, and draped. The operating room was set up for any surgery which might be indicated on a compound fracture.


The wound was then thoroughly visualized by gentle retraction. All residual dead tissue was excised, and the depths of the wound were cleansed of old blood clot. The fracture site was exposed. In the great majority of cases, reduction was attempted by traction or manipulation, but internal fixation could be employed at this time, as part of the first procedure of reparative surgery, if its advantages were evident, as, for instance, in condylar fractures of joints, oblique fractures of long bones which were easily reducible, or segmental defects due to bone loss. Wound closure, usually with drainage, completed this stage of reparative surgery.

If later roentgenograms showed that adequate reduction had not been achieved by traction or manipulation and if the contour of the fracture permitted, internal fixation could be performed at another operation, perhaps after wound healing.

Delayed internal fixation was not reserved for the ideal case. On the contrary, it was frequently employed in compound fractures that were major problems under any plan of management, as, for instance, an avulsion of the soft parts of the arm which exposed the shaft of the humerus for several inches or a grossly displaced, infected fracture of the femur, with a hinge soft-part wound, 66 days after surgery. These are situations hardly included in the realm of reparative surgery.


After the program of reparative surgery of compound fractures, including internal fixation on special indications, had been extended to all general hospitals in the Mediterranean theater, the results of this procedure were closely checked. In the majority of cases, they were regarded as satisfactory. The incidence of wound infection had apparently not been increased. Drainage had usually not been prolonged. Ultimate scar formation was expected to be less. Fracture reduction was certainly improved. There seemed, therefore, every reason to hope that there would be considerable improvement in both anatomic and functional end results in the fractures in which this method was employed.

One disturbing fact was that, in spite of frequent warnings to the contrary, a few surgeons in the theater had accomplished internal fixation by plating in a relatively large number of fractures of the tibia. The consultant in orthopedic surgery had repeatedly pointed out that anatomic conditions in this area do not permit satisfactory coverage of the denuded bone by vascular soft parts. The risk of massive sequestration after the application of a plate had also been emphasized. This was particularly true because of the stripping of the periosteum necessary when the plate was applied to the anteromedial surface of the bone. The surgeons who, despite these warnings, attempted to manage fractures of the tibia by plating often found that they had created for themselves major problems in closure of the wound, even when fixation had been carried out through a separate surgical incision. Results observed in the Mediter-


ranean Theater of Operations in the management of compound fractures of the tibia by internal fixation were far less satisfactory than the results achieved by this method in fractures of any other long bone.

On the whole, the results observed in cases in which internal fixation had been used seemed highly satisfactory when the patients were examined before evacuation to the Zone of Interior. The results, in fact, seemed particularly good when they were evaluated in the light of the problems which had to be solved in these cases. Overseas observation, however, did not settle the question of end results. For that, investigation in the Zone of Interior was necessary. Previous attempts to secure followup data had not been successful. The Surgeon, Mediterranean Theater of Operations, therefore requested that the consultant in orthopedic surgery for the theater be ordered to the United States, to carry out, among other observations, a survey of casualties with compound fractures who had been treated in the Mediterranean theater by delayed internal fixation and who were then in Zone of Interior hospitals. This survey was carried out in the spring of 1945, with the cooperation and assistance of the Surgical Consultants Division, Office of The Surgeon General.

Materials and Methods

The survey was conducted in 24 general hospitals in the United States between 16 March and 26 April 1945. The 332 fractures surveyed had been managed in 18 general hospitals in the Mediterranean theater, by approximately 50 orthopedic surgeons. The material thus represented a cross section of the battle experience and surgical proficiency of the theater. The majority of the patients had undergone operations in which delayed internal fixation had been performed between 4 June 1944, the date of the fall of Rome, and 1 November 1944, the date of the approximate conclusion of the fighting on the Gothic Line.

Two hundred and ninety-five case reports, representing three hundred fractures, were assembled as follows:

1. From personal examination of patients still hospitalized or still required to report for observation to outpatient clinics.

2. From a study of records of patients still hospitalized but presently absent on furlough or pass.

3. From a study of the records of patients who had been referred to convalescent hospitals, returned to duty, or given a Certificate of Disability discharge.

In each instance, the total record was studied, from wounding to the date of the investigation, and all roentgenograms were examined. The case reports made up as a result of these investigations included, as far as possible, the following data: The precise diagnosis of the bone injury; the diagnoses of associated injuries which might have influenced fracture management; an appraisal of the indication for internal fixation; the type of fixation (plate, screw, wire suture); the time interval between wounding and fixation; the


status of the wound at the time the fixation operation was performed; the surgical approach to the fracture (through the compounding wound or through a separate surgical incision); the presence of wound infection at the time of the fixation operation; the presence of associated nerve injuries; the results achieved from the standpoint of wound healing and bony union; the time at which wound healing and bony union were achieved in relation to the time of the fixation operation; and whether removal of metal and sequestra had been necessary.

Another group of 32 patients consisted of soldiers who had been demobilized, were in convalescent hospitals which were not visited, or had been returned to duty. Their status indicated, with reasonable assurance, that their wounds were healed and their fractures united, which meant that optimal results had been accomplished. Their records, however, were lacking in certain basic data, including the kind of fixation employed and the indication on which operation had been done. These cases, therefore, are necessarily omitted from some of the tabulated presentation of results.

The material secured from this survey is presented chiefly in the form of tabular data, with special emphasis on the results obtained to the date of the investigation in relation to (1) the indications for internal fixation, (2) the technique of fixation employed, and (3) the area of injury. Certain other data are also discussed.

No true controls exist for thus series. In all the hospitals surveyed, a serious effort was made to secure data on comparable cases treated by methods other than internal fixation, but the information had not been compiled and could not be obtained. A number of patients were observed, however, whose results might have been improved if internal fixation had been used to stabilize their fractures. In a number of instances of malunion and nonunion in femoral fractures studied roentgenologically, the contour of the fractures suggested that this technique might have been feasible and might have given better results. In a number of fractures of the humerus with segmental defects, it was also thought that union might have been achieved if bony apposition had been maintained by metallic internal fixation.

Definition of Terms

Indications. – Indications for the internal fixations performed in this group of fractures were classified as obligate and elective. The terms carry their own definitions.

Obligate indications included the following:

1. Bone loss which either had produced a segmental defect without contact of the fragments (figs. 38 and 43) or was associated with persistent distraction of the fragments (figs.30 and 50). In either event, union could riot be expected without corrective measures.


2. Fractures about the joints, particularly condylar fractures of the knee or elbow. Reparative measures were necessary in this type of fracture to accomplish joint congruity (fig. 53).

3. Massive soft-tissue loss which precluded routine measures of closure and required staged procedures. Fixation of the fracture was part of the attempt to achieve wound healing.

4. Associated nerve injuries which required, for optimum results, early stabilization of the fracture, sometimes with deliberate shortening of tire bone to permit approximation of the severed nerve ends (fig.39).

5. Failure to achieve arid maintain adequate reduction by manipulative measures or traction (figs. 30 and 50). Malunion, delayed union, or nonunion was inevitable in such cases in the absence of corrective measures.

The operations performed under the last mentioned indication were obviously performed after failure of other methods of treatment. In all other operations done on obligate indications, internal fixation was sometimes performed at the first operation of reparative surgery and sometimes later, in many instances after the wound had healed.

All indications which were not obligate were regarded as elective. Elective operations were carried out at the first procedure of reparative surgery.

Results. – Results were classified as favorable, unfavorable, and incomplete, as follows:

Favorable. This category was further divided into--

1. Group A. In this group of optimum results, the fractures united in perfect, or almost perfect, anatomic alignment, and the wounds healed solidly and promptly, without sequestration and without removal of tile metal used in the fixation (figs.31, 34, 35, and 43). In several fractures of the femur and of tile tibia, the metal was removed later, either as a prophylactic measure or because slight evidences of absorption were detected around the screws. Its removal, however, was not necessary to accomplish either wound healing or bony union (figs.30, 36, and 52).

2. Group B. In this group, the fractures united promptly in adequate reduction, and wound healing occurred satisfactorily after metal and sequestra had been removed (figs.32, 50, and 63).

Unfavorable. – This category was further divided into–

1. Group C. – In this group, union of the fracture occurred promptly but with massive sequestration, and metal and large sequestra had to be removed before wound healing was obtained (figs.64 and 65). All cases in the C group were fractures of the tibia, a bone in which, as already pointed out, plating is always associated with some risk. The massive sequestration which occurred threatened bony continuity and introduced the risk of refracture when weight bearing was resumed. The risk thins introduced, because of loss of bony strength, made it necessary to use braces for several months (fig. 65).

2. Group D.  – In this group, wounds healed without sequestration or removal of metal, but bony union had not occurred at the time of tile survey.


The results are classified as unfavorable for this reason.   From the standpoint of wound healing, management was successful. The fact of nonunion, of course, must be evaluated against the chances of the same result if internal fixation had not been performed (fig.44).

3. Group E. – In this group of fractures, wound healing did not occur, and in most instances fracture union was not achieved, until both sequestra and metal had been removed. The time required for both wound healing and bony union therefore exceeded normal expectancy.

4. Group F.  – In this group, fractures did not unite, and wounds did not heal. In several instances, wound healing occurred after removal of metal and sequestra, but it was expected that bone-grafting procedures would be necessary in all cases to accomplish bony union (fig.64). As in group D, these results must be evaluated against the probability of similarly unsatisfactory results if internal fixation had not been employed.

Incomplete. – In these fractures (group G), it was thought that satisfactory end results would eventually be secured, as the fractures were well united. Wound healing, however, had not occurred at the time of the survey. In some instances, plastic procedures on the wound would obviously be necessary, in addition to sequestrectomy and removal of metal, before healing could be expected to occur. In other instances, it was felt that, if metal and sequestra were removed, the same excellent results would follow this secondary procedure as had followed it in another group (B) in the series (fig. 52).

Results in relation to indications for internal fixation are presented for the whole series in tables 17 and 18, and for regional injuries in tables 19, 20, 21, and 22.

Results in Relation to Technique

Results of internal fixation in relation to technique are presented in table 23.  As it shows, plating was used in slightly over half of the 332 compound fractures of the long bones included in the survey. Tire largest proportion of entirely satisfactory results (groups A and B) was obtained with multiple screws; results were excellent in 82 of 95 such cases. There were 21 absolute failures (failure of wound healing and nonunion of the fracture) at the time of the survey in the 168 fractures treated by plating, against 4 absolute failures in 69 fractures treated by wiring. There were no absolute failures in the 95 fractures fixed by screws. There were a number of instances in the unfavorable categories in which it was thought that with the passage of time better results than were then apparent might be achieved.

1 This category of results was classified as satisfactory in the original report, on the ground that, although the fracture was ununited, the wound was healed and that the chief purpose of the survey was to determine the status of wound healing in fractures treated by delayed internal fixation. Because of the possibility of misunderstanding (since an ununited fracture cannot be considered a satisfactory result), this category of results has been moved to the unsatisfactory group of cases, in the exercise of the editorial function. The author of the volume, who conducted the survey, does not regard the results achieved as entirely unsatisfactory, since this group of cases demonstrated that the use of delayed internal taxation in open or compound fractures due to gunshot wounds did not cause a high incidence of infection. In the great majority of cases, the wound and fracture both healed in due time, it should he remembered that this survey does not represent a study of end results but is rather a progress report; in some instances, sufficient time had not elapsed from the date of internal fixation to permit final evaluation of either wound healing or fracture healing. [Editor's note.]


TABLE 17. – Results of internal fixation performed on obligate indications in 135 compound fractures 1

TABLE 18. – Results of internal fixation performed on elective indications in 165 compound fractures 1

TABLE 19. – Results of internal fixation in relation to indications in 67 compound fractures of humerus 1


TABLE 20. – Results of internal fixation in relation to indications in 31 compound fractures of radius and ulna 1

TABLE 21. – Results of internal fixation in relation to indications in 146 compound fractures of femur 1

TABLE 22. – Results of internal fixation in relation to indications in 88 compound fractures of tibia and fibula 1


TABLE 23. – Composite results of internal fixation in relation to technique and location of fracture in 332 compound fractures of long bones 1

Sequestration was apparently the chief hazard associated with internal fixation (fig. 64). It occurred in slightly over a third of the cases surveyed and, case for case, seemed more extensive than might have been expected in the light of cases in which this procedure had not been used. Experience overseas had shown that in the ordinary course of events sequestration was practically always limited to areas of bone which had, presumably, been denuded at wounding; these areas were increased by the periosteal stripping necessary when internal fixation by plating was used. In some cases surveyed in the Zone of


Interior, failure of wound healing was clearly attributable to delay in removal of sequestra as well as metal. Since the fractures were well united in all of these cases, there seemed no reason for having postponed the secondary surgery which was obviously necessary.

Sequestration was only occasionally massive when screw or wire fixation was used. When it occurred with these techniques, it did not seem to interfere with union of the fracture. In such cases, it would probably have occurred no matter what method of fracture management had been employed. There was no doubt, however, of the tendency toward sequestra formation when plating was employed. It occurred in 69 of the 168 fractures treated by this technique and was sometimes so massive that the resulting bony defect could be compensated for only by bone grafting.

Sequestration seemed less likely to occur, particularly when plates were used, if the internal fixation was done through a separate incision rather than through the compounding wound. It was also thought that wound healing occurred more promptly if the operation was performed through a separate incision.

Other Data

The details of 29 fractures in which internal fixation was performed in the presence of established wound infection are presented in table 24. The average time from wounding to operation in these cases was 36 days.

Only 37 nerve injuries were recorded in these 332 fractures, 21 in fractures of the humerus, 6 in fractures of the radius and ulna, 9 in fractures of the femur, and 1 in a fracture of the tibia and fibula. Only 1 nerve injury in 88 fractures of the tibia and fibula seems unlikely; other nerve injuries probably occurred and were not recorded.

TABLE 24. – Results of internal fixation in 2.9 compound fractures with established wound infection 1


Refracture, which in each instance involved the femur, occurred 7 times in these 332 fractures. In 6 of the 7 refractures, the second fracture occurred at a point at which there had been bone loss at the site of the original injury. In the seventh refracture, the bone loss had been caused by sequestration. In four instances, there was no obvious cause for the second fracture. In two instances, the refracture followed falls, and in the remaining instance it followed manipulations to secure motion of the knee joint. In spite of the complication of the second injury, the end results were good to excellent in all seven refractures.

Generally speaking, although no statistics were collected to prove it, it was thought that the range of knee motion following rigid internal fixation (that is, by plates and screws) of fractures of the femur, when surgery was supplemented by a program designed to achieve maximum functional results, exceeded the range achieved in comparable injuries managed by other methods. The range of knee motion, however, varied with the level of the fracture. Knee motion was excellent in some fractures of the upper half of the femur. It was less good in many fractures of the lower half and was particularly unsatisfactory in fractures of the lower third. In many cases, the explanation of less than satisfactory knee function was failure to institute a program of knee motion after the patient reached the Zone of Interior in a hip spica. The omission of this program simply failed to capitalize on one of the decided advantages of internal fixation.

It was interesting to observe the results in patients who had come from the 21st General Hospital in the Mediterranean theater, which was known to have an excellent program of postoperative knee motion in effect on its fracture wards. Whether the patients with fractured femurs from this hospital had or had not been managed by internal fixation, the range of knee motion, case for case, was superior to the range of motion observed in patients from other hospitals.


It was concluded from this survey of internal fixation carried out in hospitals in the Zone of Interior that, when this procedure is used on correct indications and is performed by the correct technique, it has a definite, if limited, place in the management of battle-incurred compound fractures of the long bones in fixed hospitals overseas. The term “correct indications” implies that the operation is performed only as an adjuvant measure and within the strict limitations of reparative surgery. Failure to institute surgical measures to forestall infection and favor wound healing invariably prejudices the results.

There were no deaths and no amputations in these 332 fractures. The overall results were satisfactory to excellent with two groups of exceptions, fractures associated with massive soft-tissue loss and fractures of the tibia and fibula managed by plating.


In the light of the results achieved in this series and evident in this survey, conclusions concerning the results possible of achievement by delayed internal fixation may be stated as follows:

1. Nonunion can be prevented in fractures with segmental defects or persistent distraction, as is evidenced by the achievement of bony union in 27 of 39 such fractures in this series.

Varying degrees of malunion, delayed union, or nonunion can be prevented in many cases in which inadequate reduction was achieved by other measures. This is evidenced by the achievement of bony union in good apposition and alinement in 64 of 75 such fractures in this series.

Bony union, with optimal reduction, can be achieved in condylar fractures about the knee and elbow. This result was achieved in all 15 such fractures in this series.

The obvious advantages of well-stabilized optimum reduction can be achieved in fractures which lend themselves to rigid stabilization, and satisfactory wound healing may be achieved in many cases, provided that fixation is accomplished by multiple screws, with minimal periosteal stripping. Union was accomplished in good position in all 95 fractures thins managed in this series. Wound healing was accomplished without sequestration or removal of metal in 71 of these fractures and was accomplished after their removal in another 14.

When plating is the method of fixation, bony union and satisfactory wound healing must be anticipated in a smaller proportion of cases. In 168 fractures managed by this technique, bony union was accomplished in 140. Wound healing was accomplished without sequestration or removal of metal in 99 fractures and was accomplished after their removal in another 34.

Improved apposition of fragments can sometimes be provided by the use of wire sutures, and favorable results may be anticipated in many instances. When this technique was used in this series, bony union was accomplished in 59 of 69 fractures. Wound healing, without sequestration and without removal of metal, was accomplished in 50 fractures, and in another 12 the wound healed after metal and sequestra were removed. Four fractures in which wire sutures were used were absolute failures, and three others were incomplete at the time of the survey.

2. The chief hazard of delayed internal fixation, namely, increased sequestration, may be explained by the periosteal stripping which the procedure entails and by its interference with readherence of soft parts to denuded bone. Other observations indicate that sequestration of bone in nonfixed battle fractures is practically always limited to bone that was probably denuded at wounding. Sequestration occurred in a little over a third of the fractures in this series, but comparable data in a control series are not available.

The sequestration which occurred in this series when screw or wire-suture fixation was used was seldom massive, did not seem to interfere with union of the fracture, and probably would have occurred in many cases if the fractures had been managed by other methods.


Massive sequestration occurred in 41 of the 168 fractures which were managed by plating, with retardation of attainment of full strength of the bone. In some fractures managed by plating, a massive defect may be created winch is reparable only by bone grafting.

3. Unless the fracture contour permits rigid fixation by screws or unless wire sutures appear advantageous, it is best to attempt reduction by traction or manipulation and strive for early wound healing. When wound healing has been accomplished, fixation by plating or some other technique is relatively safe. Wound healing occurred by this plan in 20 of 21 fractures in this series, without sequestration or removal of the metal.

Excellent results may be expected in fixations of the long bones of the upper extremity if severity of bone loss or of the soft-tissue injury does not prejudice the chances of union and wound healing. Massive soft-tissue loss had occurred in this series in 5 of 7 failures of internal fixation of fractures of the upper extremity, and bone loss had occurred in the other 2 fractures.

Fixations of the femur, performed on correct indications, by means of multiple screws or wire sutures and with minimal periosteal stripping, may be expected to give excellent results. In this series, there were no failures in fractures of the femur fixed by screws. Bone loss was responsible for tine two failures with wire-suture fixation. One of the patients, in addition, had suffered a massive loss of soft tissue. The risk of the periosteal stripping necessary when plating is employed makes it preferable to delay fixation until after wound healing unless the indications and anticipated advantages overshadow the hazard.

Fixation of the tibia by multiple screws or wire suture may be expected to give very satisfactory results. Periosteal stripping should be kept to a minimum. There were only two unfavorable results in the fractures in this series fixed by screws; in both, heavy sequestration occurred. The single entirely unfavorable result in the fractures fixed by wire sutures was expected, ultimately, to be favorable; only 4 months had elapsed since wounding, and it was thought that the fracture would eventually unite. Plating of the tibia should be reserved until after wound healing. Eight of the nine failures in fixations of the tibia, and six of the eight massive sequestrations, occurred after plating of this bone. A more extensive use of wire sutures to maintain approximation might have improved results.

When both tibia and fibula are fractured, plating of the fibula may be a useful procedure, and one which is relatively safe. It maintains length and alignment, aids in achieving apposition of tibial fragments, and provides some degree of immobilization of the fracture of the tibia.

4. Internal fixation, when used as an adjuvant to the management of unreduced, infected compound battle fractures, may aid in the control of infection and in achieving the best result which can be obtained in the circumstances. In the 29 infected cases in this series (table 24), bony union was achieved in good alignment in 20, and wound healing occurred without further


sequestration in 13. In 13 additional cases, wound healing occurred after removal of sequestra and metal.

This followup study of 332 compound fractures, all but 13 of which were combat incurred, revealed satisfactory to optimal results in 258 cases (77.7 percent). These 258 fractures illustrate the optimum results which could have been achieved. They include 21 fractures in which removal of metal and sequestra had not yet been performed but in which excellent results could be expected after secondary surgery had been carried out. They do not include eight fractures of the femur in which massive sequestration had occurred but in which there seemed reason to hope that satisfactory weight-bearing extremities would eventually be secured.


The 53 fractures classified as unsuccessful fall into 4 groups:

Eight operations, all performed on elective indications (group C). In this group, union of the fracture occurred promptly, but metal and massive sequestra had to be removed before wound healing was obtained. All cases in this group were fractures of the tibia, in which plating is always associated with some risk. The massive sequestration which occurred threatened bony continuity and introduced the risk of refracture when weight bearing was resumed. This risk made it necessary to use braces for several months longer than was necessary in the A and B groups.

Fourteen operations, nine performed on obligate and five on elective indications (group D). In this group, union had not occurred at the time of the survey, and, for this reason only, the results are classified as unfavorable.  2   From the standpoint of wound healing, the result was successful. The fact of nonunion, furthermore, must be evaluated against the chances of the same result if internal fixation had not been performed.

Six operations, three performed on obligate and three on elective indications, in which wound healing did not occur and, in most instances, union of the fracture was not achieved, until both sequestra and metal were removed (group E).  The time required for both wound healing and bony union therefore exceeded normal expectancy.

Twenty-five operations, sixteen performed on obligate and nine on elective indications, in which fractures did not unite and wounds did not heal (group F). The fractures in this group represent the only absolute failures in the series, and, even in it, wound healing occurred in several instances after removal of metal and sequestra. It was expected, however, that bone-grafting procedures would be necessary in all cases to accomplish bony union. As in group D, these results must be evaluated against the probability of similarly unsatisfactory results if internal fixation had not been employed.

2 See footnote 1, p.192 .


These unfavorable results, to consider the fractures from the separate standpoints of wound healing and bony union, include 14 fractures in which wound healing occurred but bony union did not; 8 fractures in which bony union occurred but wound healing was obtained only after metal and massive sequestra were removed (all fractures of the tibia); 6 fractures in which wound healing did not occur and bony union was usually not achieved until both sequestra and metal were removed; and 25 fractures in which neither union of the fractures nor wound healing was accomplished.

In the 53 unfavorable cases, 19 operations were performed on the femur, 20 on the tibia and fibula, 9 on the humerus, and 5 on the radius and ulna. Twenty-five were performed on elective and twenty-eight on obligate indications.

Certain of these cases warrant special comment:

In 2 operations on the femur performed on obligate indications, 1 managed by plating and 1 by wiring, the indication was segmental bone loss so extensive that nonunion was regarded as probably inevitable by more conservative methods. In a third fracture, which was managed by plating, previous management had been unsuccessful. The fracture was double, and the large rotated central fragment could not be reduced. When the patient was surveyed, the proximal fracture was well united, and it was thought that the distal fracture might still unite. In spite of the unsuccessful result in this case, the use of internal fixation was regarded as both justified and advantageous.

In one of the elective operations on the femur, the fracture, which was mildly comminuted, was only partly stabilized by screws and wire sutures; the operation had been performed in air infected field. The patient had already undergone amputation of the foot on this side and amputation of the contra-lateral leg for infection. He was observed only 4 months after wounding, and it was thought that the fracture might still unite. In another unsatisfactory operation, the major fragments of the severely comminuted femur were sutured in apposition. The wound was infected, the infection extending into the joint, and the fracture site had to be drained by dependent drainage. The failure of union in this case could not be attributed to internal fixation.

In one operation on the tibia and fibula, classified as unfavorable because of nonunion although the wound healed promptly, the failure must be charged against poor technique; examination of the roentgenograms taken overseas showed that the bone had been plated with the fragments distracted. In another operation, a 4-inch defect in the tibia had been strutted by plating the fibula, in expectation of a later bone-grafting operation. In a third operation, nonunion was explained by bone loss and unsatisfactory contact of the fragments.

In three unfavorable operations on the humerus, all managed by wiring, there was segmental bone loss. It is interesting that in 10 similar operations in the series union was achieved by the same technique. In another operation, plating was performed 25 days after wounding, when wound healing was complete, because conservative measures of fracture management had failed to achieve satisfactory reduction.


In a fracture of the bones of the forearm in which internal fixation failed to produce union even through the wound healed, the indication for operation was bone loss. The ulna was intact, but there was a l-inch segmental defect in the radius. A wire suture was applied to help maintain alignment, without any real expectation that union could be accomplished. In another patient, with a fracture of the radius and a double fracture of the ulna, the indication was elective. A Steinmann pin which had been passed down the medullary canal of the ulna had to be removed at the end of 3 weeks, and the radius was plated at this time. When the patient was observed in a Zone of Interior hospital, neither fracture of the ulna had united and reduction was poor, but it was thought possible that union of the radius might still be accomplished.

In the 25 operations regarded as absolute failures, 16 were performed on obligate and 9 on elective indications. The failure in each case must be evaluated in the light of the severity of the problem and the indications for fixation. When the 25 fractures are analyzed from this standpoint, the conclusion is reached that 2 fractures had been improved by internal fixation; 8 had not been, through progress had not been retarded; and 13 (3.9 percent of the 332 fractures in the series) had been harmed by the operation. In the two remaining fractures, a final evaluation from the standpoint was not possible at the time the survey was made.

EDITOR'S NOTE – It should be emphasized again, as the author has noted already (p. 191), that fixation of the tibia with plates and screws yielded the highest percentage and the greatest number of unfavorable results in the open fractures treated by internal fixation. Unfortunately, these techniques are the easiest to accomplish. These facts should be brought home to every medical officer. Otherwise, the tibia will continue to serve as a boobytrap for the unwary, incompletely trained surgeon, with the wounded soldier as the victim.