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 XX



Injuries of the Knee Joint

Mather Cleveland, M. D.

In the European theater, as in all other theaters, wounds of the knee joint presented a major problem. They were relatively frequent battle injuries, but none of the surgeons who were obliged to handle them had seen, in civilian practice, a sufficient number to standardize treatment, nor had they learned from personal experience the precautions and safeguards to prevent the development of infection. Furthermore, the devastating type of battle-incurred injury of the knee joint observed in military surgery does not occur in civilian practice. Suppuration of the joint was always the most dreaded complication of wounding, and all treatment was directed toward its prevention. Improper or inadequate care could mean the loss of the limb, and even death from sepsis.

Nature of the Injury

The most frequent cause of wounds of the knee joint was shell fragments, followed in order of frequency by bullets, fragments of rock or gravel, wooden splinters, and pieces of plastic from a certain type of German land mine. Other varieties of plastics, as well as wood fragments, were seldom encountered. The latter varieties of foreign bodies were readily overlooked because of the difficulty of visualizing them by roentgenograms. Rock and gravel were seldom driven into the joint by force of an explosion or the impact of a missile, unless the knee was in close contact with the ground.

The amount of initial damage which a penetrating fragment inflicted upon the joint depended in large measure not only on the size of the object but also on the velocity of its penetration. A small fragment, which had barely penetrated the joint and which had produced, according to roentgenologic evidence, little damage to the bony structures sometimes caused fulminating suppuration, depending upon the material which it carried in with it and the judgment with which the injury was managed.

Penetrating injuries of the knee joint were either direct or indirect. They were direct when the missile itself entered or perforated the joint. They were indirect when the missile fractured the femur, the tibia or the patella and thus established a communication between the synovial cavity of the joint and the compound fracture. Single or multiple structures could be damaged in either type of injury.

1The material in this chapter was secured from two reports made to the senior consultant in orthopedic surgery in the European Theater of Operations, one by Lt. Col. James E. Thompson, MC, Maj. William H. Cassebaum, MC, and Capt. (later Maj.) Charles F. Stewart, MC, of the 9th Evacuation Hospital; and the other by Maj. (later Lt. Col.) Louis A. Goldstein, MC, of the 19th General Hospital.


Indirect injuries took a wide variety of forms. Sometimes a single long fracture in a remote part of the shaft extended through the condyles into the joint. Frequently the indirect injury was a compound fracture of the patella, which was likely to be overlooked because of the difficulty of demonstrating, by roentgenograms, fracture lines extending into the joint surface. Unsuspected penetration of the joint was often discovered at operation when the compounding wound was at a considerable distance from the knee joint and the missile also lay at some distance.

Although the capsule and collateral ligaments of the medial and lateral aspects of the joint, together with some of its tendons, were frequently damaged, the damage was seldom of considerable extent. On the anterior aspect of the knee, there was greater opportunity for penetration. The capsule in that area is normally much laxer, to allow greater mobility, and the suprapatellar pouch extends for some distance above the level of the joint.

Small and compact patellar ligaments were often severed or almost completely destroyed by the passage of a fragment of any size. The lateral expansions of the larger quadriceps tendon, which blend with the capsule, made damage to this tendon less of a problem of management. Because of its structure, retraction was minimized and future function more readily resumed.

A missile which entered the knee on the mediolateral aspect and lodged in the condyle of the femur might not have traversed the joint before reaching its destination. An accurate knowledge of the anatomic reflection of the synovial membrane at that point was necessary before any opinion could justifiably be expressed. A fragment which had passed through one condyle might damage it only slightly but might completely demolish the other condyle.

The main vessels and nerves to the lower extremity, because they lie in the popliteal fossa, are safely protected from ordinary forms of injury, but they were so frequently injured by penetrating wounds about the knee that their integrity was suspect in any wound in the area. This was particularly true of the popliteal artery, injury to which took precedence over the treatment of any other injury. Damage to the vessels almost invariably meant loss of the limb; reparative procedures were seldom satisfactory. Damage to the nerves, if the injury was irreparable, resulted in a paralyzed limb, even if the blood supply remained intact.


Diagnosis of injury of the knee joint was based on a combination of clinical and roentgenologic evidence. The experience of the surgeon, combined with his knowledge of the regional anatomy, often provided a considerable part of the proof. Many times the question of whether or not the joint was penetrated was not settled until operation. In such cases, exploration was actually part of the diagnostic routine.

Motion in the wounded joint was likely to be limited and painful. Local tenderness could be elicited at various points. There was always an evident


excess of intra-articular fluid, which was usually bloodstained on aspiration. The amount present was usually in direct proportion to the severity of the injury.

Roentgenograms could be misleading because they showed only bony damage. They did not demonstrate cartilaginous coverings of the condyles, and it was therefore necessary to bear in mind that the profile of the femur or tibia observed in the joint was the representation of the cortex of the underlying bone and provided no information about the state of the cartilage. When adequate exposure was obtained at operation, it was often surprising to see the extent of the damage to important cartilaginous structures that had been completely invisible on the roentgenograms.

Evaluation of the damage to the knee joint was based on the relative importance of the structures damaged. Damage to the cartilaginous structures of the tibial and femoral condyles and the inner surfaces of the patella was of major importance. Damage to the menisci was of almost equal importance. Damage to the synovial membrane and the cancellous bone of the condyles was of lesser importance. These structures are rich in blood supply and therefore offer more resistance to infection than cartilage, which is notoriously poor in blood supply, and which, when it was detached and devitalized, served as an excellent nidus of infection.

Initial Wound Surgery

Treatment of injuries of the knee joint fell into three categories, as follows: (1) First aid, given in forward aid stations; (2) initial wound surgery, performed in the evacuation hospital or, occasionally, in the field hospital; and (3) definitive surgery, performed in the general hospital in the rear.

First aid consisted simply in the control of bleeding, if that was necessary; the application of a Thomas splint, if it was available, otherwise of some other form of immobilization; and, until it was forbidden late in the war, the local application of sulfanilamide crystals to the wound. Until late in the war, each soldier was provided with sulfonamide pills which he was supposed to take, if he was conscious, as soon as he was injured.

If the knee injury was associated with a vascular injury, which always demanded immediate attention, the casualty was cared for in a field hospital unless there was an evacuation hospital in the immediate vicinity. In the great majority of injuries of the knee joint, initial wound surgery was performed in an evacuation hospital, in which these wounds had a high priority. Consultants and experienced traumatic surgeons lost no opportunity to emphasize to more inexperienced officers that the ultimate outcome of the injury was decided in the forward hospital. The treatment the soldier received at this point would determine whether he would recover with a normal joint or a stiff joint, whether he would lose his limb, or whether he would lose his life as the result of prolonged sepsis.

If a vascular injury was present, it was always attacked first, regardless


of the amount of damage to the knee joint. Otherwise, whatever the nature of the injury, the first procedure was debridement of the soft tissues and the joint. An occasional officer believed, until the end of the war, that perforating wounds caused by high-velocity missiles, with only small wounds of entrance and exit and no demonstrable bone damage, did not require arthrotomy. This was not official policy, and it was not sound policy. The only safe plan, when penetration of the knee joint had obviously occurred or was suspected, was to open the joint and determine, by direct inspection, the conditions present. This policy held whether the injury was direct or indirect and regardless of the size of the wound.

Technique of debridement-Operation on the joint was best performed with a tourniquet in place, so that the intra-articular structures would not be obscured by troublesome bleeding. The incision into the joint was sometimes made through the original wound; when it was on the medial, lateral, or posterior aspect of the knee, this was frequently possible. Most often, however, the original wound was merely debrided, either at the beginning or the end of the operation, and the exploratory incision was separate and elective.

The incision of choice was a long medial or lateral parapatellar incision, which was regarded as giving the most satisfactory exposure. If a satisfactory operation could not be performed through one incision, another was made on the opposite aspect of the patella. At times it was necessary to combine a medial or lateral parapatellar incision with a posteromedial, posterolateral, or posterior midline incision, in order to care for damage in the posterior compartment of the knee joint. The trauma caused by multiple incisions was insignificant in comparison with the bad effects of inadequate surgery.

If the foreign body lay in the condyle of the femur, the approach of choice was through the lateral extracartilaginous portion. If it had entered the cartilage, the direct approach to it was usually best.

When the joint was entered, fluid and blood clots were evacuated, and the cavity was thoroughly irrigated with warm physiologic salt solution until the fluid returned clear. Then the extent of the damage was determined by inspection and palpation. If exposure was not adequate, the incision was extended or another was made.

Thorough debridement of the joint was the keystone of the whole operation. It had to be both careful and complete. The initial irrigation removed only blood clots, free bits of cartilage, and other debris from the recesses of the synovial cavity. Then all loose and devitalized portions of articular cartilage, bone, and menisci were removed. Traumatized surfaces were debrided down to healthy tissue with a knife or curette; leaving devitalized cartilage and bone fragments in the joint was an invitation to suppurative arthritis.

Whether penetration of the knee joint was direct or indirect, the damaged articular elements required exactly the same careful excision.

Snug closure of all synovial defects, whether traumatic or surgical, was mandatory. Leakage after operation invited ascending infection. The syno-


vial membrane, like the peritoneum, has a great capacity for resisting infection, but once infection is established in the joint, this resistance is no longer effective. With these principles in mind, the policy was to close the synovial capsule and membrane so tightly that leakage could not occur.

Large defects in the capsule, especially in the medial and lateral aspects, introduced special problems in closure which demanded special techniques. Anything was preferable to leaving the wound in the joint open. If large capsular defects were not closed initially, the risk of infection was always present, and it was usually impossible to close them successfully later by any form of surgical procedure.

Plastic procedures often made prompt reconstructive surgery possible and allowed the patient to recover such joint function as was compatible with the degree of articular damage. Frequently, the best that could be done was to approximate the synovial membrane and capsule as completely as possible, then cover the remaining defect with one or more sliding grafts of skin and subcutaneous tissue from the adjacent area. Results were usually excellent with this technique, while closure under tension inevitably resulted in breakdown of the wound. If capsular substance had been lost and not enough deep tissue was available to cover the defect, closure with skin was attempted, relaxing counterincisions being made if they seemed necessary.

After closure, penicillin solution (10,000 units in 5 cc. of physiologic salt solution) was injected into the joint cavity, preferably through a needle previously passed through the anterolateral aspect beneath the patella and known to be in the joint cavity. A less desirable technique was to close the synovial membrane around a small catheter or cannula, make the injection through it, and then, after it was removed, seal the opening with a purse-string suture.

The soft tissues and skin were left open, as in the usual debridement, and the wound edges were separated with a light, loose packing of dry, fine-mesh gauze.

After operation, the knee was immobilized in a hip spica. Many surgeons at this time cut a large anterior window into the plaster to permit direct access to the joint for inspection and treatment. A modified Tobruk splint was also satisfactory for transportation splinting.

Whenever a penetrating wound of the knee joint was diagnosed or suspected, penicillin in 25,000-unit doses was instituted routinely every 3 hours. Sulfadiazine was usually added after initial wound surgery. Treatment by this routine was continued until after definitive surgery had been performed in a general hospital.

Other techniques-Penetrating wounds through the patella which had produced extreme comminution were best treated by complete excision of the patella. After debridement of the joint, the synovial membrane and the capsule were closed in the midline.

When destruction of the femoral and tibial condyles, or both, was so complete that restoration of the joint was obviously impossible, the wisest plan was to proceed at once with primary resection. This course was justified,


however, only if the structures were so hopelessly destroyed that reparative surgery offered no possibility of securing a functioning joint.

If the injury was such that it was not possible to seal off the joint cavity without internal fixation of the fracture, a hip spica or Tobruk splint was applied and the patient was dispatched to the rear as soon as his condition permitted. This type of problem was not suitable for management in an evacuation hospital but was, instead, the responsibility of a general hospital.

Treatment in the General Hospital

As a rule, a patient in good general condition could be evacuated in plaster to a general hospital within 48 hours after debridement. Here, the next phase of treatment was instituted. The knee was examined for excess fluid, which was usually present. The fluid was aspirated and the joint irrigated with physiologic salt solution, introduced through one or two needles. When the irrigation had been concluded, penicillin solution (10,000 units in 5 cc. of physiologic salt solution) was introduced into the joint. This procedure was repeated at 24- to 48-hour intervals, depending upon the temperature reaction, the degree of edema in the tissues, the character of the fluid from the joint, and the patient's general condition. If for any reason the patient could not be evacuated, the procedure just described was carried out in the evacuation hospital.

Delayed primary closure of the skin was performed anywhere between the 4th and 10th days, depending upon the local condition of the wound and the presence or absence of joint complications.

If the patient could not be evacuated within 2 weeks, a remedial regimen was begun while he was still in traction, the exact time depending upon the circumstances of the individual case. In the occasional case in which the joint had not been opened, movement was begun within a few days after wounding. If arthrotomy had been performed but no fracture was present, exercise was begun as soon as the wound had healed, which was usually within 2 or 3 weeks. If a fracture was present, movement usually had to be deferred for a somewhat longer period, depending upon the type and location of the fracture and its current status.

The chief emphasis was on quadriceps-setting exercises. Failure to employ them for all patients with any kind of disability of the knee accounted for many inferior results. This was not, however, always the fault of the orthopedic staff. When admissions were at a peak, casualties with bone and joint injuries had to be dispersed to other surgical wards and to medical wards, and trained personnel was not usually sufficient to supervise them.

Suppurative Arthritis

If an infection of the soft tissues did not involve the joint capsule or communicate with the joint, the end results of joint surgery were unlikely to be


affected by it. Suppurative arthritis, however, was a constant hazard in military practice in all joint surgery and was the most dreaded complication of penetrating wounds of the knee joint. A normal joint was almost impossible of attainment after it had occurred. If the infection persisted for any length of time, extensive destruction of cartilage and bone was almost inevitable. In earlier wars, amputation was often required to save the patient's life. With the improved technical methods and efficient drugs available in World War II, deaths were rare and amputation was almost never necessary, but the functional results were none too good.

Suppurative arthritis usually appeared within the first week after operation or injury. The patient first complained of extreme discomfort in the joint, then of excruciating pain on the slightest movement. Swelling and tenderness were progressive. The rise in temperature was often abrupt. If the rise was gradual, the fever reached a peak of 103 to 104 F. within a few days and then rapidly assumed septic (picket fence) characteristics, with the trend generally upward.

All military experience has proved that there is no satisfactory method of draining the knee joint. Some surgeons attempted it by the use of two large parapatellar incisions, with a third incision behind the medial ligament and a fourth just posterior to the lateral ligament. These relieved the tension of the purulent exudate, but pools of it still remained in the deep recesses of the joint and a film was maintained between the surfaces in close contact, where it was held by capillary attraction between the condyles and between the menisci and the condyles. It is at these points of contact between bones that the earliest destruction of cartilage occurs in suppurative arthritis of the knee joint and that the greatest damage occurs as the lesions advance. The medial femoral condyle is usually the first affected.

What the French termed secondary resection, to distinguish it from primary resection of the knee joint practiced immediately after injury, was employed in an occasional case of suppurative arthritis of the knee joint. The French had employed this procedure even before World War I, but American surgeons were slow to take it up in World War II. To cut through infected soft tissue, bone, and cartilage, and to disturb areas of resistance seemed a violation of surgical principles, even though the objective of the operation was to shorten the septic course by removing feebly resistant cartilage, as well as infected bone and synovia. Once these structures had become infected, infection was likely to be persistent. Life and limb could therefore be saved by accepting the fact that suppurative arthritis results in a permanently damaged joint which is either ankylosed or is extremely painful on motion. Resection eliminated tissues of low vitality which harbored infection and at the same time secured really adequate drainage. French surgeons were more liberal in their indications than American surgeons, who limited the method to suppurative arthritis of the knee joint which was so extensive that there was no hope of regaining joint function. The optimum time for its performance was within the first 2 weeks.


Of the various incisions through which resection could be done, the most satisfactory was an elliptical transverse incision, through which the patella could be removed. Operation was preferably performed with a tourniquet in place. The anterior and posterior cruciate ligaments were divided and the knee joint was flexed and supported with a rolled towel in the bend. The femoral condyle was sawed off at the highest possible level, that is, just proximal to the intercondylar fossa, and the remaining cartilage was stripped free from the underlying bone. The tibial condyles were then sawed off, the excision being as conservative as was consistent with complete removal of all joint surfaces. Cuts were made so that flexion of 5 to 10 degrees resulted at the knee. The synovial membrane was completely excised.

If a tourniquet had been used, it was removed at this stage of the operation. Drains were placed at the angle of the incision. A single heavy wire suture was sometimes placed through the anterior margins of the bones, to prevent their slipping forward and backward. A few sutures were also inserted to approximate the skin loosely and prevent retraction. If the original wound was posterior, it was enlarged and used for dependent drainage. After the wound had been dressed, the limb was put up in a plaster-of-paris spica.

If posterior drainage had not been instituted, the patient was placed in the prone position at intervals. Drains were removed in from 5 to 10 days. The plaster spica was changed as often as necessary. If abscess formation developed, additional drainage was instituted.

Analysis of Cases

The management of 50 penetrating and perforating wounds of the knee joint in 49 patients treated at the l9th General Hospital during a 3-month period in the spring of 1945 is typical of the problems presented by this type of injury in the late months of World War II in the European theater.2 Sixteen of the patients were United States Army personnel, and 33 were prisoners of war. From that standpoint, as already noted (p. 63), conditions of treatment were less favorable than they probably would have been in a group of patients which did not include prisoners of war.

In 19 of the 50 cases, there were no associated fractures of the patella or the articulating ends of the tibia and femur. From that standpoint, conditions of treatment were somewhat more favorable than they frequently were in such injuries. In the remaining 31 cases, the fractures were distributed as follows: Femoral condyles, 12; patella, 6; tibial condyles, 4; patella and femoral condyles, 4; patella and tibial condyles, 3; patella, femoral and tibial condyles, 1; and femoral and tibial condyles, 1.

Initial surgery-In 40 of the 50 injuries, formal arthrotomy had been done at the time the wounds were debrided, most often through an incision separate from the original wound; in the remaining cases, the arthrotomy incision was incorporated in the battle-incurred wound. In 34 instances, the

2Semiannual Report, 19th General Hospital, European Theater of Operations, 1 January-30 June 1945.


joint was explored through either a medial or a lateral parapatellar approach; in 6 instances, both a medial and a lateral incision were necessary.

In 7 of the 10 cases in which arthrotomy was omitted at initial wound surgery, the patients were prisoners of war. In 2 other cases, penetration of the joint was not recognized in the forward installation, and the injury was thought to be only a soft-tissue wound. In the other case, the omission was deliberate.

While the records were not always clear on these points, it seemed likely that all but 2 or 3 of the 40 arthrotomized patients had been treated by the instillation of penicillin (10,000 to 50,000 units) into the joint at the time of debridement and that most of them had received both penicillin and sulfadiazine at this time. At the general hospital, prisoners of war received sulfadiazine systemically both before and after reparative surgery but were given penicillin only when necessary to control severe infections. United States Army patients received both penicillin and sulfadiazine systemically before and after operation.

Management of these 50 injuries in the 19th General Hospital is analyzed from several standpoints. There is necessarily some overlapping of figures.

Delayed primary closure of soft-tissue wounds-Delayed primary closure of the soft-tissue wounds was undertaken in 32 cases in which initial arthrotomy with closure of the capsule had been done. In 30 cases, there was no special problem. The edges of the wound could be brought together without tension and very little manipulation was necessary. Drainage was seldom necessary because dead space could be obliterated. The interval between closure of the wound and initial wound surgery ranged from 4 days in 5 cases to 16 days in 1 case. In all but 8 of the 32 cases, closure was effected between the 5th and 10th days. After operation, the knee was immobilized in a single plaster-of-paris spica. If the patient had not been evacuated at the end of 2 weeks, knee motion was started, usually while skin traction was still in effect.

In 20 of these 32 cases, primary healing followed reparative surgery. In 3 instances, the wounds became superficially infected. In the other cases, evacuation was necessary in from 4 to 6 days, and evaluation of the end results of wound closure was not possible within this period; there was, however, no evidence of infection in any instance when the patient was evacuated. In the cases in which the patients were under observation long enough to permit some evaluation of end results, it seemed likely that satisfactory function would be secured.

Delayed primary closure of capsule-In 10 cases, delayed primary closure of the capsule of the knee joint was carried out in the general hospital, in 2 instances with closure of the skin wound at the same time. In 4 other cases, the soft-tissue wound was closed 4 to 5 days after closure of the capsule. In all 6 cases, the immediate results were good, the wounds being completely healed within 2 to 3 weeks after the original injury had been sustained. In 2 other cases, the patients had to be evacuated before closure of the soft-tissue wound could be accomplished.


In the 2 remaining cases, the attempt at delayed closure of the capsule failed. Both patients were prisoners of war.

Case 29.-In this case the synovial membrane had been sutured but the capsule had been left open at initial wound surgery. The synovial closure did not heal well, and, when the patient was admitted to the general hospital 6 days after wounding, there were 3 openings into the joint. Penicillin had been instilled into the joint at debridement, but systemic penicillin therapy had apparently not been instituted.

Although the soft-tissue wound appeared clean and ready for closure when the patient was received, doubt was felt about the status of the joint. The openings in the synovial membrane did not permit adequate visualization, and wide arthrotomy was done 3 days later. The degree of intra-articular destruction revealed made it clear that initial wound surgery had not been adequate. Six days after this operation, infection in the joint was evident, and drainage of the wound had to be instituted.

There are three lessons in this case: (1) Complete removal of all intra-articular devitalized tissue and foreign material is essential in all wounds of the knee joint. This is not possible if arthrotomy is not carried out. (2) Closure of the joint cavity is best accomplished by closure of the capsule, and not only of the delicate synovial membrane, which frequently permits leaks at the suture line. (3) Penicillin, no matter how administered, is not a substitute for adequate surgery.

Case 42.-This patient had reparative surgery 7 days after wounding. The capsule was closed under moderate tension, and the very extensive soft-tissue wound was partly closed at the same time. The newly sutured portion of the capsule was left exposed, for subsequent coverage with a skin graft. The capsular wound broke down in part, and a second attempt at suture also failed. When the third attempt was made, the edges of the capsular defect were found to be so friable that a third failure was regarded as inevitable.

When the patient was evacuated, there seemed little doubt that the end result would be loss of function, partly because of the original bone and joint wound and partly because of failure to close the joint capsule at initial wound surgery. Regardless of the status of the articulation, the open capsule, with contamination and inevitable low-grade infection, was an additional liability. In this case, successful closure might possibly have been accomplished without tension by mobilization of muscle, fascia, and some capsular substance, the capsular mobilization being accomplished through partial tissue thickness in order to avoid a new opening into the joint.

In 3 cases, all in prisoners of war, closure of the joint cavity was accomplished by the use of skin; the considerable loss of capsular tissue would have made closure impossible otherwise. In the first case, a portion of the skin wound was sutured over the portion of the defect which could not be closed; the portion of the skin wound overlying the sutured capsular defect was left open. In the second case, flaps of skin were brought down over the capsular defect, the remainder of the skin wound being left open. In this case, there was an associated comminuted fracture of the lateral femoral condyles, but most of the articular surface was intact and extensive redebridement was not necessary. In the third case, the same technique was used. In all 3 cases, the joint capsule was successfully sealed off, and the portions of the skin wounds which had been left open were closed between the fifth and seventh days after closure of the capsule. In the first and second cases, healing was satisfactory. In the third, superficial necrosis of the skin flaps occurred but was not serious. It did, however, delay recovery. The other patient whose


capsular wound had been closed by means of skin flaps was discharged 45 days after operation, with function from 180-degree extension to 100-degree flexion.

As all these cases show, the first essential of management of a wound of the knee joint is adequate debridement of the joint as well as of the soft-tissue wound. The second essential is the complete, tight closure of the joint capsule. If capsular defects do not permit this kind of closure, closure must be effected by mobilization of whatever deep structures may be available. If this is not possible, then closure of the defect with skin must be attempted, counter-relaxing skin incisions being used as necessary. If large capsular defects are left open at the initial operation, they cannot be closed as successfully at a later date.

Omission of initial arthrotomy-In 6 of the 10 cases in which formal arthrotomy was omitted at debridement, the wounds were penetrating and in 4 they were perforating. Associated fractures were present in 4 cases (2 comminuted fractures of the patella, 1 fracture of the medial femoral condyle, 1 fracture of the patella and tibial plateau).

In the 6 cases in the nonarthrotomized group in which fractures were not associated with the joint injuries, conditions were favorable for uncomplicated recovery. In all these cases, the joint injury was caused by small shell fragments of high velocity, and intra-articular damage was apparently slight. Aspiration of blood from the joint, with instillation of 10,000 to 20,000 units of penicillin into the joint and a short course of systemic penicillin, was the treatment of choice. Early motion (within the first week after wounding) in traction resulted in rapid return of function.

In 3 of the 4 instances in which fractures were associated and in which free fragments of bone were found in the joint, infection did not occur, although in all of them soft-tissue damage was quite extensive. In the fourth case, an instance of a comminuted compound fracture of the patella and tibial plateau in a prisoner of war, the joint was not opened widely when the original debridement was done, and foreign bodies were permitted to remain in situ. The patient had received systemic penicillin, with local instillations into the joint, and sulfadiazine by mouth. The knee joint became infected, and a spreading cellulitis of the thigh developed. Treatment consisted of penicillin and sulfadiazine therapy, local heat, and traction. When arthrotomy was subsequently carried out, metallic foreign bodies and loose bone fragments were removed from the joint. The synovial membrane was closed, and early joint motion was started in traction. Two months later, the wounds were healed, and the range of motion was from 180-degree extension to 24-degree flexion.

The results in these 10 cases were, on the whole, better than might have been expected in the light of the omission of arthrotomy at initial wound surgery. They must not, however, be assumed to be the rule when this practice is followed. Battle wounds of the knee joint are best treated by (1) formal arthrotomy, which permits adequate debridement through an incision which provides visualization of the entire interior of the joint, and (2) tight closure of the capsule. Failure to close the capsule is always an invitation to infection.


Wound infections-Eight of the fifty wounds of the knee joint in this series were infected when the patients were received in the general hospital. In 1 instance, the infection was limited to a small, localized collection of purulent exudate in a medial wound. The infection did not communicate with the joint, and the lateral wound in the same knee was clean. Delayed primary closure of both wounds was promptly accomplished, and, while the end result is not known, it was thought that the infection would not influence the outcome of the injury in any way.

In the other 7 cases, in all of which compound comminuted fractures were also present, the infection was serious and involved the joint. In 5 cases, debridement and arthrotomy had been performed from 24 to 72 hours after wounding. In the 2 other cases, debridement of the soft-tissue wound had been performed 48 hours after wounding, but in each instance arthrotomy had been omitted. The amount of necrotic debris found in these wounds and joints when reparative surgery was undertaken in the general hospital was a clear indication of the inadequacy of the initial wound surgery. The basic cause of each of these 7 infections was a severe joint injury, considerable soft-tissue and intra-articular damage, late and inadequate debridement or both, and, in 2 cases, omission of arthrotomy.

In 6 of the 7 infected cases, the records indicated that chemotherapy and antibiotic therapy had been administered both locally and systemically, according to a satisfactory regimen. In the seventh case, the patient had been operated on in a German hospital and had certainly received no penicillin; the record did not indicate whether he had had sulfadiazine. In the eighth case, the record contained no information at all about chemotherapy.

There were 4 postoperative infections. One, in a United States soldier, was a trivial-stitch abscess, readily controlled by removal of the sutures and the application of heat. Two of the 3 infections in prisoners of war were also superficial and readily controlled. In the remaining case (case 29, already described), a severe infection of the joint and a spreading periarticular infection required radical surgery. The infection at one time threatened the vitality of the limb, but it was eventually controlled by combined surgery and penicillin therapy.

Immediate functional results-The final outcome of these 50 wounds of the knee joint is not known. Under the theater holding policy, particularly in the last days of the war, patients had to be evacuated too promptly to permit the necessary observation. In 10 cases, however, early functional results could be evaluated. In 1, in which there was a compound comminuted fracture of the femoral condyle, movement 42 days after wounding was from 180-degree extension to 100-degree flexion. The immediate results in the other 9 cases, in none of which fractures were present, are summarized in table 19.


TABLE 19.-Immediate functional results in 9 penetrating wounds of the knee joint

Case No.

Days after wounding

Range of motion in degrees (expressed in extension to flexion)









All movements free; to duty1






180-55 To duty.













1Information secured by followup letter.