Other formats

    TEI XML file   ePub eBook file  


    mail icontwitter iconBlogspot iconrss icon

War Surgery and Medicine



The following types of cases were treated:


Fractures of the facial skeleton, often associated with soft-tissue defects.


Large traumatic soft-tissue defects, often associated with compound fractures of the long bones.

Each of these will be discussed in turn.

Mandibular Fractures.—The principles used:

Civilian Type: In mandibular fractures of the civilian type with a good number of standing teeth, the principle used was to extract any teeth in the line of fracture, and then to use simple eyelet wiring with inter-maxillary fixation. This is a simple method of treating this type of fracture, produces excellent occlusion, and does not impose any work on the prosthetic laboratory.


Battle Casualties associated with Soft-tissue Damage: In this type of case the simple method of eyelet wiring was found to be quite unsuitable, so cast-metal cap splints, made in sections and then joined by connecting bars after reduction of the fracture, were used throughout. Splints were frequently put on both upper and lower jaws, and inter-maxillary fixation by means of a precision lock gave added immobilisation. Where the lower buccal sulcus was obliterated, or was thought likely to become obliterated, the splint was planned in such a way as to provide for an appliance to carry an epithelial inlay mould at a later date.


Battle Casualty Fractures associated with Bone Loss: In these cases cast-metal cap splints as described above were used to preserve the occlusion of the remaining teeth, page 362 and maintain the gap between the bone ends. The cases were then evacuated to New Zealand for bone grafting as a long-range measure.


Condylar Fractures: Generally speaking, these fractures were treated by immobilising them in correct occlusion for one month.

Edentulous Cases:

Fractures Near the Angle with a Short Edentulous Posterior Fragment: By the time we started treating these cases the method of controlling these hitherto extremely difficult cases by means of extra-oral pins had been applied in England. The principle in its turn had been borrowed from that introduced by the American, Roger Anderson, into orthopaedic surgery. No previous method had given adequate control of the short edentulous posterior fragment which was adducted and rotated by the pull of the internal pterygoid muscle. Where there were standing teeth on the anterior fragment the principle that we used was to apply a cast-metal cap splint to these teeth fitted with a removable extra-oral bar. If there was a tooth in the line of fracture, this was extracted, and then two Kirschner wires of 5/64 inch diameter were drilled into the posterior fragment at an angle to each other. These two were then clamped to each other by means of universal joints and a T-piece. The fracture was then reduced, and by means of two more universal joints and a connecting bar the T-piece and the extra-oral bar were joined firmly together. If there were standing teeth in the maxilla, inter-maxillary fixation was added by means of another metal cap splint, and a precision lock. If not, an acrylic bite block was used for the upper jaw, and the lower jaw was kept in correct relation to it by means of a plaster head cap and universal joints connected to the extra-oral bar on the lower splint. Immobilisation was continued from three weeks to a month.


Completely Edentulous Mandibular Fractures: These were treated by Roger Anderson pinning of fragments and joining them by means of universal joints and connecting bars, using Gunning-type splints for intermaxillary fixation, with a plaster head cap where necessary.

[See Illustration Section (Figures 1 and 5)]

page 363
Fractures of the Middle Third of the Face.—It was in these cases that a paper published by W. James and B. W. Fickling in February 1941 was of the greatest possible help. Their work was really a translation of a paper by Renée le Fort, written as long ago as 1901. He carried out certain experiments to find out the lines of weakness of the facial skeleton, and found that there were three main lines of weakness in the middle third of the face:

Transversely above the tooth sockets and through the floor of the nose.


Transversely at a somewhat higher level below the malar bones, through the infra-orbital foramina, and across the mid-line through the nasal processes of the maxillae.


Transversely at a still higher level through the fronto-malar sutures laterally and the root of the nose in the mid-line. This line separates the whole of the facial skeleton from the base of the skull.

These three fracture lines, which have since been named, respectively, low transverse, pyramidal, and horizontal, constitute the three main lines of weakness, and all fractures in this situation are found to follow these lines in part, wholly, or in various combinations, unilaterally, or bilaterally. For example, the ordinary depressed fracture of the malar bone is a combination of the pyramidal and the horizontal of unilateral type.

It is convenient to discuss fractures of the middle third of the face under the following headings:


Low transverse fracture.


Pyramidal fracture.


Horizontal fracture.


Depressed fracture of the malar bone.

It is clear that the first three fractures, if severe enough in degree, can upset the dental occlusion, damage the nasal airways, or cause diploplia, or various other severe long-range effects. The essence of treatment was found to be the securing of immediate reduction of the fragments and their immobilisation.

Transverse Fracture: This fracture usually caused an upset of the dental occlusion by presenting a condition of open bite. It was treated by applying a cast-metal cap splint, fitted with an extra-oral bar, to the mandibular teeth. A plaster head-cap was applied, and by means of universal joints the mandible was immobilised in its normal plane of occlusion. A similar cast-metal cap splint was attached to the maxillary teeth and, using hooks fitted at strategic points to both splints, strong rubber bands (made by cutting sections from a rubber catheter) were fitted to secure inter-maxillary fixation. It was found that even in neglected cases of impacted fracture with page 364 open bite, twenty-four hours was sufficient to effect reduction. This immobilisation was continued until there was no tendency for the deformity to recur.

Pyramidal Fracture: This was treated in the same way as the low transverse, but there was frequently a nasal fracture associated, and this required reduction as well. If this was done early it was found that no other fixation was usually necessary. Malar fracture with displacement was sometimes associated, and this was reduced by means of the Gillies temporal approach.

Horizontal Fracture: This was sometimes associated with fracture of the base of the skull, and cerebro-spinal fluid rhinorrhoea. In these cases the whole facial skeleton was floating at the level of the base of the skull, and was immobilised in exactly the same way as above. Later, however, these cases were treated by exposing the fronto-malar sutures, and wiring them with stainless steel wire, in addition to inter-maxillary fixation if dental occlusion was grossly upset.

Depressed Fracture of the Malar Bone: This was a very common fracture among the troops, and was treated almost entirely by the Gillies temporal approach. It was seldom found necessary to pack the antrum if the cases were got early.


In 1943, following the severe fighting of the Alamein period, a number of cases with severe mutilating wounds were being treated in our base hospitals in Egypt. Amongst these were cases of compound fracture of the long bones in which there was also a great loss of skin and soft tissue.

It was by no means rare to see exposed bone ends in the depths of a skin defect 18 inches by 10 inches. This might be due to the missile causing the injury actually carrying away the skin completely, or by so avulsing it that large portions sloughed owing to the inadequacy of the remaining blood supply. Obviously the treatment of the fracture took precedence over everything else, but some attempt had to be made to secure the earliest possible healing of the associated wound. If left to heal spontaneously, some of these wounds would have certainly taken months, and possibly years, to heal. During this time activity of the limb would necessarily have been limited so that the best possible functional result would not be obtained. Furthermore, scars resulting from spontaneously healed wounds of such size are often unstable, cause contractures, and limit page 365 function considerably. While lacerations in which there was no tissue loss healed rapidly under closed plaster, wounds in which there was actual tissue loss healed much more slowly.

At this time the New Zealand Plastic Surgical Unit had been formed and an elaborate saline bath unit had been set up at Helwan hospital, primarily for the treatment of burns. The plastic team at the hospital was then given the opportunity of treating these difficult and unsatisfactory cases, and made the attempt in spite of gross suppuration at the site of fracture to try to secure complete healing of these wounds in one operation.

An account of the valuable work carried out is given by Major Manchester, who was in charge of the unit.

Preparation of the Wound for Grafting

Enough was known at that time about wound healing and bacteriology in general to enable us to aim at skin grafting to cover tissue loss and to secure a complete take of the grafted skin.

The following remarks apply to the grafting of granulating surfaces in general, and we will later see how the fracture cases differ.

The preliminary wound treatment was based on the assumption that a wound would heal extremely rapidly from the edges provided there was nothing to inhibit it. It was not a question of ‘stimulating’ healing, but of removing the inhibitory factors. These inhibitory factors were exactly those which prevented grafts from taking. There were three main ones:


Bacterial infection.





Each of these will be discussed in turn.

Bacterial Infection

In studying the bacteriology of a wound as a preliminary to grafting, a most thorough investigation had to be made and repeated at frequent intervals.

It was necessary to take direct smears at the bedside from all parts of the wound, and also from the exudate on the dressing as soon as it was removed, in addition to the usual culture. It was not uncommon to find organisms in direct smears which did not grow on culture owing to the inhibitory effects of other organisms.

It was obvious, therefore, that examination of the swab taken from the wound and perhaps left for an hour or more to dry was page 366 hopelessly inadequate. The common organisms found infecting granulating wounds were as follows:


Haemolytic Streptococcus: The presence of this organism in even small numbers in a granulating wound caused marked inhibition of healing and prevented the complete taking of grafts. It was especially difficult to deal with, because when present in small numbers it was easily overlooked in the direct smear as it often occurred in diplococcal form, when the amount of exudate was not great. A graft might easily be performed under the mistaken impression that the organism was absent, but it could be recovered in chain form in almost pure culture from the exudate after perhaps two days, when it is obvious that the graft had failed. Moreover, on culture, the organism when present in small numbers was rapidly overgrown by other bacteria, especially Pseudomonas Pyocyanea and Bacillus Proteus. This difficulty could be overcome, however, by using a medium which was favourable to the growth of the pyogenic cocci but unfavourable to that of the gram-negative bacilli. Such media could be made by incorporating gentian violet or 8 per cent agar instead of the usual 3 per cent agar. When haemolytic streptococcus was present in a wound it was extremely difficult to get rid of, in spite of the common belief that all that was necessary was to apply sulphanilamide powder. Some strains were sensitive to the sulphonamides and these disappeared after a few days' application of the powder locally; but many were sulphonamide-resistant, and these presented a real problem in preparing a granulating surface for grafting. It was almost impossible to secure a complete ‘take’ when this organism was present.


Staphylococcus Aureus was not nearly so difficult to deal with as the haemolytic streptococcus. It was possible for a graft to take completely in its presence, provided the numbers were not too great.


Pseudomonas Pyocyanea produced a large amount of green exudate and overgrew cultures, masking the presence of other organisms. It was commonly stated that a wound infected with Pseudomonas Pyocyanea would not accept a graft. It was, however, possible for the largest graft to take in a wound infected with it.

[See Illustration Section (Figure 8)]

In our opinion its bad reputation in this respect was due to the presence of haemolytic streptococci whose growth it masked in culture, with the result that grafting was done under the mistaken impression that the wound was free from the streptococcus. Hence the necessity for direct smears and differential culture media. Its growth in wounds could be checked by using 1 per cent acetic page 367 acid dressings, but it was likely that this merely stopped growth in the exudate and did not affect the organism in the granulation tissue itself.


Bacillus Proteus resembled Pseudomonas Pyocyanea in that it did not appreciably affect wound healing nor the acceptance of grafts, but in culture it rapidly overgrew the pyogenic cocci unless measures were taken to prevent it.


Non-Haemolytic Streptococcus and Staphylococcus Albus: Their presence in a wound was of no practical significance so far as rapid healing and graft taking were concerned.


Diphtheroids: These organisms were commonly found infecting granulating surfaces, but were of no practical significance. Morphologically they resembled the Corynebacterium Diphtheriae, but could easily be distinguished by laboratory methods. The clinical behaviour of the wound, however, left little doubt which organism was present.


Corynebacterium Diphtheriae: No account of wound bacteriology would be complete without mentioning this organism. It was undoubtedly the most virulent of all wound infections and the most difficult to treat. Grafting in its presence was out of the question.

The only common organism, therefore, which caused much anxiety when preparing a case for grafting was the haemolytic streptococcus, but it was not so easy to be sure of its presence or absence as was generally believed.


The granulation tissue of a wound, when given a fair chance, has by virtue of its own natural defence a much more sustained bacteriostatic and bacteriocidal power than any antiseptic which can be applied to its surface. A wound is thus best able to combat infection when it is subjected to no trauma, and when the tissue cells are bathed in a physiological and isotonic medium. Trauma is applied to a wound either mechanically or chemically.

Mechanical Trauma: A dressing which stuck to the wound and had to be forcibly removed caused bleeding and opened up fresh tissue to bacterial infection. It also damaged the proliferating epithelial edge, and the repair work of hours or days might be undone.

Chemical Trauma: Strong antiseptics and coagulants were often applied to wounds. It was extremely doubtful whether these usefully inhibited bacterial growth in vivo, and it was almost certain that they damaged the living tissues and reduced their own page 368 natural resistance. Some antiseptics such as the acridine compounds did not, however, inflict much trauma, and if antiseptics had to be used these were the ones of choice.


Unless a wound has a good blood supply it cannot possibly heal rapidly for two reasons: Firstly, it is unable to resist and overcome bacterial infection; and, secondly, it is unable to support epithelial proliferation. An ischaemic wound presents a typical appearance. [See Illustration Section (Figure 6).] It has a densely scarred base and periphery through which the vessels are unable to maintain a sufficient blood supply. It has a rolled edge and a greyish base. Good examples are varicose ulcers which are due to inadequate circulation in the leg. They frequently heal rapidly when their blood supply is improved by recumbency and with treatment of the underlying cause, the varicose veins.

If bacterial infection and traumatic dressings are allowed to persist over a long period, wounds commonly develop a dense plaque of scar tissue in their base and periphery, and then grafting is out of the question unless the wound and scar tissue are ruthlessly excised until healthy tissue is exposed. Ischaemia should be prevented by securing rapid healing.

The wound treatment employed in preparation for grafting was based on these principles and aimed at mitigating the three inhibitory factors by reducing the number of infecting bacteria, by the use of an atraumatic dressing technique, and by preventing the development of ischaemia.

Dressings: Until the introduction of penicillin, dressings had to be frequent. Their aim was to absorb the exudate which formed a culture medium for the organisms which were discharged from the wound. By changing them frequently the smallest amount of exudate remained in contact, and thus by mechanical means it was possible to reduce the number of bacteria present.

By the use of the tulle gras and normal saline technique, both mechanical and chemical, trauma were avoided. First a layer of tulle gras was applied direct to the granulation tissue. Well-made tulle gras could be removed without trauma after many hours in contact with the wound. Its advantage over vaseline gauze was that it allowed the exudate to escape through its meshes. Outside this, a dressing made of several thicknesses of gauze wrung out in normal saline was applied and the dressing completed by a layer of cotton wool and a firm bandage.

During the early stages the dressings were changed four times daily. The day before the operation the frequency was increased page 369 to eight times daily. These frequent dressings had to be carried out in a place properly equipped to avoid nose and throat and dust-borne infection, and by a properly trained staff alive to these dangers. It was possible by this means to reduce the number of bacteria in a wound to such an extent that it was difficult to inoculate a culture from it.

This general routine for preparation of the granulating wound could be supplemented by the use locally of a bacteriostatic. The sulphonamide drugs were disappointing, because by no means all strains of haemolytic streptococci were sensitive to them. If after one week's treatment the haemolytic streptococci were still present, the drug was discontinued, as the organisms almost certainly belonged to a resistant strain.

When to Operate

Deciding when to operate with a good chance of success was perhaps the most difficult part of the whole procedure. Many factors had to be taken into consideration.


Pyrexia: When even a very large wound is sufficiently free from infection to accept a graft there should be no pyrexia. At the very most, the temperature should not rise above 99 degrees F.


Exudate: There should be very little exudate on a dressing after it has been left in situ for twelve hours.


Proliferation: The epithelial edge should be purple and rapidly advancing.


Colour: Healthy vascular granulations should be bright red and bleed at the slightest trauma.


Surface: The surface should be quite smooth, flat, and velvety. It should show the impressions of the tulle gras upon it for a while after this has been removed. If it is irregular and knobbly the granulations are oedematous, and this means that they are infected.


Pain: The granulations should be painless to touch. Painful granulations are septic granulations.


Bacteriology: The wound should be free from haemolytic streptococci. It is possible to secure a good take if they are present in infinitesimal numbers, and it is justifiable to make an attempt only after a strenuous effort to remove them has failed.

page 370

If the direct smears and cultures show only small numbers of other organisms and all provisos are satisfied, a graft may be performed and 100 per cent take can confidently be expected.

Technique in Fracture Cases

We will now see how the fracture cases differ, and how the technique had to be modified accordingly. They differ in that in the centre of the soft tissue is a mass of dead and infected bone which pours pus rich in bacteria over the rest of the granulating surface. Obviously, then, the infected cavity must be kept separate as far as possible from the rest of the wound, both during the preliminary dressing period and subsequent management. During the preliminary period the infected cavities were lightly packed with ribbon gauze in sufficient quantity to absorb all the exudate thrown off during the period between dressings. This part of the dressing was done first, and only when it had been attended to was the rest of the dressing performed. It was essential that the pus from the bony cavity should not be allowed to reinfect the wound.

Cases of compound fracture of the femoral shaft in a Thomas splint were first chosen for grafting because of the ease of carrying out the wound toilet without interfering with the immobilisation of the fracture. In these cases union was already firm but not yet solid at the time of operation.

It seems hardly necessary to add that the cases were chosen and treated in full collaboration with the orthopaedic surgeon, he and the plastic surgeon working together as a team.

Operation: The aim was to secure complete healing in one operation of all but the infected bone area, which must have free drainage and be accessible for frequent dressing at all times.

The first stage was to produce the skin. This was done first as a clean operation after estimating the amount required. It was cut in pieces as large as possible, and not too thick as the thinner grafts were more viable.

A pattern was then made by laying two or three layers of gauze over the wound and marking the periphery in Bonney's Blue. The infected bone cavity was marked out in the same way. The gauze was then laid on a table, wound side uppermost, and covered over by a single piece of tulle gras, the pattern readily showing through it. The skin was then carefully arranged to cover the whole marked area except for the part corresponding to the infected cavity. Here even the tulle gras was cut away, leaving a window. Great care was taken to arrange the junction between any two adjacent pieces of skin accurately, and with a clear-cut overlapping edge, otherwise local failure would result.

page 371

The tulle gras, together with the graft, was now accurately applied to the wound so that the window corresponded to the bony cavity. A few anchoring stitches were placed into healthy skin around the periphery and a pressure dressing built up of paraffin wool and cotton wool, but again so as to leave a window opposite the bone cavity. A crepe bandage was then applied, avoiding the window. The whole thickness of the window was waterproofed by saturating it with collodion or mastisol and the cavity lightly packed with ribbon gauze.

So far as the post-operative treatment is concerned, the only part that required any attention was the uncovered bony cavity, to which access was got through the window in the dressing. The ribbon gauze with which the cavity was lightly packed was removed every four hours and renewed, in order to keep the exudate from overflowing and making the dressing messy.

At the end of a week the dressings were removed and any redundant pieces of skin clipped away, and any junctional areas which still required to heal were dressed four-hourly as before the operation. In this way the whole wound, except for the actual cavity, was quickly and stably healed.

If the technique had been faithfully and accurately carried out, the skin formed an adequate covering for all purposes, with a few possible exceptions. If, however, it was found at some later stage to be unsatisfactory, it could be excised and repaired by some other method, such as by the use of tubed pedicles. This could be done at leisure and as a clean operation throughout.

The healing of these large wounds was usually followed by a rapid improvement in the patient's general condition. His weight usually rose rapidly and his morale improved.

Later in the war when penicillin was introduced much of the work above described in preparing granulating surfaces was rendered superfluous. It was found that by using systemic penicillin the frequency of dressings could be reduced to one every few days, and the period of preparation correspondingly reduced; but, apart from that, the technique remained essentially as stated above.

[See–Illustration Section (Figures 89)]

[For long-range plastic surgery see Illustration Section (Figures 1025) following p. 374]

Mandibular Bone Grafting

Certain conditions are necessary for success in bone grafting the mandible. They are:


The immobilisation of the fragments to be grafted.


The control of infection.


The use of suitable bone.

page 372

In the early part of the war much had yet to be learnt about all these three points but as the war drew to a close great improvements had been made in all three. We will now discuss each of the points in turn:

1. The Immobilisation of the Fragments

In cases with standing teeth. In these cases the method of immobilisation was as follows. A cast metal cap splint was attached to the standing teeth on each fragment. Provision was made for the subsequent union of these two pieces by means of a fish plate and screws. A similar but single splint was then applied to the maxillary teeth. The two mandibular fragments were then brought into their correct occlusional relationship with the splint on the upper. A fish plate was then designed to join the two halves of the lower splint and when it was screwed into position the lower splint was thereby made single. The lower was then united to the upper splint by means of a removable precision lock. In this way it was possible to discontinue the inter-maxillary fixation at any time but to leave the mandibular fixation intact. When it seemed obvious that a bone graft would become necessary in any particular case, the splints were always designed in the first instance with this object in view. In this way the correct relationships were maintained between the fragments and then it simply became a question of bridging the gap.


In edentulous cases. In these cases the method used was that of extra-oral pin fixation. The two fragments involved were securely pinned and after the correct occlusional relationship was established the two fragments were joined by means of universal joints and bars.


Cases where only one fragment was edentulous. In the case of the fragment which had standing teeth the immobilisation was secured by using cast metal cap splints in the manner described above. Provision was made, however, for a removable bar to be attached to the splint which was capable of extrusion through the mouth. The edentulous fragment was then pinned and after suitable reduction the two fragments were joined by means of universal joints and connecting bars. Wherever it was possible to use inter-maxillary fixation as well, we used it, but frequently in edentulous cases no inter-maxillary fixation was used. Towards the end of the war we seldom had cause for anxiety regarding fixation. We found that when we had cases of successful initial grafting followed by absorption of the graft it was nearly always due to dispensing with the fixation too soon.

2. The Control of Infection

In the early part of the war we had difficulty in controlling infection and, when a bone graft seemed likely to be necessary, we page 373 got rid of all the teeth in the neighbourhood of the projected bone graft from the beginning. In this way all foci of infection were removed. We also made it a rule that no case was bone grafted until six months after complete healing following the last trace of sepsis. Great care was also taken with regard to sepsis when working near the mouth.

With the advent of penicillin, however, we found it possible to modify this programme and we were able to do our bone grafts under penicillin cover as soon as the patient was in fit condition. Indeed, in some cases it was done even before the wounds were completely healed. Penicillin, therefore, made bone grafting possible much earlier and made it much more certain.

3. The Use of Suitable Bone

At the beginning of the Second World War the standard source of bone in many English clinics was the crest of the ilium. In the early days an attempt was made to get the bone ends cut to some sort of accurate shape such as may be used in cabinet-making and to bridge the gap with a solid piece of bone shaped to fit it, and as near as possible conform to the normal contour of the mandible. Later, however, owing largely to the work of Mr Rainsford Mowlem at Hillend Hospital, St. Albans, a change was made to the use of cancellous chips, also taken from the ilium. It was found that these had a greater resistance to infection, besides eliminating elaborate fitting in the operative technique. While we used this method, nevertheless we also used a large block of ilium as well and used the chips to fill in any defects and to make the contour accurate.

The advances that took place during the Second World War then were: (i) Much-improved methods of fixation, particularly in edentulous cases. (ii) Better control of infection by using antibiotics. (iii) The use of cancellous bone, which is more viable under conditions of infection.