RECURRENT

DISLOCATION

OF

SHOULDER

23

THE PATHOLOGY AND TREATMENT OF

RECURRENT DISLOCATION OF THE
SHOULDER-JOINT
BY A. S. BLUNDELL BANKART, LONDON

RECURRENT
dislocation of the shoulder-joint is an uncommon condition, the real
nature of which appears even now to be little understood. It may almost be said
to be peculiar to athletes and epileptics-a rather curious association which, as I
shall show, is not without aetiological significance. Most of the former class are
powerful, healthy, athletic young men to whom the frequently recurring dislocation
from trivial causes is a great and serious disability. In epileptics the dislocation
sometimes recurs with every fit. The dislocation is nearly always anterior. I
have only seen one case of posterior recurrent dislocation of the shoulder.

ALLEGED CAUSES
The condition has been attributed to abnormal laxity of the capsule and to
weakness of the surrounding muscles. The abnormal laxity of the capsule is
supposed to be due to stretching or imperfect healing of that structure after the
reduction of an ordinary traumatic dislocation, and it has been thought that
too early and too vigorous use of the arm may be the cause of the defect.
Alternatively, or in addition to these factors, recurrent dislocation has been
attributed to fracture of the glenoid cavity, fracture of the humeral tuberosities,
deformity of the head of the humerus, and to contracture of some of the muscles
around the joint.
It may be said at once that none of these alleged causes has been present in
any of the 27 consecutive cases which form the basis of this communication.
I have never seen recurrent dislocation associated with fracture of the glenoid
cavity or with any other bony abnormality, and I would suggest that, if such an
association occurs at all, it must be very rare indeed.

VARIOUS OPERATIONS
Five types of operations have been performed for the relief or cure of this
disability :I. Operations designed to diminish the size of the capsule. These comprise
various pleating or overlapping procedures, with or without incision or excision
of a portion of the capsule.
2 . Operations designed to give support to the capsule, particularly at its lowest
part, where the dislocation is believed to take place during abduction of the arm.
The most favoured of these operations is that of Clairmont and Erhlich, in which
a strip of the deltoid muscle is transplanted in the form of a sling beneath the joint,
and is supposed to contract and hold up the head of the humerus when the arm is
abducted.

24

THE

BRITISH

JOURNAL

OF

SURGERY

3. Operations designed to hold the head of the humerus in place by means

of an artificial ligament constructed either from the tendon of the long head of the
biceps (Nicola) or from a free transplant of tendon or fascia (Henderson).
4. Operations which aim at the construction of a bony block in front of the
head of the humerus.
5. The operation described below for the repair of the essential anatomical
defect.
All these operations, with the exception of the last, are based upon erroneous
ideas of the pathology of recurrent dislocation of the shoulder, or they ignore the
pathology and attempt to deal empirically with the resulting clinical condition.
In the first place, the capsule in these cases is not unduly lax. The capsule of the
shoulder-joint is normally a lax structure, but more than one surgeon has commented
on the difficulty of raising a satisfactory fold when doing a plication operation for
recurrent dislocation. Secondly, the muscles are not weak. There may be some
wasting of the muscles after this, as after any other, injury to the shoulder-joint,
but the powerful musculature that is usually exhibited by these patients is a sufficient
commentary on this point. Thirdly, the joint does not need support below, for
it is not here that the dislocation takes place. And lastly, the unnatural tethering
of the head of the humerus by an artificial ligament, and its obstruction by means
of an abnormal excrescence of bone, are equally crude and irrational methods of
dealing with a straightforward anatomical defect.
It must be admitted that these operations have sometimes been successful in
preventing recurrence of the dislocation. Indeed, it would seem that almost
anything done to the shoulder may have this effect. It has even been claimed that
simple manipulation will cure recurrent dislocation (Barker), though how this
procedure can bring about the closure of a gross and demonstrable anatomical
defect it is impossible to imagine. It is probable that most of these operations
produce this effect by limiting the normal movements of the shoulder-joint, while
the patient contributes to the result by avoiding any movement which is liable to
produce the dislocation.
THE MECHANISM AND CAUSE OF RECURRENT
DISLOCATION
It has been too readily assumed that recurrent dislocation is an unfortunate
sequel of ordinary traumatic dislocation of the shoulder-joint ; and I have known
medical men reproach themselves quite unnecessarily for their treatment of the
original injury. Recurrent dislocation has nothing whatever to do with ordinary
traumatic dislocation. It is from the first an entirely different injury, and it is
produced in an entirely different manner.
Ordinary dislocation of the shoulder is the commonest of all dislocations of
the joints. It is caused by a fall on the abducted arm. In extreme abduction the
neck of the humerus impacts against the acromion process, and then by leverage
the head is forced through the lowest and weakest part of the capsule between the
subscapularis and triceps muscles. When such a dislocation is reduced, the rent
in the capsule heals rapidly and soundly, and the dislocation never recurs.
But the dislocation which afterwards becomes recurrent is caused, not by
a fall on the abducted arm, but by a fall either directly on the back of the shoulder

RECURRENT

DISLOCATION

OF

SHOULDER

25

or on the elbow which is directed backwards and slightly, if at all, outwards. The
head of the humerus is forced out of the joint, not by leverage, but by a direct drive
from behind forwards. In its passage forwards the head shears off the fibrous or
fibrocartilaginous glenoid ligament from its attachment to the bone. The detachment occurs over practically the whole of the anterior half of the glenoid margin.
The reason why the dislocation recurs after reduction is that, whereas a rent in the
fibrous capsule heals readily and soundly, there is no tendency whatever for the
detached glenoid ligament to re-attach itself to the bone. The defect in the joint
is therefore permanent, and the head of the humerus is free to move forwards over
the anterior margin of the glenoid cavity on the slightest provocation.
In quite a number of cases the original injury is stated to have occurred at
football, and it may be noted that in this game, at least, a player is seldom sent
sprawling with his arms out, but, when he falls or is thrown to the ground, it is
usually in some moment of strenuous action, when his muscles are tense and his
arms are more or less close to his sides. Usually he is too preoccupied to make
much attempt to save himself, and in falling backwards he is likely to strike either
the back of the shoulder or the point of the elbow which is directed backwards.
The epileptic, too, falls with his muscles tensely contracted, and he, of course,
makes no effort to save himself. Frequently his arms are drawn backwards and
little, if at all, abducted, so that if he falls backwards he is likely to strike either
the back of the shoulder or the point of the elbow. Thus, both athletes and
epileptics are liable to sustain dislocation of the shoulder-joint by direct violence
rather than by indirect force or leverage.
I have now exposed the typical lesion of recurrent dislocation-namely,
detachment of the glenoid ligament from the anterior margin of the glenoid
cavity-at operation in 27 consecutive cases. It is a constant, straightforward,
uncomplicated anatomical condition, which can be observed at any time by any
competent surgeon who cares to expose the anterior margin of the glenoid cavity
in any typical case of recurrent dislocation of the shoulder-joint. It is easy then
to demonstrate the mechanism of the recurrent dislocation, for on taking hold of
the arm and pushing its upper end forwards, the head of the humerus can be made
to pass freely over the anterior margin of the glenoid cavity. No one who has ever
seen this typical lesion exposed at operation could possibly doubt that the only
rational treatment is to reattach the glenoid ligament (or the capsule) to the bone
from which it has been torn.
It may be noted that the glenoid margin cannot be exposed by the anterior
incision ordinarily employed for arthrotomy and plastic operations on the shoulderjoint, and no doubt this is the reason why the anatomy of recurrent dislocation
has not been generally recognized. The operation here described has been planned
to expose completely the anterior aspect of the shoulder-joint, and to deal directly
with this typical injury.

AUTHORS OPERATION TO REPAIR ESSENTIAL

ANATOMICAL DEFECT
Position of Patient.-The patient lies on the operating table with a small
pillow or sandbag between the shoulder-blades, and with the arm by his side.
A sandbag should be placed beneath the elbow on the affected side so as to
support the shaft of the humerus in a horizontal plane parallel to the table.

26

THE

BRITISH

JOURNAL

OF

SURGERY

Technique.I. Make an incision about 5 in. long from the lower border of the clavicle
immediately above the coracoid process in a direction downwards and outwards
along the anterior border of the deltoid muscle
(Fig. 27).
2. Find the cephalic vein, which is the guide
-_
to the interval between the deltoid and the pectoralis
CORAC*/D
major, and separate these two muscles throughout
. PPOC1555
the whole length of the wound. The cephalic vein
should be drawn inwards with the pectoralis major.
A branch of the acromio-thoracic artery will be
divided in the upper angle of the wound.
3. Hold the wound open with retractors and
define the coracoid process and the three muscles
attached to it, viz., the short head of the biceps,
coracobrachialis, and pectoralis minor (Fig. 28).
4. With a sharp thin-bladed osteotome divide
the coracoid process from above downwards about
FIG.27.-The incision.
half an inch behind its free extremity, and pull
the detached portion downwards with the three muscles attached to it.

fl
i\

FIG. 28.-The deltoid and the pectoralis major

muscles have been separated, showing the coracoid
process and the three muscles attached to it. An
osteotome has been applied to the upper surface of
the coracoid process.

FIG.29.-The
coracoid process has been divided
and the detached portion has been drawn downwards
with the muscles attached to it. T h e tendon of the
subscapularis is seen with an aneurysm. needle .passed
beneath it. The tendon has been partially divlded.

RECURRENT

DISLOCATION

OF

SHOULDER

27

5 . Rotate the humerus outwards so as to bring the subscapularis muscle into

view. Pass an aneurysm needle under the tendon of the subscapularis, and divide
the tendon near its insertion into the lesser tuberosity of the humerus (Fig. 29).
Retract the subscapularis inwards.
6. Inspect the front of the shoulder-joint. In some cases there is a wide
opening into the joint, and the glenoid ligament is found lying loose, either on the
head of the humerus or over the margin of the glenoid cavity. I n this case fix the
free edge of the capsule to the glenoid margin, as described below. If the capsule
appears to be intact, incise it over the glenoid margin, raise the narrow portion
on the inner side of the incision, and proceed to fix the outer cut edge to the glenoid
margin. This edge may include more or less of the glenoid ligament which remains
attached to the capsule. In every case the anterior margin of the glenoid cavity

FIG. 30.-The
tendon of the subscapularis has
been divided and the muscle has been retracted inwards.

A transverse opening is seen in the capsule of the joint,

the outer edge of which lies on the head of the humerus,

while the inner edge, which includes the glenoid ligament, has been raised by a hook from the anterior
margin of the glenoid cavity. Note that the glenoid
margin is smooth and rounded.

FIG. 31.-A thin shaving of bone has been raised

from the anterior margin of the glenoid cavity and the
neck of the scapula. Three holes have been made in
the glenoid margin, and a pair of vulsellum forceps is
seen in the act of making a fourth hole. A suture has
been passed through the two holes in the lower half of
the glenoid margin, and the two ends of this suture are
seen emerging from the raw surface on the front of the
neck of the scapula.

will be found to be smooth, rounded, and free of any attachments, and a blunt
instrument can be passed freely inwards over the bare bone on the front of the
neck of the scapula (Fig. 30).
7. Apply a broad chisel to the glenoid margin and raise a thin shaving of bone
from the front of the neck of the scapula. If the shaving of bone remains attached
to the scapula, preserve it ; otherwise, remove it.
8. Draw the head of the humerus away from the glenoid cavity with a sharp
hook or a special retractor.
9. Apply a pair of sharp-pointed vulsellum forceps to the anterior margin
of the glenoid cavity. Take a good bite and gently work the sharp points through
the bone. Do not be rough, or the forceps will tear out through the bone. In
this way make two pairs of holes leading from the raw surface on the front of the
neck of the scapula to just inside the glenoid cavity.

28

THE

BRITISH

JOURNAL

OF

SURGERY

10.With a small curved needle pass a silkworm gut suture through the lowermost hole from the raw surface into the joint and then back through the next hole
to the raw surface again. I n the same way pass another suture through the second
pair of holes (Fig.3:).
11. Thread the four ends of these two sutures on needles and pass them
through corresponding points near the free edge of the capsule.
12. Press the edge of the capsule down on to the raw surface of bone, and tie
each suture firmly over it with forceps. Cut the ends of the sutures short (Fig.32).
13. If the shaving of bone has been preserved, replace it over the suture line
and fix it down with a catgut stitch.

FIG.32.-The free edge of the capsule has been

drawn inwards and placed on the raw surface of bone
in front of the glenoid margin and neck of the scapula.
T h e two sutures have been passed through the edge of
the capsule. T h e lower suture has been tied, and is
shown with its ends cut short. T h e upper suture is
being tied with two pairs of forceps.

FIG. 33.-The
subscapularis tendon has been
sutured with catgut. T h e detached portion of the
coracoid process has been replaced, and is shown held
in place by two sutures of silkworm gut.

14. Suture the divided subscapularis tendon with catgut.

15. Replace the tip of the coracoid process and fix it in place with one or two
silkworm gut sutures passed either through the bone or through the tendinous
insertion of the muscles (Fig.33).
16. Unite the deep fascia with a continuous catgut suture, and close the skin
with a subcuticular stitch.
17. Apply dressings, and bandage the arm to the side with the elbow well
forward.
18. Remove the subcuticular stitch on the tenth day. Keep the arm bandaged
to the side for a month. Then begin active movements. Full movements should
be regained in about a month. The dislocation does not recur.

SUMMARY AND CONCLUSIONS

Attention is directed to the pathological anatomy of recurrent dislocation of
the shoulder-joint.

RECURRENT

DISLOCATION

OF

SHOULDER

29

The difference between recurrent dislocation and ordinary dislocation of the

shoulder is pointed out.
Recurrent dislocation is due to a wide detachment of the glenoid ligament
from the anterior margin of the glenoid cavity. This anatomical defect is quite
constant, and can be demonstrated in every typical case of recurrent dislocation
of the shoulder.
An operation is described for the repair of this defect. This operation practically restores the shoulder-joint to its normal anatomical state, and it invariably
prevents recurrence of the dislocation.
These statements are based upon an operative experience of 27 consecutive
cases of recurrent dislocation of the shoulder-joint, in every one of which the
typical injury was demonstrated. All these cases recovered full movements of the
joint, and in no case has there been any recurrence of the dislocation.
A brief account of the pathology and treatment of this condition was first
published in the British Medical Journal on December 15, 1923.