448
shoulder was immobilized in a sling for 6 weeks. The rehabilita-
tion program was started the first day after surgery with supported
active elevation to 90 deg without external rotation. External rota-
tion was allowed to 10 deg after 3 weeks and to a maximum, pain-
free external rotation position after 6 weeks. Muscle strengthening
exercises were started 6 weeks postoperatively.
Subjective result. All but 2 patients said they were satis-
fied with the operation (88%). One patient who was dis-
satisfied was unable to play sports because of recurrent
dislocations; the other didn’t notice any difference in the
shoulder condition and complained about pain and sublux-
ations. Both patients were in the instability group.
Results
Complications. There were no infections, neurovascular
complications or symptomatic posterior instability.
Pain. Eleven patients were totally free of shoulder pain at
follow-up. Of the 6 patients who reported pain postopera-
tively, 3 did not improve, 2 patients did improve, and 1 pa- Discussion
tient was worse compared with the preoperative status. In
the group with secondary subacromial impingement (6 pa- MDI of the shoulder affects mostly younger patients be-
tients), 3 patients were free of pain, 2 patients improved, tween 19 and 40 years of age. Pre-existent asymptomatic
and 1 patient did not improve. In the group with instabil- MDI may become symptomatic after an adequate trauma
ity (11 patients), 8 patients are free of pain, 2 patients did to the shoulder, resulting in a labral lesion and subsequent
not improve, and 1 patient was worse (Table 1).
increased instability. Of the 17 patients we investigated, 5
had suffered an adequate trauma preoperatively, 6 had an
Instability. At the time of follow-up, 2 patients (18%) had inadequate trauma, and 6 had no preoperative trauma. Al-
suffered a recurrent dislocation of the shoulder, and 2 pa- though excellent results may be obtained with conserva-
tients (18%) experienced subluxation. Of the 2 patients tive treatment, one in every 6 patients will not respond to
with a recurrent dislocation, 1 experienced dislocations such treatment [2, 7]. For the non-responders, several op-
during sports after a new trauma to the shoulder, and the erative techniques have been described [5, 9, 10, 11]. In
other had had 2 recurrent episodes during pregnancy. The short-term follow-up studies of arthroscopic capsular shift,
latter case was only temporary instability and disappeared the recurrence rate ranged from 0% to 45% [1, 10]. Arthro-
after the delivery. Of the 2 patients with recurrent sublux- scopic capsular shift is a technically demanding procedure
ations, one did not notice any difference compared with with a long learning curve, while with the open procedure
the preoperative situation; the other experienced fewer under direct vision, the extent of the shift can be easily de-
subluxations, which started after a fall down the stairs termined. In most reports of open capsular shift for MDI,
3.5 months after surgery.
the recurrence rate for instability ranged from 0% to 39%
[1, 8, 11, 12], which is comparable to the 36% in our study.
In studies of open capsular shift for secondary im-
Motion. There was an average loss of 18.5 (range 5–55)
deg of external rotation in 10 patients, and an average im- pingement syndrome, the results were good, with recur-
provement of 12 (range 5–20) deg in 3 patients at follow- rent impingement ranging from 13% to 36% [8]. In our
up. In 4 patients there was no difference in external rota- study two patients (33%) in the impingement group still
tion. In 5 patients a painful arc was present before the op- experienced some residual pain at follow-up.
eration, which was absent at follow-up. Twelve patients
In studies of arthroscopic capsular shrinkage [9] for
had full motion equal to that of the opposite side for active MDI, the results were good, with recurrent instability
flexion: 180 deg. Four patients had a loss of flexion of be- ranging from 3% to 7%. However, little is known about
tween 10 and 90 deg (mean 45 deg). Thirteen patients had the long-term effects of the shrinkage, the effect of the
full motion equal to that of the opposite side for abduc- thermal damage to the proprioceptive nerve endings within
tion: 180 deg. Four patients had a loss of abduction of be- the capsule, and the time for them to fully reform and
tween 10 and 90 deg (mean 31 deg).
function normally. If the applied temperature exceeds a
critical limit, destruction of the entire capsular matrix may
Strength. The mean difference in abduction strength be- occur, with devastating results [9].
tween the operated shoulder and the opposite side was
1.4 kg (range 0–6 kg).
No complications (infection, nerve injury or haematoma)
occurred in our study. In other studies dystrophy, thoracic
outlet syndrome, degenerative arthritis, haematoma, axil-
Overall score. The mean Rowe score improved from 57 lary neuropraxis and thrombophlebitis of the cephalic vein
(range 36–82) points preoperatively to 87 (range 39–100) were described [5, 8, 12]. The OACS may lead to over-
points at follow-up. The mean Constant score at follow- tightening of the anterior capsule with subsequent symp-
up was 77 (range 60–87) points for the patients who were tomatic posterior instability or degenerative arthritis [5].
operated on for secondary subacromial impingement and None of our patients demonstrated this phenomenon, prob-
83 (range 47–100) points for the patients who were oper- ably because the capsule was closed in 15 deg of external
ated on for instability. The mean Dawson questionnaire rotation. Most patients regained a sufficient range of ex-
score at follow-up was 26 (range 14–40) in the overall pa- ternal motion during their rehabilitation. In our study 15
tient group, 32 (range 27–43) in the impingement group of 17 patients (88%) thought the operation was success-
and 25 (range 14–40) in the instability group.
ful.