Abstract
Background: Some surgeons believe that they can identify patients
who are at high risk for shoulder redislocation and that these patients are
best served by immediate surgical stabilization. This natural history study
was performed to examine the validity of this concept and to determine whether
it is possible to predict the need for future surgery at the time of the index
injury and examination.
Methods: One hundred and thirty-one patients were followed for an
average of four years after their first shoulder dislocation. An extensive
history was recorded and a thorough physical examination was performed on each
patient. Final evaluation consisted of a physical examination, radiographic
evaluation, and determination of three outcome measurements.
Results: Twenty-nine (22%) of the 131 patients requested surgery
during the follow-up period. There were twenty Bankart repairs and nine
rotator cuff repairs. Forty-three patients (33%) had at least one recurrent
dislocation. Thirty-nine of these patients were in the group of ninety
patients under the age of forty years. Thirty-seven of these thirty-nine
patients either participated in contact or collision sports or used the arm at
or above chest level in their occupation. Eighteen (49%) of these thirty-seven
patients had surgery. Only two of the more sedentary patients had
redislocation, and none had surgery. Four (10%) of the forty-one patients over
the age of forty had a redislocation, but none required a Bankart repair.
However, eight (20%) of the forty-one patients required a rotator cuff repair.
Eighty-eight (67%) of the 131 patients never had a redislocation. Their
outcome scores were high and equivalent to those of the cohort of patients who
had had a successful Bankart repair of an unstable shoulder. Patients who had
redislocation but chose to cope with the instability rather than have surgery
had lower outcome scores. Twenty-two (51%) of the forty-three patients who had
recurrent instability had only one redislocation during the entire follow-up
period, whereas some patients had as many as twelve complete
redislocations.
Conclusions: Younger patients involved in contact or collision
sports or who require overhead occupational use of the arm are more likely to
have a redislocation of the shoulder than are their less active peers or older
persons. However, even in the highest-risk groups, only approximately half of
patients with shoulder redislocation requested surgery within the follow-up
period. Early surgery based on the presumption of future dislocations,
unhappiness, and disability cannot be justified.
Level of Evidence: Prognostic Level I. See Instructions
to Authors for a complete description of levels of evidence.
Patients who experience a traumatic anterior shoulder dislocation have
diverse characteristics. They differ in terms of age, sex, activity level,
occupation, sports participation, physical characteristics, and
genetics1,2.
Following a first-time dislocation, the shoulder generally takes one of three
separate clinical paths: it never redislocates, it redislocates but the
patient chooses to cope with the dislocation, or it redislocates and the
patient chooses surgery. In the past, age and sports involvement were thought
to be somewhat predictive of redislocation, but most surgeons allowed their
patients to delay surgery until their symptoms warranted
it3-7.
Recently, as arthroscopic stabilization techniques have improved, some
surgeons have proposed that young, highly athletic patients would be better
served by acute-phase
surgery8-10.
If it were possible to predict, at the time of a patient's first
dislocation, whether that patient was going to require stabilization surgery,
there might be some benefits for both the patient and society. The patients
would be spared not only the pain of multiple redislocations but also the loss
of time from school, work, and sports. In addition, they might avoid the
expense of multiple doctor visits and multiple tests that typically take place
prior to a decision to undergo surgery. The benefit for society would be that
a physician's ability to make rapid and accurate decisions would optimize the
use of medical resources. Thus, the ability to predict, at the time of a first
office visit, whether a patient will need surgery is a desirable goal.
For similar reasons, it is equally desirable to avoid performing
acute-phase surgery on patients who would not have required or desired surgery
had they been given the time for the natural history of their condition to
play out. It is possible that these patients can also be identifiable at the
time of initial presentation. This study was designed to test whether patients
who ultimately requested shoulder stabilization surgery had, at the time of
the initial injury, differed from their peers who did not eventually request
surgery in ways that allowed both the need and the lack of a need for surgery
to be predicted.
This study was approved by our institutional review board. All patients
with a first-time shoulder dislocation who presented to the emergency
department between 1995 and 1999 were referred to an "acute shoulder
clinic" in the orthopaedic department. Patients in this clinic were seen
by our sports medicine fellows, who were supervised by one of us (R.A.S.).
Patients were usually examined within one week after the injury and always
within two weeks. Consent was obtained at the initial visit, and all of the
patients who were asked to participate agreed to be entered into a prospective
study designed to follow them for approximately five years. The inclusion
criterion was a first-time traumatic anterior shoulder dislocation as verified
by a radiograph, a treating emergency department physician, or in some cases a
typical history of injury followed by a reduction maneuver. Any patient with a
previous dislocation, a previous shoulder injury, or an atraumatic dislocation
was excluded. None of our patients had an associated fracture of the greater
tuberosity. Twelve patients had a small fracture of the glenoid rim.
At the intake visit, the patients had a thorough physical examination and
radiographic evaluation. Body type and hyperlaxity signs (thumb to wrist,
elbow hyperextension, knee hyperextension, and hyperextension of the fifth
metacarpophalangeal joint) were recorded. A detailed report that included the
patient's age, sex, occupation, occupational use of the arm, other unstable
joints, and family history of unstable joints was created. The dislocation
event was described in detail and graded in terms of the degree of trauma,
score on a pain scale, method of reduction, duration of use of pain
medication, time missed from work, and duration of use of a sling. Sports
activity was detailed in terms of sports hours per year, which was also
divided into hours of collision sports, overhead or throwing sports, and all
others.
Following reduction of the dislocation, patients were allowed to wear or
not wear a sling as comfort
dictated3,5.
All patients were prohibited from resuming sports activity for a minimum of
six weeks and until they could demonstrate a full range of
motion6.
Follow-up during the five-year study period was carried out by telephone
every six months by a research assistant (M.K.). The purpose of contacting
patients every six months was twofold. First, as young patients in Southern
California tend to be a mobile population, it was deemed necessary to contact
patients often to verify their location and keep them in the study. Second, it
was thought to be more accurate to document instability events more or less as
they happened, rather than relying on the patient's memory at the final
follow-up visit years later.
Patients were asked to return for final follow-up if at all possible and
were given a monetary inducement of $100 to compensate them for travel,
inconvenience, or time off from work. A sports medicine fellow and a physical
therapist, neither of whom was otherwise involved in the intake portion of the
study, gave these patients a thorough examination with use of the American
Shoulder and Elbow Surgeons
format11. The
physical examination was augmented by strength testing of the rotator cuff
muscles with use of a Nicholas handheld dynamometer (Lafayette Instruments,
Lafayette, Indiana). The contralateral side was used as the control. Patients
were also asked to fill out three self-assessment scales: the Western Ontario
Shoulder Instability Index
(WOSI)8, the
American Shoulder and Elbow Society (ASES)
scale11, and the
Constant-Murley
Scale12. If
patients were unable to return for this evaluation, we relied on data from all
of their previous and current telephone contacts.
Patient Demographics
Of the 139 patients who were entered into the study, three were excluded
because they ultimately recalled prior instability events. Five patients were
lost to follow-up. Of the 131 patients who completed the study, twenty-nine
were followed until they underwent surgery. Twenty of these patients had a
Bankart repair, and nine had a rotator cuff repair. Of the remaining 102
patients, who were treated nonoperatively, fifty-one were followed with a
telephone interview only and fifty-one returned for the final evaluation
(Fig. 1).
The average age of the 131 patients in the study was thirty-three years
(range, twelve to eighty-two years). One hundred and two (78%) of the patients
were male. Fifty-two (40%) of the injuries were on the dominant side, and
seventy-nine were on the nondominant side. All patients experienced a complete
dislocation, which was verified either radiographically or with the report of
a physical examination and a reduction maneuver. Fifteen patients were able to
reduce the dislocation themselves, eighteen were able to reduce it with the
help of someone on the scene, and ninety-eight patients required reduction in
the emergency room. Use of a sling was left to the choice of the patient and
eighty-five patients removed the sling one week or less after it was provided
to them. Only forty-six patients continued to wear the sling for more than one
week.
Patient Follow-up
The mean duration of follow-up of the patients who were managed
nonoperatively was four years (range, two to seven years). Those who returned
for a final evaluation in the physician's office were followed for a mean of
five years (range, two to seven years), whereas those who were only
interviewed by telephone were followed for mean of three years (range, two to
five years). No patient was followed for less than two years.
Statistical Analysis
Independent t tests and Mann-Whitney tests were used to assess gender and
patient-group differences in demographic characteristics, functional
parameters, results of the physical examination, and shoulder outcome scores.
Paired t tests and Wilcoxon signed-ranks tests were used to compare preinjury
and postinjury activity levels as well as to compare the physical examination
parameters between the index and the contralateral shoulder. Chi-square and
Fisher exact tests were used to assess differences between categorical
variables. Survival curves were calculated with use of the Kaplan-Meier
method, with surgical repair as the end point. Backward stepwise logistic
regression analyses with removal at p = 0.15 or above was used to examine the
variables associated with subsequent instability. The threshold for
significance in this study was set at p = 0.05.
During the five-year study period, eighty-eight shoulders (67%) remained
stable, whereas forty-three patients (33%) had recurrent instability of the
shoulder. Nineteen of the forty-three unstable shoulders and ten of the stable
shoulders were treated with surgery (Fig.
2). Most of the surgical procedures were performed within the
first two years of the study. Kaplan-Meier survival curves with surgical
repair as the end point are demonstrated in Figures
3 and
4. We could not identify a
difference in survival according to sex.
Of the twenty-nine patients who had shoulder surgery, twenty had a Bankart
repair and nine had a rotator cuff repair.
Figure 5 shows the ages of all
patients who had surgery and the division between Bankart repairs and rotator
cuff repairs. None of the patients had both a Bankart repair and a rotator
cuff repair, and none who had a rotator cuff repair experienced a subsequent
episode of shoulder instability.
The patients were divided into two groups on the basis of their age. The
older group included forty-one patients who were at least forty years of age
(mean, fifty-four years of age) at the time of the first dislocation.
Thirty-seven (90%) of the forty-one shoulders remained stable within the
follow-up period, whereas four had a recurrent dislocation. One of the four
patients also had a rotator cuff tear, which was seen at the time of the
surgery to be a typical rotator interval tear involving the supraspinatus and
the subscapularis. The pain and instability resolved following repair of this
tear. Nine (22%) of the forty-one patients underwent surgery within the study
period, but eight of the nine operations were rotator cuff repairs
(Fig. 6). Only one patient who
was older than forty underwent an actual repair for shoulder instability
involving a Bankart lesion. While the pathological findings in the patients
undergoing a rotator cuff repair were as expected—involving either the
supraspinatus or the subscapularis, or both—we do not know the
prevalence of rotator cuff tears in our asymptomatic patients over the age of
forty. We found that patients over the age of forty commonly had weakness on
the initial examination after the dislocation. This weakness was most notable
on testing of the subscapularis. However, when all pain and weakness had
resolved by the six-week follow-up examination, we did not do any additional
studies. Patients who continued to have either pain or weakness underwent
magnetic resonance imaging, and surgical decisions were based on the
findings.
The younger group included ninety patients under the age of forty (mean
age, twenty-three years). Fifty-one (57%) of the ninety shoulders stayed
stable within the follow-up period, and the mean age of the patients with
these shoulders was twenty-six years. Thirty-nine (43%) of the ninety patients
experienced recurrent dislocations, and their mean age was twenty-one years (p
= 0.004). Two of the fifty-one stable shoulders underwent surgery. One patient
had a rotator cuff repair and one had a Bankart repair to treat chronic pain.
The latter patient never had any documented dislocations or subluxations, but
laxity as well as a Bankart lesion were demonstrated at the time of the
surgery and chronic subtle instability was thought to be the genesis of the
pain. After the Bankart repair, the pain resolved. Eighteen (46%) of the
thirty-nine patients with an unstable shoulder eventually had surgery. All
eighteen had a Bankart repair (Fig.
7).
At the end of the study, we reviewed our intake data to determine whether
there were any differences between the twenty-nine patients who ultimately
underwent surgery and the 102 patients who did not. We then analyzed the group
of ninety patients under the age of forty to ascertain whether the twenty
patients who eventually underwent surgery were different in any way from the
seventy patients who did not. Finally, we analyzed our group of ninety
patients who were under the age of forty to look for any differences between
the thirty-nine patients with recurrent dislocation and the fifty-one patients
in whom the shoulder remained stable.
With the numbers studied, the patients who underwent surgery did not
differ, in most ways, from those who did not. Specifically, sex, dominant arm,
percentage of injuries to dominant arm compared with that to the nondominant
arm, a family history of instability, a patient history of previous
instability of other joints, the type or degree of injury that caused the
shoulder dislocation, the amount of time for which the shoulder was
dislocated, the method of reduction, the place of reduction, and the time that
the shoulder was in the sling were similar in the two groups. On physical
examination, the two groups showed no difference in the initial range of
motion, strength, hyperlaxity signs, or radiographic findings.
Glenoid rim fracture also did not prove to be a significant finding. A
glenoid rim fracture was found in twelve patients, including six over the age
of forty and six under the age of forty. Of the six patients who were older
than forty, only one had surgery to address instability. The remaining five
shoulders remained stable within the follow-up period. Of the six patients
with a glenoid rim fracture who were younger than forty, three had no
additional episodes of instability within the follow-up period. Only one of
the three patients who did have recurrent instability had stabilization
surgery. Thus, within the limits of this small sample, a glenoid rim fracture
did not appear to be a risk factor for instability requiring surgery.
None of our patients who were younger than forty had evidence of
substantial degenerative changes on radiographs made either immediately after
the injury or at the time of follow-up. However, the study period was too
short for us to say anything meaningful about the development of degenerative
arthritis.
The patients who had surgery did differ from those who did not in some
respects. The patients who ultimately had surgery tended to have higher pain
scores at the initial visit (Fig.
8). However, the increases in the scores were most dramatic in the
patients who had rotator cuff repair. These patients had significantly higher
pain scores at their intake visit than those who selected nonoperative care (p
< 0.01). Patients who underwent rotator cuff surgery also reported a longer
time for the pain to resolve after reduction of the shoulder (p = 0.036).
In the group of patients under the age of forty, the total number of sports
hours alone was not predictive of who would have a recurrent dislocation. The
patients in whom the shoulder became stable after the initial dislocation
reported an average of 251 yearly sports hours at their intake examination,
whereas those who eventually had at least one recurrence reported an average
of 230 sports hours. We then looked at the hours spent specifically in contact
or collision sports. Eighteen (38%) of forty-eight patients who did not
participate in contact or collision sports had a redislocation, and eight
(17%) of those forty-eight patients requested surgery. In comparison,
seventeen (55%) of the thirty-one patients who participated in contact or
collision sports had a redislocation, and ten (32%) of the thirty-one patients
requested surgery. These trends did not reach the level of significance.
Patients who used the arm at or above chest level in their occupation were
more likely to have a subsequent instability event. Seven (28%) of the
twenty-five patients who did not use the arm at or above chest level had a
subsequent dislocation, whereas thirty-two (51%) of the sixty-three patients
who did use the arm at or above chest level had at least one subsequent
shoulder dislocation (p = 0.05). (Data regarding this factor were missing for
two patients.) The patients who used the arm at or above chest level were also
younger than the other group, with an average age of twenty-two compared with
27.4 years (p < 0.005).
Only two of the thirty-nine patients who had redislocation neither
participated in contact or collision sports nor used the arm at or above chest
level in their occupation. Neither of these patients had surgery. Thus, the
absence of occupational use of the arm at or above chest level and the absence
of participation in contact or collision sports was predictive of future
stability.
Age, sex, a pain rating of >8, an occupation requiring use of the arm at
or above chest level, and participation in collision or contact sports were
entered into the preliminary multivariate logistic regression model. The
adjusted odds ratios for the final multivariate logistic regression model are
displayed in Table I. An age of
less than twenty-five years at the time of presentation was the strongest
predictor of subsequent instability.
Because our logistic regression analysis showed that age was the dominant
factor in predicting instability, we further analyzed our patients by
subdividing them according to their decade of life. Most surgeons would not
expect the natural history of an eighteen-year-old patient to be the same as
the natural history of a thirty-two-year-old patient, and this was verified by
our study.
Ages Twelve to Nineteen Years (Thirty-nine Patients): Seventeen
shoulders (44%) remained stable during the follow-up period, and twenty-two
(56%) had at least one episode of recurrent instability. Eleven patients (50%
of those with an unstable shoulder and 28% of the patients in this age group)
chose to have surgery within the follow-up period. Eighteen of these patients
also participated in collision or contact sports involving the injured arm.
Seven (39%) of the eighteen had no shoulder instability within the follow-up
period, and eleven (61%) had at least one recurrent dislocation. Seven
patients (64% of those with an unstable shoulder who participated in collision
or contact sports and 39% of the patients in this age group who participated
in collision or contact sports) chose to have surgery within the follow-up
period.
Ages Twenty to Twenty-nine Years (Twenty-nine Patients): Eighteen
shoulders (62%) remained stable during the follow-up period, whereas eleven
(38%) had a least one episode of recurrent instability. Four of these eleven
patients (14% of the twenty-nine patients in this age group) chose to have
surgery within the follow-up period. Ten patients in this group also
participated in contact or collision sports. Five of the ten had at least one
episode of recurrent instability, and two chose to have surgery.
Ages Thirty to Thirty-nine Years (Twenty-two Patients): Sixteen
shoulders (73%) remained stable during the follow-up period, and six (27%) had
at least one episode of recurrent instability. Five chose to have surgery
within the follow-up period. Four had a Bankart repair (one for pain and three
for instability) and one had a rotator cuff repair. Three patients in this
group participated in contact or collision sports. One of the three had
recurrent instability and chose surgery.
The patients who had recurrence of the shoulder instability were a diverse
group. The number of instability events during the follow-up period ranged
from one to twelve. The majority of patients (51%) had only one event.
Interestingly, patients who ultimately chose surgery had had an average of
only 1.6 instability events before coming to that decision. Patients who chose
to cope with the instability had an average of 3.4 instability events during
the follow-up period, or slightly less than one episode per year of
follow-up.
Outcome Scores
At the time of final follow-up, the patients were asked to complete three
outcome scores: the Constant-Murley scale, the ASES, and the WOSI. These three
outcome scores were compared among three groups: patients in whom the shoulder
became stable after the first dislocation, patients who had a recurrent
shoulder dislocation but chose to cope with the instability, and patients who
had a recurrence and underwent a successful Bankart repair
(Table II). The patients whose
shoulder became stable after the first dislocation had outcome scores
equivalent to those of the patients with an unstable shoulder who had a
successful Bankart
repair13. Thus,
patients with successful nonoperative treatment had an outcome that was much
the same as that of patients with successful surgical treatment. It is of
interest that patients who chose to cope with an unstable shoulder had lower
ASES (p = 0.01), Constant-Murley (p = 0.03), and WOSI (p = 0.004) scores than
did patients who had a stable shoulder.
Kaiser Permanente is a nonprofit health maintenance organization with more
than 500,000 members in the greater San Diego area. The patients in our study
were drawn from this entire population, which included athletes of all types.
Because of the proximity of San Diego to mountains, desert, and ocean and the
favorable year-round weather, the health plan covers a highly athletic
population. However, Kaiser Permanente in San Diego does not have contracts to
cover either professional or National Collegiate Athletic Association
Division-I athletes. Therefore, our study may not be relevant to this small
subgroup of the population.
Interestingly, our results can be seen to mimic those found by Hovelius et
al. in their landmark 1996 study of the population in
Sweden4. In that
study, at ten years after the index dislocation, 129 (52%) of 247 shoulders
had remained stable while fifty-eight patients (23%) had requested operative
shoulder stabilization. In a study in the Netherlands, te Slaa et al. followed
107 shoulders from the time of the initial dislocation; fifty-six percent of
the shoulders in patients under the age of forty years were stable at four
years14. In our
study, in which the duration of follow-up was only five years, the shoulders
of fifty-one (57%) of ninety patients under the age of forty remained stable
and eighteen patients (20%) requested surgical stabilization. On the basis of
these few studies, it is probably safe to conclude that the natural history of
shoulder instability is similar from nation to nation. The similarity of these
studies also speaks to the fact that most patients with recurrent shoulder
instability experience this recurrence and make decisions regarding treatment
of the instability within the first few years following the index event.
Certainly, some patients make their decisions on a markedly delayed basis, but
these patients, considered in the context of a large population, do not change
outcome statistics by any significant degree. Thus, we believe that our study,
while relatively short-term, provides valuable insight into the true natural
history of patients with a first-time shoulder dislocation.
When reporting statistics related to shoulder dislocation rates, authors
often overlook the general population as they focus on subgroups that fall
outside of the norm. Many published papers deal with only the very old, the
very young, or the very athletic. Pevny et al. reported on fifty-two patients
over the age of forty years who sustained a first-time shoulder dislocation
while skiing15. At
a minimum of two years following this dislocation, only two patients had had a
recurrence of anterior instability, but eighteen had had a rotator cuff tear,
not all of which were treated surgically. At the time of the surgery in the
two patients with instability, the subscapularis and capsule were seen to have
ruptured from the lesser tuberosity and neither patient had a Bankart lesion.
Neviaser and Neviaser reported on twelve patients with an average age of
sixty-three years who had recurrent anterior
instability16. At
the time of surgery, the subscapularis and capsule were seen to have ruptured
from the lesser tuberosity in all of these patients. No patient had a Bankart
lesion. We had similar findings. Of our forty-one patients over the age of
forty, only four had recurrence of instability but nine (22%) underwent
surgery, primarily for rotator cuff repair.
The great diversity of opinion regarding early surgical treatment is
centered on the very young and athletic. Some authors have stated their belief
that there are subgroups of patients who are so young and athletic that
aggressive operative treatment can be justified after the initial dislocation.
Wheeler et al. observed a 92% rate of recurrence of instability after
nonoperative treatment of United States Military Academy Cadets and believed
that arthroscopic surgical intervention "should be considered as a
treatment option" in young
athletes10.
DeBerardino et al. explored this concept over a ten-year period in a study of
individuals at the United States Military
Academy9. They found
an 85% recurrence rate in nonoperatively treated cadets and demonstrated that
early arthroscopic repair could dramatically lower the prevalence of
instability at the time of an approximately two-year follow-up.
Kirkley et al. explored the role of early arthroscopic stabilization in a
group of forty patients who had been randomly assigned to be treated with
either immobilization or surgery following their first
dislocation8. At two
years, they observed a 47% redislocation rate in the nonoperatively treated
group and a 15.9% redislocation rate in the surgically treated group. In
addition, the WOSI scores of the surgical group were better than those of the
nonoperatively treated group.
In a study of thirty athletes who had had nonoperative treatment of an
anterior dislocation, Buss et al. found that twenty-six of them were able to
return to sports participation at an average of 10.2 days after the injury and
were able to complete the
season17. Eleven
athletes experienced recurrent instability during the season, and sixteen of
the thirty athletes chose to have surgical stabilization in the
off-season.
The studies of military cadets by Wheeler et
al.10 and
DeBerardino et al.9
may be criticized for dealing with a patient population that has severely
limited choices of a nature not found in any other patient population.
However, those authors do appear to have identified a subgroup that falls
outside the norm. The study by Kirkley et
al.8, which dealt
with a more general population, brings up the same issues as found in our
study. In their study, only 47% of the patients had experienced a
redislocation at two years, but the WOSI scores for the surgically treated
patients were better than those for the nonoperatively treated patients. The
WOSI scores for our patients are in agreement with theirs, with one important
difference. When we divided our nonoperatively treated patients into those
with a stable shoulder and those with an unstable shoulder, we found that only
the half that had an unstable shoulder had worse WOSI scores than the surgical
group. Our patients with a stable shoulder had scores equivalent to those in
the group with surgical stabilization. Kirkley et al. did not make this
distinction.
The study by Buss et
al.17 dealt with a
typical profile of young, highly committed athletes and offers a valuable
glimpse of what happens when you give patients choices. Even in this young
active group of athletes, only sixteen of thirty individuals chose surgery at
the end of the season.
In our study, young athletes involved in collision or contact sports were
found to be most at risk for recurrent instability. These athletes had a 55%
redislocation rate. Yet, given a choice, only 32% decided to have surgery.
Given the improvement in the outcome scores in the patients who had surgery,
some might question whether our patients made the right choice. It is not
atypical for surgeons to believe, with some certainty, that they can make
their patients better with surgery. However, there are no compelling studies
to indicate that patients with recurrent instability suffer dire consequences
from a decision not to have surgery or that the condition cannot be treated
successfully on a chronic basis if they change their minds. Thus, while we
tell our patients that we can offer them the probability of an excellent
outcome with surgery, we feel no need to push for an early decision and we
believe that patient choice is a reasonable outcome variable.
Virtually all studies of which we are aware have shown that, regardless of
the redislocation rate and the patient population, if patients with
instability are given a choice, only half will choose surgery. We have seen
from our study and that of Kirkley et
al.8 that some
patients who choose to cope with the instability will have WOSI scores that
are worse than those of surgically treated patients. Yet there appears to be
no reason to discourage patients from exercising their right of choice.
In summary, the results of our five-year study are similar to those of
other natural history studies of large populations, confirming that
approximately half of all patients who have a traumatic shoulder dislocation
will not have an additional instability event within a five-year follow-up
period and the other half will have one or more redislocations. However, the
half that do have a redislocation is a heterogeneous group, with greatly
varying degrees of instability and symptoms, and only half of this group will
ultimately request surgery.
Higher-risk patients who are more likely to have a redislocation do have
certain characteristics in common. They are young, have jobs requiring
overhead use of the arm, or are involved in contact or collision sports.
However, even in this high-risk subgroup, only half of the patients request
surgery while the other half choose to cope with the instability.
Lower-risk patients under the age of forty years who may not require
surgery may be easy to identify. A patient who does not use the arm for
collision or contact sports or at or above shoulder level at work is highly
unlikely to have a redislocation and even more unlikely to request surgery.
Patients in whom the shoulder stabilized after the first dislocation had
excellent scores on all three outcome scales. These scores were almost
identical to those of patients who had a successful Bankart repair. Patients
whose shoulder redislocated but who chose to cope with the instability had
lower outcome scores on all three scales.
The need for immediate surgery was not predictable with any accuracy in our
population. Even in our group of highest-risk young patients, 65% never
requested surgery during the follow-up period despite being contacted by our
department every six months. Thus, although surgery can result in good outcome
scores, it remains difficult to justify it in the acute setting. ?
Dowdy PA, O'Driscoll SW. Shoulder
instability. An analysis of family history. J Bone Joint Surg
Br. 1993;75:
782-4.75782
1993
[PubMed]
O'Driscoll SW, Evans DC. Contralateral
shoulder instability following anterior repair. An epidemiological
investigation. J Bone Joint Surg Br.
1991;73:
941-6.73941
1991
[PubMed]
Henry JH, Genung JA. Natural history of
glenohumeral dislocation—revisited. Am J Sports Med.
1982;10:
135-7.10135
1982
[PubMed][CrossRef]
Hovelius L, Augustini BG, Fredin H,
Johansson O, Norlin R, Thorling J. Primary anterior dislocation of the
shoulder in young patients. A ten-year prospective study. J Bone Joint
Surg Am. 1996;78:
1677-84.781677
1996
Rowe CR, Sakellarides HT. Factors
related to recurrences of anterior dislocations of the shoulder. Clin
Orthop. 1961;20:
40-8.2040
1961
Simonet WT, Cofield RH. Prognosis in
anterior shoulder dislocation. Am J Sports Med.
1984;12:
19-24.1219
1984
[PubMed][CrossRef]
te Slaa RL, Brand R, Marti RK. A
prospective arthoscopic study of acute first-time anterior shoulder
dislocation in the young: a five-year follow-up study. J Shoulder Elbow
Surg. 2003;12:
529-34.12529
2003
[CrossRef]
Kirkley A, Griffin S, McLintock H, Ng L.
The development and evaluation of a disease-specific quality of life
measurement tool for shoulder instability. The Western Ontario Shoulder
Instability Index (WOSI). Am J Sports Med.
1998;26:
764-72.26764
1998
[PubMed]
DeBerardino TM, Arciero RA, Taylor DC.
Arthroscopic stabilization of acute initial anterior shoulder dislocation: the
West Point experience. J South Orthop Assoc.
1996;5:
263-71.5263
1996
[PubMed]
Wheeler JH, Ryan JB, Arciero RA,
Molinari RN. Arthroscopic versus nonoperative treatment of acute shoulder
dislocations in young athletes. Arthroscopy.
1989;5:
213-7.5213
1989
[PubMed][CrossRef]
Richards RR, An KN, Bigliani LU,
Friedman RJ, Gartsman GM, Gristina AG, Iannotti JP, Mow VC, Sidles JA,
Zuckerman JD. A standardized method for the assessment of shoulder function.
J Shoulder Elbow Surg.
1994;3:
347-52.3347
1994
[CrossRef]
Constant CR, Murley AH. A clinical
method of functional assessment of the shoulder. Clin Orthop Relat
Res. 1987;214:
160-4.214160
1987
Sachs RA, Williams B, Stone ML, Paxton
L, Kuney M. Open Bankart repair: correlation of results with postoperative
subscapularis function. Am J Sports Med.
2005;33:
1458-62.331458
2005
[PubMed][CrossRef]
te Slaa RL, Wijffels MP, Brand R, Marti
RK. The prognosis following acute primary glenohumeral dislocation. J
Bone Joint Surg Br. 2004;86:
58-64.8658
2004
Pevny T, Hunter RE, Freeman JR. Primary
traumatic anterior shoulder dislocation in patients 40 years of age and older.
Arthroscopy. 1998;14:
289-94.14289
1998
[PubMed][CrossRef]
Neviaser RJ, Neviaser TJ. Recurrent
instability of the shoulder after age 40. J Shoulder Elbow
Surg. 1995;4:
416-8.4416
1995
[CrossRef]
Buss DD, Lynch GP, Meyer CP, Huber SM,
Freehill MQ. Nonoperative management for in-season athletes with anterior
shoulder instability. Am J Sports Med.
2004;32:
1430-3.321430
2004
[PubMed][CrossRef]