Abstract
Background:
The treatment of primary traumatic anterior shoulder dislocation varies widely from no immobilization, to two or three weeks of immobilization in internal rotation with the arm in a sling, to treatment with a brace in external rotation. The aim of the present clinical trial was to compare immobilization in internal and external rotation after anterior shoulder dislocation.
Methods:
One hundred and eighty-eight patients with a primary anterior traumatic dislocation of the shoulder were randomly assigned to treatment with immobilization in either internal rotation (ninety-five patients) or external rotation (ninety-three patients) for three weeks. The primary outcome measure was a recurrent dislocation within twenty-four months of follow-up.
Results:
The follow-up rate after a minimum period of two years was 97.9% (ninety-three of ninety-five) in the internal rotation group and 97.8% (ninety-one of ninety-three) in the external rotation group. The compliance rate with the immobilization was 47.4% (forty-five of ninety-five) in the internal rotation group and 67.7% (sixty-three of ninety-three) in the external rotation group. The intention-to-treat analyses showed that the recurrence rate was 24.7% (twenty-three of ninety-three) in the internal rotation group and 30.8% (twenty-eight of ninety-one) in the external rotation group (p = 0.36).
Conclusions:
Immobilization in external rotation does not reduce the rate of recurrence for patients with first-time traumatic anterior shoulder dislocation.
Level of Evidence:
Therapeutic Level I. See Instructions to Authors for a complete description of levels of evidence.
A major problem after a primary traumatic anterior shoulder dislocation is the high risk of recurrence among young patients. Several studies have indicated that the age of the patient at the time of primary dislocation is the most important prognostic factor in determining the risk of recurrent instability1-5. Reduction followed by a period of two to three weeks of immobilization in internal rotation with the arm in a sling or a brace has been the traditional method of treatment. The literature has shown no consensus regarding the benefit of this treatment, but earlier studies suggested that recurrent dislocations are not prevented1-6.
A new concept of treatment with immobilization in external rotation was introduced by Itoi et al.7-10. On the basis of the findings of a cadaveric study and magnetic resonance imaging (MRI), they postulated that, in external rotation, the anterior shoulder soft tissue would tighten, potentially prohibiting the development of a hematoma and promoting better coaptation of the Bankart lesion on the glenoid neck7,8. In the MRI study, the position of the torn labrum after shoulder dislocation was measured and evaluated in different positions of rotation, resulting in the conclusion that the labrum was less displaced from the anatomical position with the shoulder in external rotation and was more displaced from the anatomical position in the internal position8. On the basis of those findings, Itoi et al. hypothesized that immobilization in external rotation would decrease the rate of recurrent dislocations and initiated a prospective randomized study10. The conclusion of that study was that immobilization in external rotation reduced the risk of recurrence compared with immobilization in internal rotation. To our knowledge, the study by Itoi et al. is the only published study that has demonstrated a difference in recurrence between patients managed with immobilization in external or internal rotation10.
Our goal was to evaluate whether the recurrence rate for patients managed with immobilization in external rotation differs from that for patients managed with immobilization in internal rotation. We designed a prospective, randomized multicenter study to compare the recurrence rate after two years of follow-up in patients with primary traumatic dislocation who were managed with immobilization in external or internal rotation.
Participants
The trial was approved by our institution's scientific board, the Regional Medical Ethics Committee, and the Norwegian Social Sciences Data Services. The study is registered in a public trial registry (REK study number 11783) and reported to ClinicalTrials.gov (ClinicalTrials.gov identifier NCT00202735). All of the participants signed informed consent.
Between January 2005 and February 2008, we recruited a total of 257 patients who were managed for primary traumatic anterior shoulder dislocation at thirteen hospitals. The inclusion criteria were (1) an age of sixteen to forty years and (2) successful reduction of primary traumatic anterior glenohumeral dislocation as documented with conventional radiographs.
The exclusion criteria were (1) a glenoid fracture with a large osseous defect (including >20% of the length of the glenoid rim)11 or an osseous glenoid defect involving more than one-third of the diameter of the glenoid fossa at the same level; (2) a fracture of the greater tubercle of the humerus, with malalignment after repositioning; (3) nerve injury related to dislocation or reduction; and (4) an unwillingness or inability to participate in the investigation.
Fourteen patients did not meet the inclusion criteria, and fifteen refused to participate. Seven patients were excluded because of alcohol abuse or intoxication at the time of injury, five patients lived abroad, and twenty-eight patients had no explanation for exclusion. Of the 257 patients who were assessed for eligibility, 188 were enrolled and randomized.
Procedures
The patients were randomly assigned to immobilization in internal rotation or external rotation. Radiographic examination was completed before and after reduction. Allocation to treatment was carried out directly after reduction. The physicians were allowed to use the method of reduction with which they were familiar or the method that they used routinely.
The physicians on duty in the emergency departments enrolled the patients. Block randomization was conducted for each hospital. At each hospital, we also stratified the patients into two age groups (sixteen to twenty-four years and twenty-five to forty years). We did not stratify for sex. Concealed randomization was performed according to the method described by Altman12, and sealed envelopes were used.
Immobilization in internal rotation was performed with use of a normal collar and cuff device or a sling and a swathe. Immobilization in external rotation involved the use of an appropriately-sized shoulder immobilizer (15° UltraSling ER; DonJoy, Vista, California) (Fig. 1-A). The shoulders were immobilized with 15° of external rotation. A line at the top of the immobilizer was used to check the position and was parallel to the frontal plane when the arm was correctly placed (Fig. 1-B). The immobilization treatment was started immediately, and the patients were instructed in the proper use of the immobilizer. The immobilization period was three weeks for both groups.
Outcome Measures
All patients had a minimum of two years of follow-up. A standardized questionnaire was mailed to the patients at the time of follow-up. The primary outcome measure was a recurrent dislocation within twenty-four months after the initial dislocation. Dislocation was defined as the humeral head being completely out of the glenoid until a reducing maneuver was performed. The patients recorded the date of the first recurrent dislocation on the questionnaire, and this date was confirmed by means of a medical record review.
Secondary outcome measures were subluxation, recurrent instability, physical activity, and surgical treatment of instability. A subluxation was diagnosed when the patient reported the humeral head as being partially out of the glenoid. Recurrent instability was defined as recurrent dislocation or subluxation. The level of physical activity before and after the primary dislocation was determined, and a history of surgical procedures subsequent to the initial shoulder dislocation was recorded. The Western Ontario Shoulder Instability Index (WOSI), with twenty-one items, each with a score ranging from 0 to 100, was included in the questionnaire at the time of follow-up after two years13. The best total score possible is 0, implying no decrease in shoulder-related quality of life. The worst total score possible is 2100, implying extreme decrease in shoulder-related quality of life.
Physical Assessment
We performed no physical assessment at the time of follow-up. At baseline, the patients were examined for signs that could be associated with generalized ligamentous laxity. The contralateral, nondislocated shoulder was tested with the inferior sulcus test. A positive test was defined as ≥2 cm of translation. Other signs associated with generalized ligamentous laxity were recorded if one of the following signs was found: hyperflexibility of the wrist, hyperextension of elbows or knees exceeding 15°, or maximum external rotation in the nondislocated shoulder exceeding 90°.
Complications related to the treatment were also registered.
Registration of Compliance
Both groups of patients received identical information orally and by means of a letter about keeping the shoulder as immobilized as possible for three weeks, day and night, except when dressing and undressing or taking a shower. The patient recorded daily use as (1) not at all, (2) up to eight hours, (3) between eight and sixteen hours, or (4) more than sixteen hours every day and night. We also recorded how many days the shoulder had been immobilized.
All patients who used the immobilizer for more than sixteen hours every day and night for at least twenty days were defined as compliant and full-time users. We also used an alternative definition with more than twenty-one hours of documented daily use of brace or sling being defined as compliance.
The patients were also asked to record daily use by filling in a detailed “time recording questionnaire for duration of immobilization,” which was provided on the day of injury and returned at the time of the three-week follow-up.
Statistical Analysis
The sample size calculation was based on the preliminary report by Itoi et al. and on earlier studies on the risk of recurrence9,14,15. The calculation was based on two assumptions. One assumption was that the new treatment reduced the recurrence rate by >50%. The required sample size was calculated to be seventy-two in each group, with alpha = 0.05 and beta = 0.2, for a probability of recurrence of 0.35 in the internal rotation group and 0.15 in the external rotation group. The other assumption was that most of our patients would be young (less than twenty-four years old) and active and that there would be only a medium-sized effect with a maximum 30% reduction in the recurrence rate. The sample size was calculated to be sixty-eight in each group, with alpha = 0.05 and beta = 0.2, for a probability of recurrence of 0.75 in the internal rotation group and 0.52 in the external rotation group. In both scenarios, we assumed that up to one-third of the patients could either be lost to follow-up or be noncompliant. We planned to include a total of 100 patients in each intervention group.
The normality distribution of the variables was examined with histograms and Q-Q plots and was tested with Kolmogorov-Smirnov and Shapiro-Wilk tests. All statistical tests were two-tailed. The significance level was set at 0.05. The primary analysis was based on the principle of intention to treat. In addition, per-protocol analysis only incorporated the compliant patients. Compliance was evaluated for both groups, and patients with more than sixteen hours of daily use of a brace or sling were defined as complaint. We also performed an additional sensitivity analysis in which patients with more than twenty-one hours of documented daily use were defined as compliant. Standard statistical methods were used for descriptive statistics. The rate of recurrent dislocation and the rate of recurrent instability were compared between the groups with use of the chi-square test. We also compared recurrence rates in different age categories. Kaplan-Meier survival curves were constructed, and the difference between treatments was statistically examined with a log-rank test. The Mann-Whitney U test was used as a nonparametric test for the WOSI score because of a lack of normality in the data.
Source of Funding
There was no external funding of this study.
Demographic Characteristics at Baseline
There was no difference between the randomized groups in terms of demographic characteristics (see Appendix). The mean age (and standard deviation) was 26.8 ± 7.1 years. Eighty-three patients (44.1%) were younger than twenty-five years old. All patients (with the exception of one 15.9-year-old patient) were between sixteen and forty years of age.
Two of ninety-three patients in both intervention groups had a positive sulcus sign. Eight patients in the internal rotation group and six patients in the external rotation group had other signs associated with generalized ligamentous laxity.
Follow-up Rate
Figure 2 shows the flow of the patients through each stage of the trial. One hundred and eighty-eight patients (153 men and thirty-five women) were allocated to treatment. Of the 188 patients who were enrolled in the study, 184 (97.9%) were followed for a minimum of two years. Four patients were lost to follow-up and were excluded from analysis: one had moved abroad, and three were not available. Five patients were enrolled and randomized, but not in accordance with the protocol. Two patients were managed with immobilization in internal rotation although they had been allocated to treatment in external rotation. Three patients (one in the internal rotation group and two in the external rotation group) had recurrent dislocation at the time when they were enrolled and randomized. These five patients are included in the intention-to-treat analysis according to randomization and in the per-protocol analysis according to actual treatment received. One hundred and seventy-four patients (92.6%) answered the follow-up questionnaire by mail, and seven patients (3.7%) answered by telephone. For three patients (1.6%), information on recurrence time was obtained from medical records. The average duration of follow-up was 29.1 months (28.9 months in the internal rotation group and 29.5 months in the external rotation group; range, twenty-four to fifty-four months).
Compliance
Forty-five patients (47.4%) in the internal rotation group and sixty-three patients (67.7%) in the external rotation group were registered as compliant (p < 0.05). Three patients (one in the internal rotation group and two in the external rotation group) were not included in the per-protocol analysis because they did not have primary dislocations.
Seventy-eight patients who were managed per protocol reported more than twenty-one hours of documented daily use of the brace or sling, including thirty-one (33.3%) of ninety-three in the internal rotation group and forty-seven (51.1%) of ninety-two in the external rotation group (p < 0.05).
One hundred and fifteen patients (61.2%) completed the time register questionnaire and returned it at three weeks of follow-up. The answers from the remaining patients were received within three months from fifty-three patients (28.2%), within one and a half years from thirteen patients (6.9%), and after two years from four patients (2.1%). Registration of compliance was missing for three patients (1.6%) (Table I).
Recurrence
Fifty-one patients (27.1%) had a recurrent shoulder dislocation within two years. The intention-to-treat analyses showed that the recurrence rate was 24.7% (twenty-three of ninety-three) in the internal rotation group and 30.8% (twenty-eight of ninety-one) in the external rotation group, with no significant difference between the groups (p = 0.36).The mean time to recurrence was 11.6 months (range, two to twenty-four months) in the internal rotation group and 10.5 months (range, one week to twenty-four months) in the external rotation group. Kaplan-Meier survival curves of the time of recurrence are shown in Figure 3. The log-rank test showed no difference between the groups.
In the per-protocol analysis, the recurrence rate was 21.7% (thirteen of sixty) in the external rotation group and 13.6% (six of forty-four) in the internal rotation group, with no significant difference between the groups (p = 0.30). One patient was lost to follow-up in the per-protocol analysis. We performed a sensitivity analysis with the assumption that the four patients who were lost to follow-up had recurrent dislocations. With this acceptance, we found that the recurrence rate was 26.3% (twenty-five of ninety-five) in the internal rotation group and 32.3% (thirty of ninety-three) in the external rotation group (p = 0.37). There was no significant difference between the internal and external rotation groups in terms of recurrence rates when the patients were stratified according to age (sixteen to twenty-four years or twenty-five to forty years) (see Appendix).
Table II shows the recurrence rates in the age groups of sixteen to twenty-two years, twenty-three to twenty-nine years, and thirty to forty years.
We also performed a per-protocol analysis that included only seventy-seven of the seventy-eight patients with more than twenty-one hours of documented daily use of the brace or sling; the remaining patient had been lost to follow-up. The recurrence rate was 19.6% (nine of forty-six) in the external rotation group and 16.1% (five of thirty-one) in the internal rotation group; this difference was not significant (p = 0.70).
Secondary Outcome Measures
One hundred and seventy-four patients (92.6%), including eighty-eight in the internal rotation group and eighty-six in the external rotation group, completed the Western Ontario Shoulder Instability Index (WOSI) questionnaire. One hundred and sixty-three patients (86.7%) answered the question as to whether they had experienced subluxation, and 169 patients (89.9%) answered the question concerning return to the preinjury level of physically activity with the affected shoulder.
The median WOSI score was 375 (interquartile range, 135 to 719) in the internal rotation group and 238 (interquartile range, 101 to 707) in the external rotation group. This difference was not significant (p = 0.32).
Sixty-seven (41.1%) of 163 patients reported that they had experienced subluxation, including 38.3% (thirty-one of eighty-one) in the external rotation group and 43.9% (thirty-six of eighty-two) in the internal rotation group. If recurrent instability was defined as either recurrent dislocation or subluxation, ninety (48.9%) of all 184 patients (forty-five in each intervention group) had recurrent instability; the difference between the groups was not significant (p = 0.89).
Seventeen of fifty-one patients with recurrent dislocation had surgery at the time of follow-up, including eight in the internal rotation group and nine in the external rotation group. Two patients with recurrent subluxation underwent surgery at the time of follow-up, including one in each group. Thirty (58.8%) of the fifty-one patients with recurrence had sustained more than two dislocations, including fourteen patients in the internal rotation group and sixteen in the external rotation group.
Fifty-two (60.5%) of eighty-six patients in the internal rotation group and fifty-one (61.4%) of eighty-three patients in the external rotation group had regained their previous level of physical activity with the injured shoulder.
Two complications were observed. One patient in the internal rotation group had diminished cutaneous sensation in the eighth cervical dermatome, and one patient in the external rotation group had hyperesthesia and moderate pain in his hand.
We found no difference in the rate of recurrent dislocation between patients randomized to three weeks of immobilization in 15° of external rotation and those randomized to conventional immobilization in internal rotation. The results were consistent in the intention-to-treat and per-protocol analyses and additional sensitivity analyses.
Few studies have tested the conclusions reported by Itoi et al.10, and the results of those studies differ. Seybold et al. found that immobilization in external rotation improved the position of the labroligamentous lesion on the glenoid rim16. Liavaag et al. concurred with this conclusion in an MRI study17. Miller et al. supported the theory of coaptation by reporting increased contact force for the labrum with increased external rotation18, whereas Limpisvasti et al. questioned the importance of coaptation19. However, the most important clinical issue is to determine if immobilization in external rotation significantly reduces the recurrence rate compared with conventional treatment.
Immobilization in internal rotation is an accepted treatment for primary shoulder dislocation, although the treatment benefit is not proven1-4. Itoi et al. concluded that immobilization in external rotation for three weeks reduced the risk of recurrence of a first-time traumatic anterior shoulder dislocation when compared with the risk associated with conventional immobilization in internal rotation10. On the contrary, in a recently published trial, Finestone et al. found no difference in the recurrence rate when comparing immobilization in internal rotation or external rotation20. To our knowledge, the studies by Itoi et al. and Finestone et al. were the only previously published randomized clinical studies comparing immobilization in internal and external rotation, and the conclusions of those studies differed.
We used a randomized design and included almost 200 patients. The study by Finestone et al.20 only included fifty-one male patients, all of whom were young and physically very active; that study population was unlike both that in the present study and that in the study by Itoi et al.10, which included 198 patients of both sexes who were twelve to ninety years old. Other strengths of the present study include the loss of only four patients to follow-up and the fact that >90% answered the follow-up questionnaire. We included almost equal numbers of patients in the two immobilization groups, and the groups were comparable in terms of sex, age, and other baseline characteristics. The mean age was ten years younger than that in the study by Itoi et al.10, and, with patients ranging from sixteen to forty years of age, the present study included age groups with high, medium, and low risk of recurrence. All of the primary dislocations were verified on radiographs. In contrast to the patients in the study by Itoi et al.10, all of our patients started the immobilization immediately or within twenty-four hours. Because subluxation is a rather vaguely defined entity, we distinguished between dislocation and subluxation and chose recurrent dislocation as the primary outcome measure.
The exact incidence of recurrent dislocation and instability is difficult to state and has varied in earlier reported studies. Some authors have reported true dislocations and others have grouped dislocations and subluxations together as instability. When studying patients managed surgically for recurrent dislocation, Moseley and MacArthur21, and later Hovelius et al.22, found that >90% of the primary dislocations had recurred within two years. In a review article, Robinson and Dobson estimated the risk of recurrent instability in unselected populations to be 75% to 80% among individuals between thirteen and twenty years of age and 50% among those between twenty and thirty years of age15. The rate of recurrence within two years in the present study is in agreement with the results reported for patients in the different age categories in the study by Hovelius23, in which the highest recurrence rate (53%) was among patients with an age of seventeen to nineteen years. This rate is very close to the 50.8% rate among the patients with an age of sixteen to twenty-two years in the present study. Even the percentage of patients with recurrent instability in the present study is in agreement with that reported by Robinson and Dobson for patients between twenty and thirty years of age. Thus, our findings indicate that we included a representative population and that results can be generalized.
To avoid any effect on our results related to differences in the degree of external rotation, the patients in the external rotation group were managed with immobilization in 15° of external rotation, which was more than the 10° used by Itoi et al.8,9,16-18,19,24.
We also included a patient-based quality-of-life assessment, the Western Ontario Shoulder Instability Index (WOSI). Although we did not assess the WOSI at baseline, we found no difference in shoulder-related quality of life between the groups at the time of follow-up.
Our study had some limitations. Like Itoi et al., we found significantly better compliance among the patients who were managed with immobilization in external rotation10. Finestone et al. reported 100% compliance in both groups, but their patients were managed within the military and 78% of the patients were soldiers20. Repeated instruction of the staff at all hospitals in how to treat and inform the patients was an integral part of that study. Oral and written information concerning the use of the sling or brace was given to all patients. We also informed the patients about the background and the aim of the study, including the preliminary results from the study by Itoi et al.10. Although the information was given as neutrally and objectively as possible, it is possible that more patients in the external rotation group had confidence in the benefit of that treatment. Hovelius et al.2 and Kralinger et al.4 found that immobilization in internal rotation did not prevent recurrence after the initial dislocation of the shoulder; therefore, the lower compliance reported by patients in the internal rotation group in the present study may not have influenced the observed recurrence rate. In the external rotation group in our study, the compliance rate (67.7%) was approximately the same as the compliance rate in the study by Itoi et al. (72%)10. Although the definitions of compliance were not identical, the results from both studies indicate that in a normal clinical setting the best practicably achievable compliance rate would be approximately two-thirds of the patients.
The definition of sixteen hours as a cutoff for compliance may be criticized, but the recurrence rate was not significantly different when only patients with more than twenty-one hours of immobilization were included. Itoi et al. asked the patients at the three-week examination how many hours a day and for how long they had worn the immobilizer10.
The optimum length of immobilization has not been clearly defined. Scheibel et al. concluded that immobilization in 30° of external rotation seems to allow a similar coaptation of the glenoid labrum regardless of the duration of immobilization25. Previous studies on patients managed with immobilization in internal rotation have indicated that the length of immobilization is not associated with the rate of recurrence1,3,5. To our knowledge, only one study has demonstrated an association between recurrence and the length of immobilization, with a lower rate of recurrence for patients managed with immobilization in internal rotation for less than one week as compared with those managed with immobilization for three weeks6.
Miller et al. reported increasing contact force between the labrum and the glenoid with increasing external rotation up to a maximum of 45°, but that study also suggested that a positive contact force was already achieved in neutral position18. However, Itoi et al. achieved their results with a 10° brace10. In a study on healthy subjects, Sullivan et al. tested four commercially available shoulder external rotation braces, including two rigid models and two soft cushion models26. All models were reported to achieve less external rotation than intended. The Don Joy UltraSling 15° model was found to achieve 9.4° on the average and was the best of the soft cushion braces. This result is still approximately equal to the best possible result that one would expect from the 10° brace used in the study by Itoi et al.
Robinson et al. performed a prospective cohort study of 252 patients with primary anterior shoulder dislocations and followed the patients for five years. They reported that 86.7% of the patients who were known to have recurrent instability developed this complication within the first two years27. Rowe reported that 70.5% of all recurrent shoulder dislocations occurred within the first two years5. Additional episodes of recurrence are expected after two years, but because our study is randomized and no significant difference in the recurrence rate between the groups was found within twenty-four months, we believe that this result will persist at later follow-up.
We conclude that immobilization in 15° of external rotation does not reduce the rate of recurrence for patients with first-time traumatic anterior shoulder dislocation.
One table presenting patient demographics and one table comparing recurrence rates in two age-stratified groups (sixteen to twenty-four years and twenty-five to forty years) are available with the online version of this article on our web site at jbjs.org.
Note: The authors thank the following individuals who participated in the patient enrollment for this study: Dr. Asgeir Furnes (Skien Hospital in Telemark [STHF] Orthopedic Department Skien), Dr. Svein Austdal and Dr. Pieter Oord (Stavanger University Hospital Orthopedic Department Stavanger), Dr. Kamaran Haji Raza and Dr. Ingve Frøytland (Sørlandet Hospital HF Kristiansand, Orthopedic Department Kristiansand), Dr. Knut Johannesen and Dr. Rune Moan (St. Olavs Hospital, Orthopedic Department Trondheim), Dr. Jorunn Holm and Dr. Ragnhild Øydna Støen (Akershus University Hospital Orthopedic Department and Lillestrøm legevakt), Dr. Thore Hinderaker (Sørlandet Hospital HF Flekkefjord, Orthopedic Department Flekkefjord), Dr. Joachim Thorkildsen (Bærum Hospital, Orthopedic Department Bærum), Dr. Geir Histøl and Dr. Marianne Olsson (Drammen Hospital, Orthopedic Department Drammen), Dr. Einar Eide and Dr. Jon Erik Engen (Tønsberg Hospital in Vestfold [SIV] Orthopedic Department Tønsberg), and Dr. Vibeke Hanch-Hansen (Kongsberg Hospital).
Henry
JH;
Genung
JA. Natural history of glenohumeral dislocation—revisited. Am J Sports Med.
1982;10:135-7.
[CrossRef][PubMed]
Hovelius
L;
Eriksson
K;
Fredin
H;
Hagberg
G;
Hussenius
A;
Lind
B;
Thorling
J;
Weckström
J. Recurrences after initial dislocation of the shoulder. Results of a prospective study of treatment. J Bone Joint Surg Am.
1983;65:343-9.
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.
Kralinger
FS;
Golser
K;
Wischatta
R;
Wambacher
M;
Sperner
G. Predicting recurrence after primary anterior shoulder dislocation. Am J Sports Med.
2002;30:116-20.
Rowe
CR. Prognosis in dislocations of the shoulder. J Bone Joint Surg Am.
1956;38:957-77.
Kiviluoto
O;
Pasila
M;
Jaroma
H;
Sundholm
A. Immobilization after primary dislocation of the shoulder. Acta Orthop Scand.
1980;51:915-9.
[CrossRef]
Itoi
E;
Hatakeyama
Y;
Urayama
M;
Pradhan
RL;
Kido
T;
Sato
K. Position of immobilization after dislocation of the shoulder. A cadaveric study. J Bone Joint Surg Am.
1999;81:385-90.
[CrossRef]
Itoi
E;
Sashi
R;
Minagawa
H;
Shimizu
T;
Wakabayashi
I;
Sato
K. Position of immobilization after dislocation of the glenohumeral joint. A study with use of magnetic resonance imaging. J Bone Joint Surg Am.
2001;83:661-7.
Itoi
E;
Hatakeyama
Y;
Kido
T;
Sato
T;
Minagawa
H;
Wakabayashi
I;
Kobayashi
M. A new method of immobilization after traumatic anterior dislocation of the shoulder: a preliminary study. J Shoulder Elbow Surg.
2003;12:413-5.
[CrossRef]
Itoi
E;
Hatakeyama
Y;
Sato
T;
Kido
T;
Minagawa
H;
Yamamoto
N;
Wakabayashi
I;
Nozaka
K. Immobilization in external rotation after shoulder dislocation reduces the risk of recurrence. A randomized controlled trial. J Bone Joint Surg Am.
2007;89:2124-31.
[CrossRef]
Burkhart
SS;
De Beer
JF. Traumatic glenohumeral bone defects and their relationship to failure of arthroscopic Bankart repairs: significance of the inverted-pear glenoid and the humeral engaging Hill-Sachs lesion. Arthroscopy.
2000;16:677-94.
[CrossRef]
Altman
DG. Practical statistics for medical research.
London: Chapman and Hall; 1991.
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.
Hovelius
L;
Lind
B;
Thorling
J. Primary dislocation of the shoulder. Factors affecting the two-year prognosis. Clin Orthop Relat Res.
1983;176:181-5.
Robinson
CM;
Dobson
RJ. Anterior instability of the shoulder after trauma. J Bone Joint Surg Br.
2004;86:469-79.
[CrossRef]
Seybold
D;
Schliemann
B;
Heyer
CM;
Muhr
G;
Gekle
C. Which labral lesion can be best reduced with external rotation of the shoulder after a first-time traumatic anterior shoulder dislocation?Arch Orthop Trauma Surg.
2008;129:299-304.
[CrossRef]
Liavaag
S;
Stiris
MG;
Lindland
ES;
Enger
M;
Svenningsen
S;
Brox
JI. Do Bankart lesions heal better in shoulders immobilized in external rotation?Acta Orthop.
2009;80:579-84.
[CrossRef]
Miller
BS;
Sonnabend
DH;
Hatrick
C;
O'leary
S;
Goldberg
J;
Harper
W;
Walsh
WR. Should acute anterior dislocations of the shoulder be immobilized in external rotation? A cadaveric study. J Shoulder Elbow Surg.
2004;13:589-92.
[CrossRef]
Limpisvasti
O;
Yang
BY;
Hosseinzadeh
P;
Leba
TB;
Tibone
JE;
Lee
TQ. The effect of glenohumeral position on the shoulder after traumatic anterior dislocation. Am J Sports Med.
2008;36:775-80.
[CrossRef]
Finestone
A;
Milgrom
C;
Radeva-Petrova
DR;
Rath
E;
Barchilon
V;
Beyth
S;
Jaber
S;
Safran
O. Bracing in external rotation for traumatic anterior dislocation of the shoulder. J Bone Joint Surg Br.
2009;91:918-21.
[CrossRef]
Moseley
HF;
MacArthur
HT. Recurrent dislocation of the shoulder.
Montreal: McGill University Press; 1961.
Hovelius
L;
Akermark
C;
Albrektsson
B;
Berg
E;
Körner
L;
Lundberg
B;
Wredmark
T. Bristow-Latarjet procedure for recurrent anterior dislocation of the shoulder. A 2-5 year follow-up study on the results of 112 cases. Acta Orthop Scand.
1983;54:284-90.
[CrossRef]
Hovelius
L. Incidence of shoulder dislocation in Sweden. Clin Orthop Relat Res.
1982;166:127-31.
Hart
WJ;
Kelly
CP. Arthroscopic observation of capsulolabral reduction after shoulder dislocation. J Shoulder Elbow Surg.
2005;14:134-7.
[CrossRef]
Scheibel
M;
Kuke
A;
Nikulka
C;
Magosch
P;
Ziesler
O;
Schroeder
RJ. How long should acute anterior dislocations of the shoulder be immobilized in external rotation?Am J Sports Med.
2009;37:1309-16.
[CrossRef]
Sullivan
LG;
Bailie
R;
Weiss
N;
Miller
BS. An evaluation of shoulder external rotation braces. Arthroscopy.
2007;23:129-34.
[CrossRef]
Robinson
CM;
Howes
J;
Murdoch
H;
Will
E;
Graham
C. Functional outcome and risk of recurrent instability after primary traumatic anterior shoulder dislocation in young patients. J Bone Joint Surg Am.
2006;88:2326-36.
[CrossRef]