Redislocation of the Shoulder During the First Six Weeks After a Primary Anterior Dislocation: Risk Factors and Results of Treatment

Robinson, C. M. BMedSci, FRCS(Ed)Orth; Kelly, M. MRCS; Wakefield, A. E. MSc, MCSP

Journal of Bone & Joint Surgery - American Volume:
Scientific Article

Background: After an anterior dislocation, shoulder instability may occur with disruption of the soft-tissue or osseous restraints, leading to early redislocation. The aim of the present study was to clarify the risk factors for this complication within the first six weeks after a first-time anterior traumatic dislocation and to assess the outcome of treatment with immediate operative stabilization.

Methods: A three-year, prospective, observational cohort study of 538 consecutive patients with a first-time anterior dislocation of the shoulder was carried out. Reassessment of shoulder function was performed at a dedicated shoulder clinic, and suspected early redislocations were assessed with additional radiographs. All medically fit patients with a confirmed acute redislocation were treated with repeat closed reduction under anesthesia. Patients with unstable reductions were treated operatively. Functional and radiographic assessment of outcome was carried out during the first year after dislocation.

Results: Seventeen (3.2%) of the 538 patients sustained an early redislocation within the first week after the original dislocation. Patients at increased risk of early redislocation included those who sustained the original dislocation as the result of a high-energy injury (relative risk = 13.7), those who had a neurological deficit (relative risk = 2.0), those in whom a large rotator cuff tear occurred in conjunction with the dislocation (relative risk = 29.8), those in whom the original dislocation was associated with a fracture of the glenoid rim (relative risk = 7.0), and those who had a fracture of both the glenoid rim and the greater tuberosity (relative risk = 33.5). Following operative reconstruction, the outcome at one year after the injury was favorable in terms of function, general health, and radiographic findings. None of the patients had a redislocation or symptoms of instability at one year.

Conclusion: All patients who have substantial pain, a visible shoulder deformity, or restriction of movement at one week after reduction of a first-time dislocation should be evaluated with repeat radiographs to exclude a redislocation. Patients in whom this complication develops usually have either (1) severe disruption of the soft-tissue envelope due to a large rotator cuff tear or (2) disruption of the normal osseous restraints to dislocation due to either an isolated fracture of the glenoid rim or fractures of both the glenoid rim and the greater tuberosity. Early operative stabilization is justified for patients in whom the dislocation is associated with these coexisting conditions and who have evidence of gross instability.

Author Information

C.M. Robinson, BMedSci, FRCS(Ed)Orth; M. Kelly, MRCS; A.E. Wakefield, MSc, MCSP; Edinburgh Orthopaedic Trauma Unit, The Royal Infirmary of Edinburgh, Lauriston Place, Edinburgh EH3 9YW, Scotland, United Kingdom. E-mail address for C.M. Robinson:

Article Outline

Recurrent dislocation of the shoulder frequently occurs in young individuals following an initial traumatic anterior dislocation 1,2. Most first-time dislocations are treated with immobilization of the shoulder in a position of internal rotation, neutral flexion, and neutral abduction for four to six weeks, during which time instability usually is not apparent. However, acute redislocation may occur during this period of immobilization as a result of severe disruption of the soft-tissue or osseous structures that stabilize the joint. Early recognition of acute glenohumeral redislocation is important because failure to treat this complication in a timely fashion can result in difficulties with reduction and possibly lead to a chronic dislocation 3.

The aim of the present study was to examine prospectively the risk factors for early redislocation within the first six weeks after injury in a nonselective consecutive cohort of patients with a first-time anterior glenohumeral dislocation. We also reviewed the results of treatment and the functional outcome at one year.

Back to Top | Article Outline

Materials and Methods

Prospective Cohort Study

During the three-year period from September 1996 to September 1999, a cohort of 538 patients with a first-time, traumatic, anterior shoulder dislocation were prospectively followed for one year at a dedicated shoulder clinic ( Table I ). The cohort was consecutive and nonselective: our affiliated emergency department provides the only acute musculoskeletal trauma service for the local adult population, and during the study period all patients with a dislocation were referred to our clinic for follow-up. In all patients, relocation of the first-time anterior glenohumeral dislocation had been confirmed with use of standard anteroposterior and modified axial 4 radiographs during the initial treatment. All patients were instructed to wear a sling for six weeks after dislocation.

Dislocations associated with Neer three-part, four-part, or head-splitting proximal humeral fractures 5 and those associated with severe glenoid fractures (Ideberg Type-II to V f ractures 6 ) were excluded from the study, as they were treated with immediate operative intervention. Patients who were not local residents of our catchment area and were unable to return for follow-up were also excluded.

The senior author (C.M.R.) assessed all patients at the time of the initial one-week appointment at the shoulder clinic. Patients with pain, visible shoulder deformity, or restriction of movement were evaluated with additional radiographs to assess for the possibility of redislocation. All patients underwent a functional and radiographic assessment at the six-week appointment. All acute redislocations were detected during the first week after the original injury; no missed redislocations were apparent on the radiographs that were made at six weeks. With the exception of one patient who had discarded his sling (Case 1), all redislocations occurred despite sling immobilization ( Table II ).

According to the standard protocol at our clinic, any patient who had evidence of rotator cuff weakness on physical examination 7 at the time of the one-week or six-week appointment was evaluated with use of ultrasound, magnetic resonance imaging, or arthroscopy. Patients with suspected neurological injury were evaluated with nerve-conduction studies.

Back to Top | Article Outline
Protocol for Treatment of Patients Who Had Acute Redislocation

Patients with a confirmed acute redislocation of the shoulder were treated with an additional closed reduction of the shoulder and an assessment of stability under general anesthesia with use of an image intensifier. A so-called soft relocation of the shoulder (easy reduction without a discrete jump or locking sensation as the shoulder was relocated) together with redislocation on external rotation or extension beyond 20 from an initial position of neutral abduction, rotation, and flexion of the shoulder were considered signs of gross instability. We adopted a policy of treating all medically fit patients who had these signs with immediate open operative stabilization under the same anesthetic without performing additional imaging studies.

The senior author performed all operative procedures through a deltopectoral approach to the shoulder. An attempt was made to restore stability to the shoulder by means of a sequential repair of the injured soft-tissue or osseous restraints. Mitek suture anchors mounted with Ethibond sutures (Ethicon, Edinburgh, United Kingdom) were used to repair all capsulolabral or rotator cuff tears and comminuted osseous avulsion fragments that were unsuitable for screw fixation. Cannulated screws (Synthes, London, United Kingdom) were used to reattach larger osseous fragments. The intraoperative criterion used for satisfactory restoration of shoulder stability after reconstruction was stability of the shoulder throughout a full passive range of movement.

All patients who had a redislocation were managed with immobilization of the shoulder in a sling in the so-called safe position for six weeks after the operation. Active assisted range-of-movement exercises were begun at two weeks after the operation, with the patient avoiding abduction of the shoulder beyond 90 or external rotation beyond the neutral position. Isometric rotator cuff exercises and graduated active range-of-movement exercises performed under the supervision of a physical therapist were commenced after removal of the sling and were continued for at least six months after the operation.

Back to Top | Article Outline
Functional Assessment

At one year, all patients underwent a functional assessment of the shoulder with use of the score of Constant and Murley 8 ; the Disabilities of the Arm, Shoulder and Hand (DASH) limb-specific questionnaire 9 ; and the Short Form-36 (SF-36) general health-status questionnaire 10,11. All patients were directly questioned about symptoms of shoulder instability at each follow-up appointment.

Back to Top | Article Outline
Statistical Analysis

All information gathered during the study was recorded and analyzed with use of the SPSS software package (SPSS, Chicago, Illinois). To compare the patients who had a redislocation with the remainder of the cohort, the putative risk factors for early redislocation were subjected to statistical evaluation with use of the Mann-Whitney U test for numeric data and the chi-square or Fisher exact test for categorical data. The SF-36 scores for patients who had a redislocation were compared with those for the remainder of the cohort and those for age-matched controls in the general population 11. A p value of ≤0.05 was considered significant.

Back to Top | Article Outline


Seventeen (3.2%) of the 538 patients in the cohort sustained an acute redislocation. The majority of these patients were middle-aged or elderly, and the sex distribution was approximately equal ( Table II ). The original dislocations in this group were more commonly produced by high-energy injury (relative risk = 13.7; 95% confidence interval, 5.0 to 38.0; p < 0.05) compared with those in the remainder of the cohort and typically were the result of traction injuries sustained during a fall from a height or during an assault ( Table I ). Patients who sustained a redislocation also had a higher prevalence of nerve palsy after the first dislocation (relative risk = 2.0; 95% confidence interval, 1.1 to 4.0; p < 0.05). Four of the seventeen patients had axillary nerve dysfunction, and two had a palsy of the lower trunk of the brachial plexus. On the basis of the clinical, radiographic, and operative findings, two groups of patients were identified: those without an associated fracture (Group I) and those with an associated fracture (Group II).

Back to Top | Article Outline
Patients with an Isolated Dislocation without an Associated Fracture (Group I)

Group I consisted of seven patients (Cases 1 through 7) who did not have an associated fracture ( Table II ). One patient (Case 1) sustained a redislocation during an early reinjury; the shoulder was stable following closed relocation and was not managed operatively. All of the other redislocations were unstable following closed reduction. One elderly female patient (Case 2) was treated with closed reduction alone, as the risks associated with open reconstructive surgery were deemed to be high because of medical comorbidities. She had clinical evidence of a large rotator cuff tear and subsequently had a second redislocation with development of a chronic dislocation.

All five patients who were treated operatively had a massive rotator cuff tear that propagated through the supraspinatus and infraspinatus tendons at their humeral insertions, with dislocation of the humeral head through a rent in the rotator interval. Three patients had a tear that propagated into the inferiorly displaced subscapularis tendon at its humeral attachment, whereas the other two had an avulsion of the inferior aspect of the capsulolabral complex from the anterior part of the glenoid. In patients who had a dislocation without an associated fracture, the presence of a rotator cuff tear was therefore associated with a much higher risk of redislocation (relative risk = 29.8; 95% confidence interval, 3.7 to 244.1; p < 0.05). Satisfactory stability was restored in all five patients following reattachment of the capsulolabral tears, rotator cuff repair, and closure of the rotator interval.

Back to Top | Article Outline
Patients with an Initial Fracture-Dislocation (Group II)

Group II consisted of ten patients (Cases 8 through 17) who had an initial fracture-dislocation of the shoulder ( Table II ). None of the eighty-one patients with an isolated tuberosity fracture had an acute redislocation within the first six weeks. Redislocation occurred in three of the sixteen patients who had an isolated Ideberg Type-Ia glenoid rim fracture and in seven of the eleven patients who had a bifocal fracture involving both the glenoid rim and the greater tuberosity. The risk of redislocation was therefore increased both in patients with an isolated glenoid fracture (relative risk, 7.0; 95% confidence interval, 2.2 to 21.9; p < 0.05) and in those with a bifocal fracture (relative risk = 33.5; 95% confidence interval, 15.7 to 71.7; p < 0.05).

All shoulders in this group were grossly unstable on assessment with the patient under anesthesia following relocation, and all were treated operatively. In all cases, the anteroinferior glenoid rim fracture fragment, along with its attached inferior glenohumeral ligament, was displaced inferiorly. The humeral head was dislocated through a rent in the rotator interval. The greater tuberosity fracture fragment (when present) was displaced posterosuperiorly ( Fig. 1 ).

During the operative procedure, internal fixation of the glenoid fracture was carried out initially, with restoration of stability in all patients with an isolated glenoid fracture and in four patients with a bifocal fracture. In the remaining three patients, stability was not restored following glenoid fixation, and internal fixation of the greater tuberosity was carried out ( Figs. 2-A , 2-B, 2-C , 2-D , and 2-E ). The rotator interval lesion was closed with use of nonabsorbable sutures in all cases.

Back to Top | Article Outline
Complications and Functional and Radiographic Outcomes

At one year, one of the two patients in Group I who had been managed nonoperatively (Case 1) had not had a redislocation and had no functional deficit or symptoms of instability. The other patient in Group I who had been managed nonoperatively (Case 2) had a chronic dislocation and poor functional scores but had only mild pain ( Table II ). Of the five patients in Group I who had been managed operatively, one (Case 3) had development of inferior subluxation of the humeral head at six weeks. Arthroscopy revealed a rerupture of the subscapularis portion of the repair, and an open repair of the rerupture was carried out. Her functional scores improved, although a partial inferior subluxation of the humeral head was visible on radiographs made at one year.

All of the remaining patients who had been managed operatively (Cases 4 through 17) recovered uneventfully. At one year, all of these patients had satisfactory functional scores ( Table II ), none had symptoms of instability, and all had a congruous glenohumeral joint as seen on radiographs. The motor component of all nerve palsies resolved spontaneously, although two patients had residual paresthesias in the cutaneous distribution of the axillary nerve.

At one year, the functional (Constant and Murley 8 and DASH 9 ) scores for the patients who had a redislocation were not significantly different from those for the remainder of the cohort after correction for age and sex. The patients who had a redislocation also were not significantly different from the remainder of the cohort or from age-matched controls in the general population with regard to any of the scores on the eight subcategories of the SF-36 questionnaire.

Back to Top | Article Outline


Acute redislocation of the shoulder is an important, if poorly recognized, complication that occurs following a first-time dislocation. The prevalence of this complication in the present study was 3.2%. In general, this dislocation occurred predominantly in middle-aged and elderly individuals who sustained the initial dislocation following a relatively high-energy injury. In most patients, there was evidence of a traction injury during a fall, resulting in anteroinferior dislocation and a higher occurrence of axillary nerve and brachial plexus injury.

Although this complication has not previously been described after anterior dislocation, there have been reports of chronic anterior dislocations 3,12,13, which may occur as a late presentation of a missed injury, as the result of neglect of an unstable dislocation, or following an early, unrecognized redislocation. As in the present study, patients with chronic dislocations are often middle-aged or elderly 12,13 and often have associated glenoid fractures 13,14. It is possible that many patients who have a chronic dislocation have been omitted from previous studies of anterior dislocation. It is also likely that other studies of anterior dislocation may have excluded a number of patients with unstable fracture-dislocations who were treated with primary operative stabilization.

Acute instability following anterior dislocation appears to occur in association with either severe disruption of the capsulolabral complex and rotator cuff envelope or fractures that compromise the normal osseous restraints to dislocation. All of these features were associated with a much higher relative risk of redislocation in our study. There is considerable surface area mismatch between the two articular surfaces in the shoulder, and anterior and inferior fractures of the glenoid are known to be associated with an increased risk of instability 1,15. However, articular geometry and osseous congruency are usually considered less important than soft-tissue factors in stabilizing the glenohumeral joint. The displacement of the inferior glenohumeral ligament, which was encountered in two of the patients in Group I and in all of the patients in Group II, was probably a contributory cause of the acute instability. In addition, injury to the rotator cuff, resulting either from the massive tears in Group I or from fractures of the greater tuberosity in Group II, probably had an additional destabilizing effect on the shoulder.

The reported occurrence of glenoid rim fractures following anterior dislocation has varied considerably, and such fractures are often difficult to identify on conventional shoulder radiographs without specialized imaging 15,16. The indications for primary operative internal reduction and fixation of glenoid fractures are controversial 17-19, and biomechanical studies are lacking. In the present series, none of the Ideberg Type-I fractures were treated with primary operative intervention and, although these fractures were associated with a higher risk of early redislocation, the majority of patients did not have this complication.

Fracture of the greater tuberosity is the most common fracture associated with anterior dislocation of the shoulder, and it normally reduces the risk of subsequent shoulder instability 1,20,21. In the present study, all dislocations that were associated with an isolated tuberosity fracture were stable during the first six weeks. However, redislocation occurred in the majority of patients who had a tuberosity fracture combined with a glenoid fracture. Examination under anesthesia, with progression to immediate stabilization in the presence of an unstable reduction, should be considered for patients with these bifocal fracture-dislocations.

Rotator cuff injury is commonly associated with anterior dislocation in the elderly 22-25 and predominantly affects the tendons attached to the greater tuberosity 26. The association of rotator cuff tear with recurrent instability is well recognized 23,25, and patients with this injury pattern often have avulsion of the subscapularis and the anterior part of the capsule from the lesser tuberosity rather than a Bankart lesion. The large tears involving the entire cuff that were encountered in the operatively treated Group-I patients are relatively uncommon 23, and their association with acute redislocation has not previously been described.

Routine examination with the patient under anesthesia following anterior dislocation of the shoulder was not undertaken in this study. We are therefore unable to evaluate the sensitivity of our criteria for assessing acute shoulder instability or the prevalence of similar signs in all patients following an acute dislocation. However, the occurrence of redislocation despite sling immobilization in our series could in itself be seen as a sign of severe primary instability. None of the shoulders that were well reduced at one week dislocated during the subsequent six weeks, suggesting that most were stable following the initial reduction.

Early stabilization following the initial redislocation was successful in the majority of our patients. An unstable reduction of the shoulder is difficult to manage nonoperatively with immobilization. A poor functional return could have been anticipated given the large rotator cuff lesions that were encountered at the time of the operation in most of our patients. Longer follow-up is required to assess the risk of rotator cuff and degenerative joint disease in our group of operatively-managed patients.

Investigation performed at The Royal Infirmary of Edinburgh, Edinburgh, Scotland, United Kingdom

The authors did not receive grants or outside funding in support of their research or preparation of this manuscript. They did not receive payments or other benefits or a commitment or agreement to provide such benefits from a commercial entity. No commercial entity paid or directed, or agreed to pay or direct, any benefits to any research fund, foundation, educational institution, or other charitable or nonprofit organization with which the authors are affiliated or associated.

A commentary is available with the electronic versions of this article, on our web site ( and on our quarterly CD-ROM (call our subscription department, at 781-449-9780, to order the CD-ROM).

1. Rowe CR. Prognosis in dislocations of the shoulder. J Bone Joint Surg Am, 1956;38: 957-77.
2. Hovelius L, Eriksson K, Fredin H, Hagberg G, Hussenius A, Lind B, Thorling J, Weckstrom J. Recurrences after initial dislocation of the shoulder. Results of a prospective study of treatment. J Bone Joint Surg Am, 1983;65: 343-9.
3. Schulz TJ, Jacobs B, Patterson RL Jr. Unrecognized dislocations of the shoulder. J Trauma, 1969;9: 1009-23.
4. Wallace WA, Hellier M. Improving radiographs of the injured shoulder. Radiography, 1983;49: 229-33.
5. Neer CS 2nd. Displaced proximal humeral fractures. I. Classification and evaluation. J Bone Joint Surg Am, 1970;52: 1077-89.
6. Ideberg R. Fractures of the scapula involving the glenoid fossa. In: Bateman JE, Walsh RP, editors. Surgery of the shoulder. Philadelphia: BC Decker; 1984. p 63-6.
7. Pisan M, Gerber C. Clinical examination of the rotator cuff. Curr Orthop, 2000;14: 155-60.
8. Constant CR, Murley AH. A clinical method of functional assessment of the shoulder. Clin Orthop, 1987;214: 160-4.
9. Hudak PL, Amadio PC, Bombardier C. Development of an upper extremity outcome measure: the DASH (disabilities of the arm, shoulder and hand) [corrected]. The Upper Extremity Collaborative Group (UECG). Am J Ind Med, 1996;29: 602-8. Erratum 1996;30:372.
10. Ware JE Jr, Snow KK, Kosinkski M, Gandek B. SF-36 health survey: manual and interpretation guide. Boston: The Health Institute, New England Medical Center; 1993.
11. Jenkinson C, Layte R, Wright L, Coulter A. The UK SF-36: an analysis and interpretation manual. Oxford: Health Services Research Unit; 1996.
12. Rowe CR, Zarins B. Chronic unreduced dislocations of the shoulder. J Bone Joint Surg Am, 1982;64: 494-505.
13. Flatow EL, Miller SR, Neer CS 2nd. Chronic anterior dislocation of the shoulder. J Shoulder Elbow Surg, 1993;2: 2-10.
14. Kummel BM. Fractures of the glenoid causing chronic dislocation of the shoulder. Clin Orthop, 1970;69: 189-91.
15. Bigliani LU, Newton PM, Steinmann SP, Connor PM, McIlveen SJ. Glenoid rim lesions associated with recurrent anterior dislocation of the shoulder. Am J Sports Med, 1998;26: 41-5.
16. Runkel M, Kreitner KF, Wenda K, Rudig L, Degreif J, Grebe P. [Nuclear magnetic tomography in shoulder dislocation]. Unfallchirurg, 1993;96: 124-8. German.
17. DePalma AF. Surgery of the shoulder. 3rd ed. Philadelphia: JB Lippincott; 1983. p 366-7.
18. Goss TP. Fractures of the glenoid cavity. J Bone Joint Surg Am, 1992;74: 299-305.
19. Papagelopoulos PJ, Koundis GL, Kateros KTw2, Babis GC, Nikolopoulos KE, Fragiadakis EG. Fractures of the glenoid cavity: assessment and management. Orthopedics, 1999;22: 956-61.
20. Rowe CR, Sakellarides HT. Factors related to recurrences of anterior dislocations of the shoulder. Clin Orthop, 1961;20: 40-8.
21. Mason BJ, Kier R, Bindleglass DF. Occult fractures of the greater tuberosity of the humerus: radiographic and MR imaging findings. AJR Am J Roentgenol, 1999;172: 469-73.
22. Hawkins RJ, Bell RH, Hawkins RH, Koppert GJ. Anterior dislocation of the shoulder in the older patient. Clin Orthop, 1986;206: 192-5.
23. Neviaser RJ, Neviaser TJ, Neviaser JS. Anterior dislocation of the shoulder and rotator cuff rupture. Clin Orthop, 1993;291: 103-6.
24. Sonnabend DH. Treatment of primary anterior shoulder dislocation in patients older than 40 years of age. Conservative versus operative. Clin Orthop, 1994;304: 74-7.
25. Gumina S, Postacchini F. Anterior dislocation of the shoulder in elderly patients. J Bone Joint Surg Br, 1997;79: 540-3.
26. Harryman DT 2nd, Mack LA, Wang KY, Jackins SE, Richardson ML, Matsen FA 3rd. Repairs of the rotator cuff. Correlation of functional results with integrity of the cuff. J Bone Joint Surg Am, 1991;73: 982-9.
Copyright 2002 by The Journal of Bone and Joint Surgery, Incorporated