 |
 |
 |
|
| home | CME | help | feedback | subscriptions | archive |
|
© 1996 The Journal of Bone and Joint Surgery, Inc. Primary Anterior Dislocation of the Shoulder in Young Patients. A Ten-Year Prospective Study* , GÄVLE , ÖREBRO , MALMÖ*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.
Two hundred and forty-five patients who had had 247 primary anterior dislocations of the shoulder were followed for ten years in a multicenter study at twenty-seven Swedish hospitals. The ages of the patients at the time of the dislocation ranged from twelve to forty years. The patients were assigned to one of three treatment groups: immobilization with the arm tied with a bandage to the torso for three to four weeks after reduction of the dislocation; use of a sling, which was discontinued after the patient was comfortable; or immobilization for various durations. At the ten-year follow-up evaluation, no additional dislocation had occurred in 129 shoulders (52 per cent). Recurrent dislocation necessitating operative treatment had developed in fifty-eight shoulders (23 per cent): thirty-four (34 per cent) of the ninety-nine shoulders in patients who were twelve to twenty-two years old, sixteen (28 per cent) of the fifty-seven shoulders in patients who were twenty-three to twenty-nine years old, and eight (9 per cent) of the ninety-one shoulders in patients who were thirty to forty years old. Twenty-four (22 per cent) of the shoulders that had had at least two recurrences during the first two or five years seemed to have stabilized spontaneously without operative intervention at ten years. Dislocation of the contralateral shoulder occurred in association with sixteen (16 per cent) of the ninety-nine shoulders in patients who were twelve to twenty-two years old, twelve (21 per cent) of the fifty-seven shoulders in patients who were twenty-three to twenty-nine years old, and only three (3 per cent) of the ninety-one shoulders in patients who were thirty to forty years old. The type and duration of the initial treatment had no effect on the rate of recurrence. Radiographs, made for 185 shoulders at the time of the primary dislocation, demonstrated an evident Hermodsson (Hill-Sachs) lesion in ninety-nine shoulders (54 per cent); this finding was associated with a significantly worse prognosis with regard to recurrence than was no evident lesion (p < 0.04). Radiographs made for 208 shoulders at the ten-year follow-up examination were evaluated for post-dislocation arthropathy. Twenty-three shoulders (11 per cent) had mild arthropathy and eighteen (9 per cent) had moderate or severe arthropathy. Some of the shoulders that had arthropathy had had no recurrence.
We previously described the two and five-year results of treatment of primary anterior dislocation in 257 shoulders (255 patients)10,12-14. The same cohort constitutes the basis for the present study, with the duration of follow-up extended to ten years for 247 shoulders (245 patients). (Nine patients died and one was lost to follow-up.) This prospective clinical and radiographic study was performed to provide additional information with regard to the results of treatment of primary anterior dislocation of the shoulder in younger patients (those twelve to forty years old).
In 1977, one of us (L. H.) initiated a prospective study to evaluate the results of treatment of primary anterior dislocation of the glenohumeral joint in patients who were forty years old or less10,14. Twenty-seven Swedish hospitals participated in the study. After reduction with manipulation, the dislocations were treated either with immobilization with the arm tied with a bandage to the torso for three to four weeks (group 1) or with the limb held in a sling until the patient was comfortable (group 2). At the larger institutions, the patients were assigned randomly to these two groups on the basis of the date of the injury. At each of the smaller institutions, they were assigned randomly to only one method of treatment. Early in the study, a third group (group 3) was established; this group included twenty patients who could not follow the assigned protocol, ten who had received inaccurate instructions, two who were alcoholics, three who had had reduction of the dislocation more than twenty-four hours after the injury, two who also needed operative treatment for a displaced fracture of the greater tuberosity, two who had an epileptic disorder, and two who had had a suspected subluxation before the primary dislocation. Of the original 257 dislocations, 205 (80 per cent) were in male patients and fifty-two (20 per cent) were in female patients. The etiology of the primary dislocation was related to a sports activity in fifty-eight (57 per cent) of the 102 shoulders in the patients who were twelve to twenty-two years old, thirty-four (57 per cent) of the sixty shoulders in the patients who were twenty-three to twenty-nine years old, and thirty-six (38 per cent) of the ninety-five shoulders in the patients who were thirty to forty years old. The dislocation was considered spontaneous when the trauma had been so trivial that the shoulder should not have dislocated, as in throwing a ball, missing a ball in tennis, or swimming. A spontaneous dislocation occurred in fourteen (14 per cent) of the shoulders in the patients who were twelve to twenty-two years old, three (5 per cent) of the shoulders in the patients who were twenty-three to twenty-nine years old, and one (1 per cent) of the shoulders in the patients who were thirty to forty years old. At the ten-year follow-up examination, individual parameters that had been noted on the radiographs at the time of the injury were evaluated with regard to their prognostic importance10,13,14. The position of the humerus in relation to the glenoid was documented. An abducted position of the humeral head was termed luxatio abducta (Fig. 1). This finding was present in nineteen of the 140 shoulders that were in the dislocated position when the radiographs were made. Two years after the injury, these patients were contacted by telephone and were asked to recall the position of the arm at the time of the dislocation. Five patients did not recall the position to be other than "hanging down," as is generally described by a patient who has an anterior dislocation. However, fourteen patients described a clear clinical picture of an erecta or abducta position, with use of terms such as "not possible to get the arm down," "locked upward," and "had to rest the arm on the head." The results for these fourteen shoulders were evaluated at ten years and were compared with those for the remaining shoulders in which the humeral position had not been unusual at the time of the dislocation.
We also analyzed the occurrence of fracture of the greater tuberosity in conjunction with the primary dislocation. Thirty-two shoulders had had this fracture, which was more common in patients who were less than fifteen years old and in those who were more than thirty years old14. The presence of a posterior defect of the humeral head (a Hermodsson7 or Hill-Sachs8 lesion) when the primary dislocation occurred14 was analyzed with respect to the ten-year result. Of the 257 shoulders, ten were lost to follow-up before ten years. One man had died between the two and five-year evaluations, eight men had died between the five and ten-year evaluations, and one woman had left the country. Thus, the results of the current study are based on 247 shoulders (245 patients) who were available for the ten-year follow-up evaluation. One hundred and thirty-one patients had a dislocation of the left shoulder, 112 had a dislocation of the right shoulder, and two had a dislocation of both shoulders. One hundred and thirty-one dislocations involved the non-dominant shoulder and 113 involved the dominant shoulder; the three remaining shoulders were in patients who were ambidextrous. The ten-year follow-up evaluation consisted of a personal interview, a physical examination, and radiographs of both shoulders. The patients were questioned as to history of pain, function of the shoulder, recurrence, dislocation of the contralateral shoulder, and any operative treatment that they had had because of symptoms of glenohumeral instability. The indications for operative intervention varied, as the dislocations were treated at many hospitals; however, the decision to operate was always made at the surgeon's discretion on the basis of the patient's subjective assessment of symptoms referable to the glenohumeral instability. Patients who had had no recurrence at ten years were questioned as to whether or not they considered the shoulder to be stable. Thirty-four patients refused to come to the hospital for the follow-up interview, so they were evaluated with use of a questionnaire, which was either sent by mail or administered on the telephone. Two hundred and eleven patients (213 shoulders) had a physical examination as well as a follow-up interview. However, as many physicians were involved in performing the physical examinations, these data were not consistent and consequently are not reported. One hundred and eighty-nine patients had ten-year radiographs of both shoulders and nineteen, of only the involved side. Two anteroposterior radiographs, one lateral radiograph, and one axillary radiograph were made for each shoulder. The radiographs were examined by one of us (L. H.) and by one independent observer. Arthropathy that developed after the dislocation was graded, according to the system described by Samilson and Prieto29, as mild (osteophytes less than three millimeters in size on the humeral head [Fig. 2-A]), moderate (osteophytes between three and seven millimeters on the humeral head or the glenoid rim [Fig. 2-B]), or severe (osteophytes of more than seven millimeters [Fig. 2-C], with or without articular incongruity [Fig. 3]). (The latter category included all shoulders in which the articular joint space was broader superiorly than inferiorly and the articular margins of the humeral head and the glenoid were not parallel.)
Several categories were established to define the status of the shoulder at the follow-up evaluations. We used the term primary dislocation to describe the initial dislocation of the previously healthy shoulder. If no additional dislocation was noted at the time of follow-up, the dislocation was considered solitary. The term recurrent dislocation was used when at least two dislocations had occurred after the primary event. Shoulders that had had recurrent dislocation at the two or five-year follow-up evaluation but had had no additional dislocation at the ten-year evaluation were considered to have stabilized spontaneously. The term occasional dislocation was used when one or two dislocations had occurred between each of the follow-up intervals but the patient considered the shoulder to have functioned normally during these periods. The term subluxation was used when there had been at least one dislocation followed by immediate, spontaneous reduction. Patients who had also had a dislocation or subluxation of the contralateral shoulder either before or after the dislocation that had brought them into this study were characterized as having had bilateral dislocation. Statistical analysis was performed with use of the chi-square test with Yates correction. A value of p < 0.05 was considered significant.
Ten years after the initial dislocation, 129 (52 per cent) of the 247 shoulders had had no additional episodes of dislocation, although eight of these shoulders were considered by the patient to be unstable. Eleven shoulders (4 per cent) had had one recurrence of the dislocation. Fifty-eight shoulders (23 per cent) had had an operative procedure for recurrent dislocation. The remaining forty-nine shoulders (20 per cent) were also classified as having recurrent dislocation but had not been treated operatively, although two were scheduled for operative treatment. Radiographs revealed mild arthropathy after the dislocation in twenty-three shoulders, moderate arthropathy in sixteen, and severe arthropathy in two (Table I). Of the eighteen shoulders that had moderate or severe arthropathy, twelve had had recurrent dislocation and six had not. With the numbers available, this difference was not significant.
Articular incongruity was seen on the radiographs of thirty-two shoulders (Fig. 3), twenty-three of which also had evidence of arthropathy (p < 0.0001). These twenty-three shoulders had had at least one recurrent dislocation, compared with seventy-nine of the 176 that had a congruent glenohumeral joint (p = 0.01). Of the 189 patients who had follow-up radiographs bilaterally, twenty-four had dislocation or subluxation of the contralateral shoulder and four of these twenty-four contralateral shoulders had evidence of moderate or severe arthropathy. Of the 165 stable contralateral shoulders, only three had evidence of mild arthropathy on the radiographs. This difference was significant (p < 0.001). The patient's subjective assessment of function of the shoulder did not differ significantly, with the numbers available, when the shoulders with and without arthropathy were compared.
Treatment
Radiographs Made at the Time of the Primary Dislocation Ninety-nine (54 per cent) of the 185 shoulders that were evaluated radiographically at the time of the primary dislocation and were followed for ten years had a posterior defect of the humeral head (a Hermodsson7 or Hill-Sachs8 lesion). Of these ninety-nine shoulders, sixty redislocated at least once during the ten-year follow-up period, compared with thirty-eight of the eighty-six that did not have such a lesion (p < 0.04), and fifty-one redislocated at least twice (recurrent dislocation), compared with thirty-seven that had no evident lesion. With the numbers available, this difference was not significant (p = 0.31). Eleven of the fourteen shoulders in which the humeral head had been in an erecta abducta position at the time of the primary dislocation (Fig. 1) had had at least one redislocation during the ten-year follow-up period. Seven of these fourteen had had an operation, a higher prevalence than in the group in which the arm had not been positioned at the side; however, with the numbers available, this difference was not significant (p = 0.08).
Age
Type of Trauma
Gender
Dislocation of the Contralateral Shoulder (Bilateral Dislocation)
Handedness
Operative Treatment
Stabilization by the Time of the Ten-Year Follow-up Evaluation
The patients were questioned with regard to their assessment of recovery of the dislocated shoulder at ten years. Of the 247 shoulders, 202 (82 per cent) were considered by the patient as having had complete or almost complete recovery of function. The corresponding values for the shoulders that had been treated operatively and for those that had not had recurrence were 81 per cent (forty-seven of fifty-eight) and 94 per cent (121 of 129). Seventeen (71 per cent) of twenty-four shoulders that had had recurrent dislocation but had stabilized spontaneously without operative treatment were considered by the patient as having had complete or almost complete recovery, compared with nine (39 per cent) of twenty-three shoulders that had had redislocation without operative treatment between the five and ten-year follow-up evaluations. Twelve of the thirteen shoulders that had stabilized over time with no additional symptoms of instability were considered by the patient as having normal or almost normal function, compared with only five of eleven that had stabilized over time but were subjectively considered unstable.
The results of this ten-year follow-up study shed light on the treatment and prognosis for these relatively common injuries. Perhaps most importantly, our findings suggest that the prognosis after a primary glenohumeral dislocation, even in younger patients, is not as devastating as was previously believed1,18,26,27. Furthermore, it should be emphasized that some shoulders with recurrent dislocation become stable over time. Three or four weeks of immobilization after reduction of the primary dislocation did not appear to influence the outcome in our patients. This finding is in contrast to the common opinion16,21,27,30 that immobilization should be used for three to six weeks after a primary dislocation. However, it is in agreement with the results of other studies1,5,18. It has been generally accepted that age and prognosis are related. In the current study, one-third of the patients who had had the initial dislocation before the age of thirty years needed operative stabilization (Fig. 4). Previous studies have shown higher rates of recurrence in younger patients. McLaughlin and MacLellan retrospectively compared two clinical groups—one comprising 265 shoulders that needed operative treatment for recurrent dislocation and the other, 315 shoulders that had not had recurrent dislocation—and found that 95 per cent of 181 primary dislocations in patients who were eleven to twenty years old recurred18. Rowe, in a series of 488 patients who had both primary and recurrent dislocation, found an 83 per cent rate of recurrence in the 107 shoulders of patients who were less than twenty years old26. Our study, in contrast to most others, was prospective, and the duration of follow-up was similar for all patients. Symptoms of recurrent instability necessitating operative treatment were evident at the five-year evaluation in patients who were younger than twenty-six years, whereas in those who were between twenty-six and forty years old the need for operative treatment increased during the five to ten-year follow-up period. The reason for the high number of shoulders (twenty-four [22 per cent] of 107) that stabilized spontaneously in the current study is unclear. Diminished outward rotation with increased age28, decreased physical activity, constitutional differences with regard to proprioception of the joint17, and post-dislocation arthropathy29 may influence the prognosis in this respect. Although about one-third of the primary dislocations in our patients who were twenty-nine years old or less needed operative treatment because of recurrence or remaining instability within ten years, the results of the current study do not support a recommendation for routine prophylactic operative treatment, even for patients in the youngest age-groups. O'Driscoll and Evans reviewed the results for 188 patients one to twenty years after operative treatment for anterior instability of the shoulder23. They found an over-all prevalence of bilateral involvement of 24 per cent, which is higher than the over-all rates in our study although similar to the rate in our twenty-three to twenty-nine-year-old age-group (twelve [21 per cent] of fifty-seven shoulders). We agree with O'Driscoll and Evans that the occurrence of bilateral instability should be reported in future studies because it might clarify the etiology of this disorder. In Rowe's series of shoulder dislocations, the non-dominant shoulder had a slightly (but not significantly) higher prevalence of dislocation26. We also noted a higher rate of involvement of the non-dominant shoulder; however, as in Rowe's series, this difference was not significant. Furthermore, there was no relationship between handedness and recurrence in either study. This is contrary to the findings in a study of Swedish ice-hockey players, in which left-grip (right-handed) players more often had operative treatment of the left shoulder and right-grip (left-handed) players more often had operative treatment of the right shoulder9. Evidence of a defect of the humeral head7,8 on the radiograph after the primary dislocation was associated with a higher rate of recurrence (one or more). This is in accordance with Rowe's finding that 50 per cent of seventy-eight shoulders without a defect of the humeral head had recurrent dislocation, compared with 83 per cent of forty-seven shoulders with such a defect26. It is possible that in our study the routine radiographs that were made after the primary dislocation were not optimum with regard to the identification of a defect of the humeral head. Thus, the prevalence of this lesion may well have been higher if the shoulders had been evaluated arthroscopically or with computerized tomography. The development of post-dislocation arthropathy was described by Samilson and Prieto29. They believed that the prevalence of osteoarthrosis after dislocation of the shoulder was unknown and that the long-term results after different operative procedures had not been well defined. Neer et al. reported the results of 273 total shoulder replacements and found that 10 per cent of the shoulders had had operative treatment for osteoarthrosis after recurrent dislocation22. Our finding that twenty-three (11 per cent) of 208 shoulders had mild arthropathy and eighteen (9 per cent) had moderate or severe arthropathy may be somewhat surprising. Even more remarkable is the fact that shoulders that had had one recurrence had approximately the same degree of arthropathy as was noted in the shoulders that had had recurrent or operatively treated dislocation (Table I). Singer et al., in a long-term follow-up study of the results of fourteen Bristow procedures, found mild arthropathy in six patients, moderate arthropathy in one, and severe arthropathy in three31. They postulated that it was the primary dislocation that had initiated the arthropathy and that later recurrences were of minor importance in this respect. This corresponds to our findings. Shoulder dislocation is a rather common disorder. In Sweden, the prevalence in the general population is about 2 per cent1 (approximately 200,000 people). Nevertheless, few Swedish orthopaedic surgeons have performed an arthrodesis or a total shoulder replacement in patients who have post-dislocation arthropathy. It is possible that more patients than we are aware of have had a shoulder replacement for this reason, as they may have forgotten, or may not have been asked, about dislocations that occurred when they were younger unless they had had operative treatment. Fourteen shoulders had unusual radiographic and clinical findings, with an abducted position of the arm, when the primary dislocation occurred (Fig. 1). We consider this condition as a special entity that is closely related to luxatio erecta, and we referred to it in the current report as luxatio abducta14. Rockwood et al. stated that in patients who have luxatio erecta the humerus is usually locked in a position between 110 and 160 degrees of abduction25. The condition is more common among the elderly. The mechanism of injury is a hyperabduction force that impinges the neck of the humerus on the acromion25. A review of the literature suggests that many shoulders previously described as having luxatio erecta2,3,6,15,19,20,24,32 were quite similar to those that had luxatio abducta in our series. Thus, luxatio erecta can be subdivided into a classic but less common form and a more common abducta type. This is also in accordance with the classification of Downey et al., who distinguished between a subglenoid (classic erecta) and a subcoracoid (abducta) type4. The higher rate of recurrence among the fourteen shoulders that had the abducta type in our series was not significant; however, seven of these shoulders needed operative repair, suggesting a trend (p = 0.08), and this may indicate a worse prognosis for these shoulders.
Orthopedic Department, Gävle Hospital, 801 87 Gävle, Sweden.
¶Orthopedic Department, Karlstad Hospital, 651 85 Karlstad, Sweden. #Orthopedic Department, University Hospital, 581 85 Linköping, Sweden. **Orthopedic Department, Falun Hospital, 791 82 Falun, Sweden.
  CiteULike  Connotea  Del.icio.us  Technorati What's this?
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Abstract
Introduction
| home | CME | help | feedback | subscriptions | archive | site index |
|
Stanford University Libraries' HighWire Press (TM) assists in the publication of JBJS Online. |