Академический Документы
Профессиональный Документы
Культура Документы
www.elsevier.com/locate/ymse
Neer Award 2008: Arthropathy after primary anterior shoulder dislocation: 223 shoulders prospectively followed up for twenty-ve years
Lennart Hovelius, MD, PhDa,b,*, Modolv Saeboe, MDc
a b
Division of Surgery and Perioperative Science, Department of Orthopedics, Umea University Hospital, Umea, Sweden Orthopedic Department, Gavle Hospital, Gavle, Sweden c Radiologic Department, Bollnas Hospital, Bollnas, Sweden
Background: Shoulder dislocation may cause arthropathy, but the natural history of this evolution is not well described. We therefore conducted a radiographic follow-up 25 years after the primary shoulder dislocation. Methods: A prospective Swedish multicenter study (1978-1979) included 257 shoulders in 255 patients (age, 12-40 years) with a rst-time anterior shoulder dislocation. After 25 years, 227 patients (229 shoulders) were alive and had follow-up. Radiographic imaging was performed in 223 shoulders (97%). Results: Shoulders were normal in 44%. Arthropathy was mild in 29%, moderate in 9%, and severe in 17%. Of the shoulders without a recurrence, 18% had moderate/severe arthropathy. The corresponding gures were 39% for shoulders that recurred once or more (without surgery) and 26% (16 of 62) for surgically stabilized shoulders. Seven of 221 patients (7 of 223 shoulders) were considered alcoholic at 25 years and all had severe arthropathy (P < .001). Other factors that correlated with moderate/severe arthropathy were age older than 25 years at primary dislocation (P .01) and primary dislocation caused by highenergy sports activity (P .009). Shoulders that had not recurred had less arthropathy than shoulders classied as recurrent (P .047) or stabilized over time (P .007). Sixty-two surgically stabilized shoulders had less arthropathy than those that became stable over time (P .047). Mild arthropathy at 10 years was associated with moderate/severe arthropathy at 25 years in 19 of 30 shoulders (63%) compared with 13 of 146 (9%) classied as normal at 10 years (P < .001). Joint incongruence at 10 years was associated with moderate/severe arthropathy at 25 years (P .001). Conclusion: Age at primary dislocation, recurrence, high-energy sports, and alcohol abuse were factors associated with the development of arthropathy. Also shoulders without a recurrence were associated with arthropathy. 2009 Journal of Shoulder and Elbow Surgery Board of Trustees. Keywords: Anterior shoulder dislocation; recurrence; alcohol abuse; arthropathy
This study received the Charles Neer Award at the annual American Shoulder and Elbow Surgeons Meeting, San Francisco, CA, March 8, 2008. *Reprint requests: L. Hovelius, Division of Surgery and Perioperative Science, Umea Department of Orthopedics, Furumov. 26 A, 80642 S-406427 Gavle, Sweden. E-mail address: hovelius@swipnet.se (L. Hovelius).
Numerous articles have addressed the prognosis of the primary anterior shoulder dislocation.15,20,23,26 The evolution of arthropathy after an anterior dislocation is mostly reported in surgical series of recurrent anterior dislocations.1,4,5,8,13,14,19,21,25,27 Samilson and Prieto24 coined the
1058-2746/2009/$36.00 - see front matter 2009 Journal of Shoulder and Elbow Surgery Board of Trustees. doi:10.1016/j.jse.2008.11.004
ARTICLE IN PRESS
2 entity dislocation arthropathy in 1983 and suggested a classication that has been commonly used. They further showed that shoulders without surgery could develop arthropathy after a solitary dislocation. This study on primary anterior dislocations has been ongoing since 1978,10 and a 10-year follow-up showed an 20% incidence of arthropathy (11% mild, 9% moderate/severe).12 When we started this study in 1977, the aim was to establish if conservative treatment with immobilization of a primary dislocation for 3 to 4 weeks changed the outcome compared with symptomatic treatment. Our earlier followup evaluations have demonstrated the same prognosis for both kinds of treatment.15 The surgical repair after the recurrent dislocation has been claimed to be the cause of the subsequent dislocation arthropathy.4,5,8,19 The present study describes the evolution of this process after 25 years. L. Hovelius, M. Saeboe
The radiographic examination was considered complete when all views were obtained at follow-up; nearly complete when 1 or 2 AP views and a subcoracoidal or axial view had been performed; and nally, incomplete if 1 or 2 AP views or just 1 or 2 undenable views had been performed. The degree of arthropathy on the radiographs was graded by both authors together as mild, osteophytes 3 mm or less on the humeral head; moderate, osteophytes between 3 and 7 mm; or severe, exceeding 7 mm, with or without articular incongruity. 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 parallel12,14 (Figure 2). The etiology of the primary dislocation with respect to arthropathy was analyzed for the following groups: 1. no athletic activity. 2. all types of athletic activities, which was divided into further 2 classes; namely, 3. sports causing high-energy injuries, including contact sports, motocross, parachute injuries, ski jumping, and horse riding; and 4. dislocations caused by anticipated low-energy sports such as slalom, wrestling, judo, boxing, racket-dominant sports, noncontact recreational activities, orienteering, swimming, diving, basket, volleyball, gymnastics, and running. Reduction of outward rotation at 10 years of follow-up for the surgically stabilized shoulders was analyzed with respect to dislocation arthropathy after 25 years as (1) shoulders with 0 to 10 restriction and (2) those exceeding 10 restriction. We dened the following stability end points at follow-up: 1. solitarydno further dislocation/subluxation occurred after the primary dislocation; 2. one redislocation/subluxationdjust 1 episode of instability occurred after the primary event during 25 years; 3. surgically stabilized during the 25 years of follow-up; 4. persistently recurrentd2 dislocation/subluxations or more occurred after the primary event, the last episode during the last 10 years of follow-up; and 5. became stable over timed2 or more recurrences occurred during the rst 15 years after the primary dislocation; however, no additional dislocation/subluxation occurred during the last 10 years of follow-up.15 The term subluxation was used when the patient described a suspected dislocation, followed by immediate, spontaneous reduction. Patients who had also sustained a dislocation or subluxation of the contralateral shoulder before or after the dislocation that had brought them into this study were characterized as having had bilateral dislocation. Luxatio abducta was found in our rst study in 20 patients.10 The arm had an abducted position anterior to the glenoid when dislocated, contrary to the erecta position, where the dislocation occurs inferior to the glenoid. This position is also described in our 10-year follow-up.12 In February 2005, the 227 patients (229 shoulders) who were alive had had the follow-up.15 Twenty-eight patients were deceased.16 Radiographic examinations were done in 223 of 229 shoulders (221 patients); of these, 7 patients (7 shoulders) were classied as alcoholics. The examination was complete in 110 index shoulders, nearly complete in 43, and incomplete in 70. A radiographic examination of the contralateral shoulder was performed in 212 of 223 shoulders.
ARTICLE IN PRESS
Arthropathy after anterior shoulder dislocation 3
Figure 1 Drawings show (A) subcoracoid and (B) axial projection (Horseeld and Jones, 1987). Radiographs show the (C) subcoracoid and (D) axial views.
Statistical methods
Statistical analyses were performed with the use of the c2 test for differences in proportions among various groups. If any cell contained 5 or fewer patients, we used the Fisher exact test. Unvariate and multiple logistic regression analysis were used to evaluate effects and control for possible confounding factors. To demonstrate differences in medians of continuous variables, we used the Kruskal-Wallis and the Mann-Whitney U tests for comparing groups. A value of P < .05 was considered signicant.
Radiographic outcome
Imaging showed 99 shoulders (44%) were normal. Arthropathy was mild in 65 shoulders (29%), moderate in 21 (9%), and severe in 38 (17%). No shoulders had proximal migration of the humeral head, indicating arthropathy related to rotator cuff disease.
Results
Prognostic outcomedstability end points
At 25 years we classied 95 shoulders as a solitary dislocation, 17 had had 1 recurrence/subluxation, and 62 were surgically stabilized. Of 49 shoulders classied as having recurrent dislocation without surgery, 18 were persistently recurrent and 31 had become stable over time.
ARTICLE IN PRESS
4 L. Hovelius, M. Saeboe
Figure 3 The percentage of arthropathy in patients of different ages when the primary dislocation occurred, rated as mild (yellow) and moderate/severe (red).
Figure 2
smokers (40%) compared with 39 of 177 shoulders (22%) in patients who did not smoke or had stopped smoking during the 25 years of observation (P .032). Solitary shoulders had less arthropathy than the group that had become stable over time (P .007) and those that were persistently recurrent (P .047; Table II). Shoulders that underwent operations did not differ from solitary dislocations with respect to moderate/severe arthropathy (P .725); however, they had less arthropathy than the shoulders that stabilized by time (P .047, Table II). The group that sustained the primary dislocation because of traumatic sports had the highest percentage of moderate/ severe arthropathy (37%) and the nontraumatic sports activity the lowest (15%; P .009; Figure 4).
ARTICLE IN PRESS
Arthropathy after anterior shoulder dislocation
Table I Dislocation arthropathy at 25 years related to smoking habits Arthropathy, No. (%) Normal, No. (%) Smokers Stopped smoking Never smoked Total
a b c d e
16 21 60 97
Table II
Arthropathy related to prognosis with respect to recurrences and performed surgery due to remaining instability Degree of arthropathy, No. (%) Normal 48 4 7 5 35 99 (51) (24) (23) (28) (56) (44) Mild 30 7 10 7 11 65 (32) (41) (32) (39) (18) (29) Moderate 7 2 5 1 6 21 (7) (12) (16) (6) (10) (9) Severe 10 4 9 5 10 38 (11)b (24) (29)b (28)b (16)c (17)d Total 95b 17 31b 18b 62c 223d Moderate/severe, %a 17 35 43 29 21 24
Degree of instability No recurrence One recurrence or sublux Healed recurrent Recurrent Operated Total
a b c d
moderate or severe arthropathy compared with 33 of 70 (47%) and 21 of 70 (30%), respectively, in the shoulders with incomplete radiographic follow-up (Table V). Patients with complete examinations did not appear to have more arthropathy than those with incomplete examinations.
Figure 4 The percentage of moderate/severe arthropathy is shown with respect to etiology when the primary dislocation occurred. 1, no sporting activity; 2, all types of sports; 3, traumatic sports; 4, nontraumatic sports.
ARTICLE IN PRESS
6 L. Hovelius, M. Saeboe
Table III Type of repair and result with respect to further surgery, recurrence or subluxations dislocation arthropathy and Disabilities of Arm, Shoulder and Hand (DASH) Revision surgery, No. Type of repaira Bristow-Latarjet Putti-Platt (simplied) Eden-Hybbinette Bankart Boytchev Nicola Du toit Arthroscopic Bankart
a b c
Recurrence or subluxation, No. Yes/No 4/23 8/ 9 1/ 6 1/4 2/1 1/0 1/0 0/1
Dislocation arthropathy, No. None/Mild/Moderate/ severe 19/4/1/3 7/4/3/3b 3/1/2/1c 3/2/0/0 1/0/0/2c 0/0/0/1 1/0/0/0 1/0/0/0
DASH Mean/ Median 4.4/0.41 12 /4.3 7.6/5.8 19.8/15.5 17.9/17.9 10.0/10.0 7.5/7.5 1.8/1.8
All repairs except the Arthroscopic Bankart were open procedures. 2 alcoholics. 1 alcoholic.
Four classes of Disabilities of Arm, Shoulder and Hand (DASH) score related to degree of arthropathy, alcoholics included Degree of arthropathy, No. (%) None 76 10 6 5 97 (49) (34) (35) (33) (47) Mild 47 8 5 4 64 (31) (28) (29) (27) (31) Moderate 17 2 1 1 21 (11) (7) (6) (7) (10) Severe 14 9 5 5 33 (9) (31)b (29) (33)a (16)
a
Moderate/severe, % 20 38 35 40 25
2 alcoholics. 1 alcoholic.
dislocation occurred), arthropathy was mild in 10 and severe in 1. The corresponding gures for 87 patients at 26 to 40 years were mild in 7 and moderate in 2. Shoulders with contralateral instability had more dislocation arthropathy than stable contralateral shoulders when all groups of arthropathy were analyzed (P .016) and also with respect to moderate/severe arthropathy (P .015)
Discussion
Moderate/severe dislocation arthropathy in 26% of the patients was a surprisingly high gure. Compared with our gures after 10 years (9%),12 moderate/severe arthropathy seemed to increase with 1 unit of percentage/year. When mild arthropathy is included, the gure for arthropathy after 25 years raises to 56%. Surprisingly, nearly 1 of 5 shoulders classied as solitary had moderate/severe arthropathy, and when mild was also included, half of the solitary shoulders had arthropathy. This shows that the trauma of dislocation has long-term biologic effects on joint physiology, although the clinical impact of radiographic arthropathy may be debated. In a study on the
Putti-Platt repair28 with a mean follow-up of 22 years, the total gures of arthropathy were identical with ours, 61% compared with 56%. Also remarkable is that all alcoholics had severe arthropathy. This may be explained by the fact that alcoholics, similar to athletes engaged in high-energy sports, sustain their dislocation(s) because of a more severe and uncontrolled trauma. Depending of course on the denition, 5% to 10% of the Swedish population is considered to be alcoholic, and reckless alcoholic behavior may be the cause of shoulder dislocation.16 The alcoholics normally represent some of the drop-outs in most studies, but in this study we tracked them all. We decided, wrong or not, to describe the alcoholics separately in some of the parts of this study. This also explains why our ndings with respect to smoking and arthropathy became insignicant (P .089). Smoking and arthropathy should be studied further, because a signicant difference was found when we compared smokers with those who had never smoked and those who stopped smoking (P .023). Our observation that primary dislocations occurring in ages younger than 25 years had less arthropathy is in accordance with our previous study on the Bristow-Latarjet repair14 in which the age limit was 23 years. This is
ARTICLE IN PRESS
Arthropathy after anterior shoulder dislocation
Table V Our classication in different degrees of arthropathy related to the number of radiographic views used at the 25-year follow-up (see text) Degree of radiographic follow-up, No. (%) Classication Complete of arthropathy None Mild Moderate Severe Total 47 36 13 14 110 (42.7) (32.7) (11.8) (12.7) (49) Nearly complete 19 13 5 6 43 (44.2) (30.2) (11.6) (14.0) (19) Incomplete 33 16 3 18 70 (47.1) (22.9) (4.3) (25.7) (31) Total 99 65 21 38 223
Figure 5 Box and whisker plot demonstrates median Disabilities of Arm, Shoulder and Hand (DASH) values for normal shoulders (1), and those with mild (2), moderate (3), and severe (4) arthropathy. The horizontal lines show the median values, and the boxes show interquartile range (the middle 50% of the values). The whiskers show maximum and minimum values in the case of no outlying and extreme values. The circles and asterisks are outliers and extreme values, respectively, indicating distance from the most frequent values.
important, because age should be taken into account when comparing series with surgical stabilization with respect to arthropathy. That the 62 shoulders that were operatively stabilized had less arthropathy than those unoperated-on shoulders with one or more recurrences is interesting. Some believe that the surgical repair always causes the arthropathy,5,19,28 and this should be true when the surgery is incorrectly done; for example, when the transferred coracoid or the metallic device protrudes into the joint.29 Also recently reported has been high gures of arthropathy after the Weber osteotomy with prosthetic replacement in 26% of the shoulders after 15 years.4 Our study, however, indicates that the surgical repair may decrease the evolution of arthropathy, because the difference between shoulders with no recurrences (solitary) and the surgically stabilized (with several recurrences before surgery) was insignicant, and our operated-on shoulders had less arthropathy than the unoperated-on recurrent shoulders. We could not show that bone-block repairs caused more arthropathy than soft-tissue surgery. Does stabilizing surgery prevent arthropathy? Should one recommend preventive surgery? Perhaps, but no patient in this study had surgery because of arthritis, and the impairment was obviously tolerable. Although the shoulders that became stable over time had more arthropathy, probably explaining why they stabilized, their median DASH values were as good as those for solitary shoulders.15 Moreover, the highest incidence of arthropathy occurred in patients aged 26 to 33 years, where the recurrence rate was lower than in ages younger than 25
Figure 6 Evolution of arthropathy between the 10- and 25-year follow-up visits.
years15 or 23 years.14 Also, we do not know what happens in the long term with different surgical procedures. In this study, 54 patients had surgery at 10 years, meaning the same follow-up time as the 34 shoulders had after the Weber osteotomy,28 where 9 underwent arthroplasty (P .0003). Furthermore, many of our arthroscopic repairs are causing a new disease, at least in Sweden, termed anchor arthritis because the placement of the anchors is interfering with joint motion. Many shoulders with moderate or severe arthropathy function very well when scored according to the DASH questionnaire (Table IV) and should mean that the disability caused by dislocation arthropathy, as diagnosed in the present article and others,28 possibly is no reason for increased prophylactic surgical intervention. An observation time of 25 years may be too short, and some of the shoulders in this group will likely undergo arthroplasty during the coming 15 years, although the analysis then will be more difcult because of an increased incidence of other disorders that limit function. Can we then predict the outcome with respect to subsequent arthropathy in a shoulder with a rst time dislocation? Obviously, a patient aged 23 to 33 years with a primary dislocation (Figure 3) caused by a high-energy
ARTICLE IN PRESS
8 trauma is at risk of developing arthropathy, especially if dislocation recurs. Shoulders with mild arthropathy had scores that were as good as those for solitary and surgically stabilized shoulders (Figure 5), but then how should the grading mild be considered? Of 30 shoulders classied as mild at 10 years, 19 (63%) had moderate/severe arthropathy at 15 years (Figure 6). Only 13 of 146 normal shoulders (9%) at 10 years developed moderate/severe arthropathy at 25 years. This should mean that the observation of mild arthropathy is important because it predicts moderate/severe arthropathy in more than 60% within the next 15 years (Figure 6). In a previous study on the Bristow-Latarjet-repair, we demonstrated a higher incidence of arthropathy when additional radiographic views were used at follow-up.14 Those ndings could not be conrmed in the present study. Future studies may reveal if just one view is sufcient to determine the degree/presence of arthropathy as this study indicates. Were any of our arthropathic shoulders related to cuff tear arthropathy? We do not believe so, because no shoulder had had cuff-related surgery after 25 years,15 and furthermore, no shoulders demonstrated any sign of proximal migration of the humeral head. The strength of this study is the complete follow-up after 25 years with radiographic examination of 97% of the shoulders. A limitation of our study may be our classication system. The Samilson-Prieto model has been questioned.3,18 Rosenberg et al22 suggested a somewhat different classication. In our opinion, however, their model resulted in classifying too many shoulders as having moderate/severe arthropathy. Therefore, as no better method exists, we have preferred to use the Samilson-Prieto model, especially because we used it in our previous studies.12-14 L. Hovelius, M. Saeboe
2. Atroshi I, Gummesson C, Andersson B, Dahlren E, Johansson A. The disabilities of the arm, shoulder and hand (DASH) outcome questionnaire. Reliability and validity of the Swedish version evaluated in 176 patients. Acta Ortop Scand 2000;71:613-8. 3. Buscayret F, Bradley Edwards T, Szabo I, Adeleine P, Coudane H, et al. Glenohumeral arthrosis in anterior instability before and after surgical treatment. Incidence and contributing factors. Am J Sports Med 2004;32:1165-72. 4. Flury M, Goldhan J, Holzman P, Simmen B. Does Webers rotation osteotomy induce degenerative joint disease at the shoulder in the long term? J Shoulder Elbow Surg 2007;16:735-41. 5. Hawkins RJ, Angelo RL. Glenohumeral osteoarthrosis. A late complication of the Putti-Platt repair. J Bone Joint Surg Am 1990;72: 1193-7. 6. Hermodsson I. Roentgenologischen Studien uber die traumatischen und habituellen schulterverrenkungen nach vorn und nach unten. Acta Radiol (Stockholm) 1934;20(Suppl):1-173. 7. Hill HA, Sachs MD. The grooved defect of the humeral head. A frequently unrecognized complication of dislocations of the shoulder joint. Radiology 1940;35:690-700. 8. Hindmarsh J, Lindberg A. Eden-Hybbinettes operation for recurrent dislocations of the humero-scapular joint. Acta Orthop Scand 1967;38: 459-78. 9. Horseld D, Jones SN. A useful projection in radiography of the shoulder. J Bone Joint Surg Br 1987;69:338. 10. Hovelius L, Eriksson K, Fredin H, Hagberg G, Hussenius A, Lind B, et al. Recurrences after initial dislocation of the shoulder. Results of a prospective study of treatment. J Bone Joint Surg Am 1983;65: 343-9. 11. Hovelius L. Anterior dislocation of the shoulder in teen-agers and young adults. Five-year prognosis. J Bone Joint Surg Am 1987;69: 393-9. 12. 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. 13. Hovelius L, Sandstrom B, Rosmark D, Saebo M, Sundgren K, Malmqvist B. Long-term results with the Bankart and Bristow-Latarjet procedures. Recurrent shoulder instability and arthropathy. J Shoulder Elbow Surg 2001;10:445-52. 14. Hovelius L, Sandstrom B, Saebo M. One hundred eighteen Bristow Latarjet repairs for recurrent anterior dislocation of the shoulder. Study II. The evolution of dislocation arthropathy. J Shoulder Elbow Surg 2006;15:279-89. 15. Hovelius L, Olofsson A, Sandstrom B, Augustini BG, Krantz L, Fredin H, et al. Nonoperative treatment of primary anterior shoulder dislocation in patients forty years of age or younger. A prospective twenty-ve year follow-up. J Bone Joint Surg Am 2008;90:945-52. 16. Hovelius L, Nilsson J-A, Nordqvist A. Increased 25 year mortality in a young 255 patient cohort with primary anterior shoulder dislocation of the shoulder. Acta Orthop Scand 2007;78:822-6. 17. Hudak PL, Amdi PC, Bombardier C. Development of an upper extremity outcome measure: the DASH (disabilities of the arm, shoulder and hand). Am J Ind Med 1996;29:602-8. 18. Ilg A, Bankes M, Emery R. The intra- and inter-observer reliability of the Samilson and Prieto grading system of glenohumeral arthropathy. Knee Surg Sports Traumatol Arthrosc 2001;9:187-90. 19. Kiss J, Merisch I, Perlaky GY, Szollas L. The results of the PuttiPlatt operation with particular reference to arthritis, pain and limitation of external rotation. J Shoulder Elbow Surg 1998;7: 495-500. 20. McLaughlin HL, MacLellan DI. Recurrent anterior dislocation of the shoulder. II. A comparative study. J Trauma 1967;7:191-201. 21. Rahme H, Wikblad L, Nowak J, Larsson S. Long-term clinical and radiologic results after Eden-Hybbinette operation for anterior instability of the shoulder. J Shoulder Elbow Surg 2003;12: 15-9.
Conclusions
Arthropathy is associated with shoulder dislocation. The impairment of this evolution is mostly inconsiderable. Operative repairs are not the cause of the arthropathy unless they interfere with joint physiology.
Acknowledgment
Professor Olle Svensson was very helpful revising this manuscript. Hans Hogberg and Marina Heiden contributed to the statistical evaluation.
References
1. Allain J, Goutallier D, Glorion C. Long-term results of the Latarjet procedure for the treatment on anterior instability of the shoulder. J Bone Joint Surg Am 1998;80:841-52.
ARTICLE IN PRESS
Arthropathy after anterior shoulder dislocation
22. Rosenberg B, Richmond J, Levine W. Long-term follow-up of Bankart reconstruction. Incidence of late degenerative glenohumeral arthrosis. Am J Sports Med 1995;23:538-44. 23. Rowe C. Prognosis in dislocation of the shoulder. J Bone Joint Surg Am 1956;38:957-77. 24. Samilson R, Prieto V. Dislocation arthropathy of the shoulder. J Bone Joint Surg Am 1983;65:456-60. 25. Singer GC, Kirkland PM, Emery R. Coracoid transposition for recurrent anterior instability of the shoulder. J Bone Joint Surg Br 1995;77:73-6.
9
26. Simonet WT, Coeld RH. Prognosis in anterior shoulder dislocation. Am J Sports Med 1984;12:19-24. 27. Toolanen G, Kjellgren A, Olsson H, Hogstrom B. The Alvik gleno plasty for the unstable shoulder. Modication of the Eden-Hybbinette operation in 66 cases. Acta Orthop Scand 1990;61:111-5. 28. Van der Zwaag H, Brand R, Oberman W, Rozig P. Glenohumeral osteoarthrosis after Putti-Platt repair. J Shoulder Elbow Surg 1999;8:252-8. 29. Zuckerman JD, Matsen F. Complications about the glenohumeral joint related to the use of screws and staples. J Bone Joint Surg Am 1984;6: 175-80.