Beruflich Dokumente
Kultur Dokumente
Investigation performed at St. Michael’s Hospital and the University of Toronto, Toronto, Ontario, Canada
Background: Recent studies have suggested benefits following primary operative fixation of substantially displaced
midshaft fractures of the clavicle. We reviewed randomized clinical trials of operative versus nonoperative treatment of these
fractures, and pooled the functional outcome and complication rates to arrive at summary estimates of these outcomes.
Methods: A systematic review of the literature was performed to identify studies of randomized clinical trials comparing
operative versus nonoperative care for displaced midshaft clavicular fractures.
Results: Six studies (n = 412 patients, mean Detsky score = 15.3) were included. The nonunion rate was higher in the
nonoperatively treated patients (twenty-nine of 200) than it was in patients treated operatively (three of 212) (p = 0.001).
The rate of symptomatic malunion was higher in the nonoperative group (seventeen of 200) than it was in the operative
group (0 of 212) (p < 0.001).
Conclusions: Operative treatment provided a significantly lower rate of nonunion and symptomatic malunion and an
earlier functional return compared with nonoperative treatment. However, there is little evidence at present to show that
the long-term functional outcome of operative intervention is significantly superior to nonoperative care.
Level of Evidence: Therapeutic Level I. See Instructions for Authors for a complete description of levels of evidence.
A
midshaft fracture of the clavicle is a common skeletal Due to the increasing recognition of suboptimal out-
injury that accounts for 2.6% to 10% of all adult frac- come following nonoperative treatment, primary operative
tures1,2. Older studies report that the nonunion rate with fixation has become increasingly popular for these injuries.
conservative treatment is low (less than 1%)3,4. Recent pro- Operative implants and techniques vary, and include in-
spective studies focusing on the nonoperative treatment of tramedullary osteosynthesis and plate fixation8-11. There are a
displaced midshaft fractures in the adult population describe number of recently published, randomized prospective clin-
nonunion rates of 15% to 20%, objective shoulder muscle ical trials12-17. Since this injury commonly afflicts a young,
strength loss of 18% to 33%, poor early functioning of the active population, rapid return to function and early union
injured shoulder, and up to 42% of patients with residual se- is a priority in addition to the more traditional outcomes,
quelae at six months after injury5-7. such as recovery of muscle strength, range of motion, and the
Disclosure: None of the authors received payments or services, ei- This article was chosen to appear
ther directly or indirectly (i.e., via his or her institution), from a third electronically on March 14, 2012, in
party in support of any aspect of this work. One or more of the advance of publication in a regularly
authors, or his or her institution, has had a financial relationship, in scheduled issue.
the thirty-six months prior to submission of this work, with an entity in
the biomedical arena that could be perceived to influence or have the A commentary by William T. Obremskey,
potential to influence what is written in this work. No author has had MD, MPH, is linked to the online version of
any other relationships, or has engaged in any other activities, that this article at jbjs.org.
could be perceived to influence or have the potential to influence
what is written in this work. The complete Disclosures of Potential
Conflicts of Interest submitted by authors are always provided with
the online version of the article.
avoidance of long-term complications (such as nonunion and Data Extraction and Analysis
symptomatic malunion). Two senior authors (M.D.M. and D.B.W.) independently extracted data from all
There are numerous studies on nonoperative and oper- included studies with the use of a prespecified data extraction form. A third
investigator (R.C.M.) then reviewed the forms for completeness and agreement
ative care, but few randomized clinical trials have prospectively
and entered the data (population demographics, mechanism of injury, operative
compared the two. In this systematic review, the primary aim methods, shoulder range of motion, and patient and surgeon outcome scores,
was to use Level-I evidence from recent randomized clinical complications [nonunion or delayed union, symptomatic malunion, infection,
trials to arrive at summary estimates of treatment effect, po- hardware removal, neurologic compromise, and refracture], and return to
tential harm, and patient outcome. activity and/or work) into a custom spreadsheet. Disagreements in data ex-
traction were resolved by discussion and consensus. Data across the six studies
Materials and Methods were pooled and summary estimates of treatment effect (relative risks with
associated 95% confidence intervals) were calculated. A random effects model
Research Questions
was used to ensure that these studies represent a random sample of all po-
W e posed three questions with regard to the treatment of acute, displaced
midshaft fractures of the clavicle: (1) Is there a significant difference in
patient and surgeon-assessed outcomes of primary operative fixation versus
24
tentially available studies . In the pooled analysis, studies were weighted by the
inverse of the variance for the reported outcome. Homogeneity across the
studies was assessed with a chi-square analysis, with p £ 0.10 being considered
nonoperative treatment? (2) What are the differences in the rate and severity 25
significant. The Review Manager (RevMan) software program (The Nordic
of a defined set of complications (nonunion or delayed union, symptomatic
Cochrane Centre, Copenhagen, Denmark), provided by The Cochrane Col-
malunion, infection, hardware removal, neurologic compromise, and refrac-
laboration, was used for graphical representation of the pooled data.
ture) for operative fixation versus nonoperative treatment at one year? and (3)
Is there evidence that, when compared with nonoperative care, primary fixation
of displaced midshaft clavicular fractures minimizes disability and improves Source of Funding
function at twelve months after injury? This study was supported by an internal educational grant from the St.
Michael’s Hospital Orthopaedic Research and Education Fund (to R.C.M.).
Research Ethics Board approval is not required for this type of study at our
Eligibility Criteria institution.
Studies were included in this review if they were randomized prospective trials
of operative versus nonoperative treatment of acute, completely displaced
midshaft fractures of the clavicle. Quasi-randomized studies (nonrandom
Results
treatment allocation) were excluded, as were studies that included delayed Literature Search and Review of the Proceedings of
union or nonunion, children (i.e., younger than sixteen years of age), or Annual Meetings
pathological fractures.
(range, 14 to 16) (Table II). All five studies that reported the
method of randomization used a sealed envelope technique, a
method considered suboptimal by modern trial standards22.
Treatment
Fixation was accomplished with a variety of small fragment
plates as the operative technique in three studies12-14 and with
pin osteosynthesis in three studies15-17 (i.e., small-diameter
[2.0, 2.5, or 3.0 mm] intramedullary titanium pins in two
studies15,17 and a modified Hagie pin in one study16). The
postoperative rehabilitation was similar in all four studies in
which it was described: a sling for comfort for ten days, fol-
lowed by early active range-of-motion exercises. At six weeks
postoperatively, strengthening and a gradual return to sport
and/or heavy lifting, overhead, or repetitive activities was al-
lowed12,14,16,17. All studies used a standard sling for nonoperative
care12-17, with institution of range-of-motion exercises, as pain
allowed, followed by physiotherapy12,14,16,17.
*CS = Constant Score, DASH = Disabilities of the Arm, Shoulder and Hand score, SANE = Single Assessment Numeric Evaluation score, NA = not
available.
sling (twenty-eight). The SANE and L’Insalata scores were sig- nonunions in the operative group (3%) and seven nonunions in
nificantly higher at week three for the operative group (p < 0.044 the nonoperative group (14%) (p = 0.042).
and p < 0.015). There were no significant differences between
these two groups at any other time. One patient in each group Witzel (2007)
developed a nonunion (one [3%] of twenty-nine in the operative The study by Witzel randomized thirty-three patients to non-
group versus one [4%] of twenty-eight in the nonoperative operative care (sling) and thirty-five to operative fixation with
group). This study is notable for the high complication rate in an elastic intramedullary pin, and it concentrated on dis-
the operative group (48%) compared with the nonoperative tinguishing differences in early return to function15. Early
group (7%). The high complication rate for the operative group (one-month) pain scores were significantly better for the
was due primarily to minor complications from pin prominence operative group (p = 0.05), as was strength (p = 0.01). It was
posterolaterally (causing pin-track infections and/or require- also noted that, on the sixtieth day, 80% of the operatively
ment for early pin removal). treated patients resumed athletic activity, while only 55% of
nonoperatively treated patients did.
Canadian Orthopaedic Trauma Society (COTS) (2007)
The COTS study randomized 132 patients to a standard sling Smith et al. (2000)
(sixty-five) or small-fragment plate fixation (sixty-seven)12. Forty- In a randomized clinical trial, Smith and colleagues random-
nine nonoperatively treated patients and sixty-two operatively ized one hundred patients to a sling or small-fragment plate
treated patients completed the one year follow-up. The CS and fixation; only thirty-five nonoperatively treated and thirty
DASH scores were significantly better at all time points for the operatively treated patients completed follow-up (mean
operative group of this study (p < 0.01). There were two follow-up, 18.5 months)14. Of the thirty operatively treated
679
TH E JO U R NA L O F B O N E & JO I N T SU RG E RY J B J S . O RG
d
O P E R AT I V E V E R S U S N O N O P E R AT I V E C A R E OF DI SP L ACE D
V O LU M E 94-A N U M B E R 8 A P R I L 18, 2 012
d d
M I D S H A F T C L AV I C U L A R F R A C T U R E S
Virtanen Blocked randomization, 1 year 15% Equal functional results at No null hypothesis 16/20
13
et al. sized randomly from one year, but nonoperative
(2010) 4-10, sealed envelope group had higher risk of
nonunion
Smekal Balanced block 2 years 13% CS scores were higher and Randomization block 16/20
17
et al. randomization with DASH scores were lower in size known by investigators,
(2009) a block size of 4, and the operative group, and so partially predictable, no
sealed envelope this group also had faster null hypothesis or sample-size
mobilization and less pain calculation stated, take-home
DASH scores not monitored
Judd Sealed envelope 1 year 0% Higher complication rate Inconsistent follow-up at all 16/21
16
et al. in the operative group, Hagie time points (lowest showing
(2009) pin is not recommended is 46%), no null hypothesis
or sample-size calculation
stated
12
COTS Permutation blocks of 1 year 16% Operative care improved Multicenter trial (8 centers), 16/20
(2007) 4 per investigational patient-oriented outcomes number of surgeons not stated
site, sealed envelope and surgeon-oriented
outcomes, patients had
earlier return to function
and decreased rates of
malunion and nonunion
15
Witzel Not stated 120 days Not stated Operative care group had Randomization technique not 14/20
(2007) improved mobility, less pain, described, no null hypothesis or
and were significantly stronger sample-size calculation stated
than the patients who received
nonoperative care
Smith Sealed envelope Until radiographic 35% Results suggest that all Unpublished article, attrition 14/20
14
et al. union or nonunion midshaft fractures should bias—only 65% for follow-up, no
(2000) be operatively fixed if null hypothesis or sample-size
displaced >2 cm, to calculation stated
prevent nonunions
*CS = Constant Score, DASH = Disabilities of the Arm, Shoulder and Hand score, COTS = Canadian Orthopaedic Trauma Society. †Detsky score is
designed to evaluate the quality of a randomized clinical trial (maximum score is 20 for surgical trial).
pin protrusion (treated with removal of hardware), and wound primarily of hardware-related complications from prominence
infection (Table III). One study was a statistical outlier in that the of the pin posterolaterally), compared with 7% in the nonop-
complication rate in the operative group was 48% (consisting erative group16. This was a significant source of heterogeneity in
Fig. 2
Graph showing complication rates comparing operative (experimental) versus nonoperative (control) groups (p = 0.18). The size of each square is
proportional to the weight of the study. Cots = Canadian Orthopaedic Trauma Society. Z = p value of weighted test for overall effect, CI = confidence interval,
df = degrees of freedom, I2 = test statistic.
681
TH E JO U R NA L O F B O N E & JO I N T SU RG E RY J B J S . O RG
d
O P E R AT I V E V E R S U S N O N O P E R AT I V E C A R E OF DI SP L ACE D
V O LU M E 94-A N U M B E R 8 A P R I L 18, 2 012
d d
M I D S H A F T C L AV I C U L A R F R A C T U R E S
Fig. 3
Graph showing comparison of nonunion and symptomatic malunion rates between operative (experimental) and nonoperative (control) groups (p <0.0001).
The size of each square is proportional to the weight of the study. Cots = Canadian Orthopaedic Trauma Society. Z = p value of weighted test for overall effect,
CI = confidence interval, df = degrees of freedom, I2 = test statistic.
the assessment of complication rates (p = 0.01). The use of this in risk after primary fixation as compared with nonoperative
intramedullary implant (Hagie pin) in this fashion is associated treatment) for operative versus nonoperative treatment was
with a high rate of requirement for removal8,16. 22% (ranging from a 23% risk with nonoperative treatment to
a 1% risk with operative care). The number needed to treat for
Nonunion and Symptomatic Malunion Rates one patient to benefit from surgical intervention (in terms of
We identified nonunion and symptomatic malunion and an- the prevention of nonunion or symptomatic malunion) was 4.6
alyzed them independently. Three of the studies classified a (n = 412 total patients). Of the 212 patients in the operative
nonunion as no evidence of osseous union radiographically group, three (1%) developed nonunion. Of the 200 patients in
at one year after injury12,13,16. Two studies used a similar defi- the nonoperative group, twenty-nine (15%) developed non-
nition at an earlier time point (five months in one study and union (p = 0.001) (Fig. 4). The absolute risk reduction for
six months in the other), and one study reported no non- nonunion for operative versus nonoperative treatment was
unions14,15,17. Radiographic malunion is ubiquitous following 14% (ranging from a 15% risk with nonoperative treatment to
nonoperative treatment of a displaced clavicular fracture: this a 1% risk with operative care). The number needed to treat for
finding was not included as a complication in these trials. one patient to benefit from surgical intervention in nonunion
Symptomatic malunion was defined as a patient with symp- rates was 7.6 (n = 412 total patients). There were no symptom-
toms severe enough to warrant corrective osteotomy. Non- atic malunions (0%) reported in the 212 patients treated
union or symptomatic malunion were significantly more operatively and seventeen malunions (9%) in the 200 patients
common in the nonoperative group (forty-six [23%] of 200 in the nonoperative group (p < 0.001). The absolute risk re-
nonoperatively treated patients versus three [1%] of 212 op- duction for symptomatic malunion for operative versus
eratively treated patients, p < 0.0001) (Fig. 3). The absolute risk nonoperative treatment was 9% (ranging from a 9% risk with
reduction for nonunion and symptomatic malunion (decrease nonoperative treatment to a 0% risk with operative care). The
Fig. 4
Graph showing comparison of nonunion rate between operative (experimental) and nonoperative (control) groups (p = 0.001). The size of each square is
proportional to the weight of the study. Cots = Canadian Orthopaedic Trauma Society. Z = p value of weighted test for overall effect, CI = confidence interval,
df = degrees of freedom, I2 = test statistic.
682
TH E JO U R NA L O F B O N E & JO I N T SU RG E RY J B J S . O RG
d
O P E R AT I V E V E R S U S N O N O P E R AT I V E C A R E OF DI SP L ACE D
V O LU M E 94-A N U M B E R 8 A P R I L 18, 2 012
d d
M I D S H A F T C L AV I C U L A R F R A C T U R E S
number needed to treat for one patient to benefit from surgical removal. Theoretically, this complication could be minimized by
intervention for symptomatic malunion was 11.7. less prominent implants or improved insertion techniques.
Our review shows that 15% of 200 patients in the nonop-
Early Return to Function erative group developed a nonunion, which is significantly greater
There were multiple indicators in these six studies that primary (p = 0.001) than the 1% rate of nonunion in the 212 patients of
fixation resulted in an earlier return to function12,17. Pain scores the operative group. The number needed to treat for one person to
at five, nineteen, and thirty-three days after injury were benefit from operative treatment in relation to nonunion was 7.6.
significantly better in the operative group in the study by Another complication significantly more common in the non-
Witzel15. A return to moderate activity on the sixtieth day after operative group is symptomatic malunion (seventeen of 200 in
injury was possible in 80% of the operatively treated patients the nonoperative group versus 0 of 212 in the operative group, p <
but only in 55% of the nonoperatively treated patients in that 0.001). However, with the data available, we were not able to draw
study15. Multiple studies demonstrated significantly superior any specific conclusions as to which patients are most likely to
functional scores early in the period after injury in the opera- suffer from one of these significant complications.
tive group; for example, there was a 15 to 20-point improve- Our meta-analysis shows that primary fixation of dis-
ment in the DASH scores in the operative group in one study12. placed midshaft clavicular fractures has an early effect on de-
creasing pain and improving function. This is demonstrated by
Discussion the significantly improved DASH scores (15 to 20 points) at six
and content experts, it is possible that we have not detected an symptomatic malunion), and marginally superior long-term
eligible existing trial, the results of which may be applicable to functional outcome scores were observed with operative in-
our meta-analysis. tervention. It is clear that there is a specific subset of individuals
Specific limitations of our study include the fact that one with a completely displaced midshaft fracture of the clavicle
of the six randomized trials we analyzed was published in ab- who will benefit from fixation. However, this information
stract form only. While we were able to obtain additional in- should not be used to justify an indiscriminate approach to
formation from this author, the lack of detailed information and surgical fixation of all clavicular fractures. Patients with a com-
the peer review process limits the assessment of this study. A pletely displaced midshaft clavicular fracture may be counseled
sensitivity analysis, using only the five peer-reviewed studies, did that they will be at a higher risk of sustaining nonunion and
not show any difference in our conclusions. The variety of dif- symptomatic malunion if the fracture is treated nonoperatively,
ferent outcome measures used limited our ability to combine but that there is no clear evidence that surgical treatment will
outcome scores and make more definitive conclusions; it may improve their long-term function in general. They should also be
also have resulted in a decrease in our ability to identify a true counseled that, approximately 75% of the time, a completely
difference where one actually existed. For example, five of six displaced clavicular fracture that is treated nonoperatively will heal
studies reported the nonunion rate and we were able to show with few, if any, long-term consequences. Surgeons can use this
a clear decrease in nonunion rate in the operative group; how- information appropriately to help patients make optimal deci-
ever, we were unable to do so as conclusively for the DASH or sions in a shared decision-making process. The patient must
Constant scores that only three of the studies employed. Two weigh this information against the risks of anesthesia, wound
operative techniques are currently used for primary fixation of infection, nonunion with an operation, and the potential need for
midshaft fractures of the clavicle (plate fixation and intramed- hardware removal.
ullary pin insertion), and both were used in the trials we studied;
it should be noted that a recent randomized clinical trial com- Appendix
paring locked intramedullary nailing versus plating for displaced A table showing population demographics is available
midshaft fractures of the clavicle showed a 100% union rate in with the online version of this article as a data supplement
both groups with similar outcome scores a mean of twelve at jbjs.org. n
months after injury26. However, the lack of a standard operative
protocol must be considered a weakness of our meta-analysis;
although both techniques have been associated with excellent
results in a variety of reviews, it may be that there are funda-
mental differences in outcome between the two techniques12-17. Robbin C. McKee
In this meta-analysis of randomized clinical trials com- Daniel B. Whelan, MD, FRCS(C)
paring the effect of primary operative fixation versus sling Emil H. Schemitsch, MD, FRCS(C)
treatment of displaced midshaft clavicular fractures in young Michael D. McKee, MD, FRCS(C)
Division of Orthopaedics, Department of Surgery,
active individuals, certain benefits of fixation are noted, espe- St. Michael’s Hospital and the University of Toronto,
cially early in the period after injury. A more rapid return to 800 – 55 Queen Street East, Toronto,
function, a decreased complication rate (especially with regard ON M5C 1R6, Canada.
to the more serious negative outcomes of nonunion and E-mail address for M.D. McKee: McKeeM@smh.ca
References
1. Robinson CM. Fractures of the clavicle in the adult. Epidemiology and classifi- 10. McKee MD, Seiler JG, Jupiter JB. The application of the limited contact dynamic
cation. J Bone Joint Surg Br. 1998;80:476-84. compression plate in the upper extremity: an analysis of 114 consecutive cases.
2. Postacchini F, Gumina S, De Santis P, Albo F. Epidemiology of clavicle fractures. Injury. 1995;26:661-6.
J Shoulder Elbow Surg. 2002;11:452-6. 11. Mueller M, Burger C, Florczyk A, Striepens N, Rangger C. Elastic stable intra-
3. Neer CS 2nd. Nonunion of the clavicle. J Am Med Assoc. 1960;172:1006-11. medullary nailing of midclavicular fractures in adults: 32 patients followed for 1-5
years. Acta Orthop. 2007;78:421-3.
4. Rowe CR. An atlas of anatomy and treatment of midclavicular fractures. Clin
Orthop Relat Res. 1968;58:29-42. 12. Canadian Orthopaedic Trauma Society. Nonoperative treatment compared with
plate fixation of displaced midshaft clavicular fractures. A multicenter, randomized
5. Hill JM, McGuire MH, Crosby LA. Closed treatment of displaced middle-third clinical trial. J Bone Joint Surg Am. 2007;89:1-10.
fractures of the clavicle gives poor results. J Bone Joint Surg Br. 1997;79:
13. Virtanen KJ, Paavola MP, Remes VM, Pajarinen J, Savolainen V, Bjorkenheim
537-9.
JM. Nonoperative versus operative treatment of midshaft clavicle fractures: a ran-
6. McKee MD, Pedersen EM, Jones C, Stephen DJ, Kreder HJ, Schemitsch EH, Wild domized controlled trial. Read at the 75th Annual Meeting of the AAOS; 2010 Mar
LM, Potter J. Deficits following nonoperative treatment of displaced midshaft cla- 9-12; New Orleans, LA. Paper no 331.
vicular fractures. J Bone Joint Surg Am. 2006;88:35-40. 14. Smith CA, Rudd J, Crosby LA. Results of operative versus nonoperative treat-
7. Nowak J, Holgersson M, Larsson S. Sequelae from clavicular fractures ment for 100% displaced midshaft clavicle fractures: a prospective randomized
are common: a prospective study of 222 patients. Acta Orthop. 2005;76: clinical trial. Read at the 16th Annual Open Meeting of the American Shoulder and
496-502. Elbow Surgeons; 2000 Mar 18; Orlando, FL. Paper no 31.
8. Crenshaw AH. Fractures of the shoulder girdle, arm and forearm. In: Crenshaw 15. Witzel K. [Intramedullary osteosynthesis in fractures of the mid-third of the
AH, editor. Campbell’s operative orthopaedics. 8th ed. St. Louis: Mosby Year Book; clavicle in sports traumatology]. Z Orthop Unfall. 2007;145:639-42. German.
1992. p 989-1053. 16. Judd DB, Pallis MP, Smith E, Bottoni CR. Acute operative stabilization versus
9. Poigenfürst J, Rappold G, Fischer W. Plating of fresh clavicular fractures: results nonoperative management of clavicle fractures. Am J Orthop (Belle Mead NJ).
of 122 operations. Injury. 1992;23:237-41. 2009;38:341-5.
684
TH E JO U R NA L O F B O N E & JO I N T SU RG E RY J B J S . O RG
d
O P E R AT I V E V E R S U S N O N O P E R AT I V E C A R E OF DI SP L ACE D
V O LU M E 94-A N U M B E R 8 A P R I L 18, 2 012
d d
M I D S H A F T C L AV I C U L A R F R A C T U R E S
17. Smekal V, Irenberger A, Struve P, Wambacher M, Krappinger D, Kralinger FS. Elastic 22. Detsky AS, Naylor CD, O’Rourke K, McGeer AJ, L’Abbé KA. Incorporating vari-
stable intramedullary nailing versus nonoperative treatment of displaced midshaft clavic- ations in the quality of individual randomized trials into meta-analysis. J Clin Epi-
ular fractures-a randomized, controlled, clinical trial. J Orthop Trauma. 2009;23:106-12. demiol. 1992;45:255-65.
18. Constant CR, Murley AH. A clinical method of functional assessment of the 23. Bhandari M, Richards RR, Sprague S, Schemitsch EH. The quality of reporting of
shoulder. Clin Orthop Relat Res. 1987;214:160-4. randomized trials in the Journal of Bone and Joint Surgery from 1988 through 2000.
19. Hudak PL, Amadio PC, Bombardier C. Development of an upper extremity out- J Bone Joint Surg Am. 2002;84:388-96.
come measure: the DASH (disabilities of the arm, shoulder and hand) [corrected]. 24. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials.
The Upper Extremity Collaborative Group (UECG). Am J Ind Med. 1996;29:602-8. 1986;7:177-88.
20. Williams GN, Gangel TJ, Arciero RA, Uhorchak JM, Taylor DC. Comparison of the 25. Review Manager (RevMan) [Computer Program]. Version 5.0. Copenhagen: The
Single Assessment Numeric Evaluation method and two shoulder rating scales. Nordic Cochrane Centre, The Cochrane Collaboration, 2008.
Outcomes measures after shoulder surgery. Am J Sports Med. 1999;27:214-21. 26. Ferran NA, Hodgson P, Vannet N, Williams R, Evans RO. Locked intra-
21. L’Insalata JC, Warren RF, Cohen SB, Altchek DW, Peterson MG. A self-administered medullary fixation vs plating for displaced and shortened mid-shaft clavicle
questionnaire for assessment of symptoms and function of the shoulder. J Bone Joint fractures: a randomized clinical trial. J Shoulder Elbow Surg. 2010;19:
Surg Am. 1997;79:738-48. 783-9.