Journal of Orthopaedic Trauma Publish Ahead of Print

DOI: 10.1097/BOT.0000000000000936

MIDSHAFT FRACURES OF THE CLAVICLE: A META-ANALYSIS

COMPARING SURGICAL FIXATION VIA ANTEROINFERIOR

PLATING VERSUS SUPERIOR PLATING

D
Alex Nourian, BSa, Satvinder Dhaliwal, MPHb, Sitaram Vangala, MSb, Peter S. Vezeridis, MDc

TE
David Geffen School of Medicine, University of California, Los Angeles
EP
b
Department of Medicine Statistics Core, University of California, Los Angeles
c
Department of Orthopaedic Surgery, University of California, Los Angeles

Corresponding author:
C

Peter S. Vezeridis, M.D.

C

University of California, Los Angeles

Department of Orthopaedic Surgery

A

Phone: (310) 825-5003

Fax: (310) 825-1311

Email: Peter.Vezeridis@gmail.com

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 Presented in part at the Annual Meeting of the Orthopaedic Trauma Association, National

Harbor, MD, October 2016.

The authors report no conflicts of interest related to this work.

D
TE
EP
C
C
A

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 1

1 ABSTRACT

2 Objective: Tto compare the outcomes of clavicle fracture fixation using anteroinferior versus

3 superior plate placement.

5 Methods: We performed a meta-analysis of studies that have reported on outcomes following

D
6 superior or anteroinferior plate fixation for acute midshaft clavicle fractures (OTA 15-B). A

7 computerized literature search in the Pubmed, Scopus, and Cochrane Library databases was

TE
8 utilized to identify relevant articles. Only full text articles without language restrictions were

9 evaluated. The inclusion criteria consisted of: 1) fracture of the midshaft clavicle; 2) surgery for

10 acute fractures (within one month of the fracture); 3) adult patients (16 years of age and older);

11 and 4) open reduction and internal fixation with plate application in either the anteroinferior or
EP
12 superior position. Studies were excluded if they did not specify plate location, evaluated multi-

13 trauma patients, investigated minimally invasive procedures, or studied operations for revision,

14 nonunion, malunion, or infection. The primary measured outcomes were symptomatic hardware
C

15 (implant prominence or irritation) and surgery for implant removal. The secondary outcomes

16 were time to union, fracture union, nonunion, malunion, DASH score, Constant score, and
C

17 implant failure. Frequencies and proportions of cases were recorded for binary outcomes, while

18 means and standard deviations were recorded for continuous outcomes. Other summary statistics
A

19 provided were used to impute means and standard deviations under the assumption of normality

20 when these were not reported. Continuous outcomes were compared between groups using linear

21 mixed effects models, while binary outcomes were compared using mixed effects logistic

22 regression models, including fixed group effects and random study effects. P-values less than

23 0.05 were considered statistically significant. All analyses were performed using SAS v. 9.4

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 2

24 (SAS Institute Inc., Cary, NC).

25

26 Results: A total of 1,428 articles were identified amongst the three databases, of which 897

27 remained after removing duplicates. From that pool, 57 relevant studies were evaluated. Articles

28 were excluded due to an inability to specify plate location (6), a subject pool not exclusively

D
29 consisting of acute fractures (4) or midshaft fractures (2), a minimally invasive surgical approach

30 (6), use of non-standard plates (1), poor reporting of functional outcomes (2), and a duplicate

TE
31 group of patients (2). This left 34 articles to be used in the meta-analysis. Of these, 8 studies

32 reported on patients with anteroinferior plating (N=390) and 27 studies reported on patients with

33 superior plating (N=1104). No significant differences were found with respect to the functional

34 shoulder scores (DASH and Constant) between the two groups. There was no significant
EP
35 difference between each group for the probability of having a union (p=0.41), malunion (p=0.28),

36 nonunion (0.29), or implant failure (p=0.39). Patients in the superior plating group had a

37 significantly higher probability of suffering from symptomatic hardware (0.17) as compared to

C

38 patients in the anteroinferior plating group (0.08), (p=0.005). Additionally, the superior plating

39 group had a significantly higher rate of surgery for implant removal (0.11 versus 0.05),
C

40 (p=0.008).

41
A

42 Conclusion: The findings of this investigation demonstrate that plating along the superior and

43 anteroinferior aspects of the clavicle lead to similar operative outcomes with respect to union,

44 nonunion, malunion, and implant failure, as well as similar functional outcomes scores. Plates

45 applied to the superior aspect of the clavicle are associated with higher rates of symptomatic

46 hardware and more frequent implant removal.

47

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 3

48 Keywords: clavicle; fracture; plate fixation; open reduction and internal fixation; hardware
49 prominence; hardware removal
50

51

52 INTRODUCTION

53 Clavicle fractures are common injuries, occurring at a rate of approximately 5.8 per

D
54 10,000 persons per year in the United States1. The middle third region of the clavicle is the most

55 frequently affected area, comprising up to 81.3% of these fractures2. In 1960, Neer reported a

TE
56 nonunion rate of 0.1% in conservatively treated patients with middle third fractures as compared

57 to a nonunion rate of 4.6% in those treated with open reduction and internal fixation3. Based on

58 these historical results, clavicle fractures have traditionally been treated conservatively with a

59 period of brief immobilization4.

EP
60 Recent evidence has challenged the notion that the majority of clavicle fractures should

61 be treated conservatively. Hill and colleagues evaluated 52 consecutive patients treated

62 nonoperatively for a completely displaced clavicle fracture and found that 15% of patients had a
C

63 nonunion and 31% were not satisfied with their results5. A systematic review of 2144 midshaft

64 clavicle fractures found a nonunion rate of 15.1% amongst conservatively treated displaced
C

65 clavicle fractures, and surgical treatment significantly reduced the risk of nonunion6. Moreover,

66 conservative treatment of midshaft fractures may result in compromised shoulder function. In

A

67 2006, McKee et al. evaluated 30 patients treated conservatively after displaced midshaft fractures

68 and found a mean Constant score of 71 points and a mean DASH score of 24.6 points, indicating

69 substantial disability7.

70 In light of this evidence, there has been a recent trend towards operative treatment of

71 displaced midshaft clavicle fractures. Studies have demonstrated that surgical treatment with

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 4

72 open reduction and plate fixation is associated with improved outcomes and a decreased risk of

73 complications, specifically a faster time to union, fewer nonunions and malunions, and better

74 shoulder function scores8,9. There are two common approaches for plate fixation, with the plate

75 applied either to the superior or anteroinferior aspect of the clavicle10. Anteroinferior plate

76 fixation may be desirable to superior fixation due to less prominence of the plate and, in turn,

D
77 fewer subsequent procedures for implant removal. In a study by Collinge et al., only 2 of 58

78 patients who underwent anteroinferior plate fixation had implant irritation that required

TE
79 removal11. Conversely, in the 2007 Canadian Orthopaedic Trauma Society (COTS) study, 11 of

80 62 patients who underwent superior plate fixation had implant irritation or prominence, and 5

81 patients required implant removal8.

82 Despite the potential advantage of utilizing anteroinferior plate fixation, there is minimal
EP
83 evidence comparing these two techniques. One investigation that directly compared these

84 approaches found that the two groups had a similar time to union and union rate, however

85 patient-reported implant prominence was nearly double in the superior group12. Implant removal
C

86 also occurred more frequently in the superior group, but the difference was not statistically

87 significant. The purpose of the present investigation was to perform a meta-analysis of studies
C

88 that have reported on outcomes following superior or anteroinferior plate fixation for acute

89 midshaft clavicle fractures. We tested the hypothesis that anteroinferior plate fixation would lead
A

90 to less implant prominence and fewer subsequent procedures for implant removal as compared to

91 superior plate fixation.

92

93

94

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 5

95 METHODS

96 This meta-analysis was performed in accordance with the Preferred Reporting Items for

97 Systematic Reviews and Meta-Analyses (PRISMA) guidelines and the Cochrane Handbook for

98 Systematic Reviews Interventions13,14.

99

D
100 Search Strategy

101 An electronic literature search was performed in the Pubmed, Scopus, and Cochrane

TE
102 Library databases to identify relevant studies between January 1960 and November 2015. The

103 following keywords were used: “clavicle” and “fracture” and “plate” or “plating” or “plated.” A

104 manual search of the references of the selected studies was also performed to identify any

105 additional potential articles that may meet inclusion criteria.

EP
106

107 Inclusion Criteria

108 Two reviewers (A.N. and P.S.V.) evaluated the titles and abstracts to identify relevant
C

109 studies. Disagreements were discussed to determine a resolution. Only full text articles without

110 language restrictions were considered. The inclusion criteria consisted of: 1) fracture of the
C

111 midshaft clavicle; 2) surgery for an acute fracture (within one month of the injury); 3) adult-aged

112 patients (16 years of age and older); and 4) open reduction and internal fixation with plate
A

113 application in either the anteroinferior or superior position. The exclusion criteria were: 1)

114 inability to specify plate location; 2) patients with multi-system trauma; 3) minimally invasive

115 procedures; and 4) operations for revision, nonunion, malunion, and/or infection. Articles with

116 different study designs such as randomized controlled trials and cohort studies (prospective or

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 6

117 retrospective) were evaluated. Case studies, meta-analyses, and systematic reviews were not

118 considered.

119

120 Data Extraction

121 The same two reviewers carefully and independently extracted data from eligible studies.

D
122 The following basic characteristics were recorded from each study: 1) first author’s name; 2) title

123 of study; 3) year published; 4) journal; 5) patients’ age; 6) number of patients; 7) plate location;

TE
124 and 8) follow-up period. The primary measured outcomes were symptomatic hardware (implant

125 prominence or irritation) and surgery to remove symptomatic hardware. The secondary outcomes

126 were time to union, fracture union, nonunion, malunion, DASH score, Constant score, and

127 implant failure (defined as plate breakage) 15,16. Both primary and secondary outcomes were
EP
128 extracted from the studies according to their availability, as it was common for studies to

129 examine some but not all of the outcomes. For continuous outcomes, the mean and standard

130 deviation were also recorded. If the mean or standard deviation was not available but a substitute
C

131 (e.g., median, IQR, range) was available, this was recorded and used to estimate the standard

132 deviation and mean.

C

133

134 Statistical Analyses

A

135 Frequencies and proportions of cases were recorded for binary outcomes, while means

136 and standard deviations were recorded for continuous outcomes. Other summary statistics

137 provided were used to impute means and standard deviations under the assumption of normality

138 when these were not reported. Continuous outcomes were compared between groups using linear

139 mixed effects models, while binary outcomes were compared using mixed effects logistic

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 7

140 regression models, including fixed group effects and random study effects. P-values less than

141 0.05 were considered statistically significant. All analyses were performed using SAS version

142 9.4 (SAS Institute Inc., Cary, NC).

143 RESULTS

144 Literature Search

D
145 The initial database search yielded 1,428 articles. 897 studies remained after removing

146 duplicate articles (Figure 1). After screening the abstracts and titles, 57 relevant studies were

TE
147 identified. From these, studies were further excluded due to an inability to specify plate location

148 (6), a subject pool not exclusively consisting of acute fractures (4) or midshaft fractures (2), a

149 minimally invasive surgical approach (6), use of non-standard plates (1), inadequate reporting of

150 functional outcomes (2), and a duplicate group of patients (2). Thirty-four articles were used in
EP
151 the meta-analysis (Table, Supplemental Digital Content 1). Of these, 8 studies reported on

152 anteroinferior plate fixation (N=390, Table, Supplemental Digital Content 2) and 27 studies

153 reported on superior plate fixation (N=1104, Table, Supplemental Digital Content 3). The studies
C

154 were all published between the years 2007 and 2015.

155
C

156 Surgical and Functional Outcomes

157 Eight studies (390 patients) in the anteroinferior group and 27 studies (1,104 patients)
A

158 reported on the clavicle union rate (Table 1). The rate of fracture union was similar in the two

159 groups: 0.97 for the anteroinferior group and 0.98 for the superior group, (p=0.41). Sufficient

160 data was available from 4 studies (125 patients) from the anteroinferior group and 11 studies

161 (376 patients) from the superior group to perform a meta-analysis on the average time to union

162 (Table 2). The anteroinferior group had an average time to union of 15.82 weeks (95% CI 11.43,

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 8

163 20.20), while the superior group had an average of 17.12 weeks (95% CI 12.73, 21.50),

164 (p<0.0001).

165 The DASH and Constant scores were the most commonly used functional outcome scores

166 reported. A low DASH score and high Constant score indicate superior shoulder function. In the

167 present analysis, adequate data was available from 3 studies (102 patients) in the anteroinferior

D
168 group and 6 studies (225 patients) from the superior group to evaluate DASH score. The mean

169 DASH score for the anteroinferior group was 5.18 (95% CI -0.60, 10.95), compared to 9.71

TE
170 (95% CI 5.24, 14.19) for the superior group, (p=0.18). Constant score data was available in 4

171 studies (121 patients) in the anteroinferior group and in 17 studies (679 patients) in the superior

172 group. There was no difference in the mean Constant score between the anteroinferior group

173 (90.90; 95% CI 86.88, 94.92) as compared to the superior group (93.34; 95% CI 91.40, 95.28),
EP
174 (p=0.27).

175

176 Postoperative Complications

C

177 Seven studies (351 patients) in the anteroinferior group and 27 studies (1,104 patients) in

178 the superior group reported on the rate of nonunion. The superior plating group and
C

179 anteroinferior group had similar nonunion rates (0.02 versus 0.03, respectively), (p=0.29). Data

180 from 5 studies (205 patients) in the anteroinferior group and 12 studies (489 patients) in the
A

181 superior group was available regarding the rate of malunion. Malunions can occur secondary to

182 poor plating technique or a loss of reduction post-operatively. The malunion rate was higher in

183 the superior group (0.01) as compared to the anteroinferior group (0.006), but this difference was

184 not significant (p=0.28) nor clinically relevant. Six studies (222 patients) in the anteroinferior

185 group and 15 studies (624 patients) in the superior group reported on the incidence of implant

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 9

186 failure. The superior plating group (0.03) and the anteroinferior group (0.02) had similar rates of

187 patients with implant failure (p=0.39).

188 Eight studies (390 patients) in the anteroinferior group reported on the rate of

189 symptomatic hardware, as did 22 studies (848 patients) in the superior group. Meta-analysis

190 found that the superior plating group had a significantly higher probability of experiencing

D
191 symptomatic hardware (0.17) as compared to the anteroinferior group (0.08), (p=0.005) (Figure

192 2). Seven studies (371 patients) in the anteroinferior plating group and 19 studies (739 patients)

TE
193 in the superior plating group reported on surgery to remove symptomatic hardware. The superior

194 plating group had a significantly higher rate of surgery for implant removal (0.11 versus 0.05),

195 (p=0.008) (Figure 3).

196
EP
197 DISCUSSION

198 There has been a trend towards surgical management of midshaft clavicle fractures in

199 efforts to reduce the incidence of nonunion and malunion and to improve shoulder function4-7.
C

200 Although surgical treatment has the ability to mitigate poor outcomes, it also has potential

201 complications. A systematic review performed by Wijdicks and colleagues found that the
C

202 majority of complications following plate fixation of clavicle fractures were related to implant

203 irritation or failure17. Subsequent surgery may be necessary in order to treat these complications.
A

204 Wang et al. performed follow up on 48 patients with middle third clavicle fractures treated with

205 pre-contoured plates and found that 88% complained of local prominence, pain, and

206 discomfort18. Ultimately 56% of the initial study group had the plates removed, and post-

207 operatively 96% of those with plates removed were satisfied and recommended plate removal.

208 This underscores the potential need for implant removal to achieve patient satisfaction.

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 10

209 Anteroinferior plate application has drawn more interest recently because it is potentially

210 associated with less frequent implant irritation and need for implant removal as compared to

211 superior plate application.

212 There are many factors that influence the approach to plate fixation for midshaft clavicle

213 fractures including the fracture pattern, associated injuries, surgeon preference for patient

D
214 positioning, and pre-existing deformity due to prior trauma. Plate fixation in the superior and the

215 anteroinferior positions are the two most common surgical approaches. The purpose of the

TE
216 present investigation was to perform a comprehensive literature search and identify studies in

217 which plate fixation was performed in either the superior or anteroinferior position in order to

218 compare the complication rates, particularly implant prominence and need for implant removal,

219 between these two groups. This analysis identified 8 studies in the anteroinferior plating group
EP
220 and 27 studies in the superior plating group. No significant differences were found between the

221 two groups with regards to the probability of having a nonunion or malunion. The anteroinferior

222 group had an average time to union of 15.82 weeks compared to 17.12 weeks for the superior
C

223 group. Although this was statistically significant, a difference of 1.3 weeks is not a clinically

224 significant difference. Furthermore, this may be an artificial outcome as it is difficult to assess
C

225 union with an anteroinferior plate, and there is inherent variability as to when patients follow up

226 for radiographic studies in order to make this assessment. There was no statistically significant
A

227 difference between the two groups when examining their DASH scores and Constant scores.

228 This result was expected, as both superior and anteroinferior plating methods are effective

229 techniques for surgical treatment of clavicle fractures. With regards to implant complications,

230 there was a significantly higher rate of patients with symptomatic hardware and patients

231 undergoing implant removal when plating was performed on the superior aspect of the clavicle.

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 11

232 This finding is consistent with other studies that have suggested that plating on the superior

233 aspect may be associated with higher rates of symptomatic hardware and implant removal 8,11,12.

234 We were only able to identify one study in the literature that had directly compared the

235 two plating methods. In a retrospective review, Formaini et al. concluded that superior plating

236 had a significantly higher incidence of symptomatic hardware compared to anteroinferior

D
237 plating12. Their study found a trend towards an increased incidence of implant removal in the

238 superior plating group (19% versus 9%), however this was not statistically significant. A recent

TE
239 randomized clinical trial comparing 37 total patients treated with minimally invasive plating

240 found no significant differences between superior versus anteroinferior plating with regards to

241 functional scores, time to union, and complications19. While this study was a randomized trial

242 that examined plates in both positions, minimally invasive plating was performed, and this
EP
243 technique was excluded from the present investigation. In another recent study, retrospectively

244 collected data of 39 patients who underwent anteroinferior plating for a displaced midshaft

245 clavicle fracture were compared with a group of 60 patients treated with superior fixation in a
C

246 prior randomized controlled trial20,21. Although the results demonstrated a significantly higher

247 rate of asymptomatic patients with the plate still in place in the anteroinferior group (46% vs.
C

248 22%), univariate and multivariate regression analysis demonstrated that plate position was not

249 significantly associated with implant-related irritation and that patient age under 40 was
A

250 associated with irritation. The results of the present meta-analysis contrast those from previous

251 studies that suggest that plate position does not correlate with symptomatic hardware and plate

252 removal. Further randomized controlled trials may help to elucidate this question.

253 The current study does have some limitations. Anteroinferior plate fixation is not as

254 frequently described in the literature, and there are few studies that directly compare

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 12

255 complications following anteroinferior versus superior plate fixation. Certain outcomes such as

256 fracture time to union and shoulder function scores are influenced by follow-up intervals and by

257 inter-examiner variability. Furthermore, studies examining various surgeons using different plate

258 types can potentially influence outcomes. Pre-contoured plates have become the standard of care,

259 and several retrospective reviews have suggested that their use may be associated with decreased

D
260 implant prominence and need for implant removal22-24. In the current analysis, the details of the

261 plate type were reported inconsistently from study to study, with many studies not specifying

TE
262 whether pre-contoured plates were utilized. However, 6 studies in our meta-analysis did specify

263 using pre-contoured plates. Despite the lack of plate specificity, there are many similarities with

264 respect to the surgical technique. All but one study was published after 2010, indicating modern

265 plating techniques. Also, the plate size was highly consistent throughout the studies, with the
EP
266 majority using a 3.5-mm thickness plate. Even though the use of different plate types is a

267 limitation to the current study, we believe there are other similarities regarding the plating

268 techniques that can allow us to make a meaningful comparison between plating on the superior
C

269 versus the anteroinferior aspects of the clavicle. The follow-up interval may also affect the

270 development of symptomatic hardware. Of the studies included in this meta-analysis, the
C

271 majority had average follow-up times of over one year. However there were a few investigations

272 that had shorter or significantly longer follow-up. Finally, meta-analysis does not eliminate or
A

273 control for bias in each of the studies, and implant removal is typically a very subjective decision

274 between surgeon and patient and subject to multiple biases.

275 Despite these limitations, the strict inclusion and exclusion criteria of the present-study

276 made the two study groups as similar as possible. The patients included in the present analysis

277 were all adults who had open reduction and internal fixation performed with standard plates in

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 13

278 either the superior or anteroinferior position for isolated acute midshaft clavicle fractures. Finally,

279 by including data from 34 different studies with 1,494 patients, this investigation provides a

280 meaningful comparison due to the large sample size.

281

282 CONCLUSION

D
283 The current literature suggests that patients who are treated with superior clavicle plate

284 fixation may be more likely to experience symptomatic hardware and undergo removal of

TE
285 implants as compared to patients who are treated with anteroinferior plate fixation. Surgeons

286 should take into consideration the potential development of symptomatic hardware and need for

287 implant removal when selecting their approach for plate fixation of midshaft clavicle fractures. A

288 properly powered, randomized clinical trial of superior versus anteroinferior fixation using pre-
EP
289 contoured plates and objective, standardized criteria for implant removal would confirm or refute

290 these findings.

291
C

292
C

293 REFERENCES

294 1. Karl JW, Olson PR, Rosenwasser MP. The Epidemiology of Upper Extremity Fractures
A

295 in the United States, 2009. J Orthop Trauma. 2015;29(8):e242-244.

296 2. Postacchini F, Gumina S, De Santis P, Albo F. Epidemiology of clavicle fractures. J

297 Shoulder Elbow Surg. 2002;11(5):452-456.

298 3. NEER CS. Nonunion of the clavicle. J Am Med Assoc. 1960;172:1006-1011.

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 14

299 4. Rowe CR. An atlas of anatomy and treatment of midclavicular fractures. Clin Orthop

300 Relat Res. 1968;58:29-42.

301 5. Hill JM, McGuire MH, Crosby LA. Closed treatment of displaced middle-third fractures

302 of the clavicle gives poor results. J Bone Joint Surg Br. 1997;79(4):537-539.

303 6. Zlowodzki M, Zelle BA, Cole PA, Jeray K, McKee MD, Group E-BOTW. Treatment of

D
304 acute midshaft clavicle fractures: systematic review of 2144 fractures: on behalf of the

305 Evidence-Based Orthopaedic Trauma Working Group. J Orthop Trauma.

TE
306 2005;19(7):504-507.

307 7. McKee MD, Pedersen EM, Jones C, et al. Deficits following nonoperative treatment of

308 displaced midshaft clavicular fractures. J Bone Joint Surg Am. 2006;88(1):35-40.

309 8. Canadian Orthopaedic Trauma Society. Nonoperative treatment compared with plate
EP
310 fixation of displaced midshaft clavicular fractures. A multicenter, randomized clinical

311 trial. J Bone Joint Surg Am. 2007;89(1):1-10.

312 9. Robinson CM, Goudie EB, Murray IR, et al. Open reduction and plate fixation versus
C

313 nonoperative treatment for displaced midshaft clavicular fractures: a multicenter,

314 randomized, controlled trial. J Bone Joint Surg Am. 2013;95(17):1576-1584.

C

315 10. Khan LAK, Bradnock TJ, Scott C, Robinson CM. Fractures of the clavicle. Journal of

316 Bone and Joint Surgery - Series A. 2009;91(2):447-460.

A

317 11. Collinge C, Devinney S, Herscovici D, DiPasquale T, Sanders R. Anterior-inferior plate

318 fixation of middle-third fractures and nonunions of the clavicle. Journal of Orthopaedic

319 Trauma. 2006;20(10):680-686+746-747.

320 12. Formaini N, Taylor BC, Backes J, Bramwell TJ. Superior versus anteroinferior plating of

321 clavicle fractures. Orthopedics. 2013;36(7):e898-e904.

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 15

322 13. Moher D, Liberati A, Tetzlaff J, Altman DG, Group P. Preferred reporting items for

323 systematic reviews and meta-analyses: the PRISMA statement. J Clin Epidemiol.

324 2009;62(10):1006-1012.

325 14. Higgins JPT, Green S, Cochrane Collaboration. Cochrane handbook for systematic

326 reviews of interventions. Chichester, England ; Hoboken, NJ: Wiley-Blackwell; 2008.

D
327 15. Hudak PL, Amadio PC, Bombardier C. Development of an upper extremity outcome

328 measure: the DASH (disabilities of the arm, shoulder and hand) [corrected]. The Upper

TE
329 Extremity Collaborative Group (UECG). Am J Ind Med. 1996;29(6):602-608.

330 16. Constant CR, Murley AH. A clinical method of functional assessment of the shoulder.

331 Clin Orthop Relat Res. 1987(214):160-164.

332 17. Wijdicks FJ, Van der Meijden OA, Millett PJ, Verleisdonk EJ, Houwert RM. Systematic
EP
333 review of the complications of plate fixation of clavicle fractures. Arch Orthop Trauma

334 Surg. 2012;132(5):617-625.

335 18. Wang J, Chidambaram R, Mok D. Is removal of clavicle plate after fracture union
C

336 necessary? International Journal of Shoulder Surgery. 2011;5(4):85-89.

337 19. Sohn HS, Shon MS, Lee KH, Song SJ. Clinical comparison of two different plating
C

338 methods in minimally invasive plate osteosynthesis for clavicular midshaft fractures: A

339 randomized controlled trial. Injury. 2015;46(11):2230-2238.

A

340 20. Hulsmans MH, van Heijl M, Houwert RM, Timmers TK, van Olden G, Verleisdonk EJ.

341 Anteroinferior versus superior plating of clavicular fractures. J Shoulder Elbow Surg.

342 2016;25(3):448-454.

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 16

343 21. van der Meijden OA, Houwert RM, Hulsmans M, et al. Operative treatment of dislocated

344 midshaft clavicular fractures: plate or intramedullary nail fixation? A randomized

345 controlled trial. J Bone Joint Surg Am. 2015;97(8):613-619.

346 22. Vanbeek C, Boselli KJ, Cadet ER, Ahmad CS, Levine WN. Precontoured plating of

347 clavicle fractures: Decreased hardware-related complications? Clinical Orthopaedics and

D
348 Related Research. 2011;469(12):3337-3343.

349 23. Schemitsch LA, Schemitsch EH, Kuzyk P, McKee MD. Prognostic Factors for Re-

TE
350 operation Following Plate Fixation of Fractures of the Midshaft Clavicle. J Orthop

351 Trauma. 2015.

352 24. Rongguang A, Zhen J, Jianhua Z, Jifei S, Xinhua J, Baoqing Y. Surgical Treatment of

353 Displaced Midshaft Clavicle Fractures: Precontoured Plates Versus Noncontoured Plates.
EP
354 J Hand Surg Am. 2016;41(9):e263-266.

355 25. Chen CE, Juhn RJ, Ko JY. Anterior-inferior plating of middle-third fractures of the

356 clavicle. Archives of Orthopaedic and Trauma Surgery. 2010;130(4):507-511.

C

357 26. Hundekar BB. Internal fixation of displaced middle third fractures of clavicle with

358 precontoured locking plate. Journal of Orthopaedics. 2013;10(2):79-85.

C

359 27. Virtanen KJ, Remes V, Pajarinen J, Savolainen V, Björkenheim JM, Paavola M. Sling

360 compared with plate osteosynthesis for treatment of displaced midshaft clavicular
A

361 fractures: A randomized clinical trial. Journal of Bone and Joint Surgery - Series A.

362 2012;94(17):1546-1553.

363 28. Sohn HS, Kim WJ, Shon MS. Comparison between open plating versus minimally

364 invasive plate osteosynthesis for acute displaced clavicular shaft fractures. Injury. 2015.

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 17

365 29. Douraiswami B, Naidu DK, Thanigai S, Anand V, Dhanapal R. Open reduction and

366 plating for displaced mid third clavicle fractures - A prospective study. Journal of

367 Clinical Orthopaedics and Trauma. 2013;4(4):174-179.

368 30. Assobhi JE. Reconstruction plate versus minimal invasive retrograde titanium elastic nail

369 fixation for displaced midclavicular fractures. J Orthop Traumatol. 2011;12(4):185-192.

D
370 31. Galdi B, Yoon RS, Choung EW, et al. Anteroinferior 2.7-mm versus 3.5-mm plating for

371 AO/OTA type B clavicle fractures: A comparative cohort clinical outcomes study.

TE
372 Journal of Orthopaedic Trauma. 2013;27(3):121-125.

373 32. Gilde AK, Jones CB, Sietsema DL, Hoffmann MF. Does plate type influence the clinical

374 outcomes and implant removal in midclavicular fractures fixed with 2.7-mm

375 anteroinferior plates? A retrospective cohort study. Journal of Orthopaedic Surgery and
EP
376 Research. 2014:55.

377 33. Jones CB, Sietsema DL, Ringler JR, Endres TJ, Hoffmann MF. Results of anterior-

378 inferior 2.7-mm dynamic compression plate fixation of midshaft clavicular fractures.
C

379 Journal of Orthopaedic Trauma. 2013;27(3):126-129.

380 34. Van Der Ven Denise JC, Timmers TK, Flikweert PE, Van Ijseldijk ALA, Van Olden
C

381 GDJ. Plate fixation versus conservative treatment of displaced midshaft clavicle

382 fractures: Functional outcome and patients' satisfaction during a mean follow-up of 5
A

383 years. Injury. 2015;46(11):2223-2229.

384 35. Choudhari P, Chhabra. Displaced mid-shaft clavicle fractures: a subset for surgical

385 treatment. Malays Orthop J. 2014;8(2):1-5.

386 36. Das A, Rollins KE, Elliott K, et al. Early versus delayed operative intervention in

387 displaced clavicle fractures. Journal of Orthopaedic Trauma. 2014;28(3):119-123.

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 18

388 37. D'Heurle A, Le T, Grawe B, et al. Perioperative risks associated with the operative

389 treatment of clavicle fractures. Injury. 2013;44(11):1579-1581.

390 38. Dhoju D, Shrestha D, Parajuli NP, Shrestha R, Sharma V. Operative fixation of displaced

391 middle third clavicle (Edinburg type 2) fracture with superior reconstruction Plate

392 Osteosynthesis. Kathmandu University Medical Journal. 2011;9(36):286-290.

D
393 39. Ferran NA, Hodgson P, Vannet N, Williams R, Evans RO. Locked intramedullary

394 fixation vs plating for displaced and shortened mid-shaft clavicle fractures: A randomized

TE
395 clinical trial. Journal of Shoulder and Elbow Surgery. 2010;19(6):783-789.

396 40. Jiang H, Qu W. Operative treatment of clavicle midshaft fractures using a locking

397 compression plate: Comparison between mini-invasive plate osteosynthesis (MIPPO)

398 technique and conventional open reduction. Orthopaedics and Traumatology: Surgery
EP
399 and Research. 2012;98(6):666-671.

400 41. Khorami M, Fakour M, Mokarrami H, Arti HR, Nasab AM, Shahrivar F. The

401 Comparison of Results of Treatment of Midshaft Clavicle Fracture between Operative

C

402 Treatment with Plate and Non-Operative Treatment. Arch Bone Jt Surg. 2014;2(3):210-

403 214.
C

404 42. Kulshrestha V, Roy T, Audige L. Operative versus nonoperative management of

405 displaced midshaft clavicle fractures: A prospective cohort study. Journal of Orthopaedic
A

406 Trauma. 2011;25(1):31-38.

407 43. Liu HH, Chang CH, Chia WT, Chen CH, Tarng YW, Wong CY. Comparison of plates

408 versus intramedullary nails for fixation of displaced midshaft clavicular fractures.

409 Journal of Trauma - Injury, Infection and Critical Care. 2010;69(6):E82-E87.

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 19

410 44. Mirzatolooei F. Comparison between operative and nonoperative treatment methods in

411 the management of comminuted fractures of the clavicle. Acta Orthopaedica et

412 Traumatologica Turcica. 2011;45(1):34-40.

413 45. Mohammed E, Mahadane H, Mechchat A, Shimi M, Elibrahimi A, Elmrini A. Functional

414 outcome of midclavicular fracture fixation utilising a reconstruction plate. Malays Orthop

D
415 J. 2013;7(3):6-9.

416 46. Ozler T, Güven M, Kocadal AO, Uluçay C, Beyzadeoğlu T, Altıntaş F. Locked anatomic

TE
417 plate fixation in displaced clavicular fractures. Acta Orthop Traumatol Turc.

418 2012;46(4):237-242.

419 47. Saha P, Datta P, Ayan S, Garg AK, Bandyopadhyay U, Kundu S. Plate versus titanium

420 elastic nail in treatment of displaced midshaft clavicle fractures: A comparative study.
EP
421 Indian J Orthop. 2014;48(6):587-593.

422 48. Shin SJ, Do NH, Jang KY. Risk factors for postoperative complications of displaced

423 clavicular midshaft fractures. Journal of Trauma and Acute Care Surgery.
C

424 2012;72(4):1046-1050.

425 49. Tarng YW, Yang SW, Fang YP, Hsu CJ. Surgical management of uncomplicated
C

426 midshaft clavicle fractures: A comparison between titanium elastic nails and small

427 reconstruction plates. Journal of Shoulder and Elbow Surgery. 2012;21(6):732-740.

A

428 50. Wenninger Jr JJ, Dannenbaum JH, Branstetter JG, Arrington ED. Comparison of

429 complication rates of intramedullary pin fixation versus plating of midshaft clavicle

430 fractures in an active duty military population. Journal of surgical orthopaedic advances.

431 2013;22(1):77-81.

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 20

432 51. Wijdicks FJ, Houwert M, Dijkgraaf M, et al. Complications after plate fixation and

433 elastic stable intramedullary nailing of dislocated midshaft clavicle fractures: A

434 retrospective comparison. International Orthopaedics. 2012;36(10):2139-2145.

435 52. Narsaria N, Singh AK, Arun GR, Seth RR. Surgical fixation of displaced midshaft

436 clavicle fractures: elastic intramedullary nailing versus precontoured plating. J Orthop

D
437 Traumatol. 2014;15(3):165-171.

438 53. Ranalletta M, Rossi LA, Bongiovanni SL, Tanoira I, Piuzzi NS, Maignon G. Surgical

TE
439 treatment of displaced midshaft clavicular fractures with precontoured plates. Journal of

440 Shoulder and Elbow Surgery. 2015;24(7):1036-1040.

441 54. Andrade-Silva FB, Kojima KE, Joeris A, Santos Silva J, Mattar R. Single, superiorly

442 placed reconstruction plate compared with flexible intramedullary nailing for midshaft
EP
443 clavicular fractures: a prospective, randomized controlled trial. J Bone Joint Surg Am.

444 2015;97(8):620-626.

445 FIGURE LEGENDS

C

446 Figure 1. Flow chart of studies evaluated in the present meta-analysis.

C

447 Figure 2. Forrest plot demonstrating rate of symptomatic hardware by study. Studies reporting

448 zero symptomatic hardware cases (Chen (2010)25 and Hundekar (2013)26) are not displayed but
A

449 were included in the meta-analysis.

450 Figure 3. Forrest plot demonstrating rate of implant removal by study. Studies reporting zero

451 removed implant cases (Chen (2010)25, Virtanen (2012)27, Sohn (2015)28, Hundekar (2013)26 and

452 Douraiswami (2013)29) are not displayed but were included in the meta-analysis.

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 Anteroinferior Superior
OR (95% CI) p-value ICC
P (95% CI) P (95% CI)

Infections 0.02 (0.003, 0.14) 0.09 (0.06, 0.13) 4.38 (-0.58, 33.37) 0.15 0.007

Implant Failure 0.02 (0.004, 0.06) 0.03 (0.01, 0.06) 1.83 (0.47, 7.18) 0.39 0.20

0.01 (0.004,
Malunion 0.006 (0.001, 0.04) 2.27 (0.51, 10.12) 0.28 0.34
0.05)

D
Nonunion 0.03 (0.01, 0.07) 0.02 (0.01, 0.03) 0.59 (0.22, 1.57) 0.29 0.16

Union 0.97 (0.94, 0.99) 0.98 (0.96, 0.99) 1.47 (0.59, 3.67) 0.41 0.15

Complication 0.08 (0.04, 0.13) 0.17 (0.13, 0.23) 2.51 (1.32, 4.78) 0.005 0.15

TE
Hardware
0.05 (0.03, 0.08) 0.11 (0.08, 0.14) 2.36 (1.26, 4.42) 0.008 0.05
Removal

Table 1. Meta-analysis results for count variables. ICC is Intraclass Correlation Coefficient.
EP
C
C
A

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 Anteroinferior Superior Difference p-value ICC

Mean (95% CI) Mean (95% CI)

Constant Score 90.90 (86.88, 94.92) 93.34 (91.40, 95.28) -2.44 (-6.90, 2.03) 0.27 0.18

DASH Score 5.18 (-0.60, 10.95) 9.71 (5.24, 14.19) -4.54 (-11.84, 2.77) 0.18 0.19

Time to union
15.82 (11.43, 20.20) 17.12 (12.73, 21.50) -1.30 (-1.61, -0.99) <.0.0001 0.64

D
(weeks)

Table 2. Meta-analysis results for continuous variables. ICC is Intraclass Correlation Coefficient.

TE
EP
C
C
A

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 D
TE
EP
C
C
A

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 D
TE
EP
C
C
A

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 D
TE
EP
C
C
A

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.