Arch Orthop Trauma Surg (2014) 134:1493–1500 DOI 10.1007/s00402-014-2077-6
ORTHOPAEDIC SURGERY
Operative versus nonoperative treatment for displaced midshaft clavicular fractures: a meta-analysis of randomized clinical trials Lingde Kong • Yingze Zhang • Yong Shen
Received: 2 March 2014 / Published online: 29 August 2014 Ó Springer-Verlag Berlin Heidelberg 2014
Abstract Introduction There is insufficient evidence to indicate whether operative or nonoperative treatment is better for treating displaced midshaft clavicular fractures. We undertook a systematic review and meta-analysis of randomized controlled trials (RCTs) to compare the effects of the two treatments. Methods We searched the PubMed, EMBASE and Cochrane Library databases, and identified RCTs to compare the Constant score, DASH score, nonunion, malunion, and adverse events between operative and nonoperative groups of patients with displaced midshaft clavicular fractures. Results A total of 507 patients from 6 RCTs were subjected to meta-analysis. Operative treatment has an effect on improving function, which is demonstrated by significantly higher Constant scores (P = 0.0003) and lower DASH scores (P = 0.03). The rate of nonunion and the rate of malunion were significantly lower in operative group compared with that in nonoperative group (both P \ 0.0001). However, the rate of adverse events was significantly higher in operative group compared with that in nonoperative group (P = 0.003). Conclusions Operative treatment provided a significantly better functional outcome, a lower rate of nonunion and malunion, but was accompanied with a higher rate of adverse events. However, the results should be interpreted with caution and further large-scale, well-designed RCTs on this topic are still needed.
L. Kong Y. Zhang Y. Shen (&) Department of Orthopedics, The Third Hospital of Hebei Medical University, 139 Ziqiang Road, Shijiazhuang 050051, Hebei, People’s Republic of China e-mail: [email protected]
Keywords Bone fracture Clavicle Operative treatment Nonoperative treatment Meta-analysis
Introduction The clavicular fracture is one of the most common skeletal injuries that accounts for 5–10 % of all fractures. Around 80 % of clavicle fractures involve the midshaft and over half of these fractures are displaced [1, 2]. Conventionally, the fractures have been treated nonoperatively with immobilization in a sling or a figure-of-eight bandage, with the expectation of a good clinical outcome [3, 4], but some studies have emphasized the high risk of nonunion or symptomatic malunion following nonoperative treatment [5–7]. The outcome of nonoperative treatment is not as favorable as thought and there has been a growing trend to treat these fractures surgically [8, 9]. However, before advocating a policy of routine open reduction and internal fixation, further evaluation of the benefits to patients is important. Nowadays, there are several published randomized controlled trials (RCTs) regarding comparison between operative and nonoperative treatment for treating clavicular fracture. A previous metaanalysis including available RCTs was performed by McKee et al. [10] and was published in the year of 2012, but their review did not analyze functional outcomes quantitatively. Moreover, several relevant studies on this topic have been published in recent years. Thus, we undertook a meta-analysis based on high-quality data from RCTs to compare functional outcome, fracture-healing and adverse events between operative and nonoperative groups of patients with displaced midshaft clavicular fractures. Our hypothesis was that operative treatment would result in better functional outcome, more satisfactory fracture-
123
1494
Arch Orthop Trauma Surg (2014) 134:1493–1500
healing, and less adverse events than nonoperative treatment.
Data were extracted independently by two authors. Any disagreements concerning paper eligibility were resolved by discussion and consensus.
Materials and methods
Assessment of methodological quality
Literature search and inclusion criteria
The methodological quality of the studies was evaluated independently by two authors, without masking the trial names. The reviewers followed the instructions provided in the Cochrane Handbook for Systematic Reviews of Interventions [25]. The following domains were assessed: random sequence generation, allocation concealment, blinding, incomplete data outcomes, revealing of selective outcomes and any remaining biases. When the information in the study was inadequate, attempts were made to contact the authors in order to ensure that the study was evaluated correctly.
We searched the PubMed, EMBASE and Cochrane Library databases for English language studies that had been published before January 2014. Key words combined with MeSH terms, including clavicle, clavicular, and fractures, were used for searching. The search was limited to human subjects, and RCTs. In addition, reference lists of all the selected articles were hand-searched to identify other potentially eligible trials. This process was performed iteratively until no additional articles could be identified. The following inclusive selection criteria were applied: (1) study design: RCTs; (2) population: patients (16 years B age B 70 years) with a middle-third clavicular fracture that was completely displaced; (3) intervention: surgical treatment including plate fixation and intramedullary fixation; (4) comparisons: conservative treatment including slings, figure-of-eight bandages, and other physical treatments; (5) outcomes: functional outcomes measured by the Disability of the Arm, Shoulder, and Hand questionnaire (DASH) or the Constant score, fracture-healing measured by the rate of nonunion or malunion, and adverse events; (6) length of follow-up: at least 1 year. Trials were excluded if they (1) were abstracts, letters, reviews, or case reports; (2) had repeated data; and (3) did not report outcomes of interest. Data extraction and outcome measures The data included the general characteristics of each study and the outcomes measured. General characteristics included first author, year of publication, study design, number of patients, types of fracture, internal fixation, nonoperative treatment, and length of follow-up. The outcomes measured included: the DASH score, the Constant score, the rate of nonunion, malunion, and adverse events. When the same population was reported in several publications, we retained only the most informative article or complete study to avoid duplication of information. We also tried to contact the authors of primary studies to request missing data. For continuous outcomes with no standard deviations reported, we calculated standard deviations from standard errors, P values, or confidence intervals, according to the methods described in the Cochrane Handbook for Systematic Reviews of Interventions [25].
123
Statistical analysis Differences were expressed as risk ratios (RRs) with 95 % confidence intervals (CIs) for dichotomous outcomes and weighted mean differences (WMDs) with 95 % CIs for continuous outcomes. Heterogeneity was analyzed with both the Chi square test and the I2 test. A P value of \0.10 for the Chi square test was interpreted as evidence of statistical heterogeneity, and I2 was used to estimate total variation across the studies. A fixed-effect model was adopted if there was no statistical evidence of heterogeneity, and a random-effect model was adopted if statistically significant heterogeneity was present. Studies with an I2 statistic of 25–50 % were considered to have low heterogeneity, those with an I2 statistic of 50–75 % had moderate heterogeneity, and those with an I2 statistic of [75 % had high heterogeneity. Because patient characteristics, types of fracture, interventions, and other confounding factors were not consistent between studies, we further conducted sensitivity analysis to identify potential sources of heterogeneity. The presence of publication bias was assessed using the Begg and Egger tests. A P value \0.05 was judged as statistically significant, except where otherwise specified. All statistical analyses were performed using Review Manager version 5.1 (The Cochrane Collaboration, Software Update, Oxford, UK).
Results Study identification and selection A total of 68 records was identified by the initial database search. Twenty-one records were excluded because of
Fig. 1 Preferred reporting items for systematic reviews and meta-analyses (PRISMA) flow diagram outlining literature search results
Identification
Arch Orthop Trauma Surg (2014) 134:1493–1500
1495
68 records identified through database searching (PubMed, EMBASE and Cochrane Library databases)
Eligibility
Screening
47 records after duplicates removed
47 potentially relevant records screened
11 full-text articles assessed for eligibility
36 records excluded based on the titles and abstracts (reviews, non-randomized studies, or not relevant to our analysis)
5 full-text articles excluded: 1 no outcomes of interest reported 3 duplicated data 1 currently ongoing
Included
6 studies included in qualitative synthesis
6 studies included in quantitative synthesis (meta-analysis)
duplicate studies and 36 were excluded for various reasons (reviews, non-randomized studies, or not relevant to our analysis) on the basis of the titles and abstracts. The remaining 11 were retrieved for full text review, and 5 of them were excluded because 1 did not report outcomes of interest [11], 3 reported duplicated data [12–14], and 1 were currently ongoing [15]. Finally, 6 RCTs that met our inclusion criteria were included in the present meta-analysis [16–21]. The selection process for RCTs included in the meta-analysis is shown in Fig. 1. Study characteristics The main characteristics of the 6 RCTs included in the meta-analysis are presented in Table 1. These studies were published between 2007 and 2013. The sizes of the RCTs ranged from 57 to 200 patients. There were 507 patients at the time of final follow-up. A total of 259 patients underwent surgery, and the remaining 248 patients received conservative treatment. All studies focused on displaced midshaft fracture of the clavicle, except one study included patients with displaced, comminuted fracture [19]. Operative techniques included plate fixation in four studies [16, 19–21] and intramedullary pin/nail fixation in two studies [17, 18]. A sling was used as conservative treatment in five studies, and a collar and cuff was used in one study [21].
All patients were followed up for 12 months, with the exception of one study having patients followed up for 24 months [18]. Assessment of risk of bias The risk of bias is demonstrated graphically in Fig. 2 and summarized in Fig. 3. The randomization technique was mentioned in all 6 trials. However, only 3 trials described the process of random sequence generation [16, 18, 20], and 5 trials stated the method of allocation concealment [16–20]. Blinding is rarely used in orthopedic surgery trials and only one study was blinded in the assessment of outcome [21]. Three trials were judged to be at high risk of attrition bias [18, 20, 21], and the other three were at low risk [16, 17, 19]. Outcome Functional outcome Five RCTs reported the Constant score in study patients. The test for heterogeneity was significant, and the studies have moderate heterogeneity (P for heterogeneity = 0.02, I2 = 67 %). Using the random-effect model, the aggregated results suggested that the Constant score was greater
123
1496
Arch Orthop Trauma Surg (2014) 134:1493–1500
Table 1 General characteristics of each study Study (author/year)
Study design
No. of patients assigneda
Range of ages (years)
Length of follow-up (months)
No. of patients assesseda
Types of fracture
Internal fixation
Nonoperative treatment
COTS/2007
Multicenter, RCT
67/65
16–60
12
62/49
Displaced
Plate
Sling
Judd/2009
RCT
29/28
17–40
12
29/28
Displaced
Pin
Sling
Smekal/2009
RCT
33/35
18–65
24
30/30
Displaced
Nail
Sling
Mirzatolooei/ 2011
RCT
29/31
18–65
12
26/24
Comminuted, displaced
Plate
Sling
Virtanen/ 2012
RCT
28/32
18–70
12
26/25
Displaced
Plate
Sling
Robinson/ 2013
Multicenter, RCT
95/105
16–60
12
86/92
Displaced
Plate
Collar and cuff
a
Operative/nonoperative treatment
Fig. 2 Risk of bias graph: a review of the authors’ judgments regarding each risk of bias item, presented as percentages across all included studies
in the operative group in comparison to the nonoperative group at a statistically significant level (WMD 5.19, 95 % CI 2.37–8.01, P \ 0.0003) (Fig. 4). Subsequently, we performed sensitivity analysis to explore potential sources of heterogeneity. Exclusion of the trial conducted by Mirzatolooei et al. [19] resolved the heterogeneity (P for heterogeneity = 0.58, I2 = 0 %), but did not materially alter the pooled results (WMD 4.05, 95 % CI 2.44–5.66, P \ 0.00001). Information on the DASH score was provided in only four studies. The test for heterogeneity was significant, and the studies have high heterogeneity (P for heterogeneity \ 0.00001, I2 = 91 %). Using the random-effect model, the aggregated results suggested that the DASH score was lower in the operative group in comparison to the nonoperative group at a statistically significant level (WMD -6.60, 95 % CI -12.68 to -0.52, P = 0.03) (Fig. 5). Similarly, exclusion of the trial conducted by Mirzatolooei et al. [19] reduced the heterogeneity (P for heterogeneity = 0.32, I2 = 12 %), without materially altering the pooled results (WMD -3.32, 95 % CI -5.45 to -1.18, P = 0.002).
123
Fracture-healing All six studies provided data on nonunion. The test for heterogeneity was not significant, and the studies have no statistical evidence of heterogeneity (P for heterogeneity = 0.48, I2 = 0 %). Using the fixed-effect model, the rate of nonunion was significantly lower in operative group compared with that in nonoperative group (RR 0.16, 95 % CI 0.07–0.38, P \ 0.0001) (Fig. 6). Malunion was referred to symptomatic malunion instead of radiographic malunion. The test for heterogeneity was not significant, and the studies have no statistical evidence of heterogeneity (P for heterogeneity = 0.76, I2 = 0 %). Using the fixed-effect model, the rate of malunion was significantly lower in operative group compared with that in nonoperative group (RR 0.15, 95 % CI 0.07–0.34, P \ 0.00001) (Fig. 7). Adverse events We collected a set of adverse events that was reported in the RCTs, such as wound infection, implant failure,
Arch Orthop Trauma Surg (2014) 134:1493–1500
neurologic symptom, and refracture, and listed them in Table 2. As nonunion and malunion were analyzed above, they were not included here. The study records of these events were relatively inconsistent among studies and the definition of complications varied. So, we just pooled the rate of overall events without performing further subgroup analysis. The test for heterogeneity was not significant, and the studies have no statistical evidence of heterogeneity (P for heterogeneity = 0.55, I2 = 0 %). Using the fixedeffect model, the rate of adverse event was significantly
1497
higher in operative group compared with that in nonoperative group (RR 2.14, 95 % CI 1.30–3.52, P = 0.003) (Fig. 8). The common operative complications consisted of local hardware irritation, implant failure, and wound infection. The predominant complications in the nonoperative group were neurologic symptoms, and stiffness or restriction of shoulder movement. Publication bias Publication bias was not assessed because of the limited number (\10) of studies included in each analysis.
Discussion
Fig. 3 Risk of bias summary: a review of the authors’ judgments regarding each risk of bias item, for each included study
The pooled results from the meta-analysis of six RCTs suggest that operative treatment provided a significantly better functional outcome, a lower rate of nonunion and malunion, but was accompanied with a higher rate of adverse events. This meta-analysis assessed the effects of operative and nonoperative treatment for displaced midshaft clavicular fractures. McKee et al. [10] have performed a similar metaanalysis and published it in the year of 2012. However, their study did not quantitatively evaluate the functional outcomes. Besides, after his review, three relevant RCTs have been reported [19–21]. In the current study, we included all available RCTs, and made a meta-analysis of the functional outcomes. The results showed that operative treatment has an effect on improving function, which is demonstrated by significantly higher Constant scores and lower DASH scores after at least one-year follow-up. In the assessment of functional outcome, the studies showed significant heterogeneity. Our sensitivity analysis found that a trial conducted by Mirzatolooei et al. [19] contributed to the heterogeneity. Exclusion of this trial resolved the heterogeneity, but did not materially alter the pooled results. This adds robustness to this finding. Different from the features of other study populations, this
Fig. 4 Forest plot showing the comparison of the Constant score between the operative group and the nonoperative group. CI confidence interval
123
1498
Arch Orthop Trauma Surg (2014) 134:1493–1500
Fig. 5 Forest plot showing the comparison of the DASH score between the operative group and the nonoperative group. CI confidence interval
Fig. 6 Forest plot showing the rate of nonunion between the operative group and the nonoperative group. CI confidence interval
Fig. 7 Forest plot showing the rate of malunion between the operative group and the nonoperative group. CI confidence interval
study focused on patients with dislocated, comminuted fractures. As shown in Figs. 4 and 5, the mean difference (MD) between groups in this study was more significant than that in any of other studies, which means that patients with dislocated, comminuted fracture benefit more from surgery than those with dislocated fracture alone. This study may provide additional interesting clues for future research on this specific fracture. Our review shows that 13.7 % patients in the nonoperative group developed a nonunion, which is significantly higher than the 1.9 % rate of nonunion in the operative group. Another complication significantly more common in the nonoperative group is malunion (25 % in the nonoperative group versus 2.8 % in the operative group). Our
123
findings contribute to the evidence that internal fixation is a reliable method for diminishing the risk of nonunion and malunion after displaced midshaft clavicular fracture [22, 23]. Previous researches have investigated the prognostic factors that may predict nonunion or malunion in nonoperatively treated patients. For example, Virtanen et al. [20] reported that displacement of more than 1.5 clavicle widths was a risk factor for nonunion. Hill et al. [5] concluded that displacement of the fracture fragments by more that 2 cm was associated with an unsatisfactory result. However, in this review, we were not able to draw any specific conclusions as to which individuals are most likely to have nonunion or malunion, as most trials did not provide relevant data.
Arch Orthop Trauma Surg (2014) 134:1493–1500 Table 2 Adverse events in both groups
Study (author/year) COTS/2007
1499
Operative group (N = 259 patients)
Nonoperative group (N = 248 patients)
Three wound infections
Seven transient brachial plexuses
Five hardware irritations (removal required) Eight transient brachial plexuses Judd/2009
Six superficial pin-tract infections
1 refracture
One partial, transient radial nerve palsy One intramedullary pin fracture Smekal/2009 Mirzatolooei/2011
Virtanen/2012
Two implant failure Five medial nail protrusion
Three transient neurogenic compromise
Two skin dysesthesia
Two shoulder imbalance
One infection Two hypertrophic scar
Two neurologic claudication in upper extremity
One plate bent
One brachial plexus irritation
One plate broken
Two refracture
One hardware irritation One refracture Robinson/2013
Two rotator cuff impingement
One rotator cuff impingement
One adhesive capsulitis One refracture One plate bent
Fig. 8 Forest plot showing the rate of adverse events between the operative group and the nonoperative group. CI confidence interval
We extracted adverse events and listed them in a table. Patients in operative group showed a higher rate of overall adverse event. Most of these complications were hardwarerelated, such as hardware irritation and implant failure. Wound infection and skin dysesthesia problems were also common after surgical treatment. However, in patients after conservative treatment, neurologic symptoms were common problems. Surgeons can use this information to help patients make optimal decisions in a shared decisionmaking process. Theoretically, for postoperative adverse events, some could be reduced by improved surgical techniques, and some could be resolved by removal of hardware.
This review showed that patients treated surgically could get a better restoration of functional outcome, a lower rate of nonunion and malunion, but our results do not provide sufficient evidence to support the routine use of open reduction and internal fixation for all displaced clavicular fractures in adults. Offering surgery to all patients would expose large numbers of patients who would otherwise have healed and regained good shoulder function without suffering potential surgical complications. Thus, a big challenge is to identify the subgroup of patients who might benefit from surgical intervention. Further investigation on this topic is necessary before advocating operative treatment widely.
123
1500
Arch Orthop Trauma Surg (2014) 134:1493–1500
Anyhow, this meat-analysis has several potential limitations. First, though we included only RCTs, all trials had methodological flaws, which include failure to bind the outcome assessor, lack of baseline outcome data, and so on. Consequently, the quantitative results of this review should be interpreted with caution. Second, two operative techniques were used in the trials for primary fixation of midshaft clavicular fractures (plate fixation and intramedullary pin/nail insertion). There may be fundamental differences in outcomes between the two techniques [24]. However, subgroup analysis was not performed because of rather small numbers of studies in both groups. Finally, we only included English language studies. Despite our best efforts in using multiple search methods, it is possible that we have omitted non-English language trials, the results of which may be applicable to our meta-analysis. In conclusion, in this meta-analysis of randomized clinical trials comparing the effect of operative versus nonoperative treatment for displaced midshaft clavicular fractures, certain benefits of internal fixation are noted, which include a good restoration of function, a low rate of nonunion and malunion. Despite these encouraging findings, patients should be informed of high rate of adverse event after surgery. Anyhow, the results should be interpreted with caution and further large-scale, well-designed RCTs on this topic are still needed. Conflict of interest
10.
11.
12.
13.
14.
15.
16.
17.
None. 18.
References 1. Pecci M, Kreher JB (2008) Clavicle fractures. Am Fam Physician 77(1):65–70 2. Postacchini F, Gumina S, De Santis P et al (2002) Epidemiology of clavicle fractures. J Shoulder Elbow Surg 11(5):452–456 3. Grassi FA, Tajana MS, D’Angelo F (2001) Management of midclavicular fractures: comparison between nonoperative treatment and open intramedullary fixation in 80 patients. J Trauma 50(6):1096–1100 4. Nordqvist A, Petersson CJ, Redlund-Johnell I (1998) Mid-clavicle fractures in adults: end result study after conservative treatment. J Orthop Trauma 12(8):572–576 5. Hill JM, McGuire MH, Crosby LA (1997) Closed treatment of displaced middle-third fractures of the clavicle gives poor results. J Bone Joint Surg Br 79(4):537–539 6. McKee MD, Pedersen EM, Jones C et al (2006) Deficits following nonoperative treatment of displaced midshaft clavicular fractures. J Bone Joint Surg Am 88(1):35–40. doi:10.2106/JBJS. D.02795 7. Nowak J, Holgersson M, Larsson S (2005) Sequelae from clavicular fractures are common: a prospective study of 222 patients. Acta Orthop 76(4):496–502. doi:10.1080/17453670510041475 8. Liu PC, Chien SH, Chen JC et al (2010) Minimally invasive fixation of displaced midclavicular fractures with titanium elastic nails. J Orthop Trauma 24(4):217–223. doi:10.1097/BOT. 0b013e3181b8ba33 9. Mueller M, Burger C, Florczyk A et al (2007) Elastic stable intramedullary nailing of midclavicular fractures in adults: 32
123
19.
20.
21.
22.
23.
24.
25.
patients followed for 1–5 years. Acta Orthop 78(3):421–423. doi:10.1080/17453670710014013 McKee RC, Whelan DB, Schemitsch EH et al (2012) Operative versus nonoperative care of displaced midshaft clavicular fractures: a meta-analysis of randomized clinical trials. J Bone Joint Surg Am 94(8):675–684. doi:10.2106/jbjs.j.01364 Yadav V, Khare GN, Singh S et al (2013) A prospective study comparing conservative with operative treatment in patients with a ‘floating shoulder’ including assessment of the prognostic value of the glenopolar angle. Bone Joint J 95-B(6):815–819. doi:10. 1302/0301-620x.95b6.31060 Schemitsch LA, Schemitsch EH, Veillette C et al (2011) Function plateaus by one year in patients with surgically treated displaced midshaft clavicle fractures. Clin Orthop Relat Res 469(12):3351–3355. doi:10.1007/s11999-011-1915-x Neuhaus V, Ring D (2013) Effect of different statistical methods on union or time to union in a published study about clavicular fractures. J Shoulder Elbow Surg 22(4):471–477. doi:10.1016/j. jse.2012.03.015 Pearson AM, Tosteson AN, Koval KJ et al (2010) Is surgery for displaced, midshaft clavicle fractures in adults cost-effective? Results based on a multicenter randomized, controlled trial. J Orthop Trauma 24(7):426–433. doi:10.1097/BOT.0b013e3181c3e505 Stegeman SA, de Jong M, Sier CF et al (2011) Displaced midshaft fractures of the clavicle: non-operative treatment versus plate fixation (Sleutel-TRIAL). A multicentre randomised controlled trial. BMC Musculoskelet Disord 12:196. doi:10.1186/ 1471-2474-12-196 Society COT (2007) Nonoperative treatment compared with plate fixation of displaced midshaft clavicular fractures. A multicenter, randomized clinical trial. J Bone Joint Surg Am 89(1):1–10. doi:10.2106/JBJS.F.00020 Judd DB, Pallis MP, Smith E et al (2009) Acute operative stabilization versus nonoperative management of clavicle fractures. Am J Orthop (Belle Mead NJ) 38(7):341–345 Smekal V, Irenberger A, Struve P et al (2009) Elastic stable intramedullary nailing versus nonoperative treatment of displaced midshaft clavicular fractures-a randomized, controlled, clinical trial. J Orthop Trauma 23(2):106–112. doi:10.1097/BOT.0b013e318190cf88 Mirzatolooei F (2011) Comparison between operative and nonoperative treatment methods in the management of comminuted fractures of the clavicle. Acta Orthop Traumatol Turc 45(1):34–40. doi:10.3944/aott.2011.2431 Virtanen KJ, Remes V, Pajarinen J et al (2012) Sling compared with plate osteosynthesis for treatment of displaced midshaft clavicular fractures: a randomized clinical trial. J Bone Joint Surg Am 94(17):1546–1553. doi:10.2106/JBJS.J.01999 Robinson CM, Goudie EB, Murray IR et al (2013) Open reduction and plate fixation versus nonoperative treatment for displaced midshaft clavicular fractures: a multicenter, randomized, controlled trial. J Bone Joint Surg Am 95(17):1576–1584. doi:10. 2106/JBJS.L.00307 Zlowodzki M, Zelle BA, Cole PA et al (2005) Treatment of acute midshaft clavicle fractures: systematic review of 2144 fractures: on behalf of the Evidence-Based Orthopaedic Trauma Working Group. J Orthop Trauma 19(7):504–507 Smekal V, Oberladstaetter J, Struve P et al (2009) Shaft fractures of the clavicle: current concepts. Arch Orthop Trauma Surg 129(6):807–815. doi:10.1007/s00402-008-0775-7 Golish SR, Oliviero JA, Francke EI et al (2008) A biomechanical study of plate versus intramedullary devices for midshaft clavicle fixation. J Orthop Surg Res 3:28. doi:10.1186/1749-799X-3-28 Higgins JPT (2011) In: Green S (eds) Cochrane handbook for systematic reviews of interventions. Version 5.1.0 [updated March 2011]. The Cochrane Collaboration. http://www.cochranehandbook.org
ORTHOPAEDIC SURGERY
Operative versus nonoperative treatment for displaced midshaft clavicular fractures: a meta-analysis of randomized clinical trials Lingde Kong • Yingze Zhang • Yong Shen
Received: 2 March 2014 / Published online: 29 August 2014 Ó Springer-Verlag Berlin Heidelberg 2014
Abstract Introduction There is insufficient evidence to indicate whether operative or nonoperative treatment is better for treating displaced midshaft clavicular fractures. We undertook a systematic review and meta-analysis of randomized controlled trials (RCTs) to compare the effects of the two treatments. Methods We searched the PubMed, EMBASE and Cochrane Library databases, and identified RCTs to compare the Constant score, DASH score, nonunion, malunion, and adverse events between operative and nonoperative groups of patients with displaced midshaft clavicular fractures. Results A total of 507 patients from 6 RCTs were subjected to meta-analysis. Operative treatment has an effect on improving function, which is demonstrated by significantly higher Constant scores (P = 0.0003) and lower DASH scores (P = 0.03). The rate of nonunion and the rate of malunion were significantly lower in operative group compared with that in nonoperative group (both P \ 0.0001). However, the rate of adverse events was significantly higher in operative group compared with that in nonoperative group (P = 0.003). Conclusions Operative treatment provided a significantly better functional outcome, a lower rate of nonunion and malunion, but was accompanied with a higher rate of adverse events. However, the results should be interpreted with caution and further large-scale, well-designed RCTs on this topic are still needed.
L. Kong Y. Zhang Y. Shen (&) Department of Orthopedics, The Third Hospital of Hebei Medical University, 139 Ziqiang Road, Shijiazhuang 050051, Hebei, People’s Republic of China e-mail: [email protected]
Keywords Bone fracture Clavicle Operative treatment Nonoperative treatment Meta-analysis
Introduction The clavicular fracture is one of the most common skeletal injuries that accounts for 5–10 % of all fractures. Around 80 % of clavicle fractures involve the midshaft and over half of these fractures are displaced [1, 2]. Conventionally, the fractures have been treated nonoperatively with immobilization in a sling or a figure-of-eight bandage, with the expectation of a good clinical outcome [3, 4], but some studies have emphasized the high risk of nonunion or symptomatic malunion following nonoperative treatment [5–7]. The outcome of nonoperative treatment is not as favorable as thought and there has been a growing trend to treat these fractures surgically [8, 9]. However, before advocating a policy of routine open reduction and internal fixation, further evaluation of the benefits to patients is important. Nowadays, there are several published randomized controlled trials (RCTs) regarding comparison between operative and nonoperative treatment for treating clavicular fracture. A previous metaanalysis including available RCTs was performed by McKee et al. [10] and was published in the year of 2012, but their review did not analyze functional outcomes quantitatively. Moreover, several relevant studies on this topic have been published in recent years. Thus, we undertook a meta-analysis based on high-quality data from RCTs to compare functional outcome, fracture-healing and adverse events between operative and nonoperative groups of patients with displaced midshaft clavicular fractures. Our hypothesis was that operative treatment would result in better functional outcome, more satisfactory fracture-
123
1494
Arch Orthop Trauma Surg (2014) 134:1493–1500
healing, and less adverse events than nonoperative treatment.
Data were extracted independently by two authors. Any disagreements concerning paper eligibility were resolved by discussion and consensus.
Materials and methods
Assessment of methodological quality
Literature search and inclusion criteria
The methodological quality of the studies was evaluated independently by two authors, without masking the trial names. The reviewers followed the instructions provided in the Cochrane Handbook for Systematic Reviews of Interventions [25]. The following domains were assessed: random sequence generation, allocation concealment, blinding, incomplete data outcomes, revealing of selective outcomes and any remaining biases. When the information in the study was inadequate, attempts were made to contact the authors in order to ensure that the study was evaluated correctly.
We searched the PubMed, EMBASE and Cochrane Library databases for English language studies that had been published before January 2014. Key words combined with MeSH terms, including clavicle, clavicular, and fractures, were used for searching. The search was limited to human subjects, and RCTs. In addition, reference lists of all the selected articles were hand-searched to identify other potentially eligible trials. This process was performed iteratively until no additional articles could be identified. The following inclusive selection criteria were applied: (1) study design: RCTs; (2) population: patients (16 years B age B 70 years) with a middle-third clavicular fracture that was completely displaced; (3) intervention: surgical treatment including plate fixation and intramedullary fixation; (4) comparisons: conservative treatment including slings, figure-of-eight bandages, and other physical treatments; (5) outcomes: functional outcomes measured by the Disability of the Arm, Shoulder, and Hand questionnaire (DASH) or the Constant score, fracture-healing measured by the rate of nonunion or malunion, and adverse events; (6) length of follow-up: at least 1 year. Trials were excluded if they (1) were abstracts, letters, reviews, or case reports; (2) had repeated data; and (3) did not report outcomes of interest. Data extraction and outcome measures The data included the general characteristics of each study and the outcomes measured. General characteristics included first author, year of publication, study design, number of patients, types of fracture, internal fixation, nonoperative treatment, and length of follow-up. The outcomes measured included: the DASH score, the Constant score, the rate of nonunion, malunion, and adverse events. When the same population was reported in several publications, we retained only the most informative article or complete study to avoid duplication of information. We also tried to contact the authors of primary studies to request missing data. For continuous outcomes with no standard deviations reported, we calculated standard deviations from standard errors, P values, or confidence intervals, according to the methods described in the Cochrane Handbook for Systematic Reviews of Interventions [25].
123
Statistical analysis Differences were expressed as risk ratios (RRs) with 95 % confidence intervals (CIs) for dichotomous outcomes and weighted mean differences (WMDs) with 95 % CIs for continuous outcomes. Heterogeneity was analyzed with both the Chi square test and the I2 test. A P value of \0.10 for the Chi square test was interpreted as evidence of statistical heterogeneity, and I2 was used to estimate total variation across the studies. A fixed-effect model was adopted if there was no statistical evidence of heterogeneity, and a random-effect model was adopted if statistically significant heterogeneity was present. Studies with an I2 statistic of 25–50 % were considered to have low heterogeneity, those with an I2 statistic of 50–75 % had moderate heterogeneity, and those with an I2 statistic of [75 % had high heterogeneity. Because patient characteristics, types of fracture, interventions, and other confounding factors were not consistent between studies, we further conducted sensitivity analysis to identify potential sources of heterogeneity. The presence of publication bias was assessed using the Begg and Egger tests. A P value \0.05 was judged as statistically significant, except where otherwise specified. All statistical analyses were performed using Review Manager version 5.1 (The Cochrane Collaboration, Software Update, Oxford, UK).
Results Study identification and selection A total of 68 records was identified by the initial database search. Twenty-one records were excluded because of
Fig. 1 Preferred reporting items for systematic reviews and meta-analyses (PRISMA) flow diagram outlining literature search results
Identification
Arch Orthop Trauma Surg (2014) 134:1493–1500
1495
68 records identified through database searching (PubMed, EMBASE and Cochrane Library databases)
Eligibility
Screening
47 records after duplicates removed
47 potentially relevant records screened
11 full-text articles assessed for eligibility
36 records excluded based on the titles and abstracts (reviews, non-randomized studies, or not relevant to our analysis)
5 full-text articles excluded: 1 no outcomes of interest reported 3 duplicated data 1 currently ongoing
Included
6 studies included in qualitative synthesis
6 studies included in quantitative synthesis (meta-analysis)
duplicate studies and 36 were excluded for various reasons (reviews, non-randomized studies, or not relevant to our analysis) on the basis of the titles and abstracts. The remaining 11 were retrieved for full text review, and 5 of them were excluded because 1 did not report outcomes of interest [11], 3 reported duplicated data [12–14], and 1 were currently ongoing [15]. Finally, 6 RCTs that met our inclusion criteria were included in the present meta-analysis [16–21]. The selection process for RCTs included in the meta-analysis is shown in Fig. 1. Study characteristics The main characteristics of the 6 RCTs included in the meta-analysis are presented in Table 1. These studies were published between 2007 and 2013. The sizes of the RCTs ranged from 57 to 200 patients. There were 507 patients at the time of final follow-up. A total of 259 patients underwent surgery, and the remaining 248 patients received conservative treatment. All studies focused on displaced midshaft fracture of the clavicle, except one study included patients with displaced, comminuted fracture [19]. Operative techniques included plate fixation in four studies [16, 19–21] and intramedullary pin/nail fixation in two studies [17, 18]. A sling was used as conservative treatment in five studies, and a collar and cuff was used in one study [21].
All patients were followed up for 12 months, with the exception of one study having patients followed up for 24 months [18]. Assessment of risk of bias The risk of bias is demonstrated graphically in Fig. 2 and summarized in Fig. 3. The randomization technique was mentioned in all 6 trials. However, only 3 trials described the process of random sequence generation [16, 18, 20], and 5 trials stated the method of allocation concealment [16–20]. Blinding is rarely used in orthopedic surgery trials and only one study was blinded in the assessment of outcome [21]. Three trials were judged to be at high risk of attrition bias [18, 20, 21], and the other three were at low risk [16, 17, 19]. Outcome Functional outcome Five RCTs reported the Constant score in study patients. The test for heterogeneity was significant, and the studies have moderate heterogeneity (P for heterogeneity = 0.02, I2 = 67 %). Using the random-effect model, the aggregated results suggested that the Constant score was greater
123
1496
Arch Orthop Trauma Surg (2014) 134:1493–1500
Table 1 General characteristics of each study Study (author/year)
Study design
No. of patients assigneda
Range of ages (years)
Length of follow-up (months)
No. of patients assesseda
Types of fracture
Internal fixation
Nonoperative treatment
COTS/2007
Multicenter, RCT
67/65
16–60
12
62/49
Displaced
Plate
Sling
Judd/2009
RCT
29/28
17–40
12
29/28
Displaced
Pin
Sling
Smekal/2009
RCT
33/35
18–65
24
30/30
Displaced
Nail
Sling
Mirzatolooei/ 2011
RCT
29/31
18–65
12
26/24
Comminuted, displaced
Plate
Sling
Virtanen/ 2012
RCT
28/32
18–70
12
26/25
Displaced
Plate
Sling
Robinson/ 2013
Multicenter, RCT
95/105
16–60
12
86/92
Displaced
Plate
Collar and cuff
a
Operative/nonoperative treatment
Fig. 2 Risk of bias graph: a review of the authors’ judgments regarding each risk of bias item, presented as percentages across all included studies
in the operative group in comparison to the nonoperative group at a statistically significant level (WMD 5.19, 95 % CI 2.37–8.01, P \ 0.0003) (Fig. 4). Subsequently, we performed sensitivity analysis to explore potential sources of heterogeneity. Exclusion of the trial conducted by Mirzatolooei et al. [19] resolved the heterogeneity (P for heterogeneity = 0.58, I2 = 0 %), but did not materially alter the pooled results (WMD 4.05, 95 % CI 2.44–5.66, P \ 0.00001). Information on the DASH score was provided in only four studies. The test for heterogeneity was significant, and the studies have high heterogeneity (P for heterogeneity \ 0.00001, I2 = 91 %). Using the random-effect model, the aggregated results suggested that the DASH score was lower in the operative group in comparison to the nonoperative group at a statistically significant level (WMD -6.60, 95 % CI -12.68 to -0.52, P = 0.03) (Fig. 5). Similarly, exclusion of the trial conducted by Mirzatolooei et al. [19] reduced the heterogeneity (P for heterogeneity = 0.32, I2 = 12 %), without materially altering the pooled results (WMD -3.32, 95 % CI -5.45 to -1.18, P = 0.002).
123
Fracture-healing All six studies provided data on nonunion. The test for heterogeneity was not significant, and the studies have no statistical evidence of heterogeneity (P for heterogeneity = 0.48, I2 = 0 %). Using the fixed-effect model, the rate of nonunion was significantly lower in operative group compared with that in nonoperative group (RR 0.16, 95 % CI 0.07–0.38, P \ 0.0001) (Fig. 6). Malunion was referred to symptomatic malunion instead of radiographic malunion. The test for heterogeneity was not significant, and the studies have no statistical evidence of heterogeneity (P for heterogeneity = 0.76, I2 = 0 %). Using the fixed-effect model, the rate of malunion was significantly lower in operative group compared with that in nonoperative group (RR 0.15, 95 % CI 0.07–0.34, P \ 0.00001) (Fig. 7). Adverse events We collected a set of adverse events that was reported in the RCTs, such as wound infection, implant failure,
Arch Orthop Trauma Surg (2014) 134:1493–1500
neurologic symptom, and refracture, and listed them in Table 2. As nonunion and malunion were analyzed above, they were not included here. The study records of these events were relatively inconsistent among studies and the definition of complications varied. So, we just pooled the rate of overall events without performing further subgroup analysis. The test for heterogeneity was not significant, and the studies have no statistical evidence of heterogeneity (P for heterogeneity = 0.55, I2 = 0 %). Using the fixedeffect model, the rate of adverse event was significantly
1497
higher in operative group compared with that in nonoperative group (RR 2.14, 95 % CI 1.30–3.52, P = 0.003) (Fig. 8). The common operative complications consisted of local hardware irritation, implant failure, and wound infection. The predominant complications in the nonoperative group were neurologic symptoms, and stiffness or restriction of shoulder movement. Publication bias Publication bias was not assessed because of the limited number (\10) of studies included in each analysis.
Discussion
Fig. 3 Risk of bias summary: a review of the authors’ judgments regarding each risk of bias item, for each included study
The pooled results from the meta-analysis of six RCTs suggest that operative treatment provided a significantly better functional outcome, a lower rate of nonunion and malunion, but was accompanied with a higher rate of adverse events. This meta-analysis assessed the effects of operative and nonoperative treatment for displaced midshaft clavicular fractures. McKee et al. [10] have performed a similar metaanalysis and published it in the year of 2012. However, their study did not quantitatively evaluate the functional outcomes. Besides, after his review, three relevant RCTs have been reported [19–21]. In the current study, we included all available RCTs, and made a meta-analysis of the functional outcomes. The results showed that operative treatment has an effect on improving function, which is demonstrated by significantly higher Constant scores and lower DASH scores after at least one-year follow-up. In the assessment of functional outcome, the studies showed significant heterogeneity. Our sensitivity analysis found that a trial conducted by Mirzatolooei et al. [19] contributed to the heterogeneity. Exclusion of this trial resolved the heterogeneity, but did not materially alter the pooled results. This adds robustness to this finding. Different from the features of other study populations, this
Fig. 4 Forest plot showing the comparison of the Constant score between the operative group and the nonoperative group. CI confidence interval
123
1498
Arch Orthop Trauma Surg (2014) 134:1493–1500
Fig. 5 Forest plot showing the comparison of the DASH score between the operative group and the nonoperative group. CI confidence interval
Fig. 6 Forest plot showing the rate of nonunion between the operative group and the nonoperative group. CI confidence interval
Fig. 7 Forest plot showing the rate of malunion between the operative group and the nonoperative group. CI confidence interval
study focused on patients with dislocated, comminuted fractures. As shown in Figs. 4 and 5, the mean difference (MD) between groups in this study was more significant than that in any of other studies, which means that patients with dislocated, comminuted fracture benefit more from surgery than those with dislocated fracture alone. This study may provide additional interesting clues for future research on this specific fracture. Our review shows that 13.7 % patients in the nonoperative group developed a nonunion, which is significantly higher than the 1.9 % rate of nonunion in the operative group. Another complication significantly more common in the nonoperative group is malunion (25 % in the nonoperative group versus 2.8 % in the operative group). Our
123
findings contribute to the evidence that internal fixation is a reliable method for diminishing the risk of nonunion and malunion after displaced midshaft clavicular fracture [22, 23]. Previous researches have investigated the prognostic factors that may predict nonunion or malunion in nonoperatively treated patients. For example, Virtanen et al. [20] reported that displacement of more than 1.5 clavicle widths was a risk factor for nonunion. Hill et al. [5] concluded that displacement of the fracture fragments by more that 2 cm was associated with an unsatisfactory result. However, in this review, we were not able to draw any specific conclusions as to which individuals are most likely to have nonunion or malunion, as most trials did not provide relevant data.
Arch Orthop Trauma Surg (2014) 134:1493–1500 Table 2 Adverse events in both groups
Study (author/year) COTS/2007
1499
Operative group (N = 259 patients)
Nonoperative group (N = 248 patients)
Three wound infections
Seven transient brachial plexuses
Five hardware irritations (removal required) Eight transient brachial plexuses Judd/2009
Six superficial pin-tract infections
1 refracture
One partial, transient radial nerve palsy One intramedullary pin fracture Smekal/2009 Mirzatolooei/2011
Virtanen/2012
Two implant failure Five medial nail protrusion
Three transient neurogenic compromise
Two skin dysesthesia
Two shoulder imbalance
One infection Two hypertrophic scar
Two neurologic claudication in upper extremity
One plate bent
One brachial plexus irritation
One plate broken
Two refracture
One hardware irritation One refracture Robinson/2013
Two rotator cuff impingement
One rotator cuff impingement
One adhesive capsulitis One refracture One plate bent
Fig. 8 Forest plot showing the rate of adverse events between the operative group and the nonoperative group. CI confidence interval
We extracted adverse events and listed them in a table. Patients in operative group showed a higher rate of overall adverse event. Most of these complications were hardwarerelated, such as hardware irritation and implant failure. Wound infection and skin dysesthesia problems were also common after surgical treatment. However, in patients after conservative treatment, neurologic symptoms were common problems. Surgeons can use this information to help patients make optimal decisions in a shared decisionmaking process. Theoretically, for postoperative adverse events, some could be reduced by improved surgical techniques, and some could be resolved by removal of hardware.
This review showed that patients treated surgically could get a better restoration of functional outcome, a lower rate of nonunion and malunion, but our results do not provide sufficient evidence to support the routine use of open reduction and internal fixation for all displaced clavicular fractures in adults. Offering surgery to all patients would expose large numbers of patients who would otherwise have healed and regained good shoulder function without suffering potential surgical complications. Thus, a big challenge is to identify the subgroup of patients who might benefit from surgical intervention. Further investigation on this topic is necessary before advocating operative treatment widely.
123
1500
Arch Orthop Trauma Surg (2014) 134:1493–1500
Anyhow, this meat-analysis has several potential limitations. First, though we included only RCTs, all trials had methodological flaws, which include failure to bind the outcome assessor, lack of baseline outcome data, and so on. Consequently, the quantitative results of this review should be interpreted with caution. Second, two operative techniques were used in the trials for primary fixation of midshaft clavicular fractures (plate fixation and intramedullary pin/nail insertion). There may be fundamental differences in outcomes between the two techniques [24]. However, subgroup analysis was not performed because of rather small numbers of studies in both groups. Finally, we only included English language studies. Despite our best efforts in using multiple search methods, it is possible that we have omitted non-English language trials, the results of which may be applicable to our meta-analysis. In conclusion, in this meta-analysis of randomized clinical trials comparing the effect of operative versus nonoperative treatment for displaced midshaft clavicular fractures, certain benefits of internal fixation are noted, which include a good restoration of function, a low rate of nonunion and malunion. Despite these encouraging findings, patients should be informed of high rate of adverse event after surgery. Anyhow, the results should be interpreted with caution and further large-scale, well-designed RCTs on this topic are still needed. Conflict of interest
10.
11.
12.
13.
14.
15.
16.
17.
None. 18.
References 1. Pecci M, Kreher JB (2008) Clavicle fractures. Am Fam Physician 77(1):65–70 2. Postacchini F, Gumina S, De Santis P et al (2002) Epidemiology of clavicle fractures. J Shoulder Elbow Surg 11(5):452–456 3. Grassi FA, Tajana MS, D’Angelo F (2001) Management of midclavicular fractures: comparison between nonoperative treatment and open intramedullary fixation in 80 patients. J Trauma 50(6):1096–1100 4. Nordqvist A, Petersson CJ, Redlund-Johnell I (1998) Mid-clavicle fractures in adults: end result study after conservative treatment. J Orthop Trauma 12(8):572–576 5. Hill JM, McGuire MH, Crosby LA (1997) Closed treatment of displaced middle-third fractures of the clavicle gives poor results. J Bone Joint Surg Br 79(4):537–539 6. McKee MD, Pedersen EM, Jones C et al (2006) Deficits following nonoperative treatment of displaced midshaft clavicular fractures. J Bone Joint Surg Am 88(1):35–40. doi:10.2106/JBJS. D.02795 7. Nowak J, Holgersson M, Larsson S (2005) Sequelae from clavicular fractures are common: a prospective study of 222 patients. Acta Orthop 76(4):496–502. doi:10.1080/17453670510041475 8. Liu PC, Chien SH, Chen JC et al (2010) Minimally invasive fixation of displaced midclavicular fractures with titanium elastic nails. J Orthop Trauma 24(4):217–223. doi:10.1097/BOT. 0b013e3181b8ba33 9. Mueller M, Burger C, Florczyk A et al (2007) Elastic stable intramedullary nailing of midclavicular fractures in adults: 32
123
19.
20.
21.
22.
23.
24.
25.
patients followed for 1–5 years. Acta Orthop 78(3):421–423. doi:10.1080/17453670710014013 McKee RC, Whelan DB, Schemitsch EH et al (2012) Operative versus nonoperative care of displaced midshaft clavicular fractures: a meta-analysis of randomized clinical trials. J Bone Joint Surg Am 94(8):675–684. doi:10.2106/jbjs.j.01364 Yadav V, Khare GN, Singh S et al (2013) A prospective study comparing conservative with operative treatment in patients with a ‘floating shoulder’ including assessment of the prognostic value of the glenopolar angle. Bone Joint J 95-B(6):815–819. doi:10. 1302/0301-620x.95b6.31060 Schemitsch LA, Schemitsch EH, Veillette C et al (2011) Function plateaus by one year in patients with surgically treated displaced midshaft clavicle fractures. Clin Orthop Relat Res 469(12):3351–3355. doi:10.1007/s11999-011-1915-x Neuhaus V, Ring D (2013) Effect of different statistical methods on union or time to union in a published study about clavicular fractures. J Shoulder Elbow Surg 22(4):471–477. doi:10.1016/j. jse.2012.03.015 Pearson AM, Tosteson AN, Koval KJ et al (2010) Is surgery for displaced, midshaft clavicle fractures in adults cost-effective? Results based on a multicenter randomized, controlled trial. J Orthop Trauma 24(7):426–433. doi:10.1097/BOT.0b013e3181c3e505 Stegeman SA, de Jong M, Sier CF et al (2011) Displaced midshaft fractures of the clavicle: non-operative treatment versus plate fixation (Sleutel-TRIAL). A multicentre randomised controlled trial. BMC Musculoskelet Disord 12:196. doi:10.1186/ 1471-2474-12-196 Society COT (2007) Nonoperative treatment compared with plate fixation of displaced midshaft clavicular fractures. A multicenter, randomized clinical trial. J Bone Joint Surg Am 89(1):1–10. doi:10.2106/JBJS.F.00020 Judd DB, Pallis MP, Smith E et al (2009) Acute operative stabilization versus nonoperative management of clavicle fractures. Am J Orthop (Belle Mead NJ) 38(7):341–345 Smekal V, Irenberger A, Struve P et al (2009) Elastic stable intramedullary nailing versus nonoperative treatment of displaced midshaft clavicular fractures-a randomized, controlled, clinical trial. J Orthop Trauma 23(2):106–112. doi:10.1097/BOT.0b013e318190cf88 Mirzatolooei F (2011) Comparison between operative and nonoperative treatment methods in the management of comminuted fractures of the clavicle. Acta Orthop Traumatol Turc 45(1):34–40. doi:10.3944/aott.2011.2431 Virtanen KJ, Remes V, Pajarinen J et al (2012) Sling compared with plate osteosynthesis for treatment of displaced midshaft clavicular fractures: a randomized clinical trial. J Bone Joint Surg Am 94(17):1546–1553. doi:10.2106/JBJS.J.01999 Robinson CM, Goudie EB, Murray IR et al (2013) Open reduction and plate fixation versus nonoperative treatment for displaced midshaft clavicular fractures: a multicenter, randomized, controlled trial. J Bone Joint Surg Am 95(17):1576–1584. doi:10. 2106/JBJS.L.00307 Zlowodzki M, Zelle BA, Cole PA et al (2005) Treatment of acute midshaft clavicle fractures: systematic review of 2144 fractures: on behalf of the Evidence-Based Orthopaedic Trauma Working Group. J Orthop Trauma 19(7):504–507 Smekal V, Oberladstaetter J, Struve P et al (2009) Shaft fractures of the clavicle: current concepts. Arch Orthop Trauma Surg 129(6):807–815. doi:10.1007/s00402-008-0775-7 Golish SR, Oliviero JA, Francke EI et al (2008) A biomechanical study of plate versus intramedullary devices for midshaft clavicle fixation. J Orthop Surg Res 3:28. doi:10.1186/1749-799X-3-28 Higgins JPT (2011) In: Green S (eds) Cochrane handbook for systematic reviews of interventions. Version 5.1.0 [updated March 2011]. The Cochrane Collaboration. http://www.cochranehandbook.org
