Early Versus Delayed Surgery for Midshaft Clavicle Fractures: A Systematic Review

Abstract

Background

Orthopaedic surgeons debate the timing of and necessity for surgical intervention when treating displaced midshaft clavicle fractures (MCFs). This systematic review evaluates the available literature regarding functional outcomes, complication rates, nonunion, and reoperation rates between patients undergoing early versus delayed surgical management of MCFs.

Methods

Search strategies were applied in PubMed (Medline), CINAHL (EBSCO), Embase (Elsevier), Sport Discus (EBSCO), and Cochrane Central Register of Controlled Trials (Wiley). Following an initial screening and full-text review, demographic and study outcome data was extracted for comparison between the early fixation and delayed fixation studies.

Results

Twenty-one studies were identified for inclusion. This resulted in 1158 patients in the early group and 44 in the delayed. Demographics were similar between groups except for a higher percentage of males in the early group (81.6% vs. 61.4%) and longer time to surgery in the delayed group (4.6 days vs. 14.5 months). Disability of the arm, shoulder, and hand scores (3.6 vs. 13.0) and Constant-Murley scores (94.0 vs. 86.0) were better in the early group. Percentages of initial surgeries resulting in complication (33.8% vs. 63.6%), nonunion (1.2% vs. 11.4%), and nonroutine reoperation (15.8% vs. 34.1%) were higher in the delayed group.

Conclusion

Outcomes of nonunion, reoperation, complications, DASH scores, and CM scores favor early surgery over delayed surgery for MCFs. However, given the small cohort of delayed patients who still achieved moderate outcomes, we recommend a shared decision-making style for treatment recommendations regarding individual patients with MCFs.

Level of Evidence: II

Introduction

Clavicle fractures account for 2.6% of all fractures, and of these, 81.3% are midshaft clavicle fractures (MCFs).¹ The necessity for and timing of surgical treatment for clavicle fractures has long been a subject of debate. Historically, orthopaedic surgeons have treated all MCFs nonoperatively with success, and early research reported increased rates of nonunion in acute fixation of MCFs.² However, later research reported increased risk of nonunion in displaced diaphyseal clavicle fractures treated nonoperatively, concluding that while most MCFs can be treated nonoperatively, comminution and displacement increase subsequent nonunion risk.³ Concordantly, studies comparing open-reduction and plate fixation of displaced MCFs with nonoperative treatment showed plate fixation significantly reduced the risk of nonunion and malunion.⁴ Several reviews have supported acute fixation for markedly displaced, shortened, or comminuted fractures to decrease nonunion and malunion rates, and increase functional outcomes.^5-7

A recent Cochrane review found no clinically significant improvement in upper arm function at one or more years following surgery for displaced or angulated MCFs, nor any difference in pain or quality of life. Low quality review evidence suggested surgery may reduce the risk of symptomatic nonunion, malunion, or other complications.⁸ With studies showing varying complication rates for surgical fixation of MCFs, unclear clinical efficacy, and no financial benefits,⁹ one might suggest all midshaft fractures be treated nonoperatively, reserving surgery for those patients who experience nonunion or malunion.

This review sought to comprehensively evaluate the outcomes of early versus delayed surgical fixation of MCFs. If delayed surgical treatment results in similar outcomes as early surgical treatment, it is possible that all clavicle fractures could be managed nonoperatively initially. Then, only those fractures that did not heal with nonoperative treatment would be surgically treated. We hypothesize that early fixation of MCFs would provide similar functional outcomes, union rates, complication rates, and reoperation rates when compared to delayed fixation.

Methods

Data Sources and Search Strategy

Search strategies were developed with the assistance of a health sciences librarian with expertise in searching for systematic reviews. Search strategies, including both index and keyword methods to maximize sensitivity, were devised for the following databases: PubMed, CINAHL (EBSCO), Embase (Elsevier), Sport Discus (EBSCO), and Cochrane Central Register of Controlled Trials (Wiley) (Table 1). Total yield for the search is illustrated in the flow diagram (Figure 1).

Table 1.

Search Strategies and Results for Each Database at Each Search Date

		Studies returned at each search
Database	Search Strategy	6/2016	3/2018	12/2018	1/2021
PubMed	"Clavicle/injuries"[Mesh] OR ("Fractures, Bone"[Mesh] AND "Clavicle"[Mesh]) OR (fracture[Text Word] OR fractured[Text Word] OR fractures[Text Word]) AND (clavicle[Text Word] OR clavicular) AND (middle[Text Word] OR mid shaft[Text Word] OR midshaft[Text Word] OR middle third[Text Word] OR mid [Text Word] OR diaphyseal) AND "Clavicle/surgery"[Mesh] OR "Fracture Fixation"[Mesh] OR "Fractures, Malunited/surgery"[Mesh] OR "Fractures, Ununited/surgery"[Mesh] OR "Internal Fixators"[Mesh] OR "Bone Transplantation"[Mesh] OR ORIF[Text Word] OR fixation[Text Word] OR conservative[Text Word] OR surgery[Text Word] OR surgical[Text Word] OR operative[Text Word] OR delayed[Text Word] OR late[Text Word] OR acute[Text Word] OR early [Text Word] OR management [Text Word] OR treatment [Text Word]…Limited to English	798	199	66	208
CINAHL	#1: MH “Clavicle Fractures+” OR (MH “Fractures+” AND MH “Clavicle”) OR TX (frac- ture* N5 clavicl*) AND TX (mid*OR diaphyseal) #2: MH “Fracture Fixation” OR MH “Internal Fixators+” OR MH “Clavicle/SU” OR MH “ Fractures+/SU” OR MH “Bone Transplantation” OR TI (ORIF OR fixation OR conservative OR surgery OR surgical OR operative OR delayed OR late OR acute OR early OR management or treatment) OR AB (ORIF OR fixation OR conservative OR surgery OR surgical OR operative OR delayed OR late OR acute OR early OR management or treatment) Limited to English	325	100	35	98
Sport Discus	#1: ZE “clavicle” OR XE “clavicle-wounds and injuries” OR TX (fracture* N5 clavic*) AND TX (mid* OR diaphyseal) #2: ZE “fractures-treatment” OR TI (ORIF OR fixation OR conservative OR surgery OR surgical OR operative OR delayed OR late OR acute OR early OR management or treatment) OR AB (ORIF OR fixation OR conservative OR surgery OR surgical OR operative OR delayed OR late OR acute OR early OR management or treatment)	89	31	7	14
Embase	#1: (‘clavicle fracture’/exp OR (‘fracture’/exp AND ‘clavicle’/exp) OR (fracture NEAR/5 clavic*):ab,ti AND mid* OR diaphaseal #2: ‘surgery’/lnk OR ‘fracture fixation’/exp OR ‘bone transplantation’/exp OR ‘internal fixator’/exp OR (ORIF OR fixation OR conservative OR surgery OR surgical OR operative OR delayed OR late OR acute OR early OR management or treatment):ab,ti Limited to English	635	231	74	327
Cochrane (Central)	**Identical to PubMed search strategy, but no English Language Filter was applied	69	48	5	39

Figure 1.

Flow diagram outlining the systematic review process from initial search results to articles included in the study. This chart was adapted from PRISMA. From: Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 2021;372:n71. doi: 10.1136/bmj.n71. For more information, visit: http://www.prisma-statement.org/.

Search Results

Our search strategy identified 3,416 articles. After removing duplicates, 1,681 articles remained for screening. Two independent reviewers screened the titles and abstracts to exclude reviews, non-English language publications, case studies, nonoperative treatment, biomechanical studies, and studies with less than one-year follow up. All clavicle fixation methods were included for this study. This screening process resulted in the identification of 646 articles (Figure 1).

Two independent authors reviewed the full texts of the remaining articles. Discrepancies in inclusion or exclusion decisions were decided upon by the senior author. Articles were excluded for the following criteria: 1) pediatric population (<18 y/o), 2) distal clavicle fractures, 3) medial clavicle fractures, 4) AC separation, 5) level of evidence of three or lower. Studies also had to specify time to surgery as less than six weeks or exclude fractures older than six weeks to be considered early intervention. Operations were considered delayed for fractures older than six weeks, persistent nonunion, or malunion. Studies were excluded if they did not describe the time from injury to surgery nor mention exclusion of old fractures. All operative techniques were considered for this study. The remaining articles and 26 systematic reviews on midshaft clavicle fracture treatment had their bibliographies reviewed to identify articles missed by our search strategy. Following this process, nine papers were identified and underwent the same screening process outlined above. The whole process identified 21 articles to review.

Data Extraction

The following factors were extracted from the 21 studies by two independent reviewers: patient age, time to surgery, complications, union rate, number of subsequent surgeries, and post-surgical outcomes at 12 or more months using the Disability of the Arm, Shoulder, and Hand (DASH) score and Constant-Murley (CM) PRO scores. Minor complications included: pain/ dysesthesia, skin irritation, skin numbness, implant prominence or irritation due to prominence, superficial infection, unspecified infection, frozen shoulder, functional limitation, dehiscence, malunion, delayed union, superficial vein thrombosis, and transient brachial plexus symptoms. Major complications included: implant failure or loosening, deep infection, refracture, nonunion, or cardiovascular events. Additively, major and minor complications made up total complications.

If subsets of patients within each paper did not meet exclusion criteria, this subset was not included in analysis while keeping the group of patients that did meet exclusion criteria. Papers describing two early surgical treatment groups had data from each group entered and kept separately in tabulations but were ultimately pooled into the final results.

Quality Assessment

To assess study quality, two independent reviewers utilized the NIH Study Quality Assessment Tools to rate each study as good, fair, or poor. For randomized controlled trials (RCT), we used the “Quality Assessment of Controlled Intervention Studies” tool, and for prospective cohort studies (PCS), we used the “Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies” tool.¹⁰ For this process, no discrepancies were encountered between the two reviewers.

Data Analyses

Excel v.1808 (Microsoft Inc, Redmond, WA) was utilized to perform basic demographic calculations such as weighted average patient age and percent of male patients. Additionally, weighted averages and percentages were calculated for outcome variables: the comparative DASH scores and CM scores, and occurrence of nonunions, complications, and reoperations among early versus delayed fixation patients.

Results

This systematic review process identified 18 RCTs and 3 PCSs that met inclusion criteria (Figure 1). Nine studies compared early fixation with a plate or nail to nonoperative management,^11-19 and we analyzed only the operatively treated patients; 10 studies compared two methods of early surgical fixation;^20-29 and two studies compared two methods of surgical fixation for delayed clavicle fractures.^30,31 This yielded 1,158 patients receiving early operative intervention, and 44 patients receiving delayed operative treatment. Notably, a delayed plate fixation intra-study cohort in the Nowak et al. paper was excluded due to inclusion of pediatric patients.³¹ Eighteen studies^11-23,25-29 reported mean age of patients and eighteen studies^11-22,24-29 reported the number of males in each cohort (Table 2). The weighted average age was 38.0 years in the early fixation group and 40.3 years in the delayed group. The percentage of male patients was 81.6% in the early fixation group and 61.4% in the delayed fixation group (Table 3). Nine early fixation studies^{12,16,20-25,28} and both delayed fixation studies^30,31 reported average time to surgery (Table 2). The weighted average time to operation in the early fixation group was 4.6 days and 14.5 months in the delayed group (Table 3).

Table 2.

Reviewed Articles with Demographic Data From Individual Study Cohorts

Author	Year	Title	LOE	Fixation	n	n Males	Age (mean)	TTS (days)
Acute
Assobhi et al.	2011	Reconstruction plate versus minimal invasive retrograde titanium elastic nail fixation for displaced midclavicular fractures	1	plate	19	17	32.6	9.5
				nail	19	16	30.3	10.2
Calbiyik et al.	2016	Prospective randomized study comparing results of fixation for clavicular shaft fractures with intramedullary nail or locking compression plate	1	plate	40	25	39.07	1.4
Calbiyik et al.	2018	Surgical treatment of displaced clavicle fractures with a novel intramedullary device; comparison of less-invasive versus standard technique	1	nail	36	23	35.22	2.11
				nail	35	21	41.82	1.85
Jiang et al.	2012	Operative treatment of clavicle midshaft fractures using a locking compression plate: comparison between mini-invasive plate osteosynthesis (MIPPO) technique and conventional open reduction	1	plate	32	NR	40	7
				plate	32	NR	45	7
Judd et al.	2009	Acute operative stabilization versus nonoperative management of clavicle fractures	1	nail	29	27	28	NR
Kulshrestha et al.	2011	Operative versus nonoperative management of displaced midshaft clavicle fractures: a prospective cohort study	2	plate	45	43	32	3.5
Kundangar et al.	2019	Clinical outcome of internal fixation of middle third clavicle fractures using locking compression plate: Comparison between open plating and MIPO	2	plate	16	16	NR	3
				plate	21	18	NR	3
Mirza-tolooei et al.	2011	Comparison Between Operative and nonoperative treatment methods in the management of comminuted fractures of the clavicle.	1	plate	26	20	36	NR
Narsaria et al.	2014	Surgical fixation of displaced midshaft clavicle fractures: elastic intramedullary nailing versus precontoured plating	1	plate	32	26	40.2	7.2
				nail	33	24	38.9	6.9
Qvist et al.	2018	Plate fixation compared with nonoperative treatment of displaced midshaft clavicular fractures: a randomized clinical trial	1	plate	75	64	40	NR
Robinson et al.	2013	Open Reduction and Plate Fixation Versus Nonoperative Treatment for Displaced Midshaft Clavicular Fractures	1	plate	95	83	32.3	NR
Sahu et al.	2018	A comparative study between plating versus titanium elastic nail system in mid-shaft clavicle fracture management	1	nail	25	18	33.28	NR
				plate	25	18	34.76	NR
Shen et al.	2008	A three-dimensional reconstruction plate for displaced midshaft fractures of the clavicle	2	plate	67	39	43.8	NR
				plate	66	36	44.7	NR
Smekal et al.	2011	Elastic stable intramedullary nailing is best for mid-shaft clavicular fractures without comminution: Results in 60 patients	1	nail	60	54	36.8	3
Sohn et al.	2015	Clinical comparison of two different plating methods in minimally invasive plate osteosynthesis for clavicular midshaft fractures: A randomized controlled trial	1	plate	19	18	46.7	4.3
				plate	18	17	50.4	4.1
Tamaoki et al.	2017	Treatment of Displaced Midshaft Clavicle Fractures: Figure-of-Eight Har-ness Versus Anterior Plate Osteosynthesis: A Randomized Controlled Trial	1	plate	59	53	30.5	NR
van der Meijden et al.	2015	Operative treatment of dislocated midshaft clavicular fractures: plate or intramedullary nail fixation? A randomized controlled trial	1	plate	58	53	38.4	NR
				nail	62	60	39.6	NR
Virtanen et al.	2012	Sling Compared with Plate Osteosynthesis for Treatment of Displaced Midshaft Clavicular Fractures	1	plate	28	24	41	NR
Woltz et al.	2017	Plate Fixation Compared with Nonoperative Treatment for Displaced Midshaft Clavicular Fractures: A Multicenter Randomized Controlled Trial	1	plate	86	80	38.3	NR
Delayed
Kabak et al.	2004	Treatment of midclavicular nonunion: comparison of dynamic compression plating and low-contact dynamic compression plating techniques	1	plate	16	10	40	10.2*
				plate	17	9	42.7	11.4*
Nowak et al.	2001	A prospective comparison between external fixation and plates for treatment of midshaft nonunions of the clavicle	2	Ex-Fix	11	8	37	25.6*

NR= not reported. ^*These time to surgery values are reported in months.

Table 3.

Pooled Results of Outcome Variables Following Systematic Review

	Acute	Delayed
Average Age	38.0	40.3
Percentage of Males	81.6	61.4
Average Time to Operation	4.6 days	14.5 months
DASH score	3.6	13.0
Constant-Murley score	88.5	86.0*
Constant-Murley score without Sahu et al.	94.0	86.0*
Percent Minor Complications	25.0	52.3
Percent Major Complications	6.4	11.4
Percent Total Complications	33.8**	63.6
Percent Nonunion	1.2	11.4
Percentage of Reoperation	21.7	34.1
Percentage of Nonroutine Reoperation	15.8	34.1

*Only one value reported by Nowak et al., this represents an average of 11 patients.³¹

**van der Meijden et al. reported some complications only as total complications without delineation of minor or major, therefore, those were included in the final comparison of total complications.²⁹

Averages Were Weighted Depending on Study Population and Percentages Were Calculated Using Pooled Study Population.

DASH Score

Nine studies^{13,15-19,21,26,29} in the early fixation group and one study³⁰ in the delayed fixation group reported mean DASH scores at 12 or more months (Table 4). Analysis found weighted average DASH scores of 3.6 in the early intervention group and 13.0 in the delayed intervention group (Table 3).

Table 4.

Outcome Data Reported in the Reviewed Articles Included in This Systematic Review

Author	DASH	Constant-Murley Score	Minor Complications	Major Complications	Total Complications	Non-Routine Reoperation	Total Reoperations
Acute
Assobhi et al.	NR	89.9	4	3	7	2	2
	NR	95.5	3	0	3	3	3
Calbiyik et al.	8.19	90.1	11	3	14	2	2
Calbiyik et al.	NR	94.38	4	2	6	NR	NR
	NR	92.85	1	2	3	NR	NR
Jiang et al.	NR	NR	2	0	2	NR	NR
	NR	NR	12	0	12	NR	NR
Judd et al.	NR	NR	17	5	22	8	31
Kulshrestha et al.	NR	NR	8	2	10	6	6
Kundangar et al.	NR	96	13	1	14	NR	NR
	NR	94	5	1	6	1	1
Mirzatolooei et al.	8.6	89.8	11	3	14	2	2
Narsaria et al.	NR	96.2	5	3	8	22	22
	NR	94.6	1	2	3	1	33
Qvist et al.	1.7	NR	44	5	49	17	17
Robinson et al.	3.4	92	40	5	45	16	16
Sahu et al.	1.87	9.36	2	0	2	NR	NR
	4.8	15.08	4	4	8	NR	NR
Shen et al.	NR	NR	1	0	1	1	1
	NR	NR	8	0	8	8	8
Smekal et al.	0.5	98	8	10	18	12	12
Sohn et al.	NR	95.7	2	2	4	6	6
	NR	97.2	1	0	1	3	3
Tamaoki et al.	3.3	NR	14	1	15	3	3
van der Meijden et al.	2.4	99.2	0	0	36	5	5
	3.9	91.3	33	0	43	10	12
Virtanen et al.	4.3	86.5	4	3	7	0	0
Woltz et al.	4.5	95.4	17	13	30	23	23
Delayed
Kabak et al.	NR	NR	9	2	11	10	10
	NR	NR	1	0	1	2	2
	N/A	N/A	5*	0*	5*	N/A	N/A
Nowak et al.	NR	86	8	3	11	3	3

NR= not reported, N/A= not applicable. *Complications not specified as to which study group they belonged to in the study by Kabak et al.³⁰

Constant-Murley Score

Thirteen early intervention papers^{13,15,16,18-22,24-26,28,29} and one delayed intervention paper³¹ reported mean CM scores at 12 or more months (Table 4). The weighted average from studies describing early surgery was 88.5. However, when Sahu et al. was excluded because of outliers, the weighted average CM score from 12 studies was 94.0.²⁶ In the delayed group, only one study reported an average CM of 86.0 (Table 3).

Nonunion

Seventeen early intervention papers,^11-26,28 and both delayed intervention papers^30,31 reported on nonunion (Table 4). In the early fixation group, there were 11 nonunions versus five in the delayed fixation group. This equates to 1.2% of initial operations resulting in nonunion in the early fixation group compared to 11.4% of initial operations resulting in nonunion in the delayed fixation group (Table 3).

Complications

All early^11-29 and delayed^30,31 studies reported on complications (Table 4). The percent of initial fixations resulting in a minor complication in the early group was 25.0% and in the delayed group was 52.3%. The percentage of patients experiencing major complication was 6.4% in the early group and 11.4% in the delayed group (Table 3). One paper by Van der Meijden et al. only reported some of their complications as “total complications,” these values were added to the final total complications but not minor or major categories.²⁹ Additionally in the Kabak et al. paper describing plate fixation of two delayed MCF cohorts, five minor complications were pooled in our analysis that were not explicitly stated to be experienced by one of the intrastudy cohorts.³⁰ Overall, the percentage of patients experiencing any complication was 33.8% in the early fixation group and 63.6% in the delayed fixation group (Table 3).

Reoperation Rates

Reoperation was reported in seventeen early intervention papers^{11-22,24,25,27-29} and both delayed intervention papers (Table 4).^30,31 When all reasons for reoperation were considered, 21.7% of all early fixation patients and 34.1% of all delayed fixation patients had reoperation. After excluding routine implant removals, 15.8% of early fixation patients and 34.1% of delayed fixation patients underwent reoperation (Table 3).

Limitations

Following evaluation of the RCTs with the NIH tool, six were found to be of good quality,^{14,17,23,28-30} nine studies were found to be of fair quality,^{11,13,15,18-20,25-27} and three studies were found to be of poor quality.^16,21,22 For the PCSs, evaluation showed all three studies to be of poor quality.^12,24,31 Notably, one delayed study and five early studies were rated as poor quality. Additionally, very few qualifying studies described delayed surgical repair of midshaft clavicle fractures, making direct statistical comparison challenging. Specifically, seventeen RCTs described early surgery,^{11,13-23,25-29} and two of the PCSs described early surgery,^12,24 while only one RCT³⁰ and one PCS³¹—two studies total—described delayed surgery. The scarcity of studies directly comparing early versus delayed surgery with a level of evidence greater than three made it difficult to conduct a high-level systematic review comparing the two surgical-timing protocols.

Discussion

In the present review, the numerical differences observed often favored early operation when comparing early and delayed surgery for MCFs. Demographically, patients in the early intervention papers and delayed intervention papers were similar with exception of percentage of males in each population. Notably, the percentage of patients in the early operative group experiencing nonunion, complication, and reoperation for any reason was lower than that seen in delayed intervention studies. Additionally, DASH scores were lower and CM scores were higher in early fixation studies.

To our knowledge, this is the first systematic review comparing level I and II evidence of early and delayed surgery in MCFs. However, some retrospective studies have sought to directly compare outcomes of early vs. delayed surgery for MCFs. In one study comparing immediate MCF fixation versus fixation of nonunion or malunion six or more months from initial fracture (mean 63 months), Potter et al. found CM scores were superior in immediate operations (89 vs 95, p= 0.04), and DASH scores were not statistically different despite being lower in patients receiving early intervention (7.2 vs. 3.0, p= 0.15).³² Das et al. prospectively compared the outcomes of early fixation within three weeks of fracture with a delayed fixation group receiving an operation from three weeks to three months after their initial fracture. In their study, they found no difference in functional outcomes between early and delayed surgery for MCFs when assessed using a shortened version of the DASH, the quickDASH.³³ Our systematic review showed improved DASH scores and CM scores for early fixation of MCFs, but CM score in the delayed group only represents the results of 11 patients from a single study. Interestingly, the Das et al. study did not find significant differences when the delayed surgeries were completed within three months, whereas our study and Potter et al.’s study showed larger differences between the functional outcome scores of early and delayed surgeries when delay was extended beyond a year. Therefore, it seems that a transition point may exist sometime between three months and one year where early and delayed surgery no longer result in equitable functional outcomes.

In the present study, post-surgical complications were common in both delayed and early MCF operations. However, there was a propensity for complications in the delayed group of almost two times the percentage in the early group when examining major, minor, and total complications. Das et al. considered nonunion, wound healing, infection, symptomatic metal work, and scars as complications. In their study, they discovered no significant difference in complication rates between early and delayed operation.³³ A study by Sawalha & Guisasola retrospectively compared the outcomes of 90 acute fixations at an average of 10 days to 20 delayed surgeries for nonunion at an average of 15 months. They found no significant difference between total complications nor major complications in their study population.³⁴ In our study, the most prevalent complications in the early intervention group were pain/dysesthesia (n= 71) and implant prominence (n= 68), while functional limitation (n= 9) and pain/dysesthesia (n=7) were highest in delayed interventions. Hence, the most common complications observed tend to be minor complications, yet these still can cause significant morbidity that may lead to reoperation.

In our study, the rate of nonunion and reoperation between early and delayed surgery for MCFs was lower in the early fixation group. Interestingly, Sawalha & Guisasola found no difference between the acute and delayed surgery groups in terms of nonunion rates and reoperation rates. In their study the most common reason for reoperation in the acute group was prominent metal work (n= 7), whereas persistent delayed union or nonunion was the most common reason in the delayed group (n= 5).³⁴ The most common reason for reoperation following delayed fixation in our study was implant removal for cosmesis or pain, compared to implant removal for infection, symptomatic implant, prominence, bending, or pain in the early group. The delayed group in our study had five reoperations for nonunion compared to four in the early group, which is substantial when considering five of 44 patients required nonunion reoperation with delayed surgery. Das et al. reported one nonunion in their early group and zero nonunions in their delayed group but did not report the number of reoperations in each group.³³ Based on the present study and these previous studies, nonunion and reoperation appear to be more common in delayed operations. Importantly, persistent symptomatic nonunion usually requires a much more involved reoperation with bone grafting to reconstruct the clavicle. The risk of nonunion and subsequent reoperation is an important consideration to discuss when recommending initial nonoperative treatment to patients with MCF.

Despite evidence of poorer outcomes in delayed fixation, some studies have performed surgical delay, opting for initial nonoperative treatment, in an attempt to elucidate the best predictors of failed nonoperative treatment. Das et al. was able to withhold operation for most patients, successfully treating them nonoperatively, reserving operation only for patients with initially high-risk fractures or fractures that did not begin the healing process. However, they acknowledge limitations because more severe fractures were probably more likely to be treated by early surgery, meaning even fractures that did not heal using nonoperative treatment and required delayed surgery were likely less complex.³³ Similarly, Nicholson et al. conducted a prospective cohort study that examined nonoperatively treated patients 6 weeks after their initial injury to determine if predictive factors of nonunion existed. They found that during this delayed assessment, combined findings from examining the patient with the quickDASH questionnaire, looking for callus formation on radiographs, and observing fracture movement during physical examination could accurately predict nonunion in these patients. They used receiver operating curve analysis with these variables to deduce an area under the curve of 87.3%, meaning they could predict nonunion with relatively good accurracy.³⁵ Given these results, it seems even short delays of a few weeks have the potential to significantly reduce the number of procedures performed while maintaining the same level of successful care provided to all patients.

For displaced midshaft clavicle fractures, this systematic review suggests early operative fixation results in improved outcomes in midshaft clavicle fractures; yet definitive conclusions about the utility of delayed surgery are difficult with data from such a small cohort. With some studies showing promise of initial delay in all patients and the present study still showing moderate results for delayed surgery, we suggest the best approach to deciding when to operate likely involves surgeons using a shared decision-making style to develop the best treatment plan for each patient. Surgeons should discuss the complications associated with early surgical fixation, nonoperative treatment, and delayed surgery with each patient so they can make an informed decision that suits their expectations. Still, the lack of prospective studies directly comparing early versus delayed surgery is remarkable; therefore, a prospective study comparing early and delayed fixation of MCFs is warranted to appropriately evaluate the two interventions and conclusively recommend the best initial treatment for displaced MCFs.