Secondary Logo

Journal Logo

FEATURED ARTICLES

What Is the Best Evidence to Guide Management of Acute Achilles Tendon Ruptures? A Systematic Review and Network Meta-Analysis of Randomized Controlled Trials

Meulenkamp, Brad MD, FRCSC1-3; Woolnough, Taylor MD1; Cheng, Wei PhD2; Shorr, Risa MLIS4; Stacey, Dawn PhD2,4,5; Richards, Megan MD, FRCSC1; Gupta, Arnav BHSc3; Fergusson, Dean PhD2; Graham, Ian D. PhD, FCAHS, FNYAM2,6

Author Information
Clinical Orthopaedics and Related Research: October 2021 - Volume 479 - Issue 10 - p 2119-2131
doi: 10.1097/CORR.0000000000001861

Abstract

Introduction

Achilles tendon ruptures are common and debilitating, and they are followed by intensive rehabilitation to regain function of plantarflexion strength [11]. They most commonly occur during activities that require explosive acceleration with movements such as jumping and sprinting. Typically most common in 30- to 40-year-old males, the incidence of these injuries continues to increase [11, 33, 37, 77]. Despite the rising incidence, there remains little consensus on how best to treat acute Achilles tendon ruptures [11, 37]. Further, both operative and nonoperative treatment strategies continue to evolve, increasing uncertainty for both patients and surgeons.

Nonsurgical treatment of Achilles ruptures once consisted of cast immobilization in plantarflexion with prolonged immobilization, allowing for apposition and healing of the ruptured tendon. Because of concerns of rerupture and calf atrophy, open surgical management often has been preferred over nonoperative management for active, healthy patients. However, complications largely unique to surgery such as wound dehiscence, infections, and other soft tissue issues occur in up to 10% to 15% of treated patients [63]. Functional rehabilitation protocols with early weightbearing and ankle mobilization have seen wider use in recent years, with studies reporting similar patient-reported outcome scores, return to sport, and rerupture risk compared with operative treatment but without subjecting patients to the risks of surgery [11, 22, 33, 37, 77]. With increasing evidence supporting functional rehabilitation, practice has shifted rather drastically with an associated reduction in surgical treatment by more than 50% in the past 20 years [3, 56, 57], although this trend has not been seen in the United States [67, 90]. Successful functional rehabilitation programs require substantial patient cooperation and supervision, which may be hindered by patient and system factors, such as lack of physiotherapy access [22, 33, 37, 77]. Despite advances in nonoperative treatment, many surgeons continue to advocate for surgical management because of increased confidence in maintaining tendon apposition, length, and strength [69], as well the traditional belief that rerupture risk is lower [11, 80]. To reduce surgical site complications, minimally invasive and percutaneous repair of the Achilles tendon have been advocated [47, 69]. Although some authors report percutaneous repair is associated with an increased risk of sural nerve complications and rerupture compared with open repair, these findings have been disputed [29, 48, 52].

Numerous meta-analyses have been performed to establish the superiority of one treatment over another [22, 23, 80, 100]. Constrained by design, traditional pairwise meta-analyses can only evaluate two treatments that have been directly compared in trials. Considering the multiple treatments available for Achilles tendon ruptures, the limitations of pairwise analysis have led to pooling of treatments into heterogeneous groups (such as, operative versus nonoperative management) and numerous overlapping meta-analyses [35, 66]. A network meta-analysis addresses this issue by facilitating simultaneous comparison of multiple treatments [6, 9, 53, 63]. Further, a network meta-analysis allows for the comparison of treatments that were not evaluated in a head-to-head manner in the original randomized controlled trials (RCTs). This approach facilitates the estimation of relative treatment effects for interventions that have not been directly compared in head-to-head trials and for treatments that have only been compared in a limited number of trials.

Our goal was to use network meta-analysis to answer the following questions: Considering open repair, minimally invasive surgery (MIS) repair, functional rehabilitation, or primary immobilization for acute Achilles tendon ruptures, (1) which intervention is associated with the lowest risk of rerupture? (2) Which intervention is associated with the lowest risk of complications resulting in surgery?

Materials and Methods

Search Strategy

We conducted a systematic review with network meta-analyses using methods guided by the Cochrane Handbook for Systematic Reviews of Interventions [31]. This review is reported in adherence with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) extension statement for incorporating network meta-analysis [36, 77]. We published a research protocol [66] and registered this study prospectively with PROSPERO (CRD42018093033). Our electronic search of medical and rehabilitation literature related to management of acute Achilles tendon rupture was performed from database inception to the search date (September 30, 2019) using Medline, Embase, CINAHL, PEDro, and Cochrane Central Register of Controlled Trials. The primary author (BM) developed the search strategy in consultation with a senior information specialist (RS). The strategy was then peer reviewed by a second medical librarian in accordance with the Peer Review of Electronic Search Strategies (PRESS) framework [76]. Previously published systematic reviews were cross-referenced for any missed studies. In addition, we manually searched relevant unpublished evidence sources (grey literature), including meeting abstracts from the Orthopaedic Trauma Association, American Academy of Orthopaedic Surgery, and American Orthopaedic Foot and Ankle Society (AOFAS) annual meetings from 2014 to 2019 to identify emerging studies nearing completion. Preprint servers and foreign-language journals not included in the specified databases were not searched. No language limits were used. The search strategy for one database is available with the published study protocol [61].

Inclusion and Exclusion Criteria

Inclusion criteria were RCTs directly comparing two or more interventions for the treatment of first-time, acute (less than 4 weeks since injury) Achilles tendon ruptures with a minimum follow-up of 6 months. This minimum follow-up was chosen to maximize study inclusion while ensuring appropriate demonstration of return to activity and complications [99]. Interventions of interest included conventional cast immobilization with delayed weightbearing for at least 6 weeks (primary immobilization), bracing and/or splinting with ROM earlier than 6 weeks (functional rehabilitation), open surgical repair, and percutaneous or minimally open surgical repair (MIS). For inclusion as a functional rehabilitation protocol, ankle ROM had to be started before 6 weeks postrupture with or without early weightbearing. MIS treatment included all surgical modalities that did not completely open and reflect the paratenon, including limited transverse incisions, suture-shuttling techniques, and device-assisted techniques. Use of primary immobilization has largely decreased in recent years in favor of functional rehabilitation; however, this treatment was included as a comparator for other treatments and therefore any expertise bias resulting from its inclusion was anticipated to have little impact on our key findings. We excluded studies investigating modifications of only one of the above treatments. For example, we did not include RCTs examining early versus late weightbearing after open surgical repair (as both treatment arms would be considered open surgical repair).

We chose exclusion criteria based on factors that may alter the natural history of tendon repair and rehabilitation: (1) patients younger than 16 years of age, (2) chronic tendon ruptures, (3) tendon rerupture, (4) inclusion of patients with preexisting Achilles tendinopathy, and (5) musculotendinous junction tears. If two or more studies reported the same information, we included only the study with most complete data (that is, the complete reporting of outcomes of interest). Studies were excluded if nonrandom loss to follow-up was greater than 20%.

Screening

The search was conducted on September 30, 2019. Studies were screened using Covidence (Veritas Health Information Ltd). Two reviewers (BM, MR) screened all titles, abstracts, and full-text articles independently and in duplicate. Disagreements at the title and abstract stages were resolved by automatic inclusion, and disagreements at the full-text stage were resolved by consensus. Study authors were contacted if eligibility criteria were unclear.

Outcomes

The primary outcomes for quantitative synthesis were (1) rerupture and (2) post-treatment complications resulting in surgery. Secondary outcomes included functional outcome score, strength, and ROM. Both outcomes were evaluated at the longest reported follow-up. Despite inclusion in the published study protocol [61], the outcomes of overall complications and return to activity were not included in the final analysis. During peer-review, it became apparent that analyzing pooled complications, while statistically robust, resulted in an outcome of unclear clinical relevance because it would have involved pooling common but relatively mild complications (such as superficial infection) with rarer but devastating events (like complex regional pain syndrome [CRPS]). Thus, the outcome of pooled complications was excluded from the final analysis. Complications resulting in surgery were chosen partly as surrogates for serious complications as there were an insufficient number of serious complications not resulting in surgery for statistical analysis. Patients with rerupture, if treated surgically, were counted in both rerupture and complications resulting in surgery. Return to function was excluded from the final analysis because of the inconsistent and heterogeneous nature of the available evidence, which precluded appropriate application of network meta-analysis methodology.

Data Extraction

Data were abstracted in duplicate by two reviewers (MR, AG) using a standardized extraction document (Microsoft Excel 16.2), which was developed and piloted a priori. Discrepancies were resolved by consensus and input from a third reviewer (BM or WC). Study authors were contacted in cases of incomplete data. Abstracted data included study author, year, country of publication, outcome data, and participant demographics (mean age, sex, risk factors for complication such as smoking status, fluoroquinolone or steroid use, diabetes, and smoking), surgical repair method including technique and suture, surgeon experience, length of immobilization, and weightbearing status.

Quality Assessment

The Cochrane Risk of Bias version 2 (ROB 2) assessment tool was used to evaluate bias in the following domains: random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting, and other bias [83]. Two reviewers (BM, AG) evaluated all studies and assigned risk as high risk, low risk, or unclear, with disagreements resolved by consensus. Risk of bias between studies (such as, small-study effects signaling publication bias) was assessed and presented as funnel plots. The overall quality of the evidence was determined and ranked per the Grades of Recommendation, Assessment, Development, and Evaluation approach for network meta-analyses [25, 71].

Study Characteristics

The search identified 630 citations; 103 studies underwent full-text review, of which 19 RCTs (1316 patients) were included in the final analysis (Fig. 1). Included studies were published between 1981 and 2018 (median = 2008). Unique pairwise comparisons included open surgery versus MIS [1, 2, 21, 41, 46, 52, 73], open surgery versus functional rehabilitation [13, 49, 64, 85, 88, 96], MIS versus functional rehabilitation [60], and open surgery versus primary immobilization [10, 42, 62, 65]. One study had three treatment arms (Supplemental Table 1; Supplemental Digital Content 1, https://links.lww.com/CORR/A584) [54]. The mean number of patients per study treatment arm was 35 ± 16 patients (Table 1). Across studies, the mean age was 41 ± 5 years, and 80% of participants were male. Mean age and sex composition were similar across treatment arms. Mean follow-up across all studies was 22 ± 12 months, which was also similar between treatment groups. Only one study had a follow-up less than 12 months [52].

Fig. 1
Fig. 1:
The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram for study screening.
Table 1. - Demographic information for included studies
Parameter Open surgery MIS Functional rehabilitation Primary immobilization
Number of studies 17 9 7 4
Patients per treatment arm 35 ± 16 28 ± 9 33 ± 20 52 ± 9
Age in years 41 ± 5 43 ± 6 39 ±2 39 ± 1
Percent male 81 ± 8 79 ± 11 75 ±13 85 ± 5
Follow-up in months 21 ± 12 22 ± 10 23 ± 11 27 ± 23
Publication year 2009 (1981-2018) 2009 (2008-2018) 2008 (1995-2006) 1997 (1981-2011)
Data presented as mean ± SD or median (range); number of studies represents the number of studies including the specified intervention as a treatment arm and therefore the sum of this row exceeds the total number of studies included in the network meta-analysis; percent male represents the pooled value for the treatment arm calculated using weighted means for each treatment group; MIS = minimally invasive surgery.

Study Quality

The cumulative risk of bias was deemed low across domains (Fig. 2). Of the 19 included studies, nine were deemed to be low risk of bias in at least five domains, and 14 studies were deemed low risk in at least four domains (Supplemental Fig. 1; Supplemental Digital Content 2, https://links.lww.com/CORR/A585). Overall, risk of selection, attrition, and reporting bias was low. However, risk of performance bias and detection bias were high, largely because of the difficulty of blinding when performing studies of surgical treatments. Risk of bias was not expected to compromise pooled results, as it appeared similar across treatment arms.

Fig. 2.
Fig. 2.:
The pooled risk of bias for all included studies divided by source of bias.

Ethical Approval

This network meta-analysis did not involve human participants and therefore was not subject to institutional review board approval.

Meta-analysis Methodology

We performed pairwise meta-analysis for all primary outcomes when direct comparisons were available. The Mantel-Haenszel random-effects model was applied to binary outcomes in the presence of sufficient clinical, methodological, and statistical homogeneity (heterogeneity I2 < 50%) [32]. Pairwise analysis results are expressed as odds ratios for dichotomous outcomes with 95% confidence intervals. Forest plots from pairwise analysis were generated using Review Manager (Version 5.3, The Cochrane Collaboration, Nordic Cochrane Centre).

Network Meta-analysis Methodology

To perform network analyses, we used the OpenBUGS software (Version 3.2.3) and the R2OpenBUGS package (Version 3.2) in R (Version 3.4.2, Open Access Online) [81, 84]. We generated network diagrams for each outcome to ensure well-connected network geometry (at least one closed loop among interventions). We assessed the validity of the transitivity assumption (that is, homogeneity/similarity across studies) by thoroughly reviewing study methods, patient characteristics, and enrollment criteria using established methods [15, 16]. All treatments were assessed to be “jointly randomizable” and could reasonably be applied to any patient in the network [74]. This assumption was supported by relatively strict inclusion criteria and the similar composition of pooled treatment groups [24]. Random-effects Bayesian network meta-analysis with vague priors was performed for each outcome. Prior distributions describe information outside of the included studies used to determine the posterior distribution from which summary measures (such as, mean and SD) are calculated [39]. We used vague priors as there appeared to be sufficient data to estimate variance appropriately without introducing subjectivity into our models, which may occur with truly informative priors [97]. Adequacy of model fit was assessed by comparing the total residual deviance with the number of unconstrained data points (the number of intervention arms across studies in the analysis) and was considered adequate if these quantities were approximately equal. Model selection was based on deviance information criteria, with smaller values being preferred and a difference of five or more points representing an important difference in fit between models. Model convergence was assessed using established methods including the Gelman-Rubin diagnostics and the Potential Scale Reduction Factor [4, 20]. Validity of the consistency assumption (the agreement between direct and indirect evidence) was assessed by fitting random effects unrelated means models to the data and comparing deviance information criterion (DIC) values and posterior mean deviance contributions with the DIC values from consistency models. Deviance residuals, the amount of deviance from each observation, were then plotted to identify inconsistency between direct and indirect evidence (Supplemental Fig. 2; Supplemental Digital Content 3, https://links.lww.com/CORR/A586). Total residual deviance values were lower than the number of unconstrained data points due to several studies with zero occurrences of the outcome of interest (Supplemental Table 2; Supplemental Digital Content 4, https://links.lww.com/CORR/A587).

Our results are presented using odds ratios (OR) and 95% credible intervals (CrI), a measure of imprecision derived using the posterior distributions, which are akin to a Bayesian equivalent of confidence intervals. Comparisons were inferred to be significant if the 95% CrI of the OR did not cross one [17]. A Surface Under the Cumulative Ranking (SUCRA) curve, a numeric representation of treatment ranking, was calculated for each intervention. As SUCRA nears one (the maximum possible value), the greater the probability a treatment is in the top ranks of treatments [75]. Values approaching zero indicate a greater probability a treatment is in the bottom ranks. Number needed to treat was calculated using the difference in mean patient-expected event rates [12]. Summary of findings tables are presented using open surgery as the reference treatment as it was most well connected to other interventions by direct evidence. Comparison-adjusted funnel plots were applied to assess for small-study effects as signals of publication bias. We performed a sensitivity analysis for risk of complications by excluding keloid scars as a complication. For the outcome of complications resulting in surgery, we performed a sensitivity analysis by excluding the study with follow-up duration stated to be “at least 6 months” [52].

Results

Rerupture

We found no difference between open surgical repair, MIS repair, and functional rehabilitation for risk of rerupture, and primary immobilization was associated with a greater risk of rerupture than open repair. Specifically, the network analysis for rerupture (19 RCTs, 1316 participants) demonstrated no difference between open surgical repair (reference treatment) and MIS repair (OR 0.96 [95% CrI 0.22 to 4.15]; p > 0.05) or functional rehabilitation (OR 2.18 [95% CrI 0.80 to 6.20]; p > 0.05) (Fig. 3). We also found no difference between functional rehabilitation and MIS (OR 0.45 [95% CrI 0.10 to 1.82]; p > 0.05) (Supplemental Fig. 3; Supplemental Digital Content 5, https://links.lww.com/CORR/A588). Compared with open surgery, primary immobilization was associated with a greater risk of rerupture (OR 4.06 [95% CrI 1.47 to 11.88]; p < 0.05) (Table 2). Open surgical repair and MIS were ranked most favorably for risk of rerupture (SUCRA 0.80 and 0.79, respectively) (Supplemental Table 3; Supplemental Digital Content 6, https://links.lww.com/CORR/A589). There were no differences between pairwise and network-derived estimates (Supplemental Fig. 4; Supplemental Digital Content 7, https://links.lww.com/CORR/A590).

Fig. 3.
Fig. 3.:
The network geometry for risk of rerupture. Node size is proportionate to the number of participants in the specified treatment arm and is indicated by n = below the treatment name. Edge (connecting line) thickness is proportionate to the number of studies informing an indicated comparison and is specified with the number adjacent the edge.
Table 2. - Summary of findings table for rerupture
Studies: 19a
Participants: 1316
Minimally invasive surgery (9 RCTsb; 250 participants) Functional rehabilitation (7 RCTsb; 254 participants) Primary immobilization (4 RCTsb; 207 participants) Open surgery (17 RCTsb; 605 participants)
Relative effect (95% CrI) 0.96 (0.22-4.15)c 2.18 (0.80-6.20)c 4.06 (1.47-11.88)d Reference
NNT 173.9 18.9e (Harm) 12.9e (Harm) Reference
GRADE evaluation Lowf,g Moderateg Lowf,g Reference
Mean ranking (95% CrI) 1.65 (1-4) 3.00 (2-4) 3.76 (2-4) 1.59 (1-3)
Interpretation of findings Possibly superior Possibly inferior Probably inferior Reference
Relative effect values are odds ratios relative to open surgery; mean rank was calculated based on surface under the cumulative ranking curve (SUCRA) values; small sample sizes were considered in the evaluation of imprecision.
aTotal number of studies across all treatments.
bNumber of RCTs including the treatment of interest.
c95% CrI for odds ratio crosses one, indicating no difference relative to reference treatment; p > 0.05.
dInferred to be statistically significant with a 95% CI for odds ratio not crossing one; p < 0.05
eTreatment associated with a relative effect indicating harm. Therefore, associated value is number needed to harm.
fDowngraded for risk of bias.
gDowngraded for imprecision; NNT = number needed to treat; GRADE = Grading of Recommendations Assessment, Development, and Evaluation; CrI = credible interval.

Complications Resulting in Surgery

We found a lower risk of complications resulting in surgery with MIS repair relative to both open surgery and functional rehabilitation, and we found no difference in risk between functional rehabilitation and open surgery. The network analysis (15 RCTs, 949 patients) demonstrated that MIS repair was associated with a lower risk of complications resulting in surgery than open surgical repair (OR 0.22 [95% CrI 0.04 to 0.93]; p < 0.05) and functional rehabilitation (OR 0.16 [95% CrI 0.02 to 0.90]; p < 0.05). We found no difference in complications resulting in surgery between functional rehabilitation and open surgery (OR 1.46 [95% CrI 0.35 to 5.36]; p > 0.05). Immobilization was associated with a greater risk of complications resulting in surgery than any other treatment. Minimally invasive surgery was ranked most highly for complications resulting in surgery (SUCRA 0.99) (Table 3). Consistency was present between pairwise and network comparisons (Supplemental Fig. 5; Supplemental Digital Content 8, https://links.lww.com/CORR/A591).

Table 3. - Summary of findings table for complications resulting in surgery
Studies: 15a
Participants: 949
Minimally invasive surgery (9 RCTsb; 250 participants) Functional rehabilitation (6 RCTsb; 182 participants) Primary immobilization (2 RCTsb; 99 participants) Open surgery (14 RCTsb; 418 participants)
Relative effect (95% CrI) 0.22 (0.04-0.93)c 1.46 (0.35-5.36)d > 100 (22.1 to > 100)c Reference
NNT 40.5 32.5e (Harm) 101.7e (Harm) Reference
GRADE evaluation Moderatef Lowf,g Very lowf,h Reference
Mean ranking (95% CrI) 1.04 (1-2) 2.71 (2-3) 4.00 (4-4) 2.25 (2-3)
Interpretation of findings Probably superior Possibly inferior Definitely inferior Reference
Relative effect values are odds ratios relative to open surgery; mean rank was calculated based on surface under the cumulative ranking curve (SUCRA) values; small sample sizes were considered in the evaluation of imprecision.
aTotal number of studies across all treatments.
bNumber of RCTs including the treatment of interest.
cInferred to be statistically significant with a 95% CI for mean difference not crossing one; p < 0.05.
d95% CrI for odds ratio crosses one, indicating no difference relative to reference treatment; p > 0.05.
eTreatment associated with a relative effect indicating harm. Therefore, associated value is number needed to harm.
fDowngraded for risk of bias.
gDowngraded for imprecision.
hDowngraded two levels for imprecision; NNT = number needed to treat; GRADE = Grading of Recommendations Assessment, Development, and Evaluation; CrI = credible interval.

Discussion

Treatment of acute Achilles tendon ruptures remains an area of uncertainty, despite an abundance of RCTs and resultant meta-analyses. Our goal was to use a network meta-analysis to compare treatments for acute Achilles tendon ruptures, including treatments infrequently compared in head-to-head RCTs, in a simultaneous and comprehensive manner not otherwise possible with simple meta-analysis. We found no difference in the rerupture risk between open surgical repair, MIS repair, and functional rehabilitation; immobilization was associated with a greater risk of rerupture than open surgical repair. We found a lower risk of complications resulting in surgery after MIS repair relative to both open surgery and functional rehabilitation.

Limitations

Network meta-analysis can be a powerful tool for comparing nearly all randomized evidence for treatment of a given pathology. However, this study has several key limitations, two of which resulted from the choice of complications resulting in surgery as a primary outcome. First, rerupture was the most common complication resulting in surgery, and most reported reruptures were treated surgically. Therefore, readers must be aware that treatments with an increased risk of rerupture will be overrepresented in our study (because rerupture was counted in both study endpoints), and any conclusions should be balanced with information gathered from existing studies on complications other than rerupture [44, 66]. Specifically, following functional rehabilitation, more than 80% of complications resulting in operation were from rerupture, a greater proportion than MIS and open surgery. Second, we settled on the outcome of complications resulting in surgery for pragmatic reasons—to facilitate robust, pooled analysis of serious complications with tangible implications for both patients and surgeons. Unfortunately, specific complications resulting in surgery other than rerupture (such as deep infection) were reported with insufficient frequency for independent network analysis [95]. Complications associated with serious morbidity not resulting in further surgery such as deep vein thrombosis and pulmonary embolism (both of which are more common after operative management) were inconsistently reported in the included studies and were therefore excluded from our analysis [70]. Readers should interpret the risk of complications resulting in surgery within the context of existing evidence that has explored serious complications not resulting in further surgery, complications that may still lead to substantial patient morbidity [59].

Another limitation of our study is that we were unable to analyze both minor, more frequent complications, such as keloid scars, skin adhesions, and superficial infection, as well as other complications such as sural nerve injury (24% of all complications after MIS) and complex regional pain syndrome (CRPS). At the outset of this study, it was our goal to analyze these complications [61]; however, during the review process it became clear that pooling in this manner would have resulted in a study endpoint that grouped relatively inconsequential complications with very serious ones, which can be misleading. For example, it would not be appropriate to pool superficial skin infection resolving with a short course of oral antibiotics—the most common complication of surgical management (31% of all complications)—with permanent sural nerve dysfunction or CRPS, which are associated with substantial morbidity [43, 60]. Although independent analysis of these complications is undoubtedly important, it was not possible to do that in the context of this network meta-analysis. For example, other than very rare instances, wound complications and infection occur only after MIS and open surgical repair, and therefore, a pairwise meta-analysis may be more appropriate [19]. Further, with a complication such as sural nerve injury (24% of all MIS-associated complications), analyzing a variety of MIS techniques together may be inappropriate secondary to a vastly different risk of sural nerve injury across the techniques [55]. Considering the above, our conclusions are based on only two facets of morbidity after treatment and should therefore be viewed in light of past analyses on specific complications. In addition to rerupture and complications resulting in surgery, patients should be counseled about the risks of other complications such as wound complications, infection, nerve injury, and deep vein thrombosis [44, 57, 70]

Patient-reported outcomes are increasingly important in Achilles tendon rupture literature; focus has shifted at least partly from rerupture risk to validated measures of patient function. Although described in our protocol, we were unable to perform an appropriate network meta-analysis for this outcome because of small sample sizes and heterogeneity of outcome measures, even after considering the use of standardized mean differences [14, 31, 45, 51]. Inferences would have been driven by limited direct comparisons and frequent third-order comparisons (two intermediary comparators are needed to form a network), resulting in very low confidence in network estimates [95]. Differences in treatment specifics within groups may have further limited the validity of pooling studies to perform this analysis. Recent evidence has found that distinct surgical variations and variations in postoperative protocols result in different degrees of tendon elongation [8, 18, 27, 30, 40, 72, 86]. Further, post-treatment tendon elongation has been associated with lower patient-reported outcomes scores and reduced plantarflexion strength [28, 79, 91]. For these reasons, surgeons must largely rely on direct RCT evidence of patient-reported outcomes rather than pooled comparisons when putting the findings of our study into context. For reference, most included studies (6 of 7) reported no difference between MIS and open surgery, and one study reported MIS was superior to open surgery. One study reported that open surgery outperformed functional rehabilitation (Supplemental Table 4; Supplemental Digital Content 9, https://links.lww.com/CORR/A592). Only the minority of studies found differences in strength parameters (Supplemental Table 5; Supplemental Digital Content 10, https://links.lww.com/CORR/A593).

Similarly, return to work and sport analyses were not performed because of very low evidence quality and poor assessment of the endpoints in question. Inconsistency in return-to-sport reporting has been previously noted by other researchers [3, 44, 66], although studies have found no difference between open repair, MIS repair, and functional rehabilitation [23, 60, 66]. In our study, most included RCTs found no difference between return to work (Supplemental Table 6; Supplemental Digital Content 11, https://links.lww.com/CORR/A594); some studies also found no difference in return to sport (Supplemental Table 7; Supplemental Digital Content 12, https://links.lww.com/CORR/A595). Of great concern regarding risk of bias, nearly all studies lacked rigorous blinding, and authors largely did not outline return to activity (such as work and sport) criteria in detail. Surgeons must be cautious when considering our findings in the context of existing, high risk-of-bias RCT evidence on inter-treatment differences in return to activity.

Finally, several studies did not report the presence of risk factors for complications such as smoking, diabetes, and fluoroquinolone use [5]. However, the treatment indications and inclusion criteria were very similar or identical between included studies so it is unlikely that there was unequal distribution of effect modifiers between pooled network groups.

Rerupture

We found no difference in the risk of rerupture between open surgical repair, MIS repair, and functional rehabilitation. We also found that immobilization was associated with a greater risk of rerupture than open surgical repair, and although the quality of evidence was low, this finding is in agreement with existing evidence [66]. Although our study did not investigate the effect of early versus late weightbearing after open surgical treatment, previous analyses have demonstrated both postoperative protocols reduce rerupture risk compared with primary immobilization [66, 80]. When a functional rehabilitation protocol is used (that is, early full weightbearing with progressive ROM), our study and others, including recent meta-analyses pooling observational data from thousands of patients, have found no difference in rerupture risk between open surgery, MIS, and nonoperative treatment [60, 66, 78, 80, 89, 99]. Our findings are congruent with and further support the paradigm shift over the past decade: Although previously, operative treatment was considered the gold standard largely because of decreased rerupture risk, nonoperative treatment with functional rehabilitation has been accepted as a viable alternative with a rerupture risk that is no different from operative treatment [44, 66, 77]. However, the question of rerupture risk remains unsettled. To further inform treatment decision-making, future research, both randomized and observational, should examine rerupture risk between treatments in populations that may be at higher risk for poor outcomes and rerupture such as those with increased tendon diastasis or more proximal ruptures [26, 92], and older patients or those with higher BMI [7, 50, 67, 82, 93]. Rerupture risk between treatments should also be investigated in high-demand groups such as younger patients and those engaged in athletics [38].

Complications Resulting in Surgery

Our analysis of reoperation found that MIS was associated with a lower reoperation risk relative to open surgery, functional rehabilitation, and primary immobilization. The comparison between MIS and open surgery was informed by moderate-quality evidence, and the difference between MIS and functional rehabilitation was supported by low-quality evidence. Interestingly, it appears that the risk of complications resulting in operation (or reoperation, in the case of operative management) is rarely, if ever, reported in existing meta-analyses. Our findings contrast with a recent retrospective study comparing 270 patients treated with either percutaneous or open repair that found no difference in risk of complications resulting in surgery [34]. Complications resulting in surgery, most frequently rerupture and deep infection, have been associated with poor patient-reported outcomes, and in many instances, lead to severe long-term functional deficits, particularly if repeat or extensive revision surgery is needed [58, 59, 68]. For this reason, we believe patients and surgeons should consider the moderate-quality evidence that MIS may be associated with a reduced risk of complications resulting in surgery (number needed to treat = 40), particularly when compared with open surgical repair. However, patients and surgeons must balance this benefit with the potential drawbacks of MIS repair [55]. Of note, reruptures may be treated with either nonoperative management or revision surgery, though in the case of revision surgery, more involved techniques such as fascial flaps and allograft are typically used [58, 59, 68, 93]. The difference we found also calls into question the conclusions of existing cost-efficacy analyses that have assumed equal reoperation between treatments [87, 94]. The implications of further surgery on total treatment cost are likely substantial, and therefore, our findings may be of interest to policy-makers. Overall, as our study is the first to demonstrate a difference in complications resulting in surgery, future studies (including meta-analyses) should include this outcome, as it is likely of interest to patients and surgeons alike.

Conclusion

Faced with acute Achilles tendon rupture, patients should be counseled that, based on current evidence, the rerupture risk likely is no different across contemporary treatments. Considering the possibly lower risk of complications resulting in surgery associated with MIS repair, patients and surgeons must balance any benefit with the potential risks of MIS techniques. As treatments continue to evolve, consistent reporting of validated patient-reported outcome measures is critically important to facilitate analysis with existing RCT evidence. Infrequent but serious complications such as rerupture and deep infection should be further explored to determine whether meaningful differences exist in specific patient populations.

References

1. Aisaiding A, Wang J, Maimaiti R, et al. A novel minimally invasive surgery combined with early exercise therapy promoting tendon regeneration in the treatment of spontaneous Achilles tendon rupture. Injury. 2018;49:712-719.
2. Aktas S, Kocaoglu B. Open versus minimal invasive repair with Achillon device. Foot Ankle Int. 2009;30:391-397.
3. Ardern CL, Glasgow P, Schneiders A, et al. 2016 Consensus statement on return to sport from the First World Congress in Sports Physical Therapy, Bern. Br J Sports Med. 2016;50:853-864.
4. Brooks SP, Gelman A. General methods for monitoring convergence of iterative simulations. J Comput Graph Stat. 1998;7:434-455.
5. Bruggeman NB, Turner NS, Dahm DL, et al. Wound complications after open Achilles tendon repair: an analysis of risk factors. Clin Orthop Rel Res. 2004;427:63-66.
6. Caldwell DM, Ades AE, Higgins JPT. Simultaneous comparison of multiple treatments: combining direct and indirect evidence. BMJ. 2005;331:897-900.
7. Cao S, Teng Z, Wang C, Zhou Q, Wang X, Ma X. Influence of Achilles tendon rupture site on surgical repair outcomes. J Orthop Surg. 2021;29:23094990211007616.
8. Carmont MR, Kuiper JH, Grävare Silbernagel K, Karlsson J, Nilsson-Helander K. Tendon end separation with loading in an Achilles tendon repair model: comparison of non-absorbable vs. absorbable sutures. J Exp Orthop. 2017;4:26
9. Catalá-López F, Tobías A, Cameron C, Moher D, Hutton B. Network meta-analysis for comparing treatment effects of multiple interventions: an introduction. Rheumatol Int. 2014;34:1489-1496.
10. Cetti R, Christensen SE, Ejsted R, Jensen NM, Jorgensen U. Operative versus nonoperative treatment of Achilles tendon rupture: a prospective randomized study and review of the literature. Am J Sports Med. 1993;21:791-799.
11. Chiodo CP, Glazebrook M, Bluman EM, et al. Diagnosis and treatment of acute Achilles tendon rupture. J Am Acad Orthop Surg. 2010;18:503-510.
12. Citrome L, Ketter TA. When does a difference make a difference? Interpretation of number needed to treat, number needed to harm, and likelihood to be helped or harmed. Int J Clin. 2013;67:407-411.
13. Costa ML, MacMillan K, Halliday D, et al. Randomised controlled trials of immediate weight-bearing mobilisation for rupture of the tendo Achillis. J Bone Joint Surg Br. 2006;88:69-77.
14. Demets DL. Methods for combining randomized clinical trials: strengths and limitations. Stat Med. 1987;6:341-350.
15. Dias S, Sutton AJ, Ades AE, Welton NJ. Evidence synthesis for decision making 2: a generalized linear modeling framework for pairwise and network meta-analysis of randomized controlled trials. Med Decis Making. 2013;33:607-617.
16. Dias S, Welton NJ, Sutton AJ, Ades AE. NICE DSU Technical support document 2: a generalised linear modelling framework for pairwise and network meta-analysis of randomised controlled trials. Available at: https://www.ncbi.nlm.nih.gov/books/NBK310366/. Accessed September 30, 2020.
17. Foote CJ, Chaudhry H, Bhandari M, et al. Network meta-analysis: users’ guide for surgeons: part I-credibility. Clin Orthop Rel Res. 2015;473:2166-2171.
18. Frosch S, Buchhorn G, Hawellek T, Walde TA, Lehmann W, Hubert J. Comparison of the double loop knot stitch and Kessler stitch for Achilles tendon repair: a biomechanical cadaver study. PLoS One. 2020;15:e0243306.
19. Gatz M, Driessen A, Eschweiler J, Tingart M, Migliorini F. Open versus minimally-invasive surgery for Achilles tendon rupture: a meta-analysis study. Arch Orthop Trauma Surg. 202;141:383-401.
20. Gelman A. Inference and monitoring convergence. In: Gilks WR, Richardson S, Spiegelhalter DJ, eds. Markov Chain Monte Carlo in Practice. Chapman & Hall; 1996:131-143.
21. Gigante A, Moschini A, Verdenelli A, Del Torto M, Ulisse S, De Palma L. Open versus percutaneous repair in the treatment of acute Achilles tendon rupture: a randomized prospective study. Knee Surg Sports Traumatol Arthrosc. 2008;16:204-209.
22. Glazebrook M, Rubinger D. Functional rehabilitation for nonsurgical treatment of acute Achilles tendon rupture. Foot Ankle Clin. 2019;24:387-398.
23. Grassi A, Amendola A, Samuelsson K, et al. Minimally invasive versus open repair for acute Achilles tendon rupture: meta-analysis showing reduced complications, with similar outcomes, after minimally invasive surgery. J Bone Joint Surg. 2018;100:1969-1981.
24. Grävare Silbernagel K, Brorsson A, Olsson N, Eriksson BI, Karlsson J, Nilsson-Helander K. Sex differences in outcome after an acute Achilles tendon rupture. Orthop J Sports Med. 2015;3:2325967115586768.
25. Guyatt GH, Oxman AD, Schunemann HJ, Tugwell P, Knottnerus A. GRADE guidelines: a new series of articles in the Journal of Clinical Epidemiology. J Clin Epidemiol. 2011;64:380-382.
26. Hansen MS, Vestermark MT, Hölmich P, Kristensen MT, Barfod KW. Individualized treatment for acute Achilles tendon rupture based on the Copenhagen Achilles rupture treatment algorithm (CARTA): a study protocol for a multicenter randomized controlled trial. Trials. 2020;21:399.
27. Hapa O, Erduran M, Havitçioğlu H, et al. Strength of different Krackow stitch configurations using high-strength suture. J Foot Ankle Surg. 2013;52:448-450.
28. Heikkinen J, Lantto I, Piilonen J, et al. Tendon length, calf muscle atrophy, and strength deficit after acute Achilles tendon rupture: long-term follow-up of patients in a previous study. J Bone Joint Surg Am. 2017;99:1509-1515.
29. Henríquez H, Muñoz R, Carcuro G, Bastías C. Is percutaneous repair better than open repair in acute Achilles tendon rupture? Clin Orthop Relat Res. 2012;470:998-1003.
30. Herbort M, Haber A, Zantop T, et al. Biomechanical comparison of the primary stability of suturing Achilles tendon rupture: a cadaver study of Bunnell and Kessler techniques under cyclic loading conditions. Arch Orthop Trauma Surg. 2008;128:1273-1277.
31. Higgins JPT, Thomas J, Chandler J, et al. Cochrane handbook for systematic reviews of interventions version 6.0. Available at: www.training.cochrane.org/handbook. Accessed March 10, 2021.
32. Higgins JPT, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med. 2002;21:1539-1558.
33. Ho G, Tantigate D, Kirschenbaum J, Greisberg JK, Vosseller JT. Increasing age in Achilles rupture patients over time. Injury. 2017;48:1701-1709.
34. Hsu AR, Jones CP, Cohen BE, Davis WH, Ellington JK, Anderson RB. Clinical outcomes and complications of percutaneous Achilles repair system versus open technique for acute Achilles tendon ruptures. Foot Ankle Int. 2015;36:1279-1286.
35. Hurley E, Yasui Y, Gianakos A, et al. Achilles tendon repair-a systematic review of overlapping meta-analyses. Foot Ankle Orthop. 2017;2:2473011417S000212.
36. Hutton B, Salanti G, Caldwell DM, et al. The PRISMA extension statement for reporting of systematic reviews incorporating network meta-analyses of health care interventions: checklist and explanations. Ann Intern Med. 2015;162:777-784.
37. Huttunen TT, Kannus P, Rolf C, Felländer-Tsai L, Mattila VM. Acute Achilles tendon ruptures: incidence of injury and surgery in Sweden between 2001 and 2012. Am J Sports Med. 2014;42:2419-2423.
38. Jallageas R, Bordes J, Daviet JC, Mabit C, Coste C. Evaluation of surgical treatment for ruptured Achilles tendon in 31 athletes. Orthop Traumatol Surg Res. 2013;99:577-584.
39. Jansen JP, Crawford B, Bergman G, Stam W. Bayesian meta-analysis of multiple treatment comparisons: an introduction to mixed treatment comparisons. Value Health. 2008;11:956-964.
40. Kangas J, Pajala A, Ohtonen P, Leppilahti J. Achilles tendon elongation after rupture repair: a randomized comparison of 2 postoperative regimens. Am J Sports Med. 2007;35:59-64.
41. Karabinas PK, Benetos IS, Lampropoulou-Adamidou K, Romoudis P, Mavrogenis AF, Vlamis J. Percutaneous versus open repair of acute Achilles tendon ruptures. Eur J Orthop Surg Traumatol. 2014;24:607-613.
42. Keating JF, Will EM. Operative versus non-operative treatment of acute rupture of tendo Achillis: a prospective randomised evaluation of functional outcome. J Bone Joint Surg Br. 2011;93:1071-1078.
43. Kemler MA, de Vet HC. Health-related quality of life in chronic refractory reflex sympathetic dystrophy (complex regional pain syndrome type I). J Pain Symptom Manage. 2000;20:68-76.
44. Khan RJ, Smith RLC. Surgical interventions for treating acute Achilles tendon ruptures. Cochrane Database Syst Rev. 2010;(9):CD003674.
45. Kitaoka HB, Alexander IJ, Adelaar RS, Nunley JA, Myerson MS, Sanders M. Clinical rating systems for the ankle-hindfoot, midfoot, hallux, and lesser toes. Foot Ankle Int. 1994;15:349-353.
46. Kołodziej Ł, Bohatyrewicz A, Kromuszczyńska J, Jezierski J, Biedroń M. Efficacy and complications of open and minimally invasive surgery in acute Achilles tendon rupture: a prospective randomised clinical study—preliminary report. Int Orthop. 2013;37:625-629.
47. Kou J. AAOS clinical practice guideline: acute Achilles tendon rupture. J Am Acad Orthop Surg. 2010;18:511-513.
48. Kristman V, Manno M, Côté P. Loss to follow-up in cohort studies: how much is too much. Eur J Epidemiol. 2004;19:751-760.
49. Lantto I, Heikkinen J, Flinkkila T, et al. A prospective randomized trial comparing surgical and nonsurgical treatments of acute Achilles tendon ruptures. Am J Sports Med. 2016;44:2406-2414.
50. Lawrence JE, Nasr P, Fountain DM, Berman L, Robinson AH. Functional outcomes of conservatively managed acute ruptures of the Achilles tendon. Bone Joint J. 2017;99:87-93.
51. Leppilahti J, Forsman K, Puranen J, Orava S. Outcome and prognostic factors of Achilles rupture repair using a new scoring method. Clin Orthop Relat Res. 1998:152-161.
52. Lim J, Dalai R, Waseem M. Percutaneous vs. open repair of the ruptured Achilles tendon—a prospective randomized controlled study. Foot Ankle Int. 2001;22:559-568.
53. Lu G, Ades AE. Combination of direct and indirect evidence in mixed treatment comparisons. Stat Med. 2004;23:3105-3124.
54. Majewski M, Rickert M, Steinbruck K. Achilles tendon rupture. A prospective study assessing various treatment possibilities. [in German]. Orthopade. 2000;29:670-676.
55. Majewski M, Rohrbach M, Czaja S, Ochsner P. Avoiding sural nerve injuries during percutaneous Achilles tendon repair. Am J Sports Med. 2006;34:793-798.
56. Mattila VM, Huttunen TT, Haapasalo H, et al. Declining incidence of surgery for Achilles tendon rupture follows publication of major RCTs: evidence-influenced change evident using the Finnish registry study. Br J Sports Med. 2015;49:1084-1086.
57. McMahon SE, Smith TO, Hing CB. A meta-analysis of randomised controlled trials comparing conventional to minimally invasive approaches for repair of an Achilles tendon rupture. Foot Ankle Surg. 2011;17:211-217.
58. Metz R, van der Heijden GJ, Verleisdonk EJ, Andrlik M, van der Werken C. Persistent disability despite sufficient calf muscle strength after rerupture of surgically treated acute Achilles tendon ruptures. Foot Ankle Spec. 2011;4:77-81.
59. Metz R, van der Heijden GJ, Verleisdonk EJ, Kolfschoten N, Verhofstad MH, van der Werken C. Effect of complications after minimally invasive surgical repair of acute Achilles tendon ruptures: report on 211 cases. Am J Sports Med. 2011;39:820-824.
60. Metz R, Verleisdonk EJ, van der Heijden GJ, et al. Acute Achilles tendon rupture: minimally invasive surgery versus nonoperative treatment with immediate full weightbearing—a randomized controlled trial. Am J Sports Med. 2008;36:1688-1694.
61. Meulenkamp B, Stacey D, Fergusson D, Hutton B, Mlis RS, Graham ID. Protocol for treatment of Achilles tendon ruptures; a systematic review with network meta-analysis. Syst Rev. 2018;7:247.
62. Möller M, Movin T, Granhed H, Lind K, Faxen E, Karlsson J. Acute rupture of tendo Achillis: a prospective, randomised study of comparison between surgical and non-surgical treatment. J Bone Joint Surg Br. 2001;83:843-848.
63. Molloy A, Wood EV. Complications of the treatment of Achilles tendon ruptures. Foot Ankle Clin. 2009;14:745-759.
64. Nilsson-Helander K, Grävare Silbernagel K, Thomee R, et al. Acute Achilles tendon rupture: a randomized, controlled study comparing surgical and nonsurgical treatments using validated outcome measures. Am J Sports Med. 2010;38:2186-2193.
65. Nistor LA. Surgical and non-surgical treatment of Achilles tendon rupture. A prospective randomized study. J Bone Joint Surg. 1981;63:394-399.
66. Ochen Y, Beks RB, van Heijl M, et al. Operative treatment versus nonoperative treatment of Achilles tendon ruptures: systematic review and meta-analysis. BMJ. 2019;7:364.
67. Olsson N, Petzold M, Brorsson A, Karlsson J, Eriksson BI, Silbernagel KG. Predictors of clinical outcome after acute Achilles tendon ruptures. Am J Sports Med. 2014;42:1448-1455.
68. Pajala A, Kangas J, Ohtonen P, Leppilahti J. Rerupture and deep infection following treatment of total Achilles tendon rupture. J Bone Joint Surg. 2002;84:2016-2021.
69. Patel MS, Kadakia AR. Minimally invasive treatments of acute Achilles tendon ruptures. Foot Ankle Clin. 2019;24:399-424.
70. Patel A, Ogawa B, Charlton T, Thordarson D. Incidence of deep vein thrombosis and pulmonary embolism after Achilles tendon rupture. Clin Orthop Relat Res. 2012;470:270-274.
71. Puhan MA, Schunemann HJ, Murad MH, et al. A GRADE working group approach for rating the quality of treatment effect estimates from network meta-analysis. BMJ. 2014;349:g5630-g5630.
72. Rosso C, Vavken P, Polzer C, et al. Long-term outcomes of muscle volume and Achilles tendon length after Achilles tendon ruptures. Knee Surg Sports Traumatol Arthrosc. 2013;21:1369-1377.
73. Rozis M, Benetos IS, Karampinas P, Polyzois V, Vlamis J, Pneumaticos SG. Outcome of percutaneous fixation of acute Achilles tendon ruptures. Foot Ankle Int. 2018;39:689-693.
74. Salanti G. Indirect and mixed‐treatment comparison, network, or multiple‐treatments meta‐analysis: many names, many benefits, many concerns for the next generation evidence synthesis tool. Res Synth Methods. 2012;3:80-97.
75. Salanti G, Ades AE, Ioannidis JPA. Graphical methods and numerical summaries for presenting results from multiple-treatment meta-analysis: an overview and tutorial. J Clin Epidemiol. 2011;64:163-171.
76. Sampson M, McGowan J, Cogo E, Grimshaw J, Moher D, Lefebvre C. An evidence-based practice guideline for the peer review of electronic search strategies. J Clin Epidemiol. 2009;62:944-952.
77. Sheth U, Wasserstein D, Jenkinson R, Moineddin R, Kreder H, Jaglal SB. The epidemiology and trends in management of acute Achilles tendon ruptures in Ontario, Canada: a population-based study of 27 607 patients. Bone Joint J. 2017;99:78-86.
78. Shi F, Wu S, Cai W, Zhao Y. Multiple comparisons of the efficacy and safety for six treatments in acute Achilles tendon rupture patients: a systematic review and network meta-analysis. Foot Ankle Surg. 2020:S1268-7731
79. Silbernagel KG, Steele R, Manal K. Deficits in heel-rise height and Achilles tendon elongation occur in patients recovering from an Achilles tendon rupture. Am J Sports Med. 2012;40:1564-1571.
80. Soroceanu A, Sidhwa F, Aarabi S, Kaufman A, Glazebrook M. Surgical versus nonsurgical treatment of acute Achilles tendon rupture: a meta-analysis of randomized trials. J Bone Joint Surg Am. 2012;94:2136.
81. Spiegelhalter D, Thomas A, Best N, Dave L. OpenBUGS User Manual, version 3.2.3. Available at: http://www.openbugs.net/w/Manuals. Accessed March 10, 2020.
82. Stavenuiter XJR, Lubberts B, Prince RM 3rd, Johnson AH, DiGiovanni CW, Guss D. Postoperative complications following repair of acute Achilles tendon rupture. Foot Ankle Int. 2019;40:679-686.
83. Sterne JAC, Savovic J, Page MJ, et al. RoB 2: a revised tool for assessing risk of bias in randomised trials. BMJ. 2019;2:l4898.
84. Sturtz S, Ligges U, Gelman A. R2WinBUGS: a package for running WinBUGS from R. Stat Softw. 2019;12:1-16.
85. Thermann H, Zwipp H, Tscherne H. Functional treatment concept of acute rupture of the Achilles tendon. 2 years results of a prospective randomized study. Der Unfallchirurg. 1995;98:21.
86. Tian J, Rui Y, Xu Y, et al. A biomechanical comparison of Achilles tendon suture repair techniques: locking block modified Krackow, Kessler, and Percutaneous Achilles Repair System with the early rehabilitation program in vitro bovine model. Arch Orthop Trauma Surg. 2020;140:1775-1782.
87. Truntzer JN, Triana B, Harris AH, Baker L, Chou L, Kamal RN. Cost-minimization analysis of the management of acute Achilles tendon rupture. J Am Acad Orthop Surg. 2017;25:449-457.
88. Twaddle BC, Poon P. Early motion for Achilles tendon ruptures: Is surgery important? A randomized, prospective study. Am J Sports Med. 2007;35:2033-2038.
89. van der Eng DM, Schepers T, Goslings JC, Schep NW. Rerupture rate after early weightbearing in operative versus conservative treatment of Achilles tendon ruptures: a meta-analysis. J Foot Ankle Surg. 2013;52:622-628.
90. Wang D, Sandlin MI, Cohen JR, Lord EL, Petrigliano FA, SooHoo NF. Operative versus nonoperative treatment of acute Achilles tendon rupture: an analysis of 12,570 patients in a large healthcare database. Foot Ankle Surg. 2015;21:250-253.
91. Wang X, Liu H, Li D, Luo Z, Li Y, Zhang F. Modified Bunnell suture repair versus bundle-to-bundle suture repair for acute Achilles tendon rupture: a prospective comparative study of patients aged < 45 years. BMC Musculoskelet Disord. 2020;21:580.
92. Westin O, Nilsson Helander K, Grävare Silbernagel K, et al. Acute ultrasonography investigation to predict reruptures and outcomes in patients with an Achilles tendon rupture. Orthop J Sports Med. 2016;4:2325967116667920.
93. Westin O, Nilsson Helander K, Grävare Silbernagel K, et al. Patients with an Achilles tendon re-rupture have long-term functional deficits in function and worse patient-reported outcome than primary ruptures. Knee Surg Sports Traumatol Arthrosc. 2018; 26:3063-3072.
94. Westin O, Svensson M, Helander KN, et al. Cost-effectiveness analysis of surgical versus non-surgical management of acute Achilles tendon ruptures. Knee Surg Sports Traumatol Arthrosc. 2018;26:3074-3082.
95. White IR, Barrett JK, Jackson D, Higgins JPT. Consistency and inconsistency in network meta-analysis: model estimation using multivariate meta-regression. Res Synth Methods. 2012;3:111-125.
96. Willits K, Amendola A, Bryant D, et al. Operative versus nonoperative treatment of acute Achilles tendon ruptures: a multicenter randomized trial using accelerated functional rehabilitation. J Bone Joint Surg. 2010;92:2767-2775.
97. Woolnough T, Axelrod D, Bozzo A, et al. What is the relative effectiveness of the various surgical treatment options for distal radius fractures? A systematic review and network meta-analysis of randomized controlled trials. Clin Orthop Relat Res. 2021;479:348-362.
98. Wu Y, Mu Y, Yin L, Wang Z, Liu W, Wan H. Complications in the management of acute Achilles tendon rupture: a systematic review and network meta-analysis of 2060 patients. Am J Sports Med. 2019;47:2251-2260.
    99. Zellers JA, Carmont MR, Silbernagel KG. Return to play post-Achilles tendon rupture: a systematic review and meta-analysis of rate and measures of return to play. Br J Sports Med. 2016;50:1325-1332.
    100. Zhang H, Tang H, He Q, et al. Surgical versus conservative intervention for acute Achilles tendon rupture. Medicine. 2015;94:e1951.

    Supplemental Digital Content

    © 2021 by the Association of Bone and Joint Surgeons