Early functional rehabilitation versus traditional immobilization for surgical Achilles tendon repair after acute rupture: a systematic review of overlapping meta-analyses

Several meta-analyses comparing early functional rehabilitation and traditional immobilization following surgical Achilles tendon repair after acute rupture have been published. However, they have led to conflicting conclusions. The aims of this systematic review were to select high-quality meta-analyses from multiple discordant meta-analyses and to provide a postoperative rehabilitation strategy following surgical repair using currently available evidence. We performed a comprehensive search using the PubMed and Embase databases and the Cochrane Library. Assessment of Multiple Systematic Reviews (AMSTAR) instrument was used to assess the methodological quality. Three investigators independently applied the Jadad decision algorithm. Their results were then compared to ensure selection of a meta-analysis that provided the highest quality of evidence. Six meta-analyses met the eligibility criteria. AMSTAR scores ranged from 6 to 10. According to the Jadad decision algorithm, a high-quality meta-analysis with a greater number of RCTs was selected. This meta-analysis showed that early functional rehabilitation was superior to cast immobilization in terms of patient satisfaction and the time to return to pre-morbid sporting levels. There were no differences regarding major complications or the time before return to prior employment and sporting activity. Thus, we recommend early functional rehabilitation as the postoperative strategy for acute Achilles tendon ruptures.

Application of the Jadad Decision Algorithm. We interpreted and selected the discordant meta-analyses using the Jadad decision algorithm 20 , which is a useful tool to differentiate overlapping systematic reviews/meta-analyses. The Jadad Decision Algorithm 20 was designed based on following questions: (1) Do the meta-analyses ask the same question? (2) Do the meta-analyses include the same studies? (3) Do the meta-analyses containing the same trials have the same methodologic quality? (4) Do the discordant meta-analyses including different trials use the same selection criteria 20 ? Three investigators independently applied this algorithm, and their results were then compared to ensure selection of the meta-analysis that provided the highest quality of evidence to develop recommendations for a postoperative rehabilitation strategy after surgical repair.

Results
Search Results. We initially identified 274 abstracts. Six studies [14][15][16][21][22][23] met the eligibility criteria and were included. A flow diagram is shown in Fig. 1. These studies were published between 2004 and 2016. Three 15,21,23 of them were published within the past two years ( Table 1). All of the studies performed a meta-analysis. Thirteen RCTs were included in the various meta-analyses. The number of trials cited in the meta-analyses ranged from 5 to 10, with a median of seven primary RCTs cited ( Table 2). The number of included patients ranged from 273 to 570, with an average of 377 patients per study.
Search Methodology. All studies comprehensively searched databases. Each meta-analysis used at least four different databases to research RCTs. All studies included meta-analyses searched from PubMed or Medline, Embase and the Cochrane Library. There was heterogeneity among the other databases that were used. Details of the search methodology are summarized in Table 3.
Study Quality and Validity. Only RCTs were included for all meta-analyses. The AMSTAR scores were assessed for the included meta-analyses and ranged from 6 to 10 points, with a median of 7.3 points (with the maximum possible score of 11 points). One Cochrane Review 22 was assessed as the best-quality study. The AMSTAR score for each meta-analysis is shown in Table 4. According to the Oxford Levels of Evidence, three meta-analyses 15,16,22 were assessed as Level I evidence, and four studies 14,21,23 were assessed as Level II evidence (Table 5).
Heterogeneity Assessment. Statistical heterogeneity was assessed in each included meta-analysis using the I 2 statistic value. Revman software was used for all meta-analyses (Table 5). Only one meta-analysis 23 performed subgroup analyses. Five other meta-analyses 14-16,21,22 did not perform sensitivity or subgroup analyses ( Results of the Jadad Decision Algorithm. Three authors independently selected the same flow path according to the Jadad decision algorithm. Because (1) all included meta-analyses analysed the same question (comparing early functional rehabilitation with traditional immobilization following surgical Achilles tendon repair after acute rupture), (2) the included meta-analyses did not include the same RCTs, and (3) they did not have the same selection criteria, we selected the highest-quality meta-analysis based on the following factors: publication status of the primary studies, methodology of the primary studies, language restrictions and the analysis of data on individual patients. None of the eligible meta-analyses stated a restriction on the publication status of primary studies. Regarding the methodology of primary studies, all meta-analyses exclusively included RCTs with superior methodologies. McCormack and Bovard 15 restricted the study quality using a minimum of 19 items of the Downs and Black-validated 27-item checklist 24 . Concerning the publication characteristics, the included meta-analyses were published over an extended period. Thus, more recent meta-analyses were preferred to less recent meta-analyses. Additionally, all meta-analyses appropriately analysed the data. Finally, three authors selected the study by McCormack and Bovard 15 as the meta-analysis offering the best current evidence (Fig. 2).  McCormack and Bovard 15 reported that early functional rehabilitation was superior to cast immobilization regarding patient satisfaction and the time interval of return to prior sporting level. Furthermore, they found no differences concerning major complications and the rate of return to prior employment and sporting activity in their study. Hence, they concluded that early functional rehabilitation following surgical repair of acute Achilles tendon rupture was safe and leaded to higher patient satisfaction and earlier return to function.

Discussion
Multiple RCTs 9-13 comparing early functional rehabilitation with cast immobilization have been reported. However, they have presented conflicting findings regarding which strategy is better. Furthermore, many meta-analyses of RCTs, representing the highest grade of evidence, have been reported to compare two postoperative rehabilitation strategies. Although three meta-analyses 15,21,23 comprehensively searched databases during the most recent two years, they included different RCTs and reached conflicting results. Such discordance has confused decision makers (such as policymakers, clinicians and patients) who make choices based on these meta-analyses.
The possible sources of discordant results among meta-analyses were analysed and reported by Jadad et al. 20 Meanwhile, they designed a decision tool to choose high-quality evidence among discordant systematic reviews. This decision tool has been widely used to find the best available evidence among overlapping systematic reviews/ meta-analyses 25,26 . Khan et al. 22 performed a Cochrane systematic review comparing early functional rehabilitation with immobilization after surgical Achilles tendon repair. The Cochrane systematic review is internationally recognized as "the highest standard" in evidence-based health care resources and should be updated every two years to provide the most current evidence to decision makers. The Cochrane review by Khan et al. 22 obtained the highest scores based on the AMSTAR instrument in the current study. However, their review was performed in 2004 and was not updated for 13 years regarding the postoperative strategy, although multiple newly available RCTs have emerged. Hence, we did not select Khan et al. 22 as the best review. McCormack and Bovard 15 included the maximum number of RCTs and appropriately analysed the RCT data. Hence, the McCormack and Bovard 15 meta-analysis was selected to represent the best available current evidence.    McCormack and Bovard 15 found that the early rehabilitation group had three times the odds of rating patient satisfaction as 'good' or 'excellent' compared to the traditional immobilization group (p = 0.01; OR, 3.13; 95% CI 1.30 to 7.53). They also reported that there were no differences concerning major complications (including tendon re-rupture, deep infection, tendon adhesions, persistent functional/neurological deficit, thrombophlebitis/compartment syndrome and wound slough) or the rate of return to prior employment and sporting activity. However, five of the six included RCTs showed that the early rehabilitation group had a shorter time interval before return to prior sporting level than the traditional immobilization group. They concluded that early functional rehabilitation resulted in higher patient satisfaction and earlier return to function, with no increase in complications 15 . Thus, early functional rehabilitation may contribute to evidence-based practice for postoperative management after Achilles tendon repair.

Author/Year
The phases of tendon healing have been divided into inflammation, proliferation and remodelling 5 . Throughout these phases, the tensile strength of the tendon gradually increases, although it remains inferior to the uninjured tissue. The new scar tissue is biomechanically inferior, which causes increased stiffness and subsequent visco-elastic properties 27 . Animal studies have demonstrated that early rehabilitation can decrease excessive adhesion formation, improve biomechanical properties of the scar tissue and subsequently enhance the gliding function of the tendon 28,29 . Active or passive ankle joint motion exercises can eliminate local oedema and prevent joint stiffness and the atrophy of calf muscles 8 . Re-rupture following surgical repair is a potential complication for early function. Many surgeons worry that early function may increase the re-rupture rate. However, the results of the current study demonstrate that early functional rehabilitation after Achilles tendon repair does not increase the re-rupture rate and also improves patient satisfaction and facilitates earlier return to activity.
Our study has several strengths. First, this study can be considered as the first systematic review of overlapping meta-analyses to evaluate the rehabilitation strategy following surgical management of acute Achilles tendon rupture. Second, our study provides high-quality evidence for a postoperative rehabilitation strategy following surgical repair based on comprehensive summarization of previously published meta-analyses on this topic. Our study is a systematic review of overlapping meta-analyses, which is used to appraise the methodological quality and quality of reporting of meta-analyses. It is different from traditional systematic reviews that analyse primary studies. In this study, we selected the best of the currently published meta-analyses and provided high-level evidence to decision makers. Hence, the results of our study have greater strength than other systematic reviews or RCTs.
There are several limitations to this systematic review. First, this systematic review was not registered on prospective registration systems for systematic reviews. Prospective registration could improve the quality of a systematic review and increase confidence in the findings. Second, the AMSTAR tool showed that the methodological quality of the included meta-analyses was low. Although we only assessed meta-analyses, including RCTs, to ensure high methodological quality, three of the studies were assessed as Level II evidence. Most included meta-analyses did not provide a priori design. Furthermore, none of them assessed the likelihood of publication bias. Third, only English language studies were researched in this systematic review. Our search strategy may have excluded non-English studies meeting eligibility criteria. Fourth, there was a clear lack of blinding in all trials, which may have influenced patient-reported outcome measures, such as patient satisfaction. However, blinding was not possible for participants and clinicians because of the nature of surgical interventions. Fifth, the surgical techniques varied between trials, which also might have presented significant residual confounding. However, we also noted that several RCTs reported no significant differences in postoperative outcomes between traditional open and minimally invasive surgery [30][31][32] .  In brief, this systematic review of meta-analyses concludes that early functional rehabilitation can improve patient satisfaction and facilitate earlier return to activity without increased complication rates. Thus, we recommend early functional rehabilitation as the postoperative strategy for acute Achilles tendon ruptures.  Table 2) No (See Table 3