Aim This systematic review and meta-analysis sought to identify return to play (RTP) rates following Achilles tendon rupture and evaluate what measures are used to determine RTP.
Design A systematic review and meta-analysis were performed. Studies were assessed for risk of bias and grouped based on repeatability of their measure of RTP determination.
Data sources PubMed, CINAHL, Web of Science and Scopus databases were searched to identify potentially relevant articles.
Eligibility criteria for selecting studies Studies reporting RTP/sport/sport activity in acute, closed Achilles tendon rupture were included.
Results 108 studies encompassing 6506 patients were included for review. 85 studies included a measure for determining RTP. The rate of RTP in all studies was 80% (95% CI 75% to 85%). Studies with measures describing determination of RTP reported lower rates than studies without metrics described, with rates being significantly different between groups (p<0.001).
Conclusions 80 per cent of patients returned to play following Achilles tendon rupture; however, the RTP rates are dependent on the quality of the method used to measure RTP. To further understand RTP after Achilles tendon rupture, a standardised, reliable and valid method is required.
- Physical activity
- Achilles tendon
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
A rupture of the Achilles tendon predominantly occurs in middle-aged males during sports activity,1–8 with increasing incidence rates of as much as 69 per 100 000.1 ,9 The greatest increases in incidence occur in the 40–60 and over 60 age groups.1 ,10 ,11
Most people with an Achilles tendon rupture are recreationally competitive or involved in social sport12 ,13 at the time of injury and report a desire to return to same activity.14 The resumption of sports and physical activity is also an essential factor for the maintenance of health and prevention of morbidities following injury.15 Despite this goal, there are reports that only half of patients return to play 1 year after injury.14 ,16
There have been a variety of proposed reasons that may contribute to why not all patients are able to return to play. Rupture of the Achilles tendon leads to muscle weakness17 and decreased endurance,18 which persists to 10 years following injury.19–22 Performance of higher level activities, such as those required for return to play (RTP), also demonstrate large variations between individuals.17 In a study by Olsson et al,17 limb symmetry indexes between injured and non-injured sides ranged from 84% to 102% with SDs ranging from 15% to 26% during two jumping tasks. This variability may be due to any number of factors: gender, method of management, plyometric strength deficits, psychological components and other physical changes indirectly related to injury and recovery course.4 ,23–25 Patients may be physically able to return to sports activities,26 but the fear of re-rupture may cause an individual to avoid the sports activity during which injury occurred.14 Patients may also develop additional musculoskeletal problems related to changes in gait biomechanics27–29 such as knee injury or contralateral Achilles tendinopathy (25%) and rupture (6%).30
To appreciate the impact of Achilles tendon rupture on the ability for an individual to RTP, it is important to understand what the RTP rates are following injury and how RTP is currently evaluated. This will allow for standardised evaluation of RTP across studies, providing a solid basis for comparisons. Standardised evaluation can enable larger, combined cohorts more generaliseable to the larger population, which will help inform treatment and rehabilitative guidelines.
The aim of this study was to perform a systematic review and meta-analysis of RTP rates following Achilles tendon rupture and evaluate what measures are used to evaluate RTP.
Potentially relevant articles were identified via a search of PubMed, CINAHL, Web of Science and Scopus databases. Search terms included two strings linked by an AND modifier. The first search string included (Achilles tendon AND injur*) OR (Achilles tendon AND rupture). The second string was designated as (Recovery of function OR performance outcome* OR athletic performance* OR treatment outcome*). The terms tendinopathy and review were linked with NOT modifiers. Results were filtered to English language, human studies and adults. Databases were searched on 1 March 2016 with no restriction of time frame. Results of the database search were exported to a reference management database for review. Preliminary searches included terms such as ‘return to play’ and ‘return to sport’; however, these searches yielded less studies, and studies known to the authors as reporting RTP following Achilles tendon rupture were missing. Therefore, the more general, inclusive terminology was used. Articles were identified through electronic database searching; no hand searching, citation tracking or reference scanning was performed. The initial search and search strategy was performed by a reference librarian with input from the authors.
To be included, articles needed to describe patients with closed, acute Achilles tendon rupture and be of randomised control trial, cohort study or case series study design. Because of the demographics of typical patients with Achilles tendon rupture, inclusion was limited to adults (at least age 18). Finally, an outcome measure regarding return to activity, sport or play needed to be included within the abstract or body of the article. Articles were excluded based on the following criteria: case studies, chronic/delayed treatment (>4 weeks), open rupture/tendon laceration, inclusion of only patients with Achilles tendon re-rupture, non-English articles, non-human studies and bilateral ruptures. Studies including cohorts of included and excluded populations (eg, a study comparing outcomes in a cohort of acute Achilles tendon rupture vs chronic Achilles tendon rupture) were included, but only data meeting inclusion for the review were included in the data analysis (ie, acute Achilles tendon rupture).
To ensure that each study met the inclusion criteria, studies were reviewed first by title. Title review primarily excluded articles not relating to Achilles tendon rupture, basic science studies not relevant to the purpose of this study or studies pertaining to chronic rupture. Articles were then reviewed by abstract, followed by a full text review to ensure that the study included RTP outcomes either in the abstract or full text. All articles were reviewed by title and abstract by two independent reviewers to determine if the study met selection criteria. If consensus was unable to be reached, a third reviewer was consulted. For full text review, two independent reviewers each independently reviewed 25% of studies. Reviewers were consistently able to agree on inclusion of studies, and, therefore, the remaining studies were divided between the two reviewers owing to the volume of studies identified and time constraints. If there were any studies that did not clearly meet selection criteria, this study was brought to the attention of the second reviewer to ensure consensus in the studies included in the review. If consensus was unable to be reached, a third reviewer was consulted.
Risk of bias
All articles included for full text review were scored for risk of bias using the scoring system described by Ardern et al.31 This scoring system comprises six criteria, and the number of criteria fulfilled have been determined for each study. To fulfil all six criteria, the study must include selection criteria for patient inclusion/exclusion, report patient demographic data, report results of a representative population, report patients' preinjury level, compare patients' postintervention/postinjury sports level to their preinjury level and be of prospective study design. Studies were rated for risk of bias by two reviewers.
Data extraction and synthesis
Data were extracted from each included study by two reviewers. The following data were extracted: number of patients, patient demographics (age, sex), study design, percentage of patients returning to sport, measures of performance used, time frame to return to sport (with measure of variance) and time of evaluation of RTP. In studies that reported RTP percentages at multiple time points, data from the final study time point were used in our analysis. In retrospective studies where a mean follow-up time was reported, this value was used for time of evaluation of RTP. Studies that included measures of sport/activity performance but did not report a RTP percentage were included in the demographics analysis but were not included in the RTP rate meta-analysis.
When there is no definition of the measurement method, the risk of bias is considered very high, therefore, we also assessed the RTP-rated methodology of each study. Studies were classified into one of two groups. Group 1 included studies in which the measure could be implied based on how return to activity/sport was reported or the measure of RTP was clearly described in the study's methods (eg, studies that specified patients had returned to play at the same level as prior to injury or provided exact wording of interview/survey questions to patients or subscales of questionnaires specific to return to sport). Group 2 studies included those that reported a RTP percentage or time frame in the results, but the measure to determine RTP was not described and unable to be inferred based on the phrasing of the results (eg, studies that included a statement in the results such as ‘X% of patients returned to sport’). Two reviewers grouped studies into methodology classifications. Initially, two reviewers independently reviewed a subset of articles to check agreement. Then, remaining articles were divided between reviewers and reviewed by a single reviewer. If grouping classification was unclear, the second reviewer was requested to provide a classification rating. If consensus was not achieved, a third reviewer was consulted.
Means and SDs were calculated for demographic data. Pooled proportion of RTP rate, 95% CIs and I2 index (a measure of heterogeneity)32 were calculated for the total group and for the two RTP methodology subgroup classifications using StatsDirect Statistical Software (StatsDirect, England). A one-way analysis of variance (ANOVA) test was used to determine whether there were differences in RTP rates between the two RTP methodology subgroups. Pearson correlations were conducted to identify relationships between the RTP rate with the year of study and the time of RTP evaluation. ANOVA and correlations were done using SPSS V.23 (IBM SPSS Statistics for Windows, Armonk, New York, USA: IBM Corp). Weighted mean (weighted by number of patients) and SD were calculated for time to RTP means across all studies reporting a time to RTP mean with measure of variance.
Search results and description of studies
The database searches yielded 552 articles. A total of 322 articles were excluded based on title. Of the 230 articles reviewed by abstract, 199 articles were included for full text review. Of those, a total of 108 articles met the inclusion criteria (figure 1 and table 1).
Of the 108 studies, 71 reported selection criteria, 106 reported patient demographic data, 24 reported patient preinjury level of play, 82 compared postintervention/postinjury sports outcome to preinjury level and 57 were prospective in design. The mode qualitative assessment score of all included articles was four criteria fulfilled. There was no difference in mode of criteria fulfilled on qualitative assessment between methodology subgroups.
Patients were a mean (SD) of 41.4 (6.8) years of age, and 79.6% of patients were male. Six studies did not report patient sex,38 ,61 ,75 ,85 ,102 ,126 and five studies did not report patient age.45 ,62 ,73 ,75 ,80 Eighty-five of the studies reported RTP measure (RTP methodology subgroup 1) and 23 were classified as lacking RTP measure (methodology subgroup 2). There were 6506 patients included in all studies—5535 in subgroup 1 and 971 in subgroup 2.
Rate of RTP
The pooled rate of RTP (based on 98 studies) was 80% (95% CI 75% to 85%, I2=95.8%) (figure 2). Three studies81 ,95 ,121 did not report a RTP percentage but reported RTP time frames. These studies were included in the analysis for subject demographics and time frame to RTP. In methodology subgroup 1 studies, the rate of RTP was 77% (95% CI 70% to 83%) (figure 2). In methodology subgroup 2 studies, the rate of RTP was 91% (95% CI 84% to 96%) (figure 3). Lower RTP rates were significantly lower in subgroup 1 versus subgroup 2 studies (p<0.001). There was no relationship between RTP rate and year of study publication (Pearson r=−0.030, p=0.765).
Relationship between RTP and time of evaluation
The time of evaluation of RTP varied between 5 and 145.2 months (mean=23.8 months, SD=22.9 months). There was no relationship between time of evaluation and RTP rate (Pearson r=−0.026, p=0.803).
Time to RTP
The mean (SD) time to RTP (based on 37 studies) with a measure of variance was 6.0 (1.8) months (figure 4).
The results of this systematic review indicate that 80% of patients with Achilles tendon rupture will RTP; however, determination of the rate of RTP is subject to bias and dependent on the measure used to evaluate RTP. In the process of identifying how different studies evaluated RTP, we found that numerous studies report a rate of RTP without including how this rate was measured. While the rate of RTP for studies describing RTP measures fell 3% below the rate for all included studies, the rate of RTP for studies without described RTP measures (and higher risk of bias) was 11% higher than the rate for all included studies (see online supplementary figure S1).
In general, while there is good consensus that RTP is a goal in this population, the literature regarding the resumption of sports activity to preinjury levels following Achilles tendon rupture is limited. RTP is reported as an outcome in a minority (19.7%) of studies on recovery of function following Achilles tendon rupture, and a smaller number, 15.4% of 552 articles, report RTP with indication of what measure was used to gauge RTP. For the individual patient, RTP is often a primary goal and is an important component of success of treatment following Achilles tendon rupture. This points to the need for better, standardised measures for gauging RTP that encompass the multiple aspects that comprise an individual's ability to RTP.
Measures to assess RTP range from patient interview to the use of sports participation scores to objective sports performance indicators.41 ,120 The measures used to assess RTP performance address multiple aspects of play, including volume/frequency of play, type of sport/sport demands, level of play and play performance. Considering these aspects will be important in developing comprehensive measures that seek to capture RTP as a multifaceted concern.
Rate of RTP
There were large variations on RTP rates and time frames among the included studies. While the results indicate that 80% of subjects were able to return to play, studies reported as few as 18.6% and as large as 100% RTP rates. In studies of patients with ACL reconstruction, increasing stringency of RTP definition results in lower percentages of athletes who return to play.31 ,148 Therefore, the more conservative estimate of 77% provided by the methodology subgroup 1 studies in our review might be a more accurate (although likely still optimistic) estimate of true RTP rates.
There was no relationship between time of evaluation of RTP and RTP rate. This suggests that, while there was variability in time of evaluation, the majority of studies were of long enough duration to capture RTP outcomes.
Time to RTP
The average time to RTP was 6.0 months, but this also varied with study means/medians as low as 2.95 months and as high as 10.4 months. This estimate seems to coincide with timing of functional capacity recovery postrupture, which has been reported to recover rapidly 3 and 12 months,69 ,83 ,116 ,149 ,150 followed by slowed rates of change between 1 and 2 years postrupture.17 Prior studies evaluating the return of functional capacity following Achilles tendon rupture have reported large variations between individuals.151 ,152 Some of the variation seen in this analysis may be due to those individual variations. Additionally, differences in study design and definition of RTP may contribute to the variance in reported time of RTP. In the future, consistency and standardisation of RTP evaluation will assist in establishing RTP time frames that allow for improved comparison of multiple cohorts across studies.
Other functional and performance measures
There are a variety of functional and performance measures that have been used in the study of patients with Achilles tendon rupture. These measures present challenges when measuring RTP, however. On the one hand, there are a variety of validated, reliable questionnaires that assess different aspects of RTP that have been developed for use in non-Achilles tendon rupture populations. For instance, the PAS90 includes measures of activity intensity and duration. The Halasi Score,147 an adaptation of the Tegner Score,48 captures factors such as type and level of sport activity. The FAOS59 and FAAM33 have a sport-specific scale, which captures perceived difficulty with sport-related manoeuvres. While these scores capture certain components of the whole RTP picture, the difficulty with using these scores is that they were not developed specifically for use in this unique population and many have not been validated in the Achilles tendon rupture population. They may or may not capture the unique needs of this specific population. On the other hand, there are Achilles tendon rupture specific questionnaires that capture some aspects of performance but may not comprehensively capture RTP. For example, the ATRS40 is a valid, reliable, reproducible and responsive questionnaire for use in individuals following Achilles tendon rupture. While this questionnaire addresses multiple concerns reported by patients with Achilles tendon rupture, it only asks three questions regarding running, jumping and physical labour capacity, which may not adequately address the RTP question.
Ultimately, while there are numerous questionnaires that measure various aspects of RTP, there are none that comprehensively address the needs of this specific population. Given the importance of RTP for these individuals and the need to standardise measures to evaluate RTP, it would be beneficial for a consistent set of currently available tests and measures to be defined or for a measure to be developed that emphasises volume as well as type of sport with consideration of Achilles tendon loading and touches on sport performance.
Suggestions for clinical practice and research
From a study design and reporting standpoint, there are ways to improve measure of RTP using the tools currently available. While a majority of studies compare postinjury to preinjury rates, this was done via patient interview or unvalidated questionnaire in the majority of studies and only 22.2% of studies report a measure for preinjury performance. Preinjury rates are often obtained at time of follow-up and rely on accurate patient recollection of activity participation status. Prospective data regarding sport performance are ideal, but not always available. It may be beneficial to have patients rate sport performance early postinjury as opposed to a follow-up several weeks to months post-injury. Another thing to consider is improving the frequency of reporting sport-specific subscales of already validated questionnaires. The AOFAS,35 FAOS59 ,153 ,154 and Thermann43 Scores include sports and activity subscales, but these subscales are infrequently reported. Finally, though all terms implying sports participation of activity were included in this systematic review, it was noted that a number of different terms were used in the literature. Activity, level, standard, performance and intensity were all used; however, these may mean that different things to patients and responses may vary accordingly. Variations in terminology further complicate search strategy when attempting to find relevant literature. On the basis of a consensus statement from 2002, the use of RTP is recommended to improve consistency in phrasing.155
Study strengths and limitations
This systematic review assessed studies based on two separate set of criteria. The criteria developed by Ardern et al31 was used to evaluate study risk of bias. We also developed a second set of criteria to evaluate study methodology with regard to measuring RTP based on reproducibility. The combination of methods of assessing study risk of bias and methodology allowed us to distinguish between quality in reporting between groups of studies.
There are some limitations to our review. Included studies covered a range of research questions with a variety of set time points for follow-up and various types of treatments. Studies incorporated data from varied populations (recreational as well as professional athletes), which could impact the RTP rate and time frame. Broad study inclusion is beneficial for understanding RTP in the larger, varied population of individuals with Achilles tendon rupture. Similarly, while many terms were included in this literature search to capture the large number of different terms currently used to describe RTP, this allows for improved generalisability to this population of individuals. Exclusion criteria did eliminate studies of portions of the population of patients with Achilles tendon rupture (Achilles tendon laceration, delayed treatment), so the findings of this study may not be generalisable to these populations. Unpublished and non-English language studies were not included in this study, which introduces the possibility of publication and language bias.
Four out of five patients returned to sport after Achilles tendon rupture, at an average of 6 months following injury. Studies with well-defined RTP measures demonstrated lower RTP rates compared to studies that only reported the RTP rate with no description of how the rate was determined.
What are the findings?
About 80% of people are able to RTP following Achilles tendon rupture.
Measures that are better described and more reproducible are associated with lower RTP rates.
Measures evaluating RTP are variable and inconsistently reported.
How might it impact on clinical practice in the future?
Provide evidence-based RTP rate estimates for patient education.
Improve consistency in use and reporting of RTP measures.
Inform the future development of comprehensive evaluative tools to measure RTP.
The authors thank Sarah Katz, reference librarian, for her assistance with the literature search and Ryan Pohlig for his guidance with statistical analysis.
Contributors JAZ was one of the primary drafters of the manuscript, performed article review, data extraction and data analysis. MRC was one of the primary drafters of the manuscript, developed study concept and performed article review and data extraction. KGS developed study concept and research design, assisted in article review, data extraction and data analysis and provided critical feedback of the manuscript.
Funding This study was supported by the National Institute of Arthritis and Musculoskeletal and Skin Diseases of the National Institute of Health under awards R21 AR067390 and the University of Delaware Research Foundation. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.