Objectives To evaluate current practice in the treatment of proximal hamstring tendon avulsions and identify decision-making preferences.
Methods An invitation to an anonymous e-survey containing 32 questions was sent to 3475 members of the International Society of Arthroscopy, Knee Surgery and Orthopaedic Sports Medicine (ISAKOS) and the European College of Sports and Exercise Physicians (ECOSEP).
Results We received 403 (12%) unique responses with a completion rate of 79%. Participants were orthopaedic/trauma surgeons (90%), sports medicine physicians (7%) or physical therapists (2%). For 83% of the participants, the preferred treatment (ie, surgical or non-operative) depends on the individual case. Participants base their decision-making process on patient- and injury-related factors (decision modifiers). The five most frequently selected decision modifiers that support the choice for surgical treatment were diminished function (84%), neurological symptoms (74%), involved tendons (82%), tendon retraction on MRI (84%) and patient preference for surgery (78%). The majority prefer early surgical repair (<2 weeks after injury) to achieve highest functional outcome (63%) and ensure a low complication risk (61%). Suture anchors are used by 93% of the participants for tendon reattachment. Estimated recovery duration (ie, time to return to sports) was a median 12 weeks (IQR: 12–20) for non-operative treatment and 17 weeks (IQR: 12–24) for surgical treatment. Estimated reinjury risk was a median 25% (IQR: 10–31.5) and 10% (IQR: 5–20), respectively.
Conclusion This survey among experienced medical professionals has summarised current practice and identified treatment decision-making preferences. The typical surgical patient has a retracted (>2 cm) two-tendon avulsion (ie, common tendon and semimembranosus tendon), is unable to engage in sports or activities of daily life, reports sciatic symptoms and prefers surgical treatment. Surgery is thought to prolong recovery and decrease reinjury risk compared with non-operative treatment and is preferably performed early.
Level of evidence Level V.
- lower extremity
- orthopaedic sports medicine
- treatment / technique
Statistics from Altmetric.com
What are the new findings?
There is a sharp contrast between the relatively high frequency of non-operatively treated patients in practice reported by participants in our study and the low number of non-operatively treated patients in the literature.
The decision for non-operative and surgical treatment should depend on the individual patient. The five most used decision modifiers are diminished function, neurological symptoms, involved tendons, extent of tendon retraction on MRI and patient preference for surgery. The typical surgical patient has a retracted (>2 cm) two-tendon avulsion (ie, common tendon and semimembranosus tendon), is unable to engage in sports or activities of daily life, reports sciatic symptoms and prefers surgical treatment.
Early surgery (within 2 weeks after injury) is preferred to achieve highest function and lowest complication risk.
Recovery following non-operative treatment is considered shorter than recovery after surgical treatment at the cost of a higher reinjury risk.
Proximal hamstring tendon avulsions are estimated to constitute around 3% to 9% of hamstring injuries.1 2 Until the late 20th century, reports of full-thickness proximal hamstring injury due to tendon avulsion or rupture were limited to a handful of case reports.3–5 In the last two decades, the number of clinical studies on proximal hamstring avulsions has increased rapidly.6–26
Despite the recent advances in knowledge on these injuries, treatment decision-making has not become easier for several reasons. First, present systematic reviews on outcome after surgical and non-operative treatment have highlighted methodological limitations (and the ensuing risk of bias) of current studies making it difficult to provide clear recommendations.27–30 Second, the comparison between surgical and non-operative treatments suffers from a paucity of data on outcome of non-operative treatment, let alone the lack of comparative studies. Currently, outcome of only approximately 60 non-operatively treated patients has been described, including two small retrospective comparative studies.15 20 23 31–34
As a result of this lack of evidence, the comparison between outcomes following surgical and non-operative treatment is very limited. Therefore, it remains difficult to advise a surgical or non-operative approach for the individual patient and provide an accurate prognosis. We aim to evaluate current practice and decision-making for proximal hamstring tendon avulsions through an international survey among medical professionals in the field of orthopaedic traumatology and sports medicine.
Materials and methods
An invitation to an online survey was sent by email to all members of the International Society of Arthroscopy, Knee surgery and Orthopaedic Sports Medicine (ISAKOS) and the European College of Sports and Exercise Physicians (ECOSEP). The email contained information about the study purpose and procedures, as well as the link to the online survey. The first invitation was sent on 1 November 2016 with a reminder 4 weeks later. The survey closed on 1 January 2017.
Procedures and survey content
The survey was anonymous and did not collect sensitive data and was therefore exempted from ethics review by the local Institutional Review Board (Anti-Doping Lab Qatar, ADLQ). Participants were informed about the nature of the survey, the responsible research group and data collection. This included a statement that the survey was voluntary and that continuing to the survey signified agreement to participate. There was no incentive for the participants.
The survey was conducted using an online survey tool (SurveyMonkey, San Mateo, California, USA). For the development of the survey content and reporting of the survey results, the Checklist for Reporting Results of Internet E-Surveys (CHERRIES)35 was used. The survey consisted of a total of 32 items and made use of adaptive questioning. All questions required an answer to continue. Participants were able to review and change their answers through a ‘back’ button. A ‘review’ step displaying a summary of the responses was not included. Each response was assigned a unique identification number. No data on IP addresses was recorded to ensure anonymity of the participants.
The survey was conceptualised by one author (ADM) to reflect topical issues regarding hamstring avulsion treatment in the current literature. All authors including three peers per author provided feedback on content which was incorporated in the definitive survey.
Survey questions were related to the participants’ experience level with proximal hamstring tendon avulsions, treatment and prognosis. Using adaptive questioning, only participants that were involved in the treatment of proximal hamstring tendon avulsions were able to complete the survey. Based on participants’ preferred treatment, further questions only enquired about details/decision modifiers for the preferred treatment. Participants that only considered non-operative treatment were asked about how non-operative treatment was organised, while participants that only considered surgical treatment were asked about how surgical treatment was organised. In addition, participants that considered surgical treatment were asked how various factors influenced the choice for surgical treatment. These included age, gender, level of sports participation, activity frequency, tendons involved, extent of retraction, complaints of pain, complaints of neurological symptoms, complaints of diminished function and patient wish to undergo surgery. Participants that considered either treatment depending on the individual patient answered all survey questions.
An overview of the questions and distribution of responses can be found in the online supplementary appendix.
Descriptive data analysis was carried out using SPSS (Released 2015, IBM SPSS Statistics for Windows V.23.0, Armonk, New York, USA: IBM Corp). Frequencies and portions were reported as N (%). Continuous variables were presented as mean ± standard deviation for parametric data and as median (interquartile range) for non-parametric data.
The invitation to participate in the survey was sent to 3475 members, and we received 403 (12%) unique responses with a completion rate of 79%. Participants were orthopaedic/trauma surgeons (90%), sports medicine physicians (7%), physical therapists (2%) or other (1%). Thirty-one participants (8%) were not involved in the treatment of proximal hamstring tendon avulsions, and their survey ended here. Thirteen per cent treated more than 50 proximal hamstring tendon avulsions. Data for this group is presented separately in the online supplementary appendix.
Treatment choice and decision modifiers
Preferred treatment and frequency of surgical treatment in practice are shown in table 1. Rationales for surgical treatment selected by the majority (ie, >50%) of participants were return to sports at preinjury level (69%) and improved recovery of hamstring strength (lower strength deficit) (59%).
Participants were subsequently asked to select which decision modifiers (ie, patient- or injury-related characteristics) support the choice for surgical treatment and how selected modifiers impact their decision-making (table 2).
When surgical treatment was chosen, 93% use suture anchors to reattach the hamstrings to the ischial tuberosity. Allograft (9%) or autograft (8%) reconstruction is also performed. Preferred timing of surgery is summarised in table 1. Following surgery, 43% always immobilise the operated leg using a cast (5%) or brace (39%). Two per cent sometimes use a cast, 24% sometimes use a brace and the remaining 31% never use a cast/brace. For postoperative rehabilitation, 79% refer to a physiotherapist. The remaining participants provide a standard exercise protocol/home exercises (15%) or focus on symptom relief (2%).
If a non-operative approach was chosen, 77% refer to a physiotherapist, 9% provide a standard exercise protocol/home exercises and 9% focus on symptom relief.
Estimated recovery duration (ie, time to return to sports) was a median 12 weeks (IQR: 12–20) for non-operative treatment and 17 weeks (IQR: 12–24) for surgical treatment. The estimated risk of reinjury was a median 25% (IQR: 10–31.5) for non-operative treatment and 10% (IQR: 5–20) for surgical treatment.
This survey among experienced medical professionals to evaluate current practice in treatment of proximal hamstring tendon avulsions has identified several treatment decision-making preferences.
First, the decision for non-operative and surgical treatment should depend on the individual patient. For the decision-making process, the five most used decision modifiers are diminished function, neurological symptoms, involved tendons, extent of tendon retraction on MRI and patient preference for surgery. Second, when a surgical treatment is chosen, early surgery (within 2 weeks after injury) is preferred. Suture anchors are typically used for reattachment. When a non-operative treatment is chosen, patients are predominantly referred to a physiotherapist. Third, median recovery duration (ie, return to sports) is estimated to be 5 weeks longer for surgical treatment than for non-operative treatment. However, reinjury risk is estimated to be higher in patients that are treated non-operatively.
Treatment choice: practice versus evidence
We observed a sharp contrast between the relatively high frequency of non-operatively treated patients in practice reported by participants in our study and the low number of non-operatively treated patients in the literature. The vast majority (83%) of participants stated that the treatment choice should be based on the individual patient. About 60% reported that less than half of their patients with a proximal hamstring tendon avulsion were treated surgically. Yet, in current literature there are over 10 surgically treated patients for every non-operatively treated patient,15 20 23 27 31–34 potentially indicating publication bias.
The uneven distribution of published data on outcome following surgical and non-operative treatment, a scarcity of controlled studies and the high risk of (eg, selection) bias as highlighted by recent systematic reviews together make it difficult to properly compare surgical and non-operative treatment. The paucity of non-operatively treated patients in the literature is problematic for another reason. Without these data, evidence-based indications for surgical treatment cannot be determined. For identifying patients that will maximally benefit from surgical treatment, it is crucial to understand which patients would do poorly without surgery.
Participants selected several decision modifiers that are relevant for their decision-making process. Among the five most used decision modifiers, diminished function and patient preference are non-specific factors for this injury and should be considered in the decision-making process for any intervention. More hamstring-specific decision modifiers that complete this top five are neurological symptoms, tendons involved and extent of tendon retraction on MRI.
For three quarters of the participants, presence of sciatic pain supports a decision for surgical treatment. The sciatic nerve lies in close proximity to the proximal hamstring tendons,36–39 resulting in potential sciatic nerve-related symptoms in case of a proximal hamstring tendon avulsion. In an early phase, symptoms may result from stretch injury or nerve compression by an accompanying haematoma. In later stages, sciatic symptoms can arise from entrapment by adhesions. In a study by Wilson et al,40 distal (ie, below the knee) sciatic symptoms occurred in 45 of 162 (28%) of patients with a proximal hamstring tendon avulsion. Sciatic symptoms were divided into motor deficits (5%), sensory deficits (7%) and pain (22%). In surgically treated patients, within 1 year after surgery, motor deficits improved in 100%, sensory deficits in 75% and pain in 89% of patients. Median time to noted initial improvement for motor deficits, sensory deficits, and pain was 87, 23 and 44 days, respectively. However, since the rate of improvement of these symptoms in non-operatively treated patients is unknown, it remains speculation whether these symptoms resolved spontaneously or because of the intervention.
Two decision modifiers were related to imaging: the involved tendons and the amount of tendon retraction on MRI. For over 80% of participants, avulsion of both the common tendon (biceps femoris and semitendinosus) and semimembranosus tendon supports the decision for surgery. In other words, such two-tendon avulsions (ie, common tendon and semimembranosus tendon) are thought to do poorly with non-operative treatment and are therefore thought to require surgery. In the first case series of non-operatively treated proximal hamstring avulsions, all patients that were unable to run or return to sports requiring agility indeed had a two-tendon avulsion.23 Conversely, a recent non-operative series demonstrated that 71% of patients with a two-tendon avulsion returned to their previous sporting level activity.15 For comparison, surgical repair of two-tendon avulsions resulted in return to sports at preinjury level rates of 55%–96%.6 9 10 17 24 41
More than 80% of the participants use the presence and extent of retraction as a decision modifier. To our knowledge, there are no studies indicating that the degree of retraction is of prognostic value in non-operatively treated patients. Nevertheless, retraction is reported to increase over time and results in a more technically challenging repair.12 16 17 21 26 41
In summary, the decision modifiers to support the choice for surgical repair used in practice are not well supported by evidence. Further research is necessary to determine the prognostic value of these variables in non-operatively treated patients and, by extension, whether these variables are to be used in the treatment decision-making process as indications for surgery.
Prognosis and timing of surgery
The estimated recovery duration was longer for surgical (median 17 weeks) than for non-operative treatment (median 12 weeks). This difference is not surprising considering that there is often some degree of delay between injury and surgery,27 as well as an initial postoperative phase that focuses on protection of the reattached hamstrings.42 The recovery durations estimated by the survey participants were notably shorter than expected. There are no controlled trials comparing time to return to sports following both treatments. However, Hofmann et al 15 recommended at least 4 months of physiotherapy as non-operative treatment, while surgically treated patients are generally allowed to return to sport from about 6 months after surgery.6 9 10 12 14 16–18 21 22
Estimated reinjury risk following surgical (10%) and non-operative (25%) treatment is notably higher than reported in aggregated literature: the risk of a reinjury following surgical treatment is estimated at around 3%,29 30 reinjury risk for non-operatively treated patients has not been reported.
For achieving the highest functional outcome and low complication risk, participants prefer early surgical repair (<2 weeks after injury). The preferred timeframe may indicate that feel that early surgery is considered easier, leads to better results and lower risk of complications or both. Two systematic reviews using 4 weeks as cut-off drew conflicting conclusions regarding differences in outcome between acute and delayed surgical treatment.29 30 A third systematic review using 8 weeks as cut-off noted significant differences in terms of satisfaction, hamstring strength, single-legged hop test and Lower Extremity Functional Scale score.27 With regard to surgical complications, no significant differences were found between acute and delayed surgery.27
There are limitations to this study. First and foremost, questions often made use of a ‘black-and-white’ scenario with fixed answers. Although such a hypothetical scenario may not reflect the diversity of clinical practice and leaves no room for nuances, this approach was taken to allow for descriptive statistics. It forces all participants to provide an answer for an identical situation rather than an answer that may only apply within a certain context.
Second, the response rate was low (12%). However, given the large number of participants stating that they have clinical experience with this injury (92%), it is likely that the invitation to participate functioned as a first selection process filtering out those that are unfamiliar with this injury.
Implications and future directions
Our survey has provided insight into current practice and preferences in regards to treatment of proximal hamstring tendon avulsions. It has exposed gaps in the literature in regards to support for some of the aspects of current practice and can therefore serve to direct future research efforts. These studies should ideally employ randomised controlled study designs. Given our findings, such studies should randomise treatment and stratify based on the involved tendons, retraction and timing of surgery. Alternatively, as some argue that the principle of equipoise is not met based on current evidence, an international multicentre collaboration in the form of a prospective registry should be developed.
Additionally, this survey might be used to guide the clinical decision-making process as evidence-based indications for surgery have not yet been established.
Clinicians dealing with proximal hamstring tendon avulsions choose a surgical or non-operative treatment depending on the individual patient. Decision modifiers include diminished function, neurological symptoms, involved tendons, extent of tendon retraction on MRI and patient preference. The typical surgical patient has a retracted (>2 cm) two-tendon avulsion, is unable to engage in any sports or activities of daily life, reports sciatic symptoms and has a preference for surgical treatment. Surgical reattachment is preferably carried out within 2 weeks after injury. Recovery duration is estimated to be longer for surgical compared with non-operative treatment. Reinjury risk is estimated to be higher in patients that are treated non-operatively.
The authors would like to thank Mr Irwin Vilaga, Mr Ronald Barangan and the Aspetar Quality Management Department for their help in developing the online survey. ISAKOS and ECOSEP are greatly appreciated for their support in distributing the survey. In particular, the authors want to extend their gratitude to Dr Nikos Malliaropoulos, Dr Nicholas Mohtadi and the ISAKOS Hip, Groin and Thigh Committee.
Contributors All authors were involved in study development and writing/editing of the manuscript.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Ethics approval Exempted from ethics review by the local IRB (Anti-Doping Lab Qatar, ADLQ, Doha, Qatar).
Provenance and peer review Not commissioned; externally peer reviewed.
Patient consent for publication Not required.
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.