Abstract
Background: The torn hamstring is a common athletic injury. The purpose of the present study was to review the clinical presentation of this injury, the diagnostic imaging findings, the surgical technique of reattachment, and the likely clinical outcome of surgery for the treatment of avulsion of the proximal hamstring origin.
Methods: Seventy-two consecutive reconstructions in seventy-one patients with avulsion of the proximal hamstring origin were performed at a single center. The mean age at the time of the operation was 40.2 years. The mean duration of follow-up was twenty-four months, and all patients with a minimum duration of follow-up of six months were included. There were no exclusions. Patients were independently reviewed, and the mean postoperative isotonic hamstring strength was compared with that on the uninjured side.
Results: Waterskiing was the most frequent cause of injury (twenty-one cases). The mean time between the injury and the operation was twelve months. The most common pathological finding was a complete avulsion of the proximal hamstring origin (sixty-three cases; 87.5%), with a mean retraction of 7 cm (range, 0 to 20 cm). The mean postoperative isotonic hamstring strength measured 84% (range, 43% to 122%) and the mean postoperative hamstring endurance measured 89% (range, 26% to 161%) when compared with the values on the contralateral side.
Conclusions: It is important to distinguish proximal hamstring origin avulsions (for which we recommend early surgical repair) from the majority of hamstring muscle injuries (which respond well to nonoperative treatment). The present study suggests that, in cases of complete avulsion with hamstring retraction, a delay in surgical repair renders the repair more technically challenging, may increase the likelihood of sciatic nerve involvement, increases the need for postoperative bracing, and reduces postoperative outcome in terms of hamstring strength and endurance. Once the nature of the injury has been established, the surgical treatment of hamstring origin avulsions has predictable and satisfactory results.
Level of Evidence: Therapeutic Level IV. See Instructions to Authors for a complete description of levels of evidence.
The torn hamstring is a common injury in athletes1-6. The majority of these injuries are strains of the muscle or myotendinous junction, which may be treated nonoperatively with a satisfactory outcome after rehabilitation7-9. Avulsion of the hamstring origin from the ischial tuberosity occurs much less commonly, and the available literature appears to support surgical reattachment of the hamstring avulsion in order to avoid long-term functional disability and to allow a return to sporting activity10-21. Despite recent attention in the literature, there remains diagnostic difficulty in the differentiation of proximal hamstring avulsions from muscle strains and, as a consequence, patients continue to present late for definitive treatment20-23.
We previously reported the results for nine patients who had undergone surgical reattachment of complex proximal hamstring avulsions14. The purpose of the present study was to review the clinical presentation of this injury, the findings of diagnostic imaging, the surgical technique of reattachment, and the likely outcome of surgery for the treatment of proximal hamstring avulsion by evaluating a series of seventy-two consecutive cases with a mean duration of follow-up of two years. We also describe a system for the classification of this heterogeneous group of injuries to aid in preoperative planning, to help to predict clinical outcome, and to allow outcome comparison.
A series of seventy-two successive proximal hamstring reattachments (including thirty-one on the right side and forty-one on the left) were performed in seventy-one patients (including fifty male patients and twenty-one female patients) between 1994 and 2006 at the North Sydney Orthopaedic and Sports Medicine Centre. This group included the nine procedures (12.5%) described our previous series14. The subsequent sixty-three procedures (87.5%) were performed by the senior author (D.G.W.). One patient who had undergone proximal hamstring reattachment twenty-five months previously sustained an avulsion of the contralateral proximal hamstring attachment (following a fall) with subsequent reconstruction. The mean age at the time of the operation was 40.2 years (range, 12.9 to 66.2 years), and the mean time between the injury and the operation was twelve months (range, eight days to 104 months). The mean duration of follow-up was twenty-four months (range, six to 156 months), with twenty-six patients having more than two years of follow-up. All patients who underwent surgical reconstruction with a minimum duration of follow-up of six months were included. There were no exclusions. The Ethics Committee of the Australian Institute of Musculoskeletal Research accepted the study as a clinical review.
Radiographic Diagnosis
Magnetic resonance imaging is the most accurate imaging modality for the diagnosis of hamstring origin injuries24,25. In order to obtain magnetic resonance images of adequate spatial resolution, a surface coil is required. A combination proton-density and fat-suppressed proton-density fast-spin-echo sequence was employed in the axial, coronal, and oblique sagittal planes; the latter plane was angled to be orthogonal to the ischial tuberosity facet for the hamstring origin. The normal hamstring origin has a homogeneously low signal on all pulse sequences, reflecting the closely packed collagen fibers of normal tendon. A similar low signal is present at the interface between the tendon and the low-signal cortical bone of the ischial tuberosity. Magnetic resonance imaging criteria for the diagnosis of hamstring avulsion in the acute setting were discontinuity of the bone-tendon unit, manifested as focal T2-signal hyperintensity (a fluid signal) intervening between the tendon edge and the ischial tuberosity. The degree of retraction of the tendon edge was calculated. Ancillary findings suggestive of a recent hamstring avulsion included a fluid collection surrounding the tendon edge, tracking distally, within the investing fascia of the hamstring muscle group and feathery intramuscular edema indicating a concomitant low-grade muscle or muscle-tendon junction strain.
Magnetic resonance imaging findings suggestive of a chronic avulsion included the absence of a fluid collection at the margins of the tendon edge, fatty infiltration, and reduction in the volume of the involved hamstring muscles and scarring of the tendon to adjacent structures; the latter finding was indicated by the absence of an intervening fat plane. The sciatic nerve was carefully assessed for evidence of tethering to the tendon edge.
Classification
A system for the classification of proximal hamstring avulsions was devised by the senior author (D.G.W.). With this system, five types of avulsions can be identified on the basis of the anatomical location of the injury, the degree of avulsion (incomplete or complete), the degree of muscle retraction (if avulsion is complete), and the presence or absence of sciatic nerve tethering. As this system has not been published previously, its reliability and validity are unknown. In our hands, the classification system has facilitated preoperative planning for the anticipated difficulty of surgery in terms of the amount of tendon retraction, sciatic nerve tethering, and the presence of a hematoma.
Type-1 injuries are osseous avulsions, which typically represent apophyseal injuries in skeletally immature patients (Fig. 1). Type-2 injuries are avulsions at the musculotendinous junction (Figs. 2-A, 2-B, and 2-C). Type-3 injuries are incomplete tendon avulsions from bone (Figs. 3-A, 3-B, and 4). These injuries may present late with tears situated at the deep margin, often with an ineffective scar response. A small focus of fluid signal intensity may be evident on magnetic resonance imaging with an overlying intact portion of the tendon edges. Type-4 injuries are complete tendon avulsions with no or minimal retraction of the tendon ends (Fig. 5). These injuries are typically acute and are readily apparent on magnetic resonance imaging, with the free end of the tendon identified in a fluid collection surrounding the tendon edge, tracking distally, within the investing fascia of the hamstring muscle group. Type-5 injuries are complete tendon avulsions from bone with retraction of the tendon ends. This group may be subdivided into Type-5a injuries, which are not associated with sciatic nerve involvement (Fig. 6), and Type-5b injuries, which are associated with sciatic nerve tethering (Fig. 7). Patients with Type-5 injuries typically present late, suggesting that in cases of complete tears, the tendon may retract with time. Magnetic resonance imaging is particularly useful for identifying the degree of the retraction.
Surgical Approach and Technique
Surgery is performed with use of general anesthesia and with the patient in the prone position with the knee draped to allow unrestricted flexion. A 5 to 7-cm longitudinal incision is made from the palpable ischial tuberosity inferiorly over the defect. The posterior cutaneous nerve of the thigh is identified and protected. A retractor is placed beneath the inferior free edge of the gluteus maximus, which is then retracted superiorly. The deep fascia is incised in the line of the incision. The sciatic nerve, which passes in close proximity deep and lateral to the proximal hamstring origin, is, when necessary, identified and protected. The incision is extended as necessary to allow for the identification and reduction of the retracted hamstring origin.
Once the tendon ends have been mobilized, the inferior and lateral aspects of the ischial tuberosity are exposed and scar tissue is removed. Three Super QuickAnchors Plus (DePuy Mitek, Raynham, Massachusetts) with additional number-5 Ethibond sutures (Ethicon, Somerville, New Jersey) threaded through the anchors are inserted into the exposed tuberosity. The sutures are passed through the tendon ends with use of a modified Mason-Allen technique and are tied when the avulsed tendons are reapproximated snugly to the ischial tuberosity.
When knee flexion is required to remove tension from the surgical repair, a hinged knee brace may be applied with the knee held in as much as 90° of flexion. Immobilization is maintained for a maximum of eight weeks. When a brace is required, the degree to which knee extension is limited may be sequentially reduced as permitted by the hamstring tension. Mobilization with partial weight-bearing on crutches is allowed for the first six weeks. Thereafter, full weight-bearing is permitted, without support. A physiotherapy program consisting of stretching and closed-chain strengthening exercises is started at three months, with a graduated return to sports activity by six months.
Routine postoperative reviews were conducted at two weeks, six weeks, three months, six months, and one year. All patients were subsequently recalled for the present study. At six months, data were collected on the ability to return to sports activity and the level of sports activity reached. These findings were compared with the preinjury levels.
Quantitative assessments of isotonic hamstring and quadriceps muscle strength and endurance were made with use of a previously published protocol14. Quadriceps muscle strength was measured by finding the maximum weight at which six repetitions of knee extension could be performed between 90° and 0° flexion, at a rate of one cycle per second, with fatigue occurring on the seventh repetition. Hamstring muscle strength was measured by finding the maximum weight at which six repetitions of knee flexion could be performed between 0° and 90° flexion, at a rate of one cycle per second, with fatigue occurring on the seventh repetition. Quadriceps and hamstring muscle endurance was quantified by measuring the number of repetitions to fatigue at 50% of this resistance, at a rate of one cycle per second. Measurements were made with use of isotonic Keiser quadriceps curl and hamstring curl equipment (Keiser Sports Health Equipment, Fresno, California). The percentage muscle strength and the percentage endurance were calculated by dividing the results for the involved limb by the results for the uninvolved limb. Standardized values for hamstring strength and endurance were calculated by dividing the percentage hamstring strength and endurance values by the percentage quadriceps strength and endurance values to take into account the effect of both leg dominance and overuse/disuse.
Statistical analyses were performed with the use of a paired t test for the comparison of hamstring strength and endurance values and standardized hamstring strength and endurance values. Unpaired t tests were used to compare the time to surgery and the amount of retraction, and the chi-square test was used to compare brace usage between patients with Type-5a and 5b injuries (complete tendon avulsion and retraction without or with sciatic nerve involvement). The Fisher exact test was used to compare brace usage between the patients with acute and chronic Type-5 injuries. Unpaired t tests were used to compare muscle strength and endurance between all patients with acute and chronic injuries and between all patients treated with and without bracing.
A six-group analysis of variance was used to compare hamstring and quadriceps strength and endurance among all groups.
The main precipitating activities are listed in Table I. The most frequent cause of injury was waterskiing (twenty-one cases; 29%).
After a minimum duration of follow-up of six months (mean, two years), seventy-one patients (seventy-two injuries) were available for review (Table II). Of the sixty-three cases of complete avulsion (Type 4 or 5), sixteen were Type 4 (no retraction) and forty-seven were Type 5 (mean retraction, 7 cm; range, 2 to 20 cm). Of the forty-seven Type-5 injuries, thirty were classified as Type 5a (mean retraction, 6.7 cm; range, 2 to 20 cm) and seventeen were classified as Type 5b (mean retraction, 10 cm; range, 4 to 14 cm). The difference in the mean retraction between Types 5a and 5b was found to be significant (p < 0.05) (see Appendix). All seventeen Type-5b injuries and one Type-4 injury required neurolysis of the sciatic nerve.
In thirty cases, postoperative bracing was required to protect the repair (Table II). In one case, bracing was used for a patient who had undergone reconstruction of a chronic Type-4 rupture 10.6 months after the injury. In another case, bracing was used for an adolescent patient who had a Type-1 osseous avulsion. In the remaining twenty-eight cases, bracing was used for patients with either Type-5a or 5b injuries (see Appendix). The difference in brace usage between Type-5a and 5b injuries was found to be significant (p < 0.001).
The difference in the mean time to surgery between Type-5a and 5b injuries was found to be significant (p < 0.01) (see Appendix). There were thirty-two acute cases (with an interval of less than three months between the injury and the operation) and forty chronic cases (with an interval of three months or more between the injury and the operation). All seven Type-3 incomplete avulsions and sixteen of seventeen Type-5b avulsions (complete avulsions with retraction of the tendon and sciatic nerve involvement) had a chronic presentation. Among the thirty Type-5a injuries, there were significant differences in brace requirements, the degree of retraction, and hamstring strength and endurance between the acute and chronic subtypes (see Appendix).
Sixty-two patients (sixty-three cases) were available for the assessment of hamstring strength and endurance. One of those patients had undergone bilateral reconstruction and was excluded, leaving strength results for sixty-one patients (sixty-one cases) (see Appendix). Overall, the mean hamstring strength of the involved limb was 84% (range, 43% to 122%) of that of the contralateral limb and the mean hamstring endurance was 89% (range, 26% to 161%) of that of the contralateral limb. The mean quadriceps strength of the involved limb was 98% (range, 62% to 120%) of that of the contralateral limb, and the mean quadriceps endurance was 102% (range, 64% to 133%) of that of the contralateral limb. The mean standardized hamstring strength was 86% (range, 47% to 127%), and the mean standardized endurance was 88% (range, 28% to 148%). There was only a borderline significant difference (p = 0.05) betweenstandardized and nonstandardized hamstring strength, and there was no significant difference (p = 0.68) between standardized and nonstandardized hamstring endurance.
Significant differences among the six subtypes (see Appendix) were demonstrated for hamstring strength (p < 0.001), hamstring endurance (p = 0.04), and standardized hamstring strength (p < 0.001). In particular, there was a highly significant decrease in the hamstring strength (p < 0.001), hamstring endurance (p = 0.01), and standardized hamstring strength (p < 0.001) when Type-5b injuries were compared with Type-5a injuries. In addition, a comparison of complete tears with retraction (Type-5a and 5b injuries) according to the time to surgery demonstrated increased mean retraction (p < 0.001), increased requirement for bracing (p < 0.001), and decreased mean hamstring strength and endurance (p < 0.001 and p < 0.05) in patients in whom repair was delayed (see Appendix). Of the nine patients who had failed to return for strength testing, four patients had moved to another state and two had moved to another country. All six patients were contacted, and all reported no adverse outcomes. Two other patients who had had surgery ten years previously were unwilling to return for strength testing but reported full return to the preinjury level of sports. One patient in whom hamstring reattachment was not possible declined to return but reported a lack of satisfaction, with ongoing weakness and an inability to sprint. This patient had presented at two years after the injury with 8 cm of muscle retraction from the ischial tuberosity; a direct repair could not be achieved, so the proximal hamstring avulsion was left unattached. A direct satisfactory repair was obtained in all other cases.
At the time of the six-month postoperative check, fifty-seven patients (80%) had returned to their preinjury level of sports. The cohort included two elite athletes, three professional athletes, and two professional dancers, all of whom were performing at the preinjury level by six months after surgery. The patient with bilateral involvement and the patient in whom a direct repair was not achieved were restricted to walking exercises. Two patients gave up their preinjury sports activities (waterskiing and skiing).
One patient had development of a superficial wound infection, which resolved with oral antibiotics. One patient with a Type-5b injury had a transient delayed sciatic nerve palsy, which appeared on the first postoperative day and fully resolved within twelve hours.
One patient who had undergone surgery (without bracing) 2.6 months after the initial injury (a complete avulsion with 5 cm of retraction) slipped and fell four weeks after surgery, recreating the original injury mechanism. Acute reoperation successfully reattached the hamstring avulsion. At six months after the reoperation, the hamstring strength was 108% and the endurance was 161% of that on the contralateral side.
The hamstring muscles are among the most commonly injured muscles in athletes2,5,6. As biarticular muscles, which are rarely stretched during normal activity, they are particularly vulnerable to eccentric loading. The majority of these injuries are muscle strains occurring at the myotendinous junction7-9. However, as many as 12% of hamstring injuries may involve a tear or avulsion of the proximal hamstring origin, and 9% may be complete avulsions (Type 4 or 5)24.
The mechanism of injury is typically an eccentric contraction of the hamstring muscles as a result of a sudden forced hyperflexion of the hip with the knee fully extended. Patients typically report an immediate disabling posterior thigh pain (from the buttock to the middle part of the thigh) with weakness or giving way of the limb. This may be accompanied by a pop or snap. The patient is unable to complete the precipitating activity. Extensive thigh swelling and dramatic bruising can develop over the next forty-eight hours; these findings gradually resolve over two to four weeks. Examination with the patient in the prone position allows objective weakness of knee flexion to be readily identified. In the case of complete avulsion, there may be a palpable gap in the proximal hamstring attachment and a prominence of the retracted muscle bellies may be apparent (Fig. 8). Sitting on the affected side can be very difficult during the acute period, and the patient may sit perched on the contralateral buttock.
In chronic cases, there may be ongoing weakness with functional impairment. Continued pain may be a dominant feature, particularly when sitting. The pain may be localized or referred in a sciatic distribution. An inability to sprint and to participate in sports due to weakness of knee flexion is typical. In all of our chronic cases, the patient had seen at least one doctor who had either misdiagnosed the condition or underestimated the clinical importance of the injury. All patients in the present series who had chronic injuries presented to this unit after prolonged and unsuccessful nonoperative rehabilitation.
In our series, proximal hamstring avulsions occurred most commonly in association with waterskiing. Frequently, these injuries followed a sudden and unanticipated pull from the boat with a high drag on the skis. In the case of a novice skier, this situation typically arises during a water-start if the knees are not kept flexed. In the case of a more experienced skier, a similar mechanism may occur during a beach or dock-start as well as during a fall and sudden deceleration12. An awareness of this injury among those who participate in waterskiing may reduce the risk of injury or allow early identification and appropriate treatment. Hamstring injuries, sporting or otherwise, typically follow a slip and fall leading to a splits-like movement11.
Our classification system allows specific patterns to be identified among what may be a heterogeneous group of injuries. This allows counseling for fully informed consent in terms of the risk of complications, the likelihood of the need for postoperative bracing to protect the repair, and the likely clinical outcome. As a research tool, we hope that it will enable a more reliable comparison of studies and their outcomes. However, a limitation of the present study is that the reliability and validity of this classification system remain unknown.
Substantially displaced osseous avulsions (Type 1) of more than 1 to 2 cm should be reduced and fixed21,26,27. The surgical repair of avulsions at the musculotendinous junction (Type 2), as at other sites, is technically demanding.
The nomenclature relating to partial tears of the proximal hamstring tendon group may be unclear. As a consequence, the term incomplete tendon avulsion is preferred to describe these (Type-3) injuries. In the present study, there were seven Type-3 injuries. All of these injuries were chronic (mean time to surgery, nine months), and all of the patients with such injuries had failed to respond to nonoperative care with appropriate rehabilitation and had failed to return to the preinjury level of sports activity. With tears situated at the deep margin of the hamstring origin, there may be difficulty localizing the defect during surgical exploration. Once the defect has been localized, reattachment is relatively straightforward even in the chronic setting as there is no retraction of the tendon edge and a tension-free repair can be achieved, allowing the patient to rehabilitate without restriction of knee extension. In our series, there was a high rate of patient satisfaction and a 100% rate of return to the preinjury level of sports activity at six months. Objective postoperative mean hamstring strength and endurance scores in this group were excellent, and these results are consistent with those in the literature20.
For complete tendon avulsions (Types 4 and 5), the appearance and ease of identification of the avulsed tendon end depend on the chronicity of the rupture. Early surgery reveals free, mobile tendon ends surrounded by a hematoma. It has been our experience that the optimum time for surgery is approximately four to six weeks after the injury, when the hematoma forms a fluid-filled capsule or sac beneath the deep fascia, enclosing the tendon with an obvious tract leading down to its origin on the lateral ischial tuberosity. At this stage, there is usually no tethering and the tendon can be repaired to bone without tension, allowing rehabilitation without restriction. From six weeks onward, the presence of this fluid-filled sac becomes more unpredictable. If there is retraction of the tendon edge (Type 5), there may be difficulty in reattaching the avulsed tendon, necessitating postoperative bracing to protect the repair.
By three months, there may be no fluid collection, with marked tethering of the tendon ends to adjacent tissue rendering the tissue planes indistinct and making identification of the tendon ends more difficult. There may be extensive scarring around the sciatic nerve, making its identification difficult and surgical neurolysis necessary14,15,28. The use of a nerve stimulator is helpful in such circumstances.
The majority of our patients presented late (mean time to surgery, twelve months), and all forty patients with chronic avulsions presented after a prolonged period of nonoperative management without a satisfactory level of recovery. Satisfactory results are possible following repair of chronic hamstring avulsion injuries, although the surgical procedure is technically challenging and the postoperative rehabilitation may be more prolonged if bracing is required to protect the repair11,16,19,20.
Overall, we found no significant difference in hamstring strength and endurance between the acute and chronic cases. However, in cases of complete avulsion with retraction (Type-5 injuries), a chronic presentation was associated with a significantly increased degree of retraction, an increased requirement for bracing, and decreased hamstring strength and endurance. In cases of complete avulsion with retraction and no sciatic nerve involvement (Type-5a injuries), a chronic presentation was associated with increased mean retraction of the muscle, a highly significant increase in the requirement for postoperative bracing, a highly significant decrease in hamstring strength, and a significant decrease in hamstring endurance. Furthermore, in cases requiring postoperative bracing following repair, the mean hamstring strength and endurance were significantly lower than those in cases not requiring bracing. We believe that the poorer outcome in patients requiring postoperative bracing reflects their more chronic presentation and increased retraction rather than being an effect of the bracing itself.
In all but one of the chronic cases, it was possible to approximate the tendon ends to the ischial tuberosity by means of adequate dissection between the tendon and its fascial envelope. Knee flexion facilitates apposition of the tendon ends to the tuberosity, which can be difficult in the case of a chronic injury. Postoperative knee stiffness does not appear to pose a long-term problem with rehabilitation. Additional steps to assist tendon-to-bone approximation, such as fractional lengthening of the hamstring muscles16 and tendon augmentation with fascia lata11,22 or synthetic materials29, were not necessary in any case as adequate mobilization of the tendon ends was achieved.
In previous studies, the recommended duration of postoperative bracing has ranged from two to twelve weeks13,18,21,29. We now recommend splinting the knee only if the repair is under tension with the knee extended. In this situation, a hinged knee brace is applied at as much as 90° for a period of as long as six weeks. If bracing is necessary, the amount of flexion may be decreased in two weekly intervals as the hamstrings stretch. Knee bracing avoids the debilitation associated with bracing of the hip in extension, as recommended by some groups30,31.
The postoperative improvement was reflected in a return to the preinjury level of sports activity in fifty-seven cases (79%) by six months. All seven professional athletes and dancers were able to return to their previous level of performance. This finding relates favorably to those of previously published series11,21, including a series of incomplete avulsions treated by Lempainen et al.20, who reported an 87% rate of return to sports activity by twelve months. Nonoperative management has previously been reported to provide acceptable levels of function23, although Sallay et al.12 reported a return to regular sports activity in the cases of only three of twelve patients, all of whom were deemed to have partial tears.
The applied surgical anatomy of the proximal hamstring attachment and its relationships was recently reported in a cadaver study by Miller et al.23. Those authors confirmed the lateral insertion of the semimembranosus in relation to the more medial conjoint tendon of semitendinosus and biceps femoris as described previously on magnetic resonance imaging25. We performed cadaver dissections to become familiar with the approach and the safe placement of retractors. We have not encountered any complications relating to the inferior gluteal nerve or artery when retracting the inferior border of the gluteus maximus to allow visualization of the ischial tuberosity23. It is crucial to identify the sciatic nerve, which lies in close proximity to the hamstring origin, and to protect it during tendon mobilization. Pain in the sciatic nerve distribution and reduced muscle power have been well documented in association with proximal hamstring injuries and appear to be related to scarring and traction on the sciatic nerve11,12,14-16,28. In the present study, eighteen patients underwent neurolysis because of clinical symptoms or because of intraoperative evidence of sciatic nerve tethering that made reduction of the retracted proximal hamstring origin impossible. On review of the cases with complete avulsion with hamstring retraction, those with sciatic nerve involvement (Type-5b injuries) had a significantly longer time to surgery, a greater degree of hamstring retraction, an increased requirement for postoperative bracing, reduced mean hamstring strength, reduced mean hamstring endurance, and reduced mean standardized hamstring strength in comparison with those without sciatic nerve involvement (Type-5a injuries).
Some authors have advocated a transverse gluteal crease incision with the patient lying in a prone position19,23,30,31. This approach may provide an optimal cosmetic wound and may allow for the repair of acute injuries in the absence of retraction. However, in common with other authors14-16,18,22, we strongly recommend a longitudinal incision. This allows adequate identification and exposure of the hamstring tendons, which may be retracted into the thigh, even in acute cases. It also may be necessary to identify the sciatic nerve in an area of normal anatomy, especially in chronic cases, in which the nerve may be surrounded by scar tissue, warranting neurolysis.
The present study represents a consecutive series of patients who had been referred to a specialist sports injury unit. As such, there is a potential for referral bias and we are unable to comment on patients who had success with nonoperative management. We acknowledge that the study group may not be representative of the entire population of patients with proximal hamstring injuries. The lack of routine postoperative radiographic assessment to determine the durability of the repair presents another limitation of the present study. The absence of preoperative strength and endurance scores and the lack of patient-determined outcome scores in the present study prevent quantification of the clinical improvement following surgery. In the present series, the mean hamstring strength and endurance on the involved side were 84% and 89% in comparison with the values on the uninjured side. These findings compare favorably with the literature12,16,18,19,21 and contrast with the results of nonoperative treatment12. Despite these limitations, we attempted to quantify the outcome of surgical repair of hamstring avulsions rather than relying solely on arbitrary outcome measures. The assessment protocol for hamstring strength and endurance, as previously published14, allowed an adjustment to be made for such factors as leg dominance and the effect of injury on the performance of the affected limb. However, no significant difference was identified between the standardized and nonstandardized hamstring strength and endurance measurements.
To our knowledge, the present study represents the largest published series of its kind. We have attempted to identify the clinical and radiographic features of proximal hamstring avulsion and the importance of distinguishing this injury from the majority of hamstring injuries, which respond well to nonoperative treatment. Once the nature of the injury has been established, the operative treatment of hamstring origin avulsion can result in a satisfactory, predictable outcome. We have described a classification system, which can be used to guide the clinician in decision-making with regard to treatment and rehabilitation. The present study suggests that, in cases of complete avulsion with hamstring retraction (Type-5 injuries), a delay in surgical treatment renders the procedure more technically challenging, may increase the likelihood of sciatic nerve involvement, increases the need for postoperative bracing, and reduces postoperative hamstring strength and endurance.
Tables comparing the injury subtypes and presenting the detailed muscle strength and endurance results are available with the electronic versions of this article, on our web site at (go to the article citation and click on "Supplementary Material") and on our quarterly CD/DVD (call our subscription department, at 781-449-9780, to order the CD or DVD). 
Note: The authors thank Dr. Mervyn Cross, Consultant Surgeon, North Sydney Orthopaedic and Sports Medicine Centre, Dr. J. Poloniecki, Department of Epidemiology and Medical Statistics, St. George's Hospital Medical School, and Mrs. C. McKinnon for their ongoing help and support.
Garrett WE Jr. Muscle strain injuries: clinical and basic aspects. Med Sci Sports Exerc.1990;22:436-43.22436
1990
[PubMed]
Bennell KL, Crossley K. Musculoskeletal injuries in track and field: incidence, distribution and risk factors. Aust J Sci Med Sport.1996;28:69-75.2869
1996
Kujala UM, Orava S, Jarvinen M. Hamstring injuries. Current trends in treatment and prevention. Sports Med.1997;23:397-404.23397
1997
[CrossRef]
Clanton TO, Coupe KJ. Hamstring strains in athletes: diagnosis and treatment. J Am Acad Orthop Surg.1998;6:237-48.6237
1998
Orchard J, Seward H. Epidemiology of injuries in the Australian Football League, seasons 1997-2000. Br J Sports Med.2002;36:39-44.3639
2002
[CrossRef]
Woods C, Hawkins RD, Maltby S, Hulse M, Thomas A, Hodson A; Football Association Medical Research Programme. The Football Association Medical Research Programme: an audit of injuries in professional football—analysis of hamstring injuries. Br J Sports Med.2004;38:36-41.3836
2004
[CrossRef]
Agre JC. Hamstring injuries. Proposed aetiological factors, prevention, and treatment. Sports Med.1985;2:21-33.221
1985
[CrossRef]
De Smet AA, Best TM. MR imaging of the distribution and location of acute hamstring injuries in athletes. AJR Am J Roentgenol.2000;174:393-9.174393
2000
Järvinen TA, Järvinen TL, Kääriäinen M, Kalimo H, Järvinen M. Muscle injuries: biology and treatment. Am J Sports Med.2005;33:745-64.33745
2005
[CrossRef]
Ishikawa K, Kai K, Mizuta H. Avulsion of the hamstring muscles from the ischial tuberosity. A report of two cases. Clin Orthop Relat Res.1988;232:153-5.232153
1988
Orava S, Kujala UM. Rupture of the ischial origin of the hamstring muscles. Am J Sports Med.1995;23:702-5.23702
1995
[CrossRef]
Sallay PI, Friedman RL, Coogan PG, Garrett WE. Hamstring muscle injuries among water skiers. Functional outcome and prevention. Am J Sports Med.1996;24:130-6.24130
1996
[CrossRef]
Kurosawa H, Nakasita K, Nakasita H, Sasaki S, Takeda S. Complete avulsion of the hamstring tendons from the ischial tuberosity. A report of two cases sustained in judo. Br J Sports Med.1996;30:72-4.3072
1996
[CrossRef]
Cross MJ, Vandersluis R, Wood D, Banff M. Surgical repair of chronic complete hamstring tendon rupture in the adult patient. Am J Sports Med.1998;26:785-8.26785
1998
Street CC, Burks RT. Chronic complete hamstring avulsion causing foot drop. A case report. Am J Sports Med.2000;28:574-6.28574
2000
Klingele KE, Sallay PI. Surgical repair of complete proximal hamstring tendon rupture. Am J Sports Med.2002;30:742-7.30742
2002
Johnson AE, Granville RR, DeBerardino TM. Avulsion of the common hamstring tendon origin in an active duty airman. Mil Med.2003;168:40-2.16840
2003
Chakravarthy J, Ramisetty N, Pimpalnerkar A, Mohtadi N. Surgical repair of complete proximal hamstring tendon ruptures in water skiers and bull riders: a report of four cases and review of the literature. Br J Sports Med.2005;39:569-72.39569
2005
[CrossRef]
Brucker PU, Imhoff AB. Functional assessment after acute and chronic complete ruptures of the proximal hamstring tendons. Knee Surg Sports Traumatol Arthrosc.2005;13:411-8.13411
2005
[CrossRef]
Lempainen L, Sarimo J, Heikkilä J, Mattila K, Orava S. Surgical treatment of partial tears of the proximal origin of the hamstring muscles. Br J Sports Med.2006;40:688-91.40688
2006
[CrossRef]
Gidwani S, Bircher M. Avulsion injuries of the hamstring origin—a series of 12 patients and management algorithm. Ann R Coll Surg Engl.2007;89:394-9.89394
2007
[CrossRef]
Lempainen L, Sarimo J, Orava S. Recurrent and chronic complete ruptures of the proximal origin of the hamstring muscles repaired with fascia lata autograft augmentation. Arthroscopy. 2007;23:441.e1-5.23441.e1
2007
[CrossRef]
Miller SL, Gill J, Webb GR. The proximal origin of the hamstrings and surrounding anatomy encountered during repair. A cadaveric study. J Bone Joint Surg Am.2007;89:44-8. Erratum in: J Bone Joint Surg Am. 2007;89:637.8944
2007
Koulouris G, Connell D. Evaluation of the hamstring muscle complex following acute injury. Skeletal Radiol.2003;32:582-9.32582
2003
[CrossRef]
Bencardino JT, Mellado JM. Hamstring injuries of the hip. Magn Reson Imaging Clin N Am.2005;13:677-90, vi.13677
2005
[CrossRef]
Wootton JR, Cross MJ, Holt KW. Avulsion of the ischial apophysis. The case for open reduction and internal fixation. J Bone Joint Surg Br.1990;72:625-7.72625
1990
Kujala UM, Orava S. Ischial apophysis injuries in athletes. Sports Med.1993;16:290-4.16290
1993
[CrossRef]
Takami H, Takahashi S, Ando M. Late sciatic nerve palsy following avulsion of the biceps femoris muscle from the ischial tuberosity. Arch Orthop Trauma Surg.2000;120:352-4.120352
2000
[CrossRef]
Thomsen NO, Jensen TT. Late repair of rupture of the hamstring tendons from the ischial tuberosity—a case report. Acta Orthop Scand.1999;70:89-91.7089
1999
[CrossRef]
Colosimo A, Wyatt H, Frank K, Mangine R. Hamstring avulsion injuries. Oper Tech Sports Med.2005;13:80-8.1380
2005
[CrossRef]
Cohen S, Bradley J. Acute proximal hamstring rupture. J Am Acad Orthop Surg.2007;15:350-5.15350
2007