Article Text

Download PDFPDF

Clinical predictors of time to return to competition and of recurrence following hamstring strain in elite Australian footballers
  1. Price Warren1,
  2. Belinda J Gabbe2,
  3. Michal Schneider-Kolsky3,
  4. Kim L Bennell1
  1. 1Centre for Health, Exercise and Sports Medicine, School of Physiotherapy, University of Melbourne, Melbourne, Australia
  2. 2Department of Epidemiology and Preventive Medicine, Monash University, Melbourne, Australia
  3. 3Department of Medical Imaging and Radiation Sciences, Monash University, Melbourne, Australia
  1. Correspondence to Professor Kim L Bennell, Centre for Health, Exercise and Sports Medicine, School of Physiotherapy, Faculty of Medicine, Dentistry and Health Sciences, University of Melbourne, Victoria 3010, Australia; k.bennell{at}unimelb.edu.au

Abstract

Objective To investigate early clinical predictors of time to return to competition and of recurrence following hamstring strain.

Design Prospective observational study.

Setting Elite level of Australian football competition.

Participant 59 players who suffered a hamstring strain in 2002 season.

Predictors Clinical assessment by a physiotherapist and questionnaire.

Main outcome measures Time taken to return to play and recurrence of hamstring injury within 3 weeks.

Results Players taking more than 1 day to walk pain-free were significantly more likely (p=0.018) to take longer than 3 weeks to return to competition (adjusted odds ratio 4.0; 95% CI 1.3 to 12.6). Nine players (15.2%) experienced an injury recurrence, all involving the biceps femoris. Recurrence was more likely in players who reported a hamstring injury in the past 12 months (adjusted odds ratio 19.6; 95% CI 1.5 to 261.0; p=0.025).

Conclusion Time to walk pain-free and previous hamstring injury are predictors of time to return to competition and recurrence, respectively, and should be included in a clinical assessment to aid in prognosis.

Statistics from Altmetric.com

Request Permissions

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.

Introduction

Hamstring muscle strains are common in sports which place high demands on running speed and power such as soccer1 2 and athletics.3 They are also common in Australian football, a game that involves frequent periods of explosive and high-intensity sprinting, extended periods of sustained running, heavy physical clashes and a combination of running and kicking.4

In the Australian Football League (AFL), the elite level of competition, hamstring strains are the most common injury, resulting in the greatest number of lost playing days.4 The recurrence rates for hamstring strain are also high. Greater than 30% of hamstring strains in the AFL recur within the season, despite concentrated rehabilitation and prevention efforts.4 Returning players are particularly at risk of reinjuring their hamstring during the first 3 weeks back from injury.4

In the initial stages of treatment of a player with a hamstring strain, clinicians are often questioned about the likely prognosis. Theoretically, the results of an early clinical examination together with other relevant information from the subjective assessment should enable the clinician to predict the expected time to return to competition and whether the injury is likely to recur. Limited evidence suggests that greater reported pain at the time of injury,5 strains in the biceps femoris component6 and strains sustained via slow sustained stretching rather than explosive activity7 are associated with the time taken to return to play. However, few studies have evaluated the accuracy of early clinical assessment findings as predictors of time to return to competition and recurrence. Therefore, the aim of this study was to assess whether clinical assessment findings could predict the time taken to return to competition, and a successful return to play as evidenced by a lack of recurrence, following hamstring injury in elite Australian football players.

Methods

Participants

Players from the 10 Victorian-based AFL clubs who sustained a hamstring strain, as clinically diagnosed by club medical personnel during the 2002 season, and missed at least one match were invited to participate (n=59). The criteria for diagnosis of a hamstring muscle strain were: acute onset of posterior thigh pain, palpable hamstring muscle tenderness; reproduction of the pain on hamstring stretch; and reproduction of pain by resisted contraction of the hamstrings. Players were excluded from the study if they: were unavailable for clinical testing within 3 days of injury onset; sustained a hamstring injury within 6 weeks of the end of their playing season, as this would preclude adequate return to play follow-up; or had clinical signs and symptoms suggestive of a diagnosis other than a hamstring strain. The study was approved by the Human Research Ethics Committee of The Alfred Hospital. Written informed consent was obtained from all participants.

Procedure

Questionnaire

Demographic information, functional progression (including time taken to walk and ascend stairs pain-free), use of non-steroidal anti-inflammatory medication (NSAIDs), mechanism and timing of injury and past history of hamstring injury in the previous 12 months were obtained by questionnaire.

Clinical assessment

The clinical assessment consisted of five tests assessing hamstring flexibility, neural mobility, pain provocation and site of the injury. All assessments were completed by one of two experienced sports physiotherapists. The order of testing was standardised, and the non-injured leg was measured first for each test. The tests selected were commonly used clinically,8 referenced in the literature9,,11 and reliable.8 12 13

The passive straight leg raise (PSLR) test provides an indication of hamstring muscle length14 with the knee in full extension. A bubble inclinometer (Fabrication Enterprises, New York) was placed on the anterior tibial border 15 cm below the tibial tuberosity. At the first point of reported stretch or discomfort/pain in the hamstring, the angle between the leg and the horizontal was recorded. The value used was the difference between the injured leg compared with the non-injured leg. This test was repeated, and the average of the two measurements was recorded. The test was considered positive if there was pain provocation and a >10° deficit in the injured leg compared with the uninjured leg.

The active knee extension (AKE) test is another measure of hamstring muscle length taken in a position of 90° hip flexion. The reliability of the test has been established by Gabbe et al,13 and the method described by these authors was used for the current study. For this test, the hip was maintained in 90° hip flexion by a specially constructed frame. At the point of maximal AKE (or onset of pain), the angle between the vertical and the tibia was recorded by an inclinometer as per the PSLR test. The value used was the difference between the injured leg compared with the non-injured leg. The test was repeated, and the average of the two measurements was recorded. The test was considered positive if there was pain provocation and a >10°deficit in the injured leg compared with the uninjured leg.

The active slump test assesses pain-sensitive neuromeningeal structures that have been suggested as a potential source of pain in the posterior thigh in hamstring injuries.15,,18 In the sitting position, the player was instructed to clasp his hands behind his back, to tuck his chin onto his chest and to slump, bringing his shoulders towards his hips with full cervical, thoracic and lumbar flexion. Next, full active dorsiflexion of the foot of the injured leg was requested and the player asked to actively extend his knee until he felt a stretch or hamstring pain. The player was then asked to extend his neck to a neutral position and describe the change in sensation that occurred in the hamstring. The test was considered positive if the player's original hamstring pain was decreased and then reproduced with cervical flexion.

A test of pain provocation evaluated whether the hamstring pain could be elicited by isometric contraction of the hamstring muscles. This was performed in prone lying with approximately 15° knee flexion. The examiner also palpated the hamstrings in prone lying to locate the region (medial or lateral hamstrings) relating to the player's pain.

Outcome measures

The time to return to competition in days from the initial injury to the time of the first match back from injury was recorded. The decision to return to competition was made by the club medical team who had no knowledge of the results of the measurements included in this study. The data were categorised into taking 3 weeks or less to return to play or taking more than 3 weeks to return to play.

A hamstring injury was considered a recurrence if the player reported pain in the same hamstring muscle compartment in the same leg which fulfilled the same criteria for a hamstring injury within 3 weeks of returning to play. Given that the majority of reinjuries in the AFL occur during the first 3 weeks,19 completing three matches without recurrence was considered a successful return from injury.

Data management and analysis

Data were analysed using SPSS (version 11.5) (SPSS, Chicago, Illinois). Height and weight were included as continuous variables. A number of variables were categorised for analysis to aid in interpretation. The outcomes of interest were time to return to play (≤3 weeks, >3 weeks) and recurrence of injury within 3 weeks (no, yes). Univariate associations between the potential predictors and outcomes were assessed using χ2 or Fisher exact tests for categorical variables, and independent t tests or their non-parametric equivalents for continuous variables. Backward, stepwise logistic regression analyses were used to identify independent predictors of time to return to competition and hamstring injury recurrence. Variables demonstrating a p value less than 0.20 on univariate testing were included in the multivariate models. Adjusted odds ratios (AORs) and 95% CIs of the ORs were calculated.

Results

Characteristics of the study cohort

The cohort of 59 players represented 95% of Melbourne-based AFL players who sustained a hamstring injury resulting in at least one missed game during 2002.20 The median (range) time from injury until examination was 2 (0 to 3) days. The mean (range) age of the players was 24 (17 to 33) years while the mean (range) height and weight were 186 (174 to 200) cm and 88 (74 to 107) kg, respectively. Twelve (20%) players reported that they had sustained a hamstring injury during the previous season.

The majority of the hamstring injuries occurred during competition (63%). Forty-one players (69%) reported a running-related mechanism for their hamstring injury. Forty-two (72%) injuries involved the biceps femoris muscle. Eighteen (31%) players (13 with biceps femoris injury and five with medial hamstring injury) were taking NSAIDs in the first 3 days following their hamstring strain with the ratio of medial and lateral hamstring strains similar (29% and 31%, respectively); nine of these (50%) were taking them at the time of injury, suggesting that nine players were specifically prescribed NSAIDs for the initial postinjury period. Forty-one players (69%) took longer than 24 h until they were able to walk at a normal pace pain-free, while 46 (78%) took longer than 24 h until they were able to ascend stairs at normal pace without pain.

The time taken to return to competition ranged from 1 to 8 weeks with a median of 26 days. Twenty-six players (44%) returned to competition within 3 weeks. Only two players took longer than 6 weeks to return to competition, one taking 7 weeks and the other 8 weeks. Nine players (15%) experienced a recurrence of their hamstring injury within the first 3 weeks of returning to play. Six of these (66%) had a recurrence in the first week of returning to competition, with two in the second week and one in the third. In all cases, players were unable to participate in at least three games (range 3–7 games) following their recurrence.

Predictors of time to return to competition following hamstring injury

On univariate testing (table 1), only time to walk pain-free was significantly associated with time to return to competition. There was no association between height, weight and age and time to return to competition following hamstring injury.

Table 1

Univariate results of the association between the variables and taking greater than 3 weeks to return to competition

Time to walk pain-free, a history of hamstring strain in the previous 12 months, the region of the hamstrings injured and the time taken to ascend stairs pain-free met the criterion for inclusion in a multivariate model. Players who reported taking more than 1 day to walk pain-free following injury were significantly more likely (AOR 4.0; 95% CI 1.3 to 12.6) to take longer than 3 weeks to return to competition than those walking pain-free within 1 day (p=0.018). Players with a previous history of hamstring injury were also at elevated risk of a delayed return to competition (AOR 4.2; 95% CI 1.0 to 18.0) but this just failed to reach significance (table 2). All other variables were eliminated by the stepwise procedure.

Table 2

Independent predictors of return to competition more than 3 weeks following injury

Risk factors for a recurrence of hamstring injury

All players with medial hamstring strains (n=17) returned to play successfully and did not experience a recurrence in the first 3 weeks. Thus, analysis of the predictors for recurrence was restricted to lateral hamstring strain cases only (n=42). Nine (21%) players with a lateral hamstring strain sustained a recurrence within the first 3 weeks of returning to play.

The univariate results are shown in table 3. A history of hamstring strain in the previous 12 months was associated with injury recurrence, while the presence of an AKE deficit of greater than 10° just failed to reach statistical significance. Time to walk pain-free, taking NSAIDs, PSLR test, pain provocation test, time to ascend stairs pain-free, slump test and mechanism of injury were not univariately associated with hamstring injury recurrence (table 3); neither was age (t=1.56, p=0.127), height (t=1.10, p=0.023) or weight (t=0.090, p=0.203).

Table 3

Univariate results of the association between the variables and the recurrence of hamstring injury

A past history of hamstring strain, time to walk pain-free, use of NSAIDs in the first 3 days and the presence of an AKE deficit were included in the multivariate model. Players who reported sustaining a hamstring injury in the previous season were more likely (AOR 19.6; 95% CI 1.5 to 261.1) to experience a recurrence within 3 weeks of return to competition than those without a previous history (table 4). The remaining variables were not significant independent predictors of recurrence.

Table 4

Independent predictors of recurrence of a lateral hamstring injury within 3 weeks of return to competition

Discussion

This study examined simple clinical assessment findings and player characteristics to identify early clinical predictors of time to return to competition and early recurrence following hamstring injury in elite male Australian footballers. Time to walk pain-free and a past history of hamstring injury were significant independent predictors of time to return to competition, and early recurrence of hamstring injury, respectively. Combined with the finding that no medial hamstring strain recurred, the findings of this study provide useful prognostic information for clinicians actively involved in the treatment of hamstring injuries.

The time taken to walk pain-free postinjury is a simple prognostic indicator for clinicians and players. A player taking longer than 24 h to achieve pain-free walking was four times more likely to need more than 3 weeks to return to competition compared with players walking pain-free within 24 h of injury. It is likely that this parameter provides an indicator of injury severity, justifying its inclusion in hamstring injury grading systems.21 Using the regression coefficients from the multivariate model, an inability to walk pain-free within 1 day combined with a past history of hamstring injury resulted in a 93% chance of taking longer than 3 weeks to return. This important information can be conveyed to the coach and player in the postacute stage.

Other known predictors for hamstring injury were not associated with time to return to play or recurrence of injury except for a past history of hamstring injury. A past history of hamstring injury is well established as a predictor of future hamstring injury,4 5 and our study confirms its importance as a predictor of short-term recurrence, highlighting the need for primary prevention. No association was found between lateral strains and taking more than 3 weeks to return to competition. This contrasts to imaging studies involving Australian football players where biceps femoris strains take longer to return to play.5 6 22 This contrast may be due to the sample size in this study where only the strongest predictors of outcome were able to be identified.

Fifteen per cent of the players experienced a recurrence of their hamstring injury within the first 3 weeks of returning to play, with the majority of these occurring in the first match, which is consistent with previous reports of hamstring injury recurrence rate in the AFL.19 All medial hamstring strains returned to competition without reinjury, providing a useful clinical prognostic indicator. This may be due to the relatively reduced stretching forces on the semitendinosis and semimembranosis during what appears to be the critical late swing phase of sprinting.23 A history of hamstring injury in the previous 12 months was a significant predictor of hamstring recurrence. This may be due to any combination of inadequate rehabilitation, the effects of ongoing pathology, healing and scar formation6 or due to previously injured muscle being more susceptible to eccentric damage than uninjured muscle,24 possibly contributed to by altered hamstring firing patterns.18 25 26 Given that lateral hamstring strains and those with a past history are more likely to recur, the presence of these factors should dictate that a thorough rehabilitation be carried out in order to reduce the possibility of recurrence in the first 3 weeks of returning to competition. It may also be debated as to whether the return to play should be delayed in these players to facilitate greater healing of the injured muscle.

In recent years, MRI and ultrasound imaging are often used in conjunction with clinical assessment to diagnose and predict the severity of hamstring strains in elite-level Australian football players.6 22 27 MRI parameters of the longitudinal length6 22 and the cross-sectional area of the injury27 show good correlation with injury severity. Several studies found that MRI negative hamstring strains have a significantly faster rehabilitation time compared with MRI positive strains and are less likely to recur.6 27 28 In the only study29 comparing the clinical and radiological estimation of the duration of rehabilitation of an acute hamstring strain in elite footballers, the clinical examination was moderately more accurate than that of MRI. As return to play is a risk-management decision,30 the factors revealed in the present study should be further highlighted in the risk-management equation about the best time to return to play.

Taking NSAIDs in the first few days following a hamstring strain was the only alterable parameter in the postacute assessment and, although not significant in this study, may be an influence in both the time taken to return to play and risk of recurrence. Despite NSAIDs being recommended immediately following a muscle strain,31,,33 their current use in AFL football appears to be a decision based on clinical experience. Eighteen (31%) players were taking NSAIDs in the days following their hamstring strain (similar ratio of medial and lateral hamstring strains) with nine of these specifically prescribed NSAIDs for the initial postinjury period. There may be merit in delaying NSAIDs for a few days following hamstring injury as their earlier use may interfere with the chemotaxis of cells which is necessary for the repair and remodelling of regenerating muscle.32 Alternatively, it might also be argued that by limiting pain and inflammation, other changes predisposing to injury at new sites such as inflammation-induced atrophy and/or pain inhibition are reduced, allowing rehabilitation to proceed at a faster rate.

The strengths of this study were its high capture of injured athletes (95% of all eligible cases), its prospective nature and the rigorous assessment methodology used. Nevertheless, there are a number of limitations that warrant mention. First, the study population was restricted to elite Australian football players. Whether the results can be extrapolated to other levels of Australian football participation or other sporting populations where hamstring injuries are prevalent cannot be ascertained but requires further study. Another limitation was that the decision of when to return to play was left up to the discretion of the medical personnel, and it is possible that they were more conservative if the player had a past history of hamstring strain.

While the cohort number for a study of hamstring injuries was relatively large compared with previously published studies,15 34,,37 the study was insufficient to identify all but the largest effects, particularly for modelling predictors of recurrence. A larger study could better examine the association between clinical assessments and outcome for variables with small to moderate effects. Nevertheless, the present study did identify useful clinical prognosticators for hamstring injury.

What is already known about this topic

  • Past history of a hamstring strain and having a lateral hamstring strain are predictors available in the early clinical assessment that appear to relate to greater time to return to competition. Apart from imaging, few studies have evaluated the accuracy of early clinical assessment findings as predictors of time to return to competition and recurrence.

What this study adds

  • This study demonstrates that in elite Australian football several clinically obtainable parameters in the postacute setting following a hamstring strain contribute to the prediction of the expected time to return to play and to the likelihood of suffering a recurrence. Past history of hamstring strain and being unable to walk at a normal pace pain-free within 24 h of injury were independent predictors of being unable to return to play in less than 4 weeks from the time of injury. All players with medial hamstring strains returned to play successfully and did not suffer a reinjury. The use of NSAIDs in the postacute period may influence both time taken to return to play and risk of recurrence. Past history was the sole independent predictor of suffering a recurrence upon returning to play.

Conclusion

This study identified important early clinical parameters for hamstring injury prognosis. Defining the severity and expected return to play timescale is important in guiding rehabilitation and in team planning. Some of these factors also present as warning signs that a player might be at risk of suffering a recurrence in the first few weeks after returning to play. Players with a past history of a hamstring strain on the same leg in the preceding season and unable to walk pain free within 24 h, have a significant chance of taking 4 weeks or more to return to play. Players with a lateral hamstring strain and a past history of a hamstring strain are at a greater risk of suffering a recurrence in the first 3 weeks of returning to competition. These parameters are important in the early assessment posthamstring injury.

References

View Abstract

Footnotes

  • Funding The authors acknowledge the Australian Football League for financial support and the assistance of the AFL doctors and physiotherapists. BJG was supported by a Career Development Award (465103) from the National Health and Medical Research Council of Australia during the preparation of this manuscript.

  • Competing interests None.

  • Provenance and peer review ed Commissioned; externally peer reviewed.

  • Patient consent Obtained.