Scolaris Content Display Scolaris Content Display

Cochrane Database of Systematic Reviews Protocol - Intervention

Interventions for preventing hamstring injuries

This is not the most recent version

Collapse all Expand all

Abstract

This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:

To review the evidence from randomised and quasi‐randomised controlled trials evaluating interventions for preventing all types of hamstring injuries in physically active people.

This review aims to compare the effects (primarily the incidence of hamstring injuries) in individuals participating in relevant sport or physical occupations of:
1. interventions targeted at preventing hamstring injuries versus no or placebo intervention;
2. different targeted interventions.

We plan to study the outcomes in different age groups (adolescent versus adult); by history (first time versus recurrent hamstring injury) and participation in professional/ occupational versus leisure activities.

Background

Description of the condition

Hamstrings are a group of muscles of the posterior thigh, that cross both the hip and knee joints and are involved in thigh extension and knee flexion of the lower limb. The hamstring muscle group comprises three separate muscles: the semitendinosus and semimembranosus and biceps femoris. Hamstring injury is defined by the anatomical site within the muscle that is affected, and injury must be present in one or more of the component muscles. Hamstring injuries are commonly classified as grades one, two and three, where severity increases with the grade of injury (Kellet 1986). It is important to note that the type of injury may differ in adolescents compared to the skeletally mature person. Injuries occur in patterns unique to the skeletally immature adolescent, given their developing bones and supporting ligamentous structures (Frank 2007). Diagnosis is based on patient history, including the cause of injury and clinical findings of local pain and loss of function, demonstrated by palpation, range of motion and muscle testing (Kujala 1997). Magnetic resonance imaging (MRI) can provide further information on the extent of injury (Brandser 1995).

Musculotendinous injuries usually occur as a result of either direct or indirect trauma (Agre 1985). Direct trauma such as from a blow results in muscle contusion (bruising). Indirect trauma can occur as a result of alteration in the intensity or synergy of contraction, resulting in a single powerful muscle contraction. Hamstrings are capable of producing large forces (Garrett 1984; Noonan 1999) that are most beneficial during periods of increased activity requiring bursts of speed or rapid acceleration and deceleration. Recent evidence has suggested that the hamstring muscles are most vulnerable to injury during the rapid change from eccentric to concentric function, where the leg decelerates to strike the ground during running (Verrall 2001). Indirect trauma can also result from a overstretch of the musculotendinous unit leading to a strain, tear or avulsion (Agre 1985). It is generally claimed that strain injuries most often occur near the musculotendinous junction (Garrett 1996).

Hamstring injuries are commonplace in many mainstream sports and occupations involving physical activity (Kroll 1997). The initial Football Association Audit of Injuries study (Hawkins 2001) found that 12% of all injuries reported over two seasons were hamstring strains. Injury prevention is a key goal of every trainer, coach, and participant of physical activity, whether pursued in an occupational or personal context. If a person is injured they are unable to participate fully and, often, not at all. A hamstring injury will cause professional footballer to miss approximately three matches or weeks of play (Woods 2004). Hamstring injuries are often serious requiring rehabilitation and they tend to recur (Croisier 2004; Orchard 2002). For instance, Orchard 2002a reported a high rate of recurrence in 34% of primary cases over the course of a Australian football season.

The causes of hamstring injuries are complicated and multifactorial (Gleim 1997). In general, a distinction in risk factors has been made between so‐called intrinsic (person‐related) and extrinsic (environment‐related) risk factors (Inklar 1994; Taimela 1990; Van Mechelen 1992). Intrinsic factors include hamstring muscle weakness, strength imbalances, fatigue, inadequate flexibility, body mechanics and disturbed posture, poor running technique and psychosocial factors (Agre 1985; Croisier 2002; Knapik 1992; Worrell 1992). Extrinsic factors include unsatisfactory warm‐up and training procedures, fatigue related to enforced excessive activity, poor playing surfaces and unsuitable training and sports specific activities (Hawkins 1999; Safran 1988). Orchard 2001 suggested intrinsic factors are more predictive of muscle strain than extrinsic factors. However, a recent systematic review (Foreman 2006) concluded that no single risk factor was found to have a significant association with hamstring injury (although the review was limited by the use of smaller cohort studies and the exclusion of non‐English papers). Understanding the individual risk factors for injury is an important basis for developing preventive measures.

Description of the intervention

The prevention of hamstring injuries is an ongoing process where intervention is necessary as long as the participant engages in physical activity placing them at risk. Many interventions are widely employed by participants, trainers, coaches and therapists to prevent such injuries. These include exercise therapy to strengthen, stabilise and lengthen the hamstring muscles, such as stretching and strengthening exercises (Croisier 2002); neuromuscular injury prevention strategies including proprioceptive balance training (Emery 2007; Turl 1998); chiropractic, spinal manipulative therapy (SMT) and correction of lumbar‐pelvic biomechanics (Hoskins 2005); muscle activation work to improve hip extension motor patterns and running technique (Hoskins 2005); massage and mobilisation to increase flexibility and range of movement directed towards soft‐tissues structures (Brosseau 2002) articular structures (Cibulka 1986) and neural tissue (Turl 1998); education, including awareness of the risks for hamstring injury and importance of training (Arnason 2005), and functional training and sport specific drills (Verrall 2005).

Why it is important to do this review

Despite the relatively high incidence of hamstring injuries in sport, evidence of the efficacy of preventive interventions is not well established. The authors are not aware of any systematic reviews specifically focused on the interventions used for the prevention of hamstring injuries.

Objectives

To review the evidence from randomised and quasi‐randomised controlled trials evaluating interventions for preventing all types of hamstring injuries in physically active people.

This review aims to compare the effects (primarily the incidence of hamstring injuries) in individuals participating in relevant sport or physical occupations of:
1. interventions targeted at preventing hamstring injuries versus no or placebo intervention;
2. different targeted interventions.

We plan to study the outcomes in different age groups (adolescent versus adult); by history (first time versus recurrent hamstring injury) and participation in professional/ occupational versus leisure activities.

Methods

Criteria for considering studies for this review

Types of studies

Randomised controlled trials and quasi‐randomised (method of allocating participants to a treatment which is not strictly random; e.g. by date of birth, hospital record number, alternation) controlled trials in which any prevention approach is compared with another or a control group. Cluster‐randomised trials, abstracts and unpublished studies where appropriate, will also be included.

Trials which involve testing of preventative interventions in 'laboratory' conditions or which only report intermediate outcome measures (muscle strength or flexibility) that have no proven relationship to clinical outcomes along with reviews will be excluded.

Types of participants

We will include trials involving physically active people of either sex from adolescence (using The World Health Organization's definition of adolescence as the period of life between 10‐19 (Goodburn 1995)) to adults of any age who are at risk of incurring hamstring injuries. Participants with and without a history of hamstring injury will be included and a distinction between them made where possible.

We will exclude trials focusing on interventions for people with existing or under treatment for lower‐limb musculoskeletal injuries. Reports focusing on children below the age of 10 will also be excluded.

Types of interventions

Any intervention or intervention programme (one or more intervention in comparison with a control or one or more alternative interventions) for the prevention of hamstring injuries will be included in the review.

These will be categorised as:
1. Stretching or mobilising exercises;
2. Strengthening exercises (eccentric and concentric);
3. Exercise to correct movement dysfunction and sport specific drills;
4. Manual therapies (articular, neural and soft tissue mobilisation and massage);
5. Neuromuscular strategies (proprioceptive balance work);
6. Educational awareness programmes;
7. General intervention programmes (warm‐up exercises, aerobic drills).

Types of outcome measures

Primary outcomes

1. The incidence of any hamstring injury will be included regardless of grade, site and chronicity. Where possible distinctions will be made between injuries that have been clinically confirmed and those that are confirmed with imaging, and furthermore are of sufficient severity to prevent resumption of full activities.

Secondary outcomes

1. Compliance with the intervention;
2. Duration of intervention;
3. Severity of hamstring injuries (grade, surgery considered);
4. Complications/adverse effects associated with chosen intervention: e.g. fitness deficit, muscle soreness, other injuries;
5. Incidence of other lower‐limb injuries and low back pain;
6. Measures of service utilisation or resource use (e.g. costs, diagnostic procedures).

Search methods for identification of studies

Electronic searches

We will search the Cochrane Bone, Joint and Muscle Trauma Group Specialised Register (to present), the Cochrane Central Register of Controlled Trials (The Cochrane Library, current issue), MEDLINE (OVID‐WEB from 1966), CINAHL (from 1982), SPORTDiscus (from 1949), PEDro ‐ The Physiotherapy Evidence Database at http://www.pedro.fhs.usyd.edu.au (to present), AMED (from 1985), EMBASE (from 1980), and reference lists of articles. We will also search Current Controlled Trials at www.controlled‐trials.com and the UK National Research Register at www.update‐software.com/national/ for ongoing and recently completed trials. No language restrictions will be applied.

In MEDLINE (OVID‐WEB) the search strategy will be combined with all three sections of the optimal MEDLINE search strategy for randomised trials (Higgins 2005) (seeAppendix 1). The search strategy will be modified for use in other databases.

Searching other resources

To identify theses and unpublished studies we will also contact experts in the field and institutions. Also references given in the trials retrieved will be further examined.

We will also include the findings from handsearches of supplements (1997 onwards) in eight journals (Am J Sports Med; Br J Sport Med; Sports Med; J Am Acad Orthop Surg; Chiropract Osteopathy; Med Sci Sports Exerc; Clin J Sport Med; J Orthop Sports Phys Ther).

Data collection and analysis

Selection of studies

Both review authors will obtain and screen abstracts and citations identified by searches. We will retrieve full text articles for any that may describe eligible studies and perform independent study selection. Any disagreement will be resolved by discussion.

Data extraction and management

Using a data extraction form, both review authors will independently extract trial details and data. The double entry facility within Review Manager 4.2 will be used for data entry. Any differences will be resolved by discussion. Where data is not provided in the text or tables, extraction of results from graphs in trial reports will be considered. Corresponding authors will be contacted to provide missing information for trial results and methods.

Assessment of risk of bias in included studies

In this review, risk of bias will be assessed indirectly in terms of different aspects of methodological quality.

Both review authors will independently assess the included studies using a piloted modification of the generic quality assessment tool developed by the Cochrane Bone, Joint and Muscle Trauma Group. Each criterion is rated as either 'Y' (Yes), '?' (uncertain/ unknown) or 'N' (No), depending on how successfully the criterion is met. The rating scheme covering 14 aspects of trial validity, plus brief notes of coding guidelines for some items, is shown in Table 1.

Open in table viewer
Table 1. Methodological quality assessment scheme

Items

Scores

Notes

A. Was an effective method of randomisation used?

Y = yes, e.g. use of random tables
? = no description of method
N = quasi‐randomised method only

To achieve 'Y', a random (unpredictable) assignment sequence is required.

B. Was the assigned treatment adequately concealed prior to allocation?

Y = method did not allow disclosure of assignment ? = small but possible chance of disclosure of assignment or unclear N = quasi‐randomised or open list/tables

For example, a method using centralised (e.g. allocation by a central office unaware of subject characteristics) or independently‐controlled randomisation.

Cochrane code (see Handbook): Clearly yes = A; Not sure = B; Clearly no = C.

C. Were the outcomes of participants who withdrew described and included in the analysis (intention to treat)?

Y = no dropouts or withdrawals well described and accounted for in analysis.
? = withdrawals described and analysis not possible
N = no mention, inadequate mention, or obvious differences and no adjustment

D. Were the outcome assessors blind to assignment status?

Y = effective action taken to blind outcome assessors
? = small or moderate chance of unblinding of outcome assessors
N = not possible, or not mentioned (unless double‐blind), or possible but not done

E. Were the treatment and control groups comparable at entry?

Y = good comparability of groups, or confounding adjusted for in analysis
? = confounding small; mentioned but not adjusted for
N = large potential for confounding, or not discussed

The principal confounders considered were sex, age, previous overuse lower‐limb injury and prior physical activity profile.

F. Were the participants blind to assignment status after allocation?

Y = effective action taken to blind participants
? = small or moderate chance of unblinding of participants
N = not possible, or not mentioned (unless double‐blind), or possible but not done

G. Were the treatment providers blind to assignment status?

Y = effective action taken to blind treatment providers
? = small or moderate chance of unblinding of treatment providers
N = not possible, or not mentioned (unless double‐blind), or possible but not done

H. Were care programmes, other than the trial options, identical?

Y = care programmes clearly identical
? = clear but trivial differences
N = not mentioned or clear and important differences in care

Examples of clinically important differences in other interventions (co‐interventions) are those which could act as active measures for prevention of hamstring injuries: training programmes, advice on activity and etc.

I. Were the inclusion and exclusion criteria clearly defined?

Y = clearly defined
? = inadequately defined
N = not defined

To achieve 'Y', the inclusion or exclusion of individuals with a) previous hamstring injuries and b) previous exposure to trial intervention needs to be confirmed.

J. Were the interventions clearly defined?

Y = clearly defined interventions are applied with a standardised protocol
? = clearly defined interventions are applied but the application protocol is not standardised
N = intervention and/or application protocol are poorly or not defined

K. Were the outcome measures used clearly defined? (by outcome)

Y = optimal
? = adequate
N = not defined, not adequate

To achieve 'Y', the method and strategy of data collection need to be clearly defined.

L. Were diagnostic tests used in outcome assessment clinically useful? (by outcome)

Y = clearly defined and best available tests are applied with a standardised protocol
? = clearly defined tests are applied but the application protocol is not standardised
N = tests and/or application protocol are poorly or not defined

M. Is the surveillance active, and of clinically appropriate duration?(by outcome)

Y = active surveillance and appropriate duration
? = active surveillance, but inadequate duration
N = surveillance not active or not defined

N. Was compliance of subjects disclosed?

Y = compliance greater than 90% in each group after randomisation
? = compliance with allocated procedure reported
N = compliance not mentioned

Measures of compliance are likely to vary and thus the way of estimating the 90% compliance will depend on the measures used in individual trials.

Measures of treatment effect

Quantitative outcome data given in individual trial reports will be presented in the text and in the analyses, using the fixed‐effect model to estimate the relative risk (RR) with 95% confidence intervals (95% CI) for dichotomous outcomes, and mean differences with 95% confidence intervals for continuous outcomes. Ordinal scales will be treated as continuous outcomes unless and until accepted meta‐analytic techniques for ordinal outcome data become available.

Unit of analysis issues

The unit of randomisation in these trials may be groups of individuals, such as a football team, rather than individuals. For such trials, unit of analysis errors are likely to result from the presentation of the outcomes for individual participants. Where cluster randomisation has been used we will seek the intra‐cluster correlation coefficient (ICC) and pool data using the generic inverse variance. Exploratory analyses using assumed values for cluster‐randomised studies which have not reported the ICC will also be considered.

Dealing with missing data

Where appropriate, we will perform intention‐to‐treat analyses to include all people randomised to the intervention groups. We will investigate the effect of drop outs and exclusions by conducting worse and best scenario analyses. If appropriate, we will impute values for missing data.

Assessment of heterogeneity

Heterogeneity will be assessed by visual inspection of the forest plot (analysis) along with consideration of the test for heterogeneity and the I² statistic (Higgins 2003).

Assessment of reporting biases

In the unlikely event that sufficient data are available, we would attempt to assess publication bias by preparing a funnel plot.

Data synthesis

If considered appropriate, results of comparable groups of trials will be pooled. Initially we will use the fixed effect model and 95% confidence intervals. We will also consider using the random‐effects model, especially where there is unexplained heterogeneity.

Subgroup analysis and investigation of heterogeneity

We plan subgroup analyses by therapeutic approach as outlined under "Types of interventions". This will include a comparison of each approach versus control protocols, as well as direct comparisons of different approaches. Also age (adolescents and adults), gender, history or not of hamstring injury, occupational or leisure activity. To test whether the subgroups are statistically significantly different from one another, we will test the interaction using the technique outlined by Altman and Bland (Altman 2003).

Sensitivity analysis

Where possible, we plan sensitivity analyses on the primary outcome examining various aspects of trial and review methodology, including the effects of missing data, study quality (specifically allocation concealment and outcome assessor blinding), and inclusion of trials only reported in abstracts or unpublished reports. Again to test whether the subgroups are statistically significantly different from one another, we will test the interaction using the technique outlined by Altman and Bland (Altman 2003).

Table 1. Methodological quality assessment scheme

Items

Scores

Notes

A. Was an effective method of randomisation used?

Y = yes, e.g. use of random tables
? = no description of method
N = quasi‐randomised method only

To achieve 'Y', a random (unpredictable) assignment sequence is required.

B. Was the assigned treatment adequately concealed prior to allocation?

Y = method did not allow disclosure of assignment ? = small but possible chance of disclosure of assignment or unclear N = quasi‐randomised or open list/tables

For example, a method using centralised (e.g. allocation by a central office unaware of subject characteristics) or independently‐controlled randomisation.

Cochrane code (see Handbook): Clearly yes = A; Not sure = B; Clearly no = C.

C. Were the outcomes of participants who withdrew described and included in the analysis (intention to treat)?

Y = no dropouts or withdrawals well described and accounted for in analysis.
? = withdrawals described and analysis not possible
N = no mention, inadequate mention, or obvious differences and no adjustment

D. Were the outcome assessors blind to assignment status?

Y = effective action taken to blind outcome assessors
? = small or moderate chance of unblinding of outcome assessors
N = not possible, or not mentioned (unless double‐blind), or possible but not done

E. Were the treatment and control groups comparable at entry?

Y = good comparability of groups, or confounding adjusted for in analysis
? = confounding small; mentioned but not adjusted for
N = large potential for confounding, or not discussed

The principal confounders considered were sex, age, previous overuse lower‐limb injury and prior physical activity profile.

F. Were the participants blind to assignment status after allocation?

Y = effective action taken to blind participants
? = small or moderate chance of unblinding of participants
N = not possible, or not mentioned (unless double‐blind), or possible but not done

G. Were the treatment providers blind to assignment status?

Y = effective action taken to blind treatment providers
? = small or moderate chance of unblinding of treatment providers
N = not possible, or not mentioned (unless double‐blind), or possible but not done

H. Were care programmes, other than the trial options, identical?

Y = care programmes clearly identical
? = clear but trivial differences
N = not mentioned or clear and important differences in care

Examples of clinically important differences in other interventions (co‐interventions) are those which could act as active measures for prevention of hamstring injuries: training programmes, advice on activity and etc.

I. Were the inclusion and exclusion criteria clearly defined?

Y = clearly defined
? = inadequately defined
N = not defined

To achieve 'Y', the inclusion or exclusion of individuals with a) previous hamstring injuries and b) previous exposure to trial intervention needs to be confirmed.

J. Were the interventions clearly defined?

Y = clearly defined interventions are applied with a standardised protocol
? = clearly defined interventions are applied but the application protocol is not standardised
N = intervention and/or application protocol are poorly or not defined

K. Were the outcome measures used clearly defined? (by outcome)

Y = optimal
? = adequate
N = not defined, not adequate

To achieve 'Y', the method and strategy of data collection need to be clearly defined.

L. Were diagnostic tests used in outcome assessment clinically useful? (by outcome)

Y = clearly defined and best available tests are applied with a standardised protocol
? = clearly defined tests are applied but the application protocol is not standardised
N = tests and/or application protocol are poorly or not defined

M. Is the surveillance active, and of clinically appropriate duration?(by outcome)

Y = active surveillance and appropriate duration
? = active surveillance, but inadequate duration
N = surveillance not active or not defined

N. Was compliance of subjects disclosed?

Y = compliance greater than 90% in each group after randomisation
? = compliance with allocated procedure reported
N = compliance not mentioned

Measures of compliance are likely to vary and thus the way of estimating the 90% compliance will depend on the measures used in individual trials.

Figures and Tables -
Table 1. Methodological quality assessment scheme