Surgical versus conservative interventions for treating anterior cruciate ligament injuries

  • Protocol
  • Intervention



This is the protocol for a review and there is no abstract. The objectives are as follows:

To assess the effects (benefits and harms) of surgical versus conservative interventions for treating anterior cruciate ligament injuries.


Description of the condition

The anterior cruciate ligament (ACL) of the knee joint plays an essential role in both static (standing or squatting) and dynamic (walking or running) knee stability. Primarily, the ACL prevents anterior translation (forward movement) of the tibia relative to the femur in the sagittal (antero-posterior) plane, aiding stabilisation of the joint from a flexed (bent) to a near full extension (straight) position of the knee.

Rupture of the ACL is a common injury, mainly affecting young, physically active individuals; with an estimated 200,000 ACL ruptures per year in the United States (Spindler 2008). It is often injured during sporting activities such as football, skiing and basketball (Bahr 2003). In over 70% of cases, the injury is caused by a non-contact mechanism such as sudden deceleration combined with changing direction or pivoting or landing with the knee in nearly full extension after a jump (Hernandez 2006). Contact (traumatic) mechanisms of injury usually involve a translational force applied to the anterior aspect of a fixed lower leg (Hewett 2006). The acute injury is frequently characterised by knee pain and an audible ‘popping’ sound at the time of injury. The injured person presents with knee pain, swelling, haemarthrosis (bleeding into the joint space), instability with further activity and painful range of motion.

In approximately 10% of cases, the ACL injury occurs in isolation; however, in the majority of cases it is combined with other injuries, typically to the collateral ligaments, subchondral bone and meniscii (Miyasaka 1991; Bowers 2005; Hernandez 2006).

Diagnostic imaging, including magnetic resonance imaging (MRI), is used for confirmation of the diagnosis of ACL injury or rupture, and evaluation of associated pathology such as articular cartilage injury and meniscal and associated ligamentous tears; all of which play a role in maintaining stability of the knee (Crawford 2007).

Chronic ACL injury can have a profound effect on the knee kinematics (movements) of those affected. Common problems include recurrent knee instability (giving way) and symptoms of associated meniscal or articular cartilage damage such as intermittent swelling or a locking sensation (Hernandez 2006). Furthermore, the injury can lead to poor reported quality of life (Spindler 2008) and decreased activity levels (Thorstensson 2009). It is also associated with increased risk of secondary osteoarthritis of the knee irrespective of treatment (Øiestad 2009; Rout 2013). These related morbidities have been shown to be associated with high healthcare expenditure (Frobell 2010).

Description of the intervention

Surgical treatment for ACL rupture has evolved from simple repair using suturing or suturing with some sort of augmentation to ACL reconstruction, which involves reconstruction of the ligament using a substitute graft of tendon or ligament fixed into position in pre-prepared drill holes. ACL reconstruction is increasingly performed as an arthroscopic procedure. Of those who undergo surgical reconstruction, 94% are performed within one year of the initial injury (Collins 2013). ACL reconstruction is the predominant method of surgery in current practice and hundreds of thousands of these operations are carried out each year.

Three types of grafts are commonly used: those from the patient's body (autograft), cadaveric human donors (allograft) or a synthetic ligament substitute. Commonly, the hamstring tendons of semitendinosus and gracilis are harvested from the limb of the ruptured ACL and this graft is removed during the reconstruction operation. Alternatively, a bone-patella tendon-bone (BPTB) construct uses a section of the middle of the patella (kneecap) tendon with bone at either end. The relative merits of hamstring and BPTB grafts have been reviewed (Mohtadi 2011).

Conservative (non-operative) treatment for people with an ACL rupture can include the use of cryotherapy (ice), continuous passive motion (movement of the joint by a machine), restrictive bracing, electrotherapy (muscle stimulation) and exercises aimed at strengthening and balance. The use of plaster casts for initial immobilisation of the knee is very uncommon nowadays (Linko 2009).

Rehabilitation regimens used for both treatment options commonly use a three-stage progressive programme: acute, recovery and functional phases (Micheo 2010). The acute stage following injury, or immediately after surgery, aims to restore range of motion and resolve inflammation. The recovery phase is from approximately three to six weeks, with the aim of improving lower limb muscle strength and functional stability. Finally, the functional stage of rehabilitation (from six weeks onwards) concentrates on returning to previous levels of activity and decreasing the risk of re-injury (Kvist 2004). There is little consensus over the most effective rehabilitation protocol for achieving these aims (Negus 2012).

Whilst surgical interventions have become commonplace for athletic individuals, initial non-operative (conservative) treatments based around physiotherapy are used more commonly in the general population (Linko 2009).

How the intervention might work

All treatments aim to reduce knee pain and instability and restore function. However, the optimal management strategy following rupture of the ACL remains controversial. In the short term, reconstructive surgery may improve knee function for those experiencing severe instability in activity or repeated episodes of ‘giving way’, or both. However, all surgery involves an increased risk of complications such as infection. In particular, for reconstruction using autograft, significant donor site morbidity can occur, including anterior knee pain with BPTB grafts and pain and weakness of knee flexors with hamstring grafts (Spindler 2004; Mohtadi 2011).

Although studies of conservative treatment have demonstrated satisfactory results with patients returning to pre-injury activity level (Kostogiannis 2007; Linko 2009; Frobell 2013), the long-term results, in particular relating to the development of early onset osteoarthritis, are still debatable. Radiographically-diagnosed osteoarthritis has been reported in 20% to 50% of ACL-deficient knees at 10 years post injury compared with 5% in uninjured knees (Lohmander 2007; Øiestad 2009; Ajuied 2013). However, surgery has not been shown to offer protection against long-term degenerative change (Øiestad 2009; Rout 2013). Moreover, recent studies have suggested structured neuromuscular rehabilitation might provide effective recovery following ACL rupture without increasing the risks of long-term degenerative change (Delincé 2012).

Why it is important to do this review

The management of ACL injuries includes both reconstructive surgery and conservative treatments. It is unclear whether stabilising the knee surgically produces any benefit for the knee in comparison with conservative interventions. The previous Cochrane review in this area (Linko 2009) found that there was insufficient evidence to determine whether surgery or conservative management was superior for the treatment of ACL rupture and highlighted the need for good quality randomised controlled trials (RCTs) of current practice, particularly ACL reconstruction. Current surgical practice has also changed in terms of the population, with an increasing number of ACL reconstructions being performed on a young athletic (adolescent) cohort (Ramski 2013). These point to the need for a systematic review of the evidence from randomised trials comparing the effects of the current surgical versus conservative treatment methods for ACL rupture.


To assess the effects (benefits and harms) of surgical versus conservative interventions for treating anterior cruciate ligament injuries.


Criteria for considering studies for this review

Types of studies

We will include randomised controlled trials comparing surgical versus conservative interventions for treating anterior cruciate ligament injuries.

Types of participants

We will include participants of any age (thus, including children) with anterior cruciate ligament rupture. Ideally, diagnosis will have been made with positive clinical examination and either a positive MRI or a positive examination under anaesthesia (EUA).

We will exclude studies whose prime focus is on the management of ACL and a concomitant knee ligament rupture (e.g. medial collateral ligament). We will exclude people with inflammatory arthropathy or end stage osteoarthritis (Grade 4 Kellgren and Lawrence). However, we will include mixed population trials if they include only a small proportion (preferably < 10% and preferably balanced between intervention groups) of participants with other major knee ligament or cartilage lesions, or if separate data are provided for participants without these additional injuries.

Types of interventions

The interventions being compared are surgery and conservative treatment for ACL rupture. We will include any trial evaluating surgery involving ACL reconstruction. Thus, we will include any method of reconstruction (e.g. open or arthroscopic), any type of reconstruction technique (e.g. single or double bundle) or graft fixation, and any type of graft. Direct repair of the ACL is increasingly rare, so we will only include this if we find any new recently conducted trials (that is, since 2000) . Otherwise, we will not repeat the analyses provided in Linko 2009.

We will include any method of conservative treatment; this is likely to include bracing and physiotherapy, or both.

Types of outcome measures

Primary outcomes
  • Subjective knee scores, e.g. Knee Injury and Osteoarthritis Outcome Score (KOOS) (Roos 1998), anterior cruciate ligament quality of life score (Mohtadi 1998) and International Knee Documentation Committee (IKDC), subjective part (Irrgang 2001).

  • Adverse events (such as donor site morbidity, failure of graft including re-rupture, infection, deep vein thrombosis and pulmonary embolism).

  • Treatment failure including re-operation (for surgery) or subsequent operation (for conservative treatment).

Secondary outcomes
  • General health-related quality of life, preferably measured using validated scales such as SF36 (Ware 1992) and EQ5D (Brooks 1996).

  • Return to activity or level of sports participation, including Lysholm (Lysholm 1982) or Tegner (Tegner 1985) scores.

  • Functional assessments (e.g. single-hop test).

  • Composite clinical examination outcomes (IKDC, objective part; Hefti 1993).

  • Knee stability (assessed using manual methods (e.g. Lachman or pivot shift tests) or using knee ligament testing devices (e.g. KT 1000)).

  • Objective measure of muscle strength (isokinetic muscle torque).

Resource and economic outcomes

Resource and economic outcomes such as those measuring service utilisation, including cost of surgery, length of inpatient stay, outpatient attendance, duration of sick leave. The results of cost or economic analysis will be reported where available.

Timing of outcome measurement

Assessments will be made at short- (less than one year), intermediate- (one to three years) and long-term (greater than three years) follow-up.

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 (CENTRAL) (The Cochrane Library, current issue), MEDLINE (1946 to present) and EMBASE (1980 to present) using tailored search strategies.

In MEDLINE, we will combine subject-specific terms with the sensitivity- and precision-maximising version of the Cochrane Highly Sensitive Search Strategy for identifying randomised trials (Lefebvre 2011) (see Appendix 1). This strategy will be modified for use in the other databases.

We will also search the World Health Organization International Clinical Trials Registry Platform for ongoing and recently completed trials.

There will be no language or publication status restrictions.

Searching other resources

We will check reference lists of relevant articles and contact individuals or organisations for further data where necessary. We will search orthopaedic and sports medicine conference abstracts to identify any unpublished trial data.

Data collection and analysis

Selection of studies

Two review authors (LD and KH) will independently screen all titles and abstracts for potentially eligible studies, for which we will obtain full text reports. The same two authors will independently perform study selection. Any disagreements regarding the inclusion or exclusion of individual studies will be resolved by discussion or, if necessary, by involvement of a third review author.

Data extraction and management

We will develop a data collection form to include all of the relevant variables for the study, including details of methods, participants, setting, interventions, outcomes, results and funding sources. Two authors (LD and KH) will pilot the data collection form, using a representative sample of studies, in order to identify any missing items or unclear coding instructions. After finalising the form, the same two authors will independently perform data extraction. Any disagreements will be resolved by discussion between two authors or, if no consensus can be achieved, a third author will act as an arbitrator. Before the data are entered into Review Manager software (RevMan 2012), any additional data management will be done in Microsoft Excel.

Assessment of risk of bias in included studies

Two review authors (LD and KH) will independently assess the risk of bias in each included study using the Cochrane 'Risk of bias' tool. Any disagreements will be resolved by consensus between the two authors or, if no consensus can be achieved, a third author will act as an arbitrator. We will assess separately the risk of bias for the following domains: sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data and selective outcome reporting, as well as other sources of bias, such as major differences in rehabilitation. Assessors will rate the risk of bias as low, high or unclear for each domain.

Measures of treatment effect

We will present risk ratios (RRs) with 95% confidence intervals (CIs) for dichotomous outcomes. We will present mean differences (MDs) with 95% CIs for continuous outcomes for the results of individual trials or for pooled data where the same outcome measure is used. Otherwise, we will present standardised mean differences (SMDs) with 95% CIs for outcomes measured using different scales. Results based on change scores will be collected only if final values are not available.

Unit of analysis issues

Bilateral involvement of the ACL is rare and the use of cluster randomisation for these trials is unlikely. Thus, we anticipate that the units of randomisation and analysis will be the individual participant in the included studies. Should either situation arise, where possible we will make appropriate adjustments to the analyses according to guidance provided in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011) or perform sensitivity analysis to explore the impact of incorrectly analysed data. We will avoid unit of analysis issues related to repeated observations of the same outcome, such as results presented for several periods of follow-up.

Dealing with missing data

We will attempt to contact trial investigators for any key missing or unclear data or information on their trial. To avoid the risk of overly positive answers, open-ended questions will be asked (e.g. "Please describe all measures used") followed up by more focused questions if further clarification is required. Where possible, we will aim to conduct intention-to-treat analysis but will base our primary analysis on the available data. We will conduct sensitivity analyses to explore the effects of missing data and inclusion of 'per-protocol' data, should only these be available. If standard deviations are not reported for continuous outcomes, we will calculate these from standard errors, confidence intervals or exact P values where possible. We will not impute missing standard deviations.

Assessment of heterogeneity

The decision about whether or not to combine the results of individual studies will depend on an assessment of clinical and methodological heterogeneity. If studies are considered sufficiently homogeneous in their study design, we will carry out a meta-analysis and assess the statistical heterogeneity. Statistical heterogeneity of treatment effects between trials will be assessed by using a Chi² test with a significance level at P < 0.1 and the I² statistic. We will base our interpretation of the I² results on that suggested by Higgins 2011: 0% to 40% might not be important; 30% to 60% may represent moderate heterogeneity; 50% to 90% may represent substantial heterogeneity; and 75% to 100% may represent very substantial ('considerable') heterogeneity.

Assessment of reporting biases

If there are more than 10 studies included the meta-analysis, we will explore potential publication bias by generating a funnel plot and statistically test this using a linear regression test. A probability (P) value of less than 0.1 could be an indication of a publication bias or small study effects.

Data synthesis

When considered appropriate, we will pool results of comparable groups of trials using both fixed-effect and random-effects models. The choice of the model to report will be guided by careful consideration of the extent of heterogeneity and whether it can be explained, in addition to other factors, such as the number and size of included studies. Ninety-five per cent confidence intervals will be used throughout. We will consider not pooling data where there is considerable heterogeneity (I² > 75%) that cannot be explained by the diversity of methodological or clinical features among trials. Where it is inappropriate to pool data, we will still present trial data in the analyses or tables for illustrative purposes and will report these in the text.

If included, we will analyse cluster randomised trials using the generic inverse variance method.

Subgroup analysis and investigation of heterogeneity

We will perform the following subgroup analyses where sufficient data are available.

  • Age < 18 years, 18 to 30 years, > 30 years.

  • Sex.

  • Type of graft used (hamstring autograft, bone-patella-bone autograft, allograft constructs or synthetic graft).

  • Time from index injury to entry to trial (immediate (up to 10 weeks) versus later; acute (less than six months) versus chronic (over six months)).

  • Participants with and without meniscal injury.

We will investigate whether the results of subgroups are significantly different by inspecting the overlap of confidence intervals and by performing the test for subgroup differences available in Review Manager (RevMan 2012).

Sensitivity analysis

Where possible, we will assess the robustness of our findings by conducting the sensitivity analyses. These will include examining the effects of: a) missing or inappropriately analysed data, such as trials including participants treated for bilateral ACL injury; b) including trials at high or unclear risk of selection bias from inadequate concealment of allocation; c) including trials with mixed population groups (such as collateral ligament injuries); d) including trials with incomplete description of the diagnosis of the ACL injury; and e) the choice of statistical model for pooling (fixed-effect versus random-effects).

Assessing the quality of the evidence

We will use the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach to assess the quality of the body of evidence (Schünemann 2011) for each outcome listed in Types of outcome measures. The quality rating 'high' is reserved for a body of evidence based on randomised controlled trials. We may ‘downgrade’ the quality rating to 'moderate', 'low' or 'very low' depending on the presence and extent of five factors: study limitations, inconsistency of effect, imprecision, indirectness or publication bias.

'Summary of findings' table

Where data are sufficient, we will present the results and the quality assessments for the main comparison and the first seven outcomes in 'Summary of findings' tables (Schünemann 2011).


Editorial staff of the Cochrane Bone, Joint and Muscle Trauma Group (Joanne Elliott and Laura MacDonald), and editors (Helen Handoll and Haris Vasiliadis) and the external referee (Nikolaos Paschos).


Appendix 1. Search strategies

MEDLINE (Ovid Online)

1 Anterior Cruciate Ligament/ or Anterior Cruciate Ligament Reconstruction/
2 ((anterior adj2 cruciate* adj2 ligament*) or acl).tw.
3 or/1-2
4 (surg* or operat* or reconstruct* or repair* or graft*).tw.
5 (non-surg* or nonsurg* or non-operat* or nonoperat* or conserv* or rehab* or physiotherapy or physical therapy).tw.
6 and/3,4,5
7 Randomized controlled
8 Controlled clinical
9 randomized.ab.
10 placebo.ab.
11 Clinical trials as topic/
12 randomly.ab.
13 trial.ti.
14 or/7-13
15 exp Animals/ not Humans/
16 14 not 15
17 6 and 16

Contributions of authors

Andrew P Monk: lead author, writing of manuscript, co-ordination of team.
Sally Hopewell: development of review and statistical expertise.
Kristina Harris: statistical expertise.
Loretta J Davies: physiotherapy treatment expertise.
David Beard: methodological expertise.
Andrew Price: clinical expertise and development of protocol.

Declarations of interest

Andrew P Monk: none known.
Sally Hopewell: none known.
Kristina Harris: none known.
Loretta J Davies: none known.
David Beard: none known.
Andrew Price: none known.