Monk_et_al-2016-Cochrane_Database_of_Systematic_Reviews

Cochrane
Library
Cochrane Database of Systematic Reviews
Surgical versus conservative interventions for treating anterior

cruciate ligament injuries (Review)
Monk AP, Davies LJ, Hopewell S, Harris K, Beard DJ, Price AJ
Monk AP, Davies LJ, Hopewell S, Harris K, Beard DJ, Price AJ.
Surgical versus conservative interventions for treating anterior cruciate ligament injuries.
Cochrane Database of Systematic Reviews 2016, Issue 4. Art. No.: CD011166.
DOI: 10.1002/14651858.CD011166.pub2.
www.cochranelibrary.com
Surgical versus conservative interventions for treating anterior cruciate ligament injuries (Review)
Copyright © 2016 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
Cochrane Trusted evidence.
Informed decisions.
Library Better health. Cochrane Database of Systematic Reviews
TABLE OF CONTENTS
ABSTRACT..................................................................................................................................................................................................... 1
PLAIN LANGUAGE SUMMARY....................................................................................................................................................................... 2
SUMMARY OF FINDINGS.............................................................................................................................................................................. 4
BACKGROUND.............................................................................................................................................................................................. 7
OBJECTIVES.................................................................................................................................................................................................. 8
METHODS..................................................................................................................................................................................................... 8
RESULTS........................................................................................................................................................................................................ 10
Figure 1.................................................................................................................................................................................................. 11
Figure 2.................................................................................................................................................................................................. 13
DISCUSSION.................................................................................................................................................................................................. 15
AUTHORS' CONCLUSIONS........................................................................................................................................................................... 16
ACKNOWLEDGEMENTS................................................................................................................................................................................ 16
REFERENCES................................................................................................................................................................................................ 17
CHARACTERISTICS OF STUDIES.................................................................................................................................................................. 19
DATA AND ANALYSES.................................................................................................................................................................................... 24
Analysis 1.1. Comparison 1 ACL reconstruction versus conservative treatment, Outcome 1 Patient-rated knee function (KOOS-4 26
score)......................................................................................................................................................................................................
Analysis 1.2. Comparison 1 ACL reconstruction versus conservative treatment, Outcome 2 Serious adverse events relating to 26
the index knee at 2 years.....................................................................................................................................................................
Analysis 1.3. Comparison 1 ACL reconstruction versus conservative treatment, Outcome 3 Treatment failure (graft rupture or 27
ACL reconstruction)...............................................................................................................................................................................
Analysis 1.4. Comparison 1 ACL reconstruction versus conservative treatment, Outcome 4 Meniscal surgery.............................. 27
Analysis 1.5. Comparison 1 ACL reconstruction versus conservative treatment, Outcome 5 General health-related quality of life 27
(SF-36 Physical and Mental scores).....................................................................................................................................................
Analysis 1.6. Comparison 1 ACL reconstruction versus conservative treatment, Outcome 6 Return to previous activity level (pre- 27
injury Tegner activity scale level)........................................................................................................................................................
Analysis 1.7. Comparison 1 ACL reconstruction versus conservative treatment, Outcome 7 Knee stability (KT-1000 test) at 2 28
years (mm).............................................................................................................................................................................................
Analysis 1.8. Comparison 1 ACL reconstruction versus conservative treatment, Outcome 8 Knee stability (normal pivot shift or 28
Lachman tests)......................................................................................................................................................................................
APPENDICES................................................................................................................................................................................................. 28
CONTRIBUTIONS OF AUTHORS................................................................................................................................................................... 30
DECLARATIONS OF INTEREST..................................................................................................................................................................... 30
DIFFERENCES BETWEEN PROTOCOL AND REVIEW.................................................................................................................................... 30
INDEX TERMS............................................................................................................................................................................................... 30
Surgical versus conservative interventions for treating anterior cruciate ligament injuries (Review) i
Informed decisions.
[Intervention Review]
Surgical versus conservative interventions for treating anterior cruciate

ligament injuries
A Paul Monk1, Loretta J Davies1, Sally Hopewell2, Kristina Harris3, David J Beard1, Andrew J Price1
1Nuffield Department of Orthopaedics, Rheumatology and Musculoskeletal Sciences, University of Oxford, Oxford, UK. 2Oxford Clinical
Trials Research Unit, University of Oxford, Oxford, UK. 3The Botnar Research Centre Institute of Musculoskeletal Sciences, University of
Oxford, Oxford, UK
Contact: A Paul Monk, Nuffield Department of Orthopaedics, Rheumatology and Musculoskeletal Sciences, University of Oxford,
Windmill Road, Headington, Oxford, OX3 7LD, UK. [email protected].
Editorial group: Cochrane Bone, Joint and Muscle Trauma Group.

Publication status and date: New, published in Issue 4, 2016.
Citation: Monk AP, Davies LJ, Hopewell S, Harris K, Beard DJ, Price AJ. Surgical versus conservative interventions for
treating anterior cruciate ligament injuries. Cochrane Database of Systematic Reviews 2016, Issue 4. Art. No.: CD011166. DOI:
10.1002/14651858.CD011166.pub2.
ABSTRACT
Background
Rupture of the anterior cruciate ligament (ACL) is a common injury, mainly affecting young, physically active individuals. The injury is
characterised by joint instability, leading to decreased activity, which can lead to poor knee-related quality of life. It is also associated with
increased risk of secondary osteoarthritis of the knee. It is unclear whether stabilising the knee surgically via ACL reconstruction produces
a better overall outcome than non-surgical (conservative) treatment.
Objectives
To assess the effects of surgical versus conservative interventions for treating ACL injuries.
Search methods
We searched the Cochrane Bone, Joint and Muscle Trauma Group Specialised Register (18 January 2016), the Cochrane Central Register of
Controlled Trials (2016, Issue 1), MEDLINE (1946 to January Week 1 2016), MEDLINE In-Process & Other Non-Indexed Citations (18 January
2016), EMBASE (1974 to 15 January 2016), trial registers (February 2016) and reference lists.
Selection criteria
We included randomised controlled trials that compared the use of surgical and conservative interventions in participants with an ACL
rupture. We included any trial that evaluated surgery for ACL reconstruction using any method of reconstruction, type of reconstruction
technique, graft fixation or type of graft.
Data collection and analysis

Three review authors independently screened all titles and abstracts for potentially eligible studies, for which we then obtained full-text
reports. Two authors then independently confirmed eligibility, extracted data and assessed the risk of bias using the Cochrane 'Risk of
bias' tool. We used the GRADE approach to assess the overall quality of the evidence.
Main results
We identified one study in which 141 young, active adults with acute ACL injury were randomised to either ACL reconstruction followed
by structured rehabilitation (results reported for 62 participants) or conservative treatment comprising structured rehabilitation alone
Surgical versus conservative interventions for treating anterior cruciate ligament injuries (Review) 1
Informed decisions.
(results reported for 59 participants). Built into the study design was a formal option for subsequent (delayed) ACL reconstruction in the
conservative treatment group, if the participant requested surgery and met pre-specified criteria.
This study was deemed at low risk of selection and reporting biases, at high risk of performance and detection biases because of the
lack of blinding and at unclear risk of attrition bias because of an imbalance in the post-randomisation exclusions. According to GRADE
methodology, the overall quality of the evidence was low across different outcomes.
This study identified no difference in subjective knee score (measured using the average score on four of the five sub-scales of the KOOS
score (range from 0 (extreme symptoms) to 100 (no symptoms)) between ACL reconstruction and conservative treatment at two years
(difference in KOOS-4 change from baseline scores: MD -0.20, 95% confidence interval (CI) -6.78 to 6.38; N = 121 participants; low-quality
evidence), or at five years (difference in KOOS-4 final scores: MD -2.0, 95% CI -8.27 to 4.27; N = 120 participants; low-quality evidence).
The total number of participants incurring one or more complications in each group was not reported; serious events reported in the
surgery group were predominantly surgery-related, while those in conservative treatment group were predominantly knee instability.
There were also incomplete data for total participants with treatment failure, including subsequent surgery. In the surgical group at two
years, there was low-quality evidence of far fewer ACL-related treatment failures, when defined as either graft rupture or subsequent ACL
reconstruction. This result is dominated by the uptake by 39% (23/59) of the participants in the conservative treatment group of ACL
reconstruction for knee instability at two years and by 51% (30/59) of the participants at five years. There was low-quality evidence of little
difference between the two groups in participants who had undergone meniscal surgery at anytime up to five years. There was low-quality
evidence of no clinically important between-group differences in SF-36 physical component scores at two years. There was low-quality
evidence of a higher return to the same or greater level of sport activity at two years in the ACL reconstruction group, but the wide 95%
CI also included the potential for a higher return in the conservative treatment group. Based on an illustrative return to sport activities of
382 per 1000 conservatively treated patients, this amounts to an extra 84 returns per 1000 ACL-reconstruction patients (95% CI 84 fewer
to 348 more). There was very low-quality evidence of a higher incidence of radiographically-detected osteoarthritis in the surgery group
(19/58 (35%) versus 10/55 (18%)).
Authors' conclusions
For adults with acute ACL injuries, we found low-quality evidence that there was no difference between surgical management (ACL
reconstruction followed by structured rehabilitation) and conservative treatment (structured rehabilitation only) in patient-reported
outcomes of knee function at two and five years after injury. However, these findings need to be viewed in the context that many
participants with an ACL rupture remained symptomatic following rehabilitation and later opted for ACL reconstruction surgery. Further
research, including the two identified ongoing trials, will help to address the limitations in the current evidence, which is from one small
trial in a young, active, adult population.
PLAIN LANGUAGE SUMMARY
Surgical versus conservative interventions for treating anterior cruciate ligament injuries
Background
Rupture of the anterior cruciate ligament (ACL) in the knee is a common injury in young, active individuals. It often results in an unstable
knee that increases the risk of further knee damage, such as to the knee meniscii. Anterior cruciate ligament injuries in athletic individuals
are often treated surgically. Surgery usually entails ACL reconstruction, that involves removing the torn ligament and replacing it with a
tendon graft, often taken from another part of the patient's knee. Conservative (non-surgical) interventions are also used as treatment for
this injury. This usually takes the form of a progressive rehabilitation programme that includes exercises aimed at improving strength and
balance. We aimed to assess the effects of surgical versus conservative interventions for treating ACL injuries.
Results of the search
We performed a systematic literature search (up to 18 January 2016) for studies that compared surgery and conservative interventions for
treating ACL injuries. This review identified one study of 121 young, active adults with an ACL injury in the preceding four weeks. The study
compared surgery (ACL reconstruction followed by structured rehabilitation) with conservative treatment (structured rehabilitation alone).
Key results
The study found there was no difference between surgery and conservative treatment in patient-reported knee scores at two or five
years. The study failed to report on the number of participants in each group who had any type of serious or non-serious complications.
However, surgery-related complications included three cases of graft rupture in the surgery group and several participants of the
conservative treatment group had unstable knees. Twenty-three of the 59 participants in the conservative treatment group (39%) had
either reconstruction of the ACL or repair of a meniscus tear within two years and 30 (51%) underwent surgery within five years. There
was some evidence that similar numbers of participants in the two groups had surgical treatment of knee meniscal injuries at five years.
There was very low-quality evidence that more participants in the surgery group had damage to the knee that could mean that they were
at greater risk of developing osteoarthritis.
Informed decisions.
Quality of the evidence
The quality of the evidence was limited by the availability of data from only one study. The study was also at high risk of bias because
the clinicians and participants were not blinded to their treatment. Overall, the quality of the evidence was low, which means that we are
uncertain of the study findings and further research may provide evidence that could change our conclusions.
Conclusions
In young, active adults being treated for acute ACL injury, we found no difference between surgery and conservative treatment in patient-
reported outcomes of knee function at two and five years. However, many participants with an ACL rupture had unstable knees after
structured rehabilitation and opted to have surgery later on.
SUMMARY OF FINDINGS
Summary of findings for the main comparison.
Library
Cochrane
ACL reconstruction followed by structured rehabilitation compared with conservative treatment comprising structured rehabilitation alone for treating anterior
cruciate ligament (ACL) injuries
Patient or population: Adults with acute anterior cruciate ligament (ACL) injuries (young, active individuals)
Better health.
Informed decisions.
Trusted evidence.
Settings: Presentation to hospital emergency department
Intervention: ACL reconstruction (within 10 weeks of injury) followed by structured rehabilitation
Comparison: Conservative treatment (comprising structured rehabilitation)1
Outcomes Illustrative comparative risks* (95% CI) Relative effect No of Partici- Quality of the Comments
(95% CI) pants evidence
Assumed risk Corresponding risk (studies) (GRADE)
Conservative ACL reconstruction2

treatment1
Subjective knee score - The mean change The mean change 121 ⊕⊕⊝⊝ Knee injury and Osteoarthritis Out-
KOOS-4 score (mean score in KOOS-4 score from baseline in (1 study) low3 come Score (KOOS).
of 4 components) at 2 from baseline in KOOS-4 score in the
years: score from 0 (ex- the conservative ACL reconstruction A similar lack of statistical and clini-
treme symptoms) to 100 treatment group group was 0.2 low- cally important difference was found
(no symptoms) was 39.4 er (6.78 lower to 6.38 in the final KOOS-4 scores at five
higher) years (MD -2.00, 95% CI -8.27 to 4.27)
Serious adverse events Not estimable4 121 Total participants with adverse
(such as donor site mor- (1 study) events not reported.
bidity, failure of graft in-

cluding re-rupture) at 2
years
Treatment failure (graft 390 per 1000 47 per 1000 RR 0.12 121 (1 study) ⊕⊕⊝⊝ Total participants with treatment
rupture or ACL reconstruc- low6 failure (including re-operation) not
tion) at 2 years5 (16 to 133) (0.04 to 0.34) reported.
In terms of applicability, considera-

tion should be given to the study de-
sign of the trial with its inbuilt option
of delayed ACL reconstruction for in-
stability.
4
Treatment failure (menis- 542 per 1000 477 per 1000 RR 0.88 (0.62 to 120 (1 study) ⊕⊕⊝⊝ Meniscal surgery may or may not be
cal surgery) at 5 years5,7 (11 to 678) 1.25) low8 carried out during the same opera-
tion as ACL reconstruction.
Library
Cochrane
General health-related The mean SF-36 The mean SF-36 121 ⊕⊕⊝⊝
quality of life: SF-36 phys- physical com- physical component (1 study) low3
ical component score ponent score in score in the ACL re-
(range 0 to 100; higher the conserva- construction group
score = better health state) tive group of the was 4.1 higher (2.76
Better health.
Informed decisions.
Trusted evidence.
at 2 years study was 78.0 lower to 10.96 high-
er)
Return to activity or level 356 per 1000 435 per 1000 RR 1.22 121 ⊕⊕⊝⊝ There was also no difference in Tegn-
of sport at 2 years (278 to 680) (0.78 to 1.91) (1 study) low9 er activity scores reported10
Osteoarthritis at 5 years 182 per 1000 328 per 1000 RR 1.80 113 ⊕⊕⊕⊝ Radiographic assessment
(168 to 641) (0.92 to 3.52) (1 study) very low11
*The assumed risk is based on data for the conservative treatment group of the only included study. The corresponding risk (and its 95% confidence interval) is based on
the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; MD: Mean difference
GRADE Working Group grades of evidence

High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.
1 Conservative treatment comprising structured rehabilitation alone. Built into the study design was a formal option for subsequent (delayed) ACL reconstruction, if chosen by
the participant and if they met pre-specified criteria. Surgery (partial resection or fixation) for meniscal injuries when indicated by MRI and clinical findings.
2 ACL reconstruction followed by structured rehabilitation. Surgery (partial resection or fixation) for meniscal injuries when indicated by MRI, clinical and surgical findings.

3 The quality of the evidence was downgraded one level because of risk of bias reflected by lack of blinding of outcome assessment and one level for imprecision because there
was just one small trial.
4 Total numbers of participants incurring serious or non-serious adverse events were not reported. In addition, low frequency events (< 5% overall or < 3% in a treatment group)
were not included. The largest groups of listed complications relating to the index knee were 'subjective or clinical instability' (2/62 versus 19/59) and 'meniscal signs and
symptoms' (2/62 versus 13/59); both of these were less frequent in the ACL reconstruction group. Other complications including non-serious adverse events and mainly surgery-
related, were more common in the ACL reconstruction group; e.g. graft rupture (3/62 versus 1/59).
5 Definition of treatment failure split into two, both are knee stability-related. The first links with ACL surgery and the second with meniscal surgery. Of the 3 participants in the ACL
reconstruction group with ACL graft rupture, 2 underwent revision ACL reconstruction. At two years, 23/59 participants in the conservative group underwent ACL reconstruction;
at 5 years, 30/59 had undergone ACL reconstruction.
was just one small trial. We have not downgraded for indirectness, but these results need to be considered in the context of participants in the conservative treatment group
who had the option of ACL reconstruction for instability.
5
7 Surgery for meniscal injuries was resection or fixation. There was a) initial surgery, when indicated by MRI findings and clinical signs in both groups, and during surgery in the
ACL reconstruction group; and b) subsequently during follow-up. Only data for the number of participants having meniscal surgery by 5 years were available. Fifteen participants
of the conservative treatment group had meniscal surgery only.
Library
Cochrane
was just one small trial. We have not downgraded for indirectness, but these results need to be considered in the context of the potential for knee damage to the other knee
and decisions for surgery.
was just one small trial. We have not downgraded for indirectness, but these results are based on the Tegner activity scale, which may not quite represent what actually happened
in practice.
10 There was no difference in return to activity level at two years (measured using the Tegner Activity Scale) between ACL reconstruction (median 6.5; IQR 3 to 8) and conservative
Better health.
Informed decisions.
Trusted evidence.
treatment (median 5; IQR 4 to 7).
11 The quality of the evidence was downgraded one level because of risk of bias reflected by lack of blinding of outcome assessment, one level for imprecision because there was
just one small trial and one level for indirectness, given the uncertain link between symptomatic and radiographic OA.

6
Informed decisions.
BACKGROUND Three types of grafts are commonly used: those from the
patient's body (autograft), cadaveric human donors (allograft)
Description of the condition or a synthetic ligament substitute. Commonly, the hamstring
tendons of semitendinosus and gracilis are harvested from the
The anterior cruciate ligament (ACL) of the knee joint plays an
limb of the ruptured ACL; this graft is removed during the
essential role in both static (standing or squatting) and dynamic
reconstruction operation. Alternatively, a bone-patella tendon-
(walking or running) knee stability. Primarily, the ACL prevents
bone (BPTB) construct uses a section of the middle of the patella
anterior translation (forward movement) of the tibia relative to the
(kneecap) tendon with bone at either end. The relative merits of
femur in the sagittal (antero-posterior) plane, aiding stabilisation
hamstring and BPTB grafts have been reviewed (Mohtadi 2011).
of the joint from a flexed (bent) to a near full extension (straight)
position of the knee. Conservative (non-surgical) treatment for people with an ACL
rupture can include the use of cryotherapy (ice), continuous
Rupture of the ACL is a common injury, mainly affecting young,
passive motion (movement of the joint by a machine), restrictive
physically active individuals, with an estimated 200,000 ACL
bracing, electrotherapy (muscle stimulation) and exercises aimed
ruptures per year in the United States (Spindler 2008). The ACL
at strengthening and balance. The use of plaster casts for initial
is often injured during sporting activities, such as football, skiing
immobilisation of the knee is rare nowadays (Linko 2009).
and basketball (Bahr 2003). In over 70% of cases, the injury is
caused by a non-contact mechanism, such as sudden deceleration Rehabilitation regimens used for both treatment options
combined with changing direction or pivoting or landing with the commonly use a three-stage progressive programme: acute,
knee in nearly full extension after a jump (Hernandez 2006). Contact recovery and functional phases (Micheo 2010). The acute stage
(traumatic) mechanisms of injury usually involve a translational following injury, or immediately after surgery, aims to restore range
force applied to the anterior aspect of a fixed lower leg (Hewett of motion and resolve inflammation. The recovery phase is from
2006). The acute injury is frequently characterised by knee pain and approximately three to six weeks, with the aim of improving lower
an audible ‘popping’ sound at the time of injury. The injured person limb muscle strength and functional stability. Finally, the functional
presents with knee pain, swelling, haemarthrosis (bleeding into the stage of rehabilitation (from six weeks onwards) concentrates on
joint space), instability with further activity and painful range of returning the individual to previous levels of activity and decreasing
motion. the risk of re-injury (Kvist 2004). There is little consensus over
the most effective rehabilitation protocol for achieving these aims
In approximately 10% of cases, the ACL injury occurs in isolation.
(Negus 2012).
In the majority of cases, however, it is combined with other
injuries, typically to the collateral ligaments, subchondral bone and Whilst surgical interventions have become commonplace
meniscii (Bowers 2005; Hernandez 2006; Miyasaka 1991). for athletic individuals, initial non-operative (conservative)
treatments, based on physiotherapy, are used more commonly in
Diagnostic imaging, including magnetic resonance imaging (MRI),
the general population (Linko 2009).
is used to confirm the diagnosis of ACL injury or rupture, and
evaluate associated pathology such as articular cartilage injury, How the intervention might work
and meniscal and associated ligamentous tears; all of which play a
role in maintaining stability of the knee (Crawford 2007). All treatments aim to reduce knee pain and instability and restore
function. Reconstructive surgery aims to restore stability to the
Chronic ACL injury can have a profound effect on the knee knee by replacing the torn ACL. In comparison, conservative
kinematics (movements) of those affected. Common problems treatments, such as rehabilitation, aim to improve the muscle
include recurrent knee instability (giving way) and symptoms function around the knee and to substitute the function of
of associated meniscal or articular cartilage damage, such as the missing ACL. However, the optimal management strategy
intermittent swelling or a locking sensation (Hernandez 2006). following rupture of the ACL remains controversial. In the short
Furthermore, the injury can lead to poor reported quality of life term, reconstructive surgery may improve knee function for
(Spindler 2008), and decreased activity levels (Thorstensson 2009). those experiencing severe instability in activity or repeated
It is also associated with increased risk of secondary osteoarthritis episodes of ‘giving way’, or both. However, all surgery involves an
of the knee, irrespective of treatment (Øiestad 2009; Rout 2013). increased risk of complications, such as infection. In particular,
These related morbidities have been shown to be associated with for reconstruction using autograft, significant donor site morbidity
high healthcare expenditure (Frobell 2010a). can occur, including anterior knee pain with BPTB grafts and pain
and weakness of knee flexors with hamstring grafts (Mohtadi 2011;
Description of the intervention Spindler 2004).
Surgical treatment for ACL rupture has evolved from simple repair
Although studies of conservative treatment have demonstrated
using suturing or suturing with some sort of augmentation to ACL
satisfactory results with patients returning to pre-injury activity
reconstruction, which involves reconstruction of the ligament using
level (Frobell 2013; Kostogiannis 2007; Linko 2009;), the long-
a substitute graft of tendon or ligament, fixed into position in pre-
term results, in particular relating to the development of early
prepared drill holes. ACL reconstruction is increasingly performed
onset osteoarthritis, are still debatable. Controversy exists about
as an arthroscopic procedure. Of those who undergo surgical
ongoing instability and the possibility of secondary joint damage
reconstruction, 94% are performed within one year of the initial
and early osteoarthritis (Smith 2014). Radiographically diagnosed
injury (Collins 2013). Anterior cruciate ligament reconstruction
osteoarthritis has been reported in 20% to 50% of ACL-deficient
is the predominant method of surgery in current practice and
knees at 10 years post injury compared with 5% in uninjured knees
hundreds of thousands of these operations are carried out each
(Ajuied 2013; Lohmander 2007; Øiestad 2009). However, surgery has
year.
Informed decisions.
not been shown to offer protection against long-term degenerative reconstruction technique (e.g. single or double bundle) or graft
change (Øiestad 2009; Rout 2013). Moreover, recent studies have fixation and any type of graft. Direct repair of the ACL is increasingly
suggested structured neuromuscular rehabilitation might provide rare, so we only planned to include this if we found any new,
effective recovery following ACL rupture without increasing the recently conducted trials (that is, since 2000). Otherwise, we did
risks of long-term degenerative change (Delincé 2012). not repeat the analyses provided in Linko 2009. We included any
method of conservative treatment; which was likely to include
Why it is important to do this review bracing, physiotherapy, or both.
The management of ACL injuries includes both reconstructive Types of outcome measures
surgery and conservative treatments. It is unclear whether
stabilising the knee surgically produces any benefit for the knee Primary outcomes
compared with conservative intervention. The previous Cochrane
• Subjective knee scores, e.g. Knee Injury and Osteoarthritis
review in this area found that there was insufficient evidence Outcome Score (KOOS; Roos 1998), anterior cruciate ligament
from two trials to determine whether surgery, involving repair, quality of life score (Mohtadi 1998) and International Knee
or conservative management was superior for the treatment Documentation Committee (IKDC), subjective part (Irrgang
of ACL rupture, and highlighted the need for good quality 2001).
randomised controlled trials (RCTs) of current practice, particularly
ACL reconstruction (Linko 2009). Surgical practice has also changed • Adverse events (such as donor site morbidity, failure of graft
in terms of the population, with an increasing number of ACL including re-rupture, infection, deep vein thrombosis and
reconstructions being performed on a young athletic (adolescent) pulmonary embolism).
cohort (Ramski 2013). These point to the need for a systematic • Treatment failure including re-operation (for surgery) or
review of the evidence from randomised trials that compare the subsequent operation (for conservative treatment).
effects of current surgical and conservative treatment methods for
Secondary outcomes
ACL rupture.
• General health-related quality of life, preferably measured using
OBJECTIVES validated scales such as SF-36 (Ware 1992), and EQ-5D (Brooks
1996).
To assess the effects (benefits and harms) of surgical versus
conservative interventions for treating anterior cruciate ligament • Return to activity or level of sports participation, including
injuries. Lysholm (Lysholm 1982), or Tegner (Tegner 1985) scores.
• Functional assessments (e.g. single-hop test).
METHODS • Composite clinical examination outcomes (IKDC, objective part;
Hefti 1993).
Criteria for considering studies for this review • Knee stability (assessed using manual methods (e.g. Lachman
Types of studies or pivot shift tests) or using knee ligament testing devices (e.g.
KT 1000)).
We included randomised controlled trials that compared surgical
• Objective measure of muscle strength (isokinetic muscle
and conservative interventions for treating anterior cruciate
torque).
ligament injuries.
Resource and economic outcomes
Types of participants
Resource and economic outcomes, such as those that measured
We included participants of any age (thus, including children) with service utilisation, including cost of surgery, length of inpatient
anterior cruciate ligament rupture. Ideally, diagnosis had been stay, outpatient attendance, duration of sick leave.
made with positive clinical examination and either a positive MRI
or a positive examination under anaesthesia (EUA). Timing of outcome measurement
We excluded studies whose prime focus was on the management Assessments were made at short- (less than one year),
of ACL and a concomitant knee ligament rupture (e.g. medial intermediate- (one to three years) and long-term (greater than
collateral ligament). We excluded people with inflammatory three years) follow-up, where data were available.
arthropathy or end stage osteoarthritis (Grade 4 Kellgren and
Lawrence). However, if identified, we would have included Search methods for identification of studies
mixed population trials if they included only a small proportion Electronic searches
(preferably less than 10% and preferably balanced between
intervention groups) of participants with other major knee We searched the Cochrane Bone, Joint and Muscle Trauma Group
ligament or cartilage lesions, or if separate data were provided for Specialised Register (18 January 2016), the Cochrane Central
participants without these additional injuries. Register of Controlled Trials (CENTRAL; 2016, Issue 1), MEDLINE
(1946 to January Week 1 2016), MEDLINE In-Process & Other
Types of interventions Non-Indexed Citations (18 January 2016) and EMBASE (1974 to
15 January 2016), using tailored search strategies. In MEDLINE,
The interventions being compared were surgery and conservative
we combined subject-specific terms with the sensitivity- and
treatment for ACL rupture. We included any trial that evaluated
precision-maximising version of the Cochrane Highly Sensitive
surgery that involved ACL reconstruction. Thus, we included any
Search Strategy for identifying randomised trials (Lefebvre 2011;
method of reconstruction (e.g. open or arthroscopic), any type of
see Appendix 1). The search strategies for CENTRAL and EMBASE
Informed decisions.
are also shown in Appendix 1. We also searched the World Health situation have arisen, we had planned, where possible, to make
Organization International Clinical Trials Registry Platform (WHO appropriate adjustments to the analyses according to guidance
ICTRP) and ClinicalTrials.gov for ongoing and recently completed provided in the Cochrane Handbook for Systematic Reviews of
trials (4 February 2016). We did not apply any language or Interventions (Higgins 2011), or to perform sensitivity analyses to
publication status restrictions. explore the impact of incorrectly analysed data. One unit of analysis
issue that arose was in the reporting of adverse events, whereby
Searching other resources data were presented for the total complications rather than the
We checked the reference lists of relevant articles and contacted total number of participants with a complication. Thus, we were
individuals and organisations for further data where necessary. only able to present descriptive data on the frequency of individual
adverse events, such as graft rupture.
Data collection and analysis
Dealing with missing data
Selection of studies
If necessary, we had planned to contact trial investigators for
All review authors, divided into two teams (PM, LD, DB and SH, KH, any key missing or unclear data or information on their trial. To
AP), independently screened all titles and abstracts for potentially avoid the risk of overly positive answers, we had planned to ask
eligible studies, for which we then obtained full-text reports. The open-ended questions (e.g. "Please describe all measures used"),
same teams then independently performed study selection. Any followed up by more focused questions if further clarification was
disagreements regarding the inclusion or exclusion of individual required. If standard deviations were not reported for continuous
studies were resolved by discussion among the review team. outcomes, we had planned to calculate these from standard errors,
confidence intervals or exact P values where possible, using the
Data extraction and management inbuilt calculator in the Review Manager software. We did not
We developed a data collection form to include all relevant impute missing standard deviations. Where possible, we aimed
variables for the study, including details of methods, participants, to conduct intention-to-treat analyses but decided that we would
setting, interventions, outcomes, results and funding sources. Two base our primary analysis on the available data. If data had
review authors (PM and LD) piloted the data collection form in been available, we had planned to conduct sensitivity analyses to
order to identify any missing or unclear items. After finalising the explore the effects of missing data and inclusion of 'per-protocol'
form, the same two review authors independently performed data data.
extraction. Any disagreement was resolved by discussion between
Assessment of heterogeneity
the two authors and, in cases where no consensus was achieved, a
third review author (SH) acted as an arbitrator. Due to a lack of data, it was not possible to perform a meta-
analysis or assess heterogeneity across studies. If data had been
Assessment of risk of bias in included studies available, the decision about whether or not to combine the results
Two review authors (PM and LD) independently assessed the risk of individual studies would have depended on an assessment of
of bias in each included study using the Cochrane 'Risk of bias' clinical and methodological heterogeneity. Where studies were
tool. Any disagreement was resolved by consensus between the considered sufficiently homogeneous in their study design, we had
two review authors, and in cases where no consensus was achieved, planned to carry out a meta-analysis and assess the statistical
a third review author (SH) acted as an arbitrator. We assessed heterogeneity. We had planned to assess statistical heterogeneity
the risk of bias for the following domains: sequence generation, of treatment effects between trials using a Chi2 test, with a
allocation concealment, blinding of participants and personnel, significance level at P less than 0.1, and the I2 statistic. We
blinding of outcome assessment, incomplete outcome data and had planned to base our interpretation of the I2 results on that
selective outcome reporting, as well as other sources of bias, such suggested by Higgins 2011: 0% to 40% might not be important;
as major differences in rehabilitation. Assessors rated the risk of 30% to 60% may represent moderate heterogeneity; 50% to 90%
bias as low, high or unclear for each domain. may represent substantial heterogeneity; and 75% to 100% may
represent very substantial ('considerable') heterogeneity.
Measures of treatment effect
Assessment of reporting biases
We calculated the risk ratios (RRs) with 95% confidence intervals
(CIs) for dichotomous outcomes and presented the mean If there had been more than 10 studies included the meta-analysis,
differences (MDs) with 95% CIs for continuous outcomes. If data had we had planned to explore potential publication bias by generating
been available, we had planned to use the mean differences (MDs) a funnel plot and statistically testing this using a linear regression
with 95% CIs to pool the results of individual trials of continuous test. A probability (P) value of less than 0.1 could be an indication
outcomes where the same outcome measure was used, or the of a publication bias or small study effects.
standardised mean differences (SMDs) with 95% CIs for outcomes
Data synthesis
measured using different scales. We used results based on change
scores where final values were not available. We had planned, but did not perform, a meta-analysis, due to a lack
of available data. Therefore, we present data for our primary and
Unit of analysis issues secondary outcomes in the analyses for illustrative purposes and
Bilateral involvement of the ACL is rare and the use of cluster report the findings descriptively in the text.
randomisation for these trials was unlikely. Thus, we anticipated
When considered appropriate (e.g. in a future update of this review,
that the units of randomisation and analysis would be the
and if more data become available), we would pool results of
individual participant in the included studies. Should either
comparable groups of trials using both fixed-effect and random-
Informed decisions.
effects models. The choice of the model would be guided by Assessing the quality of the evidence
careful consideration of the extent of heterogeneity and whether
We used the Grading of Recommendations Assessment,
it could be explained by known factors, such as the number and
Development and Evaluation (GRADE) approach to assess the
size of included studies; 95% CIs would be used throughout. We
quality of the body of evidence for each outcome listed in Types of
would consider not pooling data where there was considerable
outcome measures (Schünemann 2011). The quality rating 'high' is
heterogeneity (I2 greater than 75%) that could not be explained by
reserved for a body of evidence based on randomised controlled
the diversity of methodological or clinical features among trials.
trials. We downgraded the quality rating to 'moderate', 'low' or 'very
If included, we would analyse cluster randomised trials using the
low' depending on the presence and extent of five factors: study
generic inverse variance method.
limitations, inconsistency of effect, imprecision, indirectness and
Subgroup analysis and investigation of heterogeneity publication bias.
We had planned, but did not perform because of insufficient data, 'Summary of findings' table
the following subgroup analyses:
We have presented the results and the quality assessment for the
• Age: younger than 18 years, 18 to 30 years, older than 30 years; main comparison, and for the primary outcomes and the top two
secondary outcomes listed in Types of outcome measures in a
• Sex;
'Summary of findings' table (Schünemann 2011). We also included
• Type of graft used (hamstring autograft, bone-patella-bone radiographically-assessed osteoarthritis, which would come under
autograft, allograft constructs or synthetic graft); 'adverse events', a primary outcome.
• Time from index injury to entry to trial (immediate (up to 10
weeks) versus later; acute (less than six months) versus chronic RESULTS
(over six months));
• Participants with and without meniscal injury. Description of studies
Results of the search
We had planned to investigate whether the results of subgroups
were significantly different by inspecting overlapping CIs and by Searches identified a total of 1273 citations from the following
performing the test for subgroup differences available in Review databases: Cochrane Bone, Joint and Muscle Trauma Group
Manager (RevMan 2012). Specialised Register (N = 16); CENTRAL (N = 431), MEDLINE (N =
358) and EMBASE (N = 468). We also searched the WHO ICTRP
Sensitivity analysis (N = 247) and ClinicalTrials.gov (N = 186). Of these, 61 full-text
We had planned, but did not perform, to assess the robustness of reports were obtained for further examination and 49 articles were
our findings by conducting sensitivity analyses. These would have initially excluded for not being randomised. Only one study (Frobell
included examining the effects of: a) missing or inappropriately 2010b), reported in a trial registration document and two separate
analysed data, such as trials including participants treated for publications with different follow-up periods, was deemed eligible
bilateral ACL injury; b) including trials at high or unclear risk for inclusion in this review. We excluded two studies (Andersson
of selection bias from inadequate concealment of allocation; c) 1991; Sandberg 1987) for not meeting the eligibility criteria (surgery
including trials with mixed population groups, such as collateral involved repair rather than reconstruction) and we identified two
ligament injuries; d) including trials with incomplete description of ongoing studies (ACL SNNAP; NTR2746). Full details are reported in
the diagnosis of the ACL injury; and e) the choice of statistical model Figure 1.
for pooling (fixed-effect versus random-effects).
Informed decisions.
Figure 1. Study flow diagram
Informed decisions.
Figure 1. (Continued)
Included studies scores, ranging from 0 (extreme symptoms) to 100 (no symptoms).
See Characteristics of included studies.
We identified one study (Frobell 2010b) in which 141 young, active
adults (aged 18 to 35 years) with an acute (rotational trauma Excluded studies
within the last four weeks) ACL injury were randomised to either
ACL reconstruction followed by structured rehabilitation (N = 69) Two trials were excluded from this review following full-text
or conservative treatment comprising structured rehabilitation eligibility assessment (Andersson 1991; Sandberg 1987). Both trials
alone (N = 72). Built into the study design was a formal option are reviewed in Linko 2009. They were excluded from this review
for subsequent (delayed) ACL reconstruction in the conservative because they were evaluating surgery that involved direct repair of
treatment group, if chosen by the participant and if pre-specified the ACL and not ACL reconstruction: see Characteristics of excluded
criteria were met. The study was carried out at two hospital sites in studies.
Lund, Sweden, recruiting between February 2002 and June 2006.
Ongoing studies
All participants had to have a score of five to nine on the We identified two ongoing randomised controlled trials; see
Tegner activity scale before their injury, and of those followed- Characteristics of ongoing studies. NTR2746 is examining the
up, all but two participants had been participating in sports clinical and cost-effectiveness of early surgery versus conservative
when injured. Anterior cruciate ligament injury was determined management (with the option for delayed surgery) for ACL rupture.
by clinical examination. An MRI was performed at the time of The criteria for inclusion in this study is broader than Frobell 2010b
randomisation; however, the results were not available until several in terms of age (18 to 65 years) and time since injury (trauma within
days later. Twelve participants (four versus eight) were excluded two months of injury). The study aimed to recruit 188 participants
because of MRI findings; of these eight (two versus six) had an and completed recruitment in February 2015. The other study
intact ACL. A further eight (three versus five) were excluded at the is designed to examine the clinical and cost effectiveness of
time of surgery; seven because of extensive meniscal fixation. Thus, surgery versus conservative management (again with the option
121 participants were included in the primary analysis of the trial for delayed surgery) in patients with non-acute (longer than four
(ACL reconstruction (N = 62) and conservative treatment (N = 59)). months) ACL deficiency (ACL SNNAP). The planned sample size is
Operative treatment involved surgical reconstruction using either 320 participants. This study is in set-up phase and due to start
the patellar-tendon or hamstring-tendon procedure. Participants recruitment in July 2016.
were evaluated at two and five years after randomisation. The
primary outcome was the change from baseline at two years in Risk of bias in included studies
subjective knee score, measured using the average score of four
(pain, symptoms, function in sports and recreation, and knee- Information on potential risk of bias for Frobell 2010b is included
related quality of life) of the five individual components of the KOOS in a 'Risk of bias' table (see Characteristics of included studies) and
summarised in Figure 2.
Informed decisions.
Figure 2. Risk of bias summary: review authors' judgements about each 'Risk of bias' item for each included study.
Allocation Incomplete outcome data

The method of random sequence generation and concealment of Twenty (14%) randomised participants were not included in the
the allocation sequence were assessed at low risk of bias. The primary analysis of the study. All 20 were post-randomisation
method used to generate the random sequence was computer exclusions, 12 because of MRI findings and eight at time of surgery,
generated, using block randomisation with a block size of 20, based on their clinical diagnosis. Although the reasons for attrition
and concealed using sequentially numbered, sealed, opaque were clearly reported by the intervention group, there was some
envelopes. imbalance in the losses between the two groups (7/69 = 10% in the
surgery group; 13/72 = 18% in the conservative group) and there is
Blinding some question whether it was appropriate to exclude for findings,
Due to the nature of the interventions it was not possible to usually resulting in extensive meniscal fixation (two versus five),
blind the participants or study personnel. However, because of at surgery. However, noting the comparability between the two
the subjective and self-reported nature of the outcomes being groups in the baseline characteristics of those who remained in the
assessed, we judged this as a potential for high risk of bias. Blinding trial, we judged the trial to be at unclear risk of attrition bias. Only
of outcome assessment was also not performed as assessors were one participant in the ACL reconstruction group was lost to follow-
aware of treatment assignment; we also judged this as a high risk up at five years.
of bias.
Selective reporting
All outcomes pre-specified in the clinical trial register (ISRCTN
84752559) were reported in the results section of the study
Informed decisions.
article(s). Therefore, we judged there was a low risk of reporting 17/59), pain, swelling or both (16/62 versus 14/59) and decreased
bias. range of motion (12/62 versus 2/59).
Other potential sources of bias The overall frequency of adverse events was not given at five
years, except for graft rupture (none had occurred between two and
We did not identify other potential sources of bias.
five years follow-up) and radiographically diagnosed osteoarthritis
Effects of interventions (19/55 versus 10/55; see Analysis 1.2).
See: Summary of findings for the main comparison Treatment failure including re-operation (for surgery) or
subsequent operation (for conservative treatment).
We identified one study reporting results for 121 young, active
Within two years, 23 (39%) out of 59 participants in the conservative
adults with acute ACL injury who had been randomised to either
treatment group underwent ACL reconstruction; subsequently,
ACL reconstruction followed by structured rehabilitation (N = 62) or
another seven participants in the conservative group underwent
conservative treatment comprising structured rehabilitation alone
ACL reconstruction. Thus, 30 (51%) participants had opted for
(N = 59; Frobell 2010b).
ACL reconstruction within five years. Defining treatment failure
Primary outcomes specifically in terms of graft rupture, whether surgically treated
or not, in the ACL reconstruction group, or subsequent ACL
Subjective knee scores reconstruction in the conservative treatment group, produced
There was no difference in subjective knee score (measured using results that strongly favoured the ACL reconstruction group at both
the KOOS-4 score (range from 0 (extreme symptoms) to 100 (no follow-up times: two years, 3/62 versus 23/59; RR 0.12, 95% CI 0.04
symptoms)) between participants with ACL reconstruction and to 0.34; and five years, 3/61 versus 30/59; low-quality evidence; see
conservative treatment in terms of change scores from baseline Analysis 1.3. Two of the three participants with graft rupture had
at two years (MD -0.20, 95% CI -6.78 to 6.38; N = 121 participants; revision ACL reconstruction.
low-quality evidence; see Analysis 1.1), or final scores at five years
In both groups of Frobell 2010b, meniscal tears were treated with
(MD -2.0, 95% CI -8.27 to 4.27; N = 120 participants; low-quality
partial resection or fixation, initially when indicated by MRI findings
evidence; see Analysis 1.1). This finding of no statistically significant
and clinical signs in both groups, and when found during ACL
or clinically important between-group differences was consistent
reconstruction in the surgery group. More participants in the ACL
across all five individual components of the final KOOS score
reconstruction group had initial meniscal surgery, with results
at two years (ACL reconstruction versus conservative treatment
provided for meniscii, not knees (34/62 versus 21/59). Conversely,
(mean values): Pain, 87.2 versus 87.7; reported P = 0.87; Symptoms,
fewer participants in the ACL reconstruction group had meniscal
78.7 versus 83.0; reported P = 0.16; Function in activities of daily
surgery during the two-year follow-up (6/62 versus 29/59; RR 0.02,
living, 93.5 versus 94.7; reported P = 0.68; Function in sports and
95% CI 0.09 to 0.44). At five years, there was little difference
recreation, 71.8 versus 71.2; reported P = 0.95; Knee-related quality
between the two groups in the numbers of participants who had
of life, 67.3 versus 63.0; reported P = 0.28). Similar findings of no
undergone meniscal surgery, whether initially or during follow-
between-group differences also applied at five years.
up: 29/61 versus 32/59; RR 0.88, 95% CI 0.02 to 1.25; see Analysis
Adverse events 1.4). Fifteen participants of the conservative treatment group had
meniscal surgery only.
A range of serious and non-serious adverse events were reported in
Frobell 2010b; however, data were presented as the total number Secondary outcomes
of complications rather than the total number of participants with
General health-related quality of life
a complication. In addition, only those events that occurred in
5% or more of participants in the trial or 3% or more participants There was no difference in health-related quality of life (measured
in one treatment group were reported. Serious adverse events using the SF-36 physical and mental components (range from
were classified as those having the potential to significantly 0 (worst health state) to 100 (best health state)) between ACL
compromise clinical outcome or result in significant disability or reconstruction and conservative treatments at two years (physical:
incapacity, requiring inpatient or outpatient care. Overall, fewer MD 4.1, 95% CI -2.76 to 10.96; mental: MD 4.50, 95% CI -0.66 to
serious adverse events involving the index knee occurred in the 9.66; N = 121 participants, low-quality evidence; see Analysis 1.5) or
ACL reconstruction group than in the conservative treatment group five years (physical: MD 1.00, 95% CI -4.54 to 6.54; mental: MD 2.00,
at two years: 26 versus 40 serious adverse events respectively 95% CI -5.06 to 9.06; N = 120 participants, low-quality evidence; see
(reported P = 0.06). Results for the different types of serious adverse Analysis 1.5).
events at two years are shown in Analysis 1.2. These show that
the excess of adverse events in the conservative treatment group Return to activity or level of sport participation
related to subjective or clinical knee instability (2/62 versus 19/59) There was no difference between the two groups in the return to
and meniscal signs and symptoms (1/62 versus 13/59). The other pre-injury activity level or higher, based on Tegner activity scale
serious adverse events, which were less frequent, favoured the (range from 1 to 10, where 1 is least strenuous activity level and 10
conservative treatment group; e.g. graft rupture (3/62 versus 1/59). corresponds to high knee-demanding activities on a professional
The overall frequency of non-serious adverse events involving the level) at either two years (27/62 versus 21/59; RR 1.22, 95% CI 0.78
index knee was higher in the ACL reconstruction group at two years to 1.91; low-quality evidence) or five years (14/61 versus 12/59; RR
(87 versus 44; reported P < 0.001). The three largest categories 1.13, 95% CI 0.57 to 2.23; low-quality evidence); see Analysis 1.6.
of these events, all of which favoured the conservative treatment
group, were: subjective or clinical knee instability (25/62 versus
Informed decisions.
There was also no between-group differences in Tegner activity ACL injuries. Overall, there was low-quality evidence indicating
scale results at two years (ACL reconstruction: median 6.5, no difference between the two groups in subjective knee score
interquartile range (IQR) 3 to 8; conservative treatment: median at either two or five years for the KOOS-4 outcome measure
5, IQR 4 to 7; N = 121 participants; reported P = 0.82; low-quality or the five separate components of the KOOS score. The total
evidence) or at five years (ACL reconstruction: median 4, IQR 2.5 number of participants incurring one or more adverse events
to 7; conservative treatment: median 4.0, IQR 2 to 7; N = 120 in each group was not reported in Frobell 2010b. The majority
participants; reported P = 0.74; low-quality evidence). of individual categories of serious adverse events were surgery-
related in the ACL reconstruction group and knee instability and
Functional assessment meniscal signs and symptoms in the conservative treatment group.
This outcome was not reported. Although there were also no data for total participants with
treatment failure, including subsequent surgery, we devised two
Composite clinical examination outcomes separate categories, one relating to ACL surgery (graft rupture and
subsequent ACL reconstruction) and the other relating to meniscal
This outcome was not reported. surgery. There was low-quality evidence of far fewer ACL-related
Knee stability treatment failures in the surgical treatment group at two years. This
result is dominated by the uptake by 39% (23/59) of participants
All three objective measures of knee stability favoured the ACL of ACL reconstruction for knee instability at two years and by 51%
reconstruction. Anterior sagittal translation of the tibia, measured (30/59) of participants at five years. There was low-quality evidence
using a knee-ligament testing device, the KT 1000 arthrometer of little difference between the two groups in participants who
performed at 134 Newtons, was lower in the ACL reconstruction had undergone meniscal surgery at anytime up to five years. There
group (MD -1.70 mm, 95% CI -2.68 to -0.72 mm; N = 121 participants; was low-quality evidence of no clinically important between-group
low-quality evidence; see Analysis 1.7). This outcome measurement differences in SF-36 physical component scores at two years. There
was not reported at five years. This result was consistent with was low-quality evidence of a higher return to the same level
the finding of greater numbers of normal knees in the surgery or greater sport activity at two years in the ACL reconstruction
group when testing the anteroposterior laxity of the knee at rest, group, but the wide 95% confidence interval also included the
measured using the Lachman's test (scores range from zero to potential for a higher return in the conservative treatment group.
three, with zero indicating normal stability and three indicating Based on an illustrative return of 382 per 1000 conservatively
severely increased laxity). Knees with normal stability (score zero) treated patients, this amounts to an extra 84 returns per 1000 ACL
at two years: RR 2.22, 95% CI 1.43 to 3.45; N = 118 participants; and reconstruction patients (95% CI 84 fewer to 348 more). There was
five years: RR: 2.37, 95% CI 1.60 to 3.51; N = 116 participants; low- very low-quality evidence of a higher incidence of radiographically-
quality evidence; see Analysis 1.8. A similar finding applied when detected osteoarthritis in the surgery group (19/58 (35%) versus
testing rotational stability at rest, using the Pivot shift test (scores 10/55 (18%)).
range from zero to three, with zero indicating normal stability
and three indicating severely increased laxity). Knees with normal Overall completeness and applicability of evidence
stability (score zero) at two years: RR 1.61, 95% CI 1.18 to 2.20; N =
Considering ACL reconstruction is relatively commonplace in some
118 participants; and five years: RR 1.96, 95% CI 1.38 to 2.77; N = 116
countries (Gianotti 2009 reported 37 per 100,000 had undergone
participants; low-quality evidence; see Analysis 1.8).
ACL reconstruction surgery per year in New Zealand) and choosing
Objective measure of muscle strength surgery is a major treatment decision, it is surprising that we only
found one trial that compared ACL reconstruction and conservative
This outcome was not reported. treatment. This small trial, which was carried out in Scandinavia,
included only young (mean age 26 years), active individuals with
Resource and economic outcomes an acute injury (not more than four weeks since injury). Further
These outcomes were not reported. potential limitations to applicability reside in the study design.
This included the formal option for subsequent (delayed) ACL
DISCUSSION reconstruction as well as separate surgery for meniscal injuries
in the conservative treatment group. Of interest, is that, of
Summary of main results the participants analysed in the conservative treatment group,
39% elected to undergo ACL reconstruction within two years
Despite a rigorous search of the literature, we only identified one
and 51% elected to undergo ACL reconstruction at five years.
randomised trial that met the inclusion criteria of this review
In our review, it was not possible to further distinguish the
(Frobell 2010b). This contrasts with the substantial literature on
outcome of the true conservatively managed patients, due to the
the different surgical techniques for anterior cruciate ligament
pragmatic aspects of the trial design. The evidence suggests that
(ACL) reconstruction. The trial compared ACL reconstruction
the present management options for ACL injury (in the acutely
followed by structured rehabilitation and conservative treatment
injured knee) are either 1) immediate surgery (within four weeks)
comprising structured rehabilitation alone, reporting results for
or 2) initial conservative care, followed by surgery if and when
121 participants at two years and 120 participants at five years. Built
conservative care fails. Notably, the definition of 'failure' was open
into the study design was a formal option for subsequent (delayed)
to interpretation and was related to participant preference and
ACL reconstruction in the conservative treatment group.
expectation. The trial did not provide full data for participants with
The Summary of findings for the main comparison presents a adverse events or treatment failure; the definition needs careful
summary of the current evidence from randomised controlled consideration for further trials.
trials for surgical versus conservative interventions for treating
Informed decisions.
Quality of the evidence examining the clinical and cost-effectiveness of early surgery versus
conservative management (with the option for delayed surgery)
Frobell 2010b was judged to be at low risk of selection and reporting for ACL rupture (NTR2746), and was due to complete enrolment in
biases, at high risk of performance and detection biases because of February 2015. The other study (ACL SNNAP) is designed to examine
the lack of blinding and at unclear risk of attrition bias because of the clinical and cost-effectiveness of surgery versus conservative
an imbalance in the post-randomisation exclusions. management (again with the option for delayed surgery) in patients
with non-acute (longer than four months) ACL deficiency. This
According to GRADE methodology, the overall quality of the
study is in set-up phase and due to start recruitment in July 2016.
evidence across different outcomes was low. Thus, further research
is very likely to have an important impact on our confidence in the
AUTHORS' CONCLUSIONS
estimate of effect and is likely to change the estimate (Summary
of findings for the main comparison). We downgraded the quality Implications for practice
of the evidence for imprecision, reflecting that only one study of
121 participants was included. As with many other studies of this We found low-quality evidence from a single trial involving 121
nature that examine surgical versus conservative interventions, young, active adults with acute ACL injuries that there was
it was not possible to blind participants or study personnel. Due no difference between surgical management (ACL reconstruction
to the subjective and self-reported nature of outcomes in this followed by structured rehabilitation) and conservative treatment
study, we judged this item as having a potential for high risk of (structured rehabilitation only) in patient-reported outcomes of
bias. An improvement in knee stability was reported in the ACL knee function at two and five years after injury. However,
reconstruction group; however, as these tests were performed by these findings need to be viewed in the context that many
unblinded outcome assessors, the results were considered to be participants with an ACL rupture remained symptomatic following
at a high risk of bias. Therefore, we downgraded the quality of rehabilitation and later opted for ACL reconstruction surgery.
the evidence one level for methodological limitations. Hence, the
results of the included trial should be interpreted with caution and Implications for research
viewed, at this stage, as requiring further investigation with studies In future updates of this review, the addition of evidence from the
of good methodological quality. two ongoing trials should help to inform the optimal management
of anterior cruciate ligament injuries.
Potential biases in the review process
There were no obvious biases within the review process. We carried Further randomised trials comparing surgery with conservative
out a comprehensive search strategy and thorough methods of management should be robust in design. Not only should such
study selection. We applied no language restrictions. The matching trials assess and report outcomes that are important to patients
of desired outcomes with those reported in trials does require an with ACL rupture (such as subjective knee function, quality of life,
element of judgement. This related particularly to the incomplete effects on daily activities, and return to activity and sport) but they
data for treatment failure in Frobell 2010b, where data were also should consider factors such as cross-over, standardisation
provided for individual procedures but not participants undergoing of interventions and treatment preferences that create additional
subsequent surgery, relating either to ACL or new knee injury. Thus, challenges in the design, conduct and interpretation of trials of this
our presentation of two categories of treatment failure, split into type.
ACL and meniscal surgery, is a compromise but one that avoids unit
of analysis issues as well as making the best use of the available
ACKNOWLEDGEMENTS
evidence. We would like to thank Helen Handoll and Haris Vasiliadis for
helpful comments about drafts of the review; and Lindsey Elstub
Agreements and disagreements with other studies or
and Joanne Elliott for support through the editorial process.
reviews
Our review is consistent with a recent systematic review by We are grateful to Helen Handoll, Nikolaos Paschos and Haris
Smith 2014 that compared ACL reconstruction and non-surgical Vasiliadis for valuable comments about the protocol; and Joanne
treatment. They included 16 studies, only one of which was a Elliott and Laura MacDonald for editorial help.
randomised trial (Frobell 2010b). The authors also concluded that This project was supported by the National Institute for Health
the evidence base is limited in methodological quality, with the Research (NIHR) via Cochrane Infrastructure funding to the
current literature insufficient on which to base clinical decision- Cochrane Bone, Joint and Muscle Trauma Group. The views and
making with respect to treatment options for people following opinions expressed therein are those of the authors and do not
ACL rupture. Our search also identified two ongoing studies with necessarily reflect those of the Systematic Reviews Programme,
a similar study design to that of Frobell 2010b. One study is NIHR, or the UK National Health Service or Department of Health.
Informed decisions.
REFERENCES
References to studies included in this review References to ongoing studies

Frobell 2010b {published data only} ACL SNNAP {published data only}
* Frobell RB, Roos EM, Roos HP, Ranstam J, Lohmander LS. Beard D. The ACL SNNAP trial: Comparison of the clinical and
A randomized trial of treatment for acute anterior cost effectiveness of two management strategies for non-acute
cruciate ligament tears. New England Journal of Medicine anterior cruciate ligament (ACL) injury: rehabilitation versus
2010;363(4):331-42. surgical reconstruction. http://www.nets.nihr.ac.uk/projects/
hta/1414063 (accessed 18 January 2016).
Frobell RB, Roos HP, Roos EM, Roemer FW, Ranstam J,
Lohmander LS. Treatment for acute anterior cruciate NTR2746 {unpublished data only}
ligament tear: five year outcome of randomised trial. BMJ Eggerding V. Cost-effectiveness of two treatment strategies of
2013;346(7895):232. an anterior cruciate ligament rupture. A randomized clinical
study. www.trialregister.nl/trialreg/admin/rctview.asp?TC=2746
Lohmander S. Surgical versus non-surgical treatment of
(accessed 30 March 2015). [Trial Registration number: NTR2746]
anterior cruciate ligament (ACL) injuries: a randomised
prospective clinical trial. http://www.controlled-trials.com/
ISRCTN84752559 (accessed 18 January 2016). Additional references
Ajuied 2013
References to studies excluded from this review Ajuied A, Wong F, Smith C, Norris M, Earnshaw P, Back D, et al.
Andersson 1991 {published data only} Anterior cruciate ligament injury and radiologic progression
of knee osteoarthritis: a systematic review and meta-analysis.
Andersson C, Odensten M Gillquist J. Knee function after
American Journal of Sports Medicine 2014;42(9):2242-52.
surgical or nonsurgical treatment of acute rupture of the
anterior cruciate ligament: A randomized study with a long- Bahr 2003
term follow up period. Clinical Orthopaedics and Related
Bahr R, Holme I. Risk factors for sports injuries: a
Research 1991;(264):255-63.
methodological approach. British Journal of Sports Medicine
Andersson C, Odensten M, Good L, Gillquist J. Surgical or non- 2003;37(5):384-92.
surgical treatment of acute rupture of the anterior cruciate
Bowers 2005
ligament. A randomized study with long-term follow-up. Journal
of Bone and Joint Surgery - American Volume 1989;71(7):965-74. Bowers AL, Spindler KP, McCarty EC, Arrigain S. Height, weight,
and BMI predict intra-articular injuries observed during ACL
Meunier A, Odensten M, Good L. Long-term results after reconstruction: evaluation of 456 cases from a prospective ACL
primary repair or non-surgical treatment of anterior cruciate database. Clinical Journal of Sport Medicine 2005;15(2):9-13.
ligament rupture: a randomized study with a 15-year follow-
up. Scandinavian Journal of Medicine and Science in Sports Brooks 1996
2007;17:230-7. Brooks R. EuroQol: the current state of play. Health Policy
1996;37:53-72.
Odensten M, Hamberg P, Nordin M, Lysholm J, Gillquist J.
Surgical or conservative treatment of the acutely torn anterior Collins 2013
cruciate ligament. Clinical Orthopaedics and Related Research Collins JE, Katz JN, Donnell-Fink LA, Martin SD, Losina E.
1985;(198):87-93. Cumulative incidence of ACL reconstruction after ACL injury
Odensten M, Hamberg P, Nordin M, Lysholm J, Gillquist J. in adults: role of age, sex and race. American Journal of Sports
Treatment of the acute torn anterior cruciate ligament: A Medicine 2013;41(3):544-9.
randomized study with a short-term follow-up [abstract]. Acta Crawford 2007
Orthopaedica Scandinavica 1984;55(4):474.
Crawford R, Walley G, Bridgman S. Magnetic resonance
Sandberg 1987 {published data only} imaging versus arthroscopy in the diagnosis of knee pathology,
Sandberg R, Balkfors B, Edwards P, Nilsson B, Westlin N. concentrating on meniscal lesions and ACL tears: A systematic
Surgical or non-surgical treatment of ligamentous knee injuries. review. British Medical Bulletin 2007;84:5-23.
A randomized controlled study - early results [abstract]. Acta Delincé 2012
Orthopaedica Scandinavica 1986;57:254.
Delincé P, Ghafil D. Anterior cruciate ligament tears:
Sandberg R, Balkfors B, Nilsson B, Westlin N. Operative versus conservative or surgical treatment? A critical review of the
non-operative treatment of recent injuries to the ligaments of literature. Knee Surgery, Sports Traumatology, Arthroscopy
the knee. Journal of Bone and Joint Surgery - American Volume 2012;20(1):48-61.
1987;69(8):1120-6.
Frobell 2010a
Frobell RB, Roos EM, Roos HP, Ranstam J, Lohmander LS.
A randomized trial of treatment for acute anterior
Informed decisions.
cruciate ligament tears. New England Journal of Medicine Linko 2009

2010;363:331-42. Linko E, Harilainen A, Malmivaara A, Seitsalo S. Surgical versus
conservative interventions for anterior cruciate ligament
Frobell 2013
ruptures in adults. Cochrane Database of Systematic Reviews
Frobell RB, Roos HP, Roos EM, Roemer FW, Ranstam J, 2005, Issue 2. [DOI: 10.1002/14651858.CD001356.pub3]
Lohmander LS. Treatment for acute anterior cruciate ligament
tear: five year outcome of randomised trial. BMJ 2013;346:f232. Lohmander 2007
Lohmander S, Englund PM, Dahl LL, Roos EW. The long-
Gianotti 2009
term consequence of anterior cruciate ligament and
Gianotti SM, Marshall SW, Hume PA, Bunt L. Incidence of meniscus injuries. American Journal of Sports Medicine
anterior cruciate ligament injury and other knee ligament 2007;35(10):1756-69.
injuries: a national population-based study. Journal of Science
and Medicine in Sport 2009;12(6):622-7. Lysholm 1982
Lysholm J, Gillquist J. Evaluation of knee ligament surgery
Hefti 1993
results with special emphasis on use of a scoring scale.
Hefti F, Muller W, Jakob RP, Staubli HU. Evaluation of knee American Journal of Sports Medicine 1982;10(3):150-4.
ligament injuries with the IKDC form. Knee Surgery, Sports
Traumatology, Arthroscopy 1993;1(3-4):226-34. Micheo 2010
Micheo W, Hernandez L, Seda C. Evaluation, management,
Hernandez 2006
rehabilitation, and prevention of anterior cruciate ligament
Hernandez L, Micheo W, Amy E. Rehabilitation update for the injury: Current concepts. American Academy of Physical Medicine
anterior cruciate ligament injured patient: Current concepts. and Rehabilitation 2010;2(10):935-44.
Boletin de la Asociacion Medica de Puerto Rico 2006;98(1):62-72.
Miyasaka 1991
Hewett 2006
Miyasaka KC, Daniel DM, Stone ML, Hirschman P. The incidence
Hewett TE, Myer GD, Ford KR. Anterior cruciate ligament of knee ligament injuries in the general population. American
injuries in female athletes: Part 1, mechanisms and risk factors. Journal of Knee Surgery 1991;4:43-8.
American Journal of Sports Medicine 2006;34(2):299-311.
Mohtadi 1998
Higgins 2011
Mohtadi N. Development and validation of the quality of life
Higgins JPT, Green S (editors). Cochrane Handbook for outcome measure (questionnaire) for chronic anterior cruciate
Systematic Reviews of Interventions Version 5.1.0 [updated ligament deficiency. American Journal of Sports Medicine
March 2011]. The Cochrane Collaboration, 2011. Available from 1998;3(26):350-9.
www.cochrane-handbook.org.
Mohtadi 2011
Irrgang 2001
Mohtadi NG, Chan DS, Dainty KN, Whelan DB. Patellar tendon
Irrgang JJ, Anderson AF, Boland AL, Harner CD, Kurosaka M, versus hamstring tendon autograft for anterior cruciate
Neyret P, et al. Development and validation of the international ligament rupture in adults. Cochrane Database of Systematic
knee documentation committee subjective knee form. American Reviews 2011, Issue 9. [DOI: 10.1002/14651858.CD005960.pub2]
Journal of Sports Medicine 2001;29(5):600-13.
Negus 2012
Kostogiannis 2007
Negus J, Fransen M, Chen JS, Parker DA, March L. Exercise-
Kostogiannis I, Ageberg E, Neuman P, Dahlberg L, Friden T, based interventions for conservatively or surgically treated
Roos H. Activity level and subjective knee function 15 years after anterior cruciate ligament injuries in adults. Cochrane
anterior cruciate ligament injury: a prospective, longitudinal Database of Systematic Reviews 2012, Issue 10. [DOI:
study of nonreconstructed patients. American Journal of Sports 10.1002/14651858.CD010128]
Medicine 2007;25(7):1135-43.
Ramski 2013
Kvist 2004
Ramski DE, Kanj WW, Franklin CC, Baldwin KD, Ganley TJ.
Kvist J. Rehabilitation following anterior cruciate ligament Anterior cruciate ligament tears in children and adolescents:
injury: current recommendations for sports participation. a meta-analysis of nonoperative versus operative treatment.
Sports Medicine 2004;34(4):269-80. American Journal of Sports Medicine 2013 Dec 4 [Epub ahead of
print].
Lefebvre 2011
Lefebvre C, Manheimer E, Glanville J. Chapter 6: Searching for RevMan 2012 [Computer program]
studies. In: Higgins JPT, Green S (editors). Cochrane Handbook The Nordic Cochrane Centre, The Cochrane Collaboration.
for Systematic Reviews of Interventions Version 5.1.0 (updated Review Manager. Version 5.2. Copenhagen: The Nordic
March 2011). The Cochrane Collaboration, 2011. Available from Cochrane Centre, The Cochrane Collaboration, 2012.
www.cochrane-handbook.org.
Informed decisions.
Roos 1998 Tegner 1985

Roos EM, Roos HP, Lohmander LS, Ekdahl C, Beynnon BD. Tegner Y, Lysholm J. Rating systems in the evaluation of knee
Knee Injury and Osteoarthritis Outcome Score (KOOS)-- ligament injuries. Clinical Orthopaedics and Related Research
development of a self-administered outcome measure. Journal 1985;(198):43-9.
of Orthopaedic & Sports Physiotherapy 1998;28(2):88-96.
Thorstensson 2009
Rout 2013 Thorstensson CA, Lohmander LS, Frobell RB, Roos EM,
Rout R, McDonnell S, Hulley P, Jayadev C, Khan T, Carr A, et al. Gooberman-Hill R. Choosing surgery: patients' preferences
The pattern of cartilage damage in antero-medial osteoarthritis within a trial of treatments for anterior cruciate ligament injury.
of the knee and its relationship to the anterior cruciate A qualitative study. BMC Musculoskeletal Disorders 2009;10:100.
ligament. Journal of Orthopaedic Research 2013;31(6):908-13.
Ware 1992
Smith 2014 Ware JE, Sherbourne CD. The MOS 36-item short-form health
Smith TO, Postle K, Penny F, McNamara I, Mann CJV. Is survey (SF-36). Conceptual framework and item selection.
reconstruction the best management strategy for anterior Medical Care 1992;30(6):473-83.
cruciate ligament rupture? A systematic review and meta-
analysis comparing anterior cruciate ligament reconstruction Øiestad 2009
versus non-operative treatment. Knee 2014;21(2):462–70. Øiestad BE, Engebretsen L, Storheim K, Risberg MA. Knee
osteoarthritis after anterior cruciate ligament injury. American
Spindler 2004 Journal of Sports Medicine 2009;37(7):1434-43.
Spindler KP, Kuhn JE, Freedman KB, Matthews CE, Dittus RS,
Harrell FE Jr. Anterior cruciate ligament reconstruction
autograft choice: bone-tendon-bone versus hamstring: does it References to other published versions of this review
really matter? A systematic review. American Journal of Sports Monk 2014
Medicine 2004;32(8):1986-95.
Monk AP, Hopewell S, Harris K, Davies LJ, Beard D, Price A.
Spindler 2008 Surgical versus conservative interventions for treating anterior
cruciate ligament injuries. Cochrane Database of Systematic
Spindler KP, Wright RW. Clinical practice. Anterior
Reviews 14, Issue 6. [DOI: 10.1002/14651858.CD011166]
cruciate ligament tear. New England Journal of Medicine
2008;359(20):135-42.
* Indicates the major publication for the study
CHARACTERISTICS OF STUDIES
Characteristics of included studies [ordered by study ID]
Frobell 2010b
Methods Randomised controlled trial
Number of centres: 2
Dates of enrolment: February 2002 to June 2006
Follow-up: 2 and 5 years
Participants 141 participants were randomised; 121 were included in the main analysis.
Assigned: 69/72 (ACL reconstruction/conservative treatment)
Assessed: 62/59 (ACL reconstruction/conservative treatment)
Inclusion criteria:
• Young adults aged 18 to 35 years presenting to the emergency department with recent knee trauma;
• Rotational trauma to a previously uninjured knee within the preceding 4 weeks;
• ACL insufficiency as determined by clinical examination;
• A score of 5 to 9 on the Tegner Activity Scale- (TAS) 12 before the injury (scores range from 1 to 10, with
a score of 5 indicating participation in recreational sports, and a score of 9 indicating participation in
competitive sports on a nonprofessional level).
Exclusion criteria:
Informed decisions.
Frobell 2010b (Continued)

• Earlier major knee injury to index knee;
• Previous knee surgery (other than diagnostic arthroscopy);
• Associated PCL injury or MCL injury Grade III;
• Concomitant severe injury to contra-lateral knee;
• Injury to lateral/posterolateral ligament complex;
• Pregnancy;
• History of deep vein thrombosis (DVT);
• Claustrophobia;
• General systemic disease affecting physical function;
• Systemic medication/abuse of steroids.
Post-randomisation exclusion criteria:
If one of the following associated injuries to the index knee were visualised on MRI, arthroscopy or
both:
• A total collateral ligament rupture;

• Full-thickness cartilage lesion;
• Bi-compartmental extensive meniscus resections or unstable longitudinal meniscus tear that re-
quired repair and specified postoperative treatment (i.e. bracing and limited ROM).
Interventions Group 1: ACL reconstruction followed by structured rehabilitation. Early ACL reconstruction was de-
fined as surgery performed within 10 weeks after the injury.
Group 2: Conservative treatment comprising structured rehabilitation alone. Built into the study design
was a formal option for subsequent (delayed) ACL reconstruction in the conservative treatment group,
if chosen by the participant reporting instability and if pre-specified criteria were met.
Outcomes Primary:
• Subjective knee scores. "The primary outcome was the change from baseline to 2 years in the average
score on four subscales of the Knee Injury and Osteoarthritis Outcome Score (KOOS) — pain, symp-
toms, function in sports and recreation, and knee-related quality of life (KOOS-4; range of scores, 0
[worst] to 100 [best])" (Frobell 2010b). Also measured as final mean KOOS-4 score at 5 years;
• Adverse events (classified as serious and non-serious);
• Treatment failure (re-operation (for surgery) or subsequent operation (for conservative treatment)).
Secondary:
• General health-related quality of life, measured using the SF-36 physical and mental components;
• Return to activity or level of sport participation, measured using the Tegner Activity Scale;
• Knee stability, measured using a knee ligament testing device KT 100, Lachman test and pivot shift
test.
Notes The study was supported by grants from the Swedish Research Council and the Medical Faculty of Lund
University, the Skåne Regional Council, the Thelma Zoegas Fund, the Stig and Ragna Gorthon Research
Foundation, the Swedish National Center for Research in Sports, and Pfizer Global Research.
Risk of bias
Bias Authors' judgement Support for judgement
Random sequence genera- Low risk Quote: "..they [patients] were randomly assigned by computer-generated ran-
tion (selection bias) dom numbers in permuted blocks of 20".
Allocation concealment Low risk Quote: "An investigator who was not involved in the randomisation procedure
(selection bias) prepared all sequentially numbered, opaque, sealed envelopes containing the
assigned interventions to ensure that the sequence was concealed."
Informed decisions.
Frobell 2010b (Continued)
Blinding of participants High risk Due to the nature of the interventions, it was impossible to blind the partici-
and personnel (perfor- pants or personnel.
mance bias)
All outcomes
Blinding of outcome as- High risk Quote: ".. assessments were performed by one or two experienced clinicians,
sessment (detection bias) both of whom were aware of the treatment assignments".
Subjective knee score
Incomplete outcome data Unclear risk Of 20 (14% of 141) post-randomisation exclusions, 12 were excluded because
(attrition bias) of MRI findings and 8 because of findings at surgery. There was some imbal-
All outcomes ance in the losses in the two groups (7/69 = 10% in the surgery group; 13/72 =
18% in the conservative group) and there is some question whether it was ap-
propriate to exclude for findings, usually resulting in extensive meniscal fixa-
tion (2 versus 5), at surgery. However, there was clear comparability between
the two groups in the baseline characteristics of the 121 kept in the trial.
Selective reporting (re- Low risk Outcomes pre-specified in the clinical trial register (ISRCTN 84752559) are re-
porting bias) ported in the results.
Other bias Low risk No additional potential sources of bias were identified.
ACL: Anterior cruciate ligament

MCL: Medial collateral ligament
PCL: Posterior cruciate ligament
Characteristics of excluded studies [ordered by study ID]
Study Reason for exclusion
Andersson 1991 Quasi-randomised trial involving 167 participants with an acute and complete ACL rupture, recruit-
ed 1980 to 1983. Reported at several follow-up times. Included in Linko 2009. Excluded because
surgery involved direct repair of the ACL and not reconstruction.
Sandberg 1987 Randomised trial involving 200 participants with acute ACL, MCL or both, injuries; recruited 1982 to
1984. Included in Linko 2009. Excluded because surgery involved direct repair of the ACL and not
reconstruction.

Characteristics of ongoing studies [ordered by study ID]
ACL SNNAP
Trial name or title ACL SNNAP study - Comparison of the clinical and cost-effectiveness of two management strategies
for non-acute anterior cruciate ligament (ACL) injury: rehabilitation versus surgical reconstruction
Participants Target number of participants: 320
Inclusion criteria:
1. Symptomatic ACL deficiency (instability-episodes of frank giving way or feeling unstable) with
ACL deficiency confirmed using clinical assessment and/or MRI scan;
Informed decisions.
ACL SNNAP (Continued)

2. Aged 18 years or above.
Exclusion Criteria:
1. Less than 4 months since injury;
2. Previous knee surgery (other than diagnostic arthroscopy) to index knee or concomitant severe
injury to contra-lateral knee;
3. Meniscal pathology considered sufficiently symptomatic to require surgery, i.e. locked knee,
large bucket handle cartilage tear;
4. Knee joint status greater than Grade 2 on Kellgren and Lawrence scale;
5. Grade 3 MCL/LCL injury, associated PCL/PLC injury;
6. Inflammatory arthropathy.
Interventions Group 1: Conservative management group: Rehabilitation with the option for later reconstruction
only if required.
Group 2: Surgical management group: ACL reconstruction.
Outcomes Primary:
• Knee Injury and Osteoarthritis Outcome Score (KOOS4) at 18 months.
Secondary:
• Failure of the intervention, defined as continued joint instability requiring further intervention.
• Return to activity/level of sport participation, measured using the Lysholm and modified Tegner
Activity Scale.
• Intervention-related complications.
• General quality of life, measured using the EuroQol EQ-5D.
• Resource-usage: data on initial treatments, subsequent healthcare contacts including re-oper-
ations, subsequent surgical reconstructions, complications, further rehabilitation and ability to
work (e.g. sickness absences/return to work, number of days off work and subjective working abil-
ity).
• Expectations of return to activity/confidence in relation to knee.
• Patient satisfaction.
Starting date July 2016
Contact information Professor David Beard
Royal College of Surgeons Surgical Intervention Trials Unit (SITU)
Nuffield Dept of Orthopaedics, Rheumatology and Musculoskeletal Sciences

University of Oxford
Notes Funding Source: National Institute for Health Research (NIHR) Health Technology Assessment
(HTA) programme (14/140/63).
NTR2746
Trial name or title Cost-effectiveness of two treatment strategies of an anterior cruciate ligament rupture. A random-
ized clinical study
Informed decisions.
NTR2746 (Continued)
Participants Target number of participants: 188
Inclusion criteria:
1. MRI proven ACL tear, physical examination on high suspicion of ACL tear, or both;
2. Age 18 to 65 years;
3. Patient agreement with randomisation;
4. Trauma within 2 months of inclusion.
Exclusion criteria:
1. Associated PCL injury or injury to the lateral or posterolateral ligament complex with significant-
ly Increased laxity;
2. Pregnancy;
3. Patient is unlikely to complete study through 2-year follow-up;
4. Insufficient command of the English language, spoken, written, or both;
5. Presence of disorder(s) that affect the activity level of the lower limb;
6. Malalignment of the knee-hip-ankle axis that requires intervention.
Interventions Group 1: ACL reconstruction: will be performed within 4 to 6 weeks after inclusion in the study, fol-
lowed by an exercise program (standardised protocol) for 9 months;
Group 2: Conservative management group: Rehabilitation training for 3 to 4 months, followed by

assessment of knee function and quality of life. If repeated episodes of giving way, in spite of re-
habilitation, occurs or the patient is not satisfied for any reason, a late reconstruction can be per-
formed (delayed surgery).
Outcomes Primary:
• Subjective knee scores, measured as change in International Knee Documentation Committee

(IKDC) questionnaire at 2 years.
Secondary:
• General health-related quality of life, measured using the SF-36 physical and mental components;
• Return to activity or level of sport participation, measured using the Tegner Activity Scale;
• Decreased productivity at paid and unpaid work and patient costs, measured by the PRODISQ,
productivity and Disease Questionnaire;
• Knee stability, measured using a knee ligament testing device KT 1000;
• Subjective knee scores, measured as change in the KOOS score;
• Adverse events.
Starting date May 2011
Contact information Dr. V. Eggerding
Erasmus Medical Centre, Department of Orthopaedics
PO Box 2040
Rotterdam
Informed decisions.
NTR2746 (Continued)
The Netherlands
Notes Funding source: ZON-MW, The Netherlands Organization for Health Research and Development.
Recruitment completed in February 2015; trial registration number: NTR2746.

LCL: Lateral collateral ligament
PCL: Posterior cruciate ligament
PLC: Posterior lateral complex
DATA AND ANALYSES
Comparison 1. ACL reconstruction versus conservative treatment
Outcome or subgroup title No. of studies No. of partici- Statistical method Effect size
pants
1 Patient-rated knee function 1 Mean Difference (IV, Fixed, 95% Totals not select-
(KOOS-4 score) CI) ed
1.1 KOOS change score at 2 years 1 Mean Difference (IV, Fixed, 95% 0.0 [0.0, 0.0]
from baseline CI)
1.2 KOOS score at 5 years 1 Mean Difference (IV, Fixed, 95% 0.0 [0.0, 0.0]
CI)
2 Serious adverse events relating to 1 Risk Ratio (M-H, Fixed, 95% CI) Totals not select-
the index knee at 2 years ed
2.1 Arthrofibrosis 1 Risk Ratio (M-H, Fixed, 95% CI) 0.0 [0.0, 0.0]
2.2 Graft rupture 1 Risk Ratio (M-H, Fixed, 95% CI) 0.0 [0.0, 0.0]
2.3 Subjective or clinical instability 1 Risk Ratio (M-H, Fixed, 95% CI) 0.0 [0.0, 0.0]
2.4 Meniscal signs and symptoms 1 Risk Ratio (M-H, Fixed, 95% CI) 0.0 [0.0, 0.0]
2.5 Pain, swelling or both 1 Risk Ratio (M-H, Fixed, 95% CI) 0.0 [0.0, 0.0]
2.6 Decreased range of motion 1 Risk Ratio (M-H, Fixed, 95% CI) 0.0 [0.0, 0.0]
2.7 Extension deficit 1 Risk Ratio (M-H, Fixed, 95% CI) 0.0 [0.0, 0.0]
2.8 'Other' 1 Risk Ratio (M-H, Fixed, 95% CI) 0.0 [0.0, 0.0]
2.9 Radiographic osteoarthritis at 5 1 Risk Ratio (M-H, Fixed, 95% CI) 0.0 [0.0, 0.0]
years
3 Treatment failure (graft rupture or 1 Risk Ratio (M-H, Fixed, 95% CI) Totals not select-
ACL reconstruction) ed
3.1 At 2 years 1 Risk Ratio (M-H, Fixed, 95% CI) 0.0 [0.0, 0.0]
Informed decisions.
Outcome or subgroup title No. of studies No. of partici- Statistical method Effect size
pants
4 Meniscal surgery 1 Risk Ratio (M-H, Fixed, 95% CI) Totals not select-
ed
4.1 Meniscal surgery at 2 years 1 Risk Ratio (M-H, Fixed, 95% CI) 0.0 [0.0, 0.0]
4.2 Meniscal surgery at any time 1 Risk Ratio (M-H, Fixed, 95% CI) 0.0 [0.0, 0.0]
over 5 years
5 General health-related quality 1 Mean Difference (IV, Fixed, 95% Totals not select-
of life (SF-36 Physical and Mental CI) ed
scores)
5.1 SF-36 physical component at 2 1 Mean Difference (IV, Fixed, 95% 0.0 [0.0, 0.0]
years CI)
5.2 SF-36 physical component at 5 1 Mean Difference (IV, Fixed, 95% 0.0 [0.0, 0.0]
years CI)
5.3 SF-36 mental component at 2 1 Mean Difference (IV, Fixed, 95% 0.0 [0.0, 0.0]
years CI)
5.4 SF-36 mental component at 5 1 Mean Difference (IV, Fixed, 95% 0.0 [0.0, 0.0]
years CI)
6 Return to previous activity level 1 Risk Ratio (M-H, Fixed, 95% CI) Totals not select-
(pre-injury Tegner activity scale lev- ed
el)
7 Knee stability (KT-1000 test) at 2 1 Mean Difference (IV, Fixed, 95% Totals not select-
years (mm) CI) ed
8 Knee stability (normal pivot shift 1 Risk Ratio (M-H, Fixed, 95% CI) Totals not select-
or Lachman tests) ed
8.1 Knee stability (normal pivot shift 1 Risk Ratio (M-H, Fixed, 95% CI) 0.0 [0.0, 0.0]
test) at 2 years
8.2 Knee stability (normal pivot shift 1 Risk Ratio (M-H, Fixed, 95% CI) 0.0 [0.0, 0.0]
test) at 5 years
8.3 Knee stability (normal Lachman 1 Risk Ratio (M-H, Fixed, 95% CI) 0.0 [0.0, 0.0]
test) at 2 years
8.4 Knee stability (normal Lachman 1 Risk Ratio (M-H, Fixed, 95% CI) 0.0 [0.0, 0.0]
test) at 5 years
Informed decisions.
Analysis 1.1. Comparison 1 ACL reconstruction versus conservative

treatment, Outcome 1 Patient-rated knee function (KOOS-4 score).
Study or subgroup ACL reconstruction Conservative treatment Mean Difference Mean Difference
N Mean(SD) N Mean(SD) Fixed, 95% CI Fixed, 95% CI
1.1.1 KOOS change score at 2 years from baseline
Frobell 2010b 62 39.2 (18.5) 59 39.4 (18.4) -0.2[-6.78,6.38]
1.1.2 KOOS score at 5 years

Frobell 2010b 61 80 (15.6) 59 82 (19.2) -2[-8.27,4.27]
Favours conservative -20 -10 0 10 20 Favours surgery
Analysis 1.2. Comparison 1 ACL reconstruction versus conservative treatment,

Outcome 2 Serious adverse events relating to the index knee at 2 years.
Study or subgroup ACL reconstruction Conservative treatment Risk Ratio Risk Ratio
n/N n/N M-H, Fixed, 95% CI M-H, Fixed, 95% CI
1.2.1 Arthrofibrosis
Frobell 2010b 1/62 0/59 2.86[0.12,68.78]
1.2.2 Graft rupture

Frobell 2010b 3/62 1/59 2.85[0.31,26.68]
1.2.3 Subjective or clinical instability

Frobell 2010b 2/62 19/59 0.1[0.02,0.41]
1.2.4 Meniscal signs and symptoms

Frobell 2010b 1/62 13/59 0.07[0.01,0.54]
1.2.5 Pain, swelling or both

Frobell 2010b 6/62 3/59 1.9[0.5,7.26]
1.2.6 Decreased range of motion

Frobell 2010b 4/62 1/59 3.81[0.44,33.08]
1.2.7 Extension deficit

Frobell 2010b 1/62 0/59 2.86[0.12,68.78]
1.2.8 'Other'
Frobell 2010b 8/62 3/59 2.54[0.71,9.11]
1.2.9 Radiographic osteoarthritis at 5 years

Frobell 2010b 19/58 10/55 1.8[0.92,3.52]
Favours surgery 0.01 0.1 1 10 100 Favours conservative
Informed decisions.

treatment, Outcome 3 Treatment failure (graft rupture or ACL reconstruction).
1.3.1 At 2 years
Frobell 2010b 3/62 23/59 0.12[0.04,0.39]
1.3.2 At 5 years
Frobell 2010b 3/61 30/59 0.1[0.03,0.3]
Analysis 1.4. Comparison 1 ACL reconstruction versus conservative treatment, Outcome 4 Meniscal surgery.
1.4.1 Meniscal surgery at 2 years
Frobell 2010b 6/62 29/59 0.2[0.09,0.44]
1.4.2 Meniscal surgery at any time over 5 years

Frobell 2010b 29/61 32/59 0.88[0.62,1.25]

Outcome 5 General health-related quality of life (SF-36 Physical and Mental scores).
1.5.1 SF-36 physical component at 2 years
Frobell 2010b 62 82.1 (19.3) 59 78 (19.2) 4.1[-2.76,10.96]
1.5.2 SF-36 physical component at 5 years

Frobell 2010b 61 85 (15.6) 59 84 (15.3) 1[-4.54,6.54]
1.5.3 SF-36 mental component at 2 years

Frobell 2010b 62 88.3 (13) 59 83.8 (15.7) 4.5[-0.66,9.66]
1.5.4 SF-36 mental component at 5 years

Frobell 2010b 61 87 (15.6) 59 85 (23) 2[-5.06,9.06]
Favours conservative -20 -10 0 10 20 Favours surgery

Outcome 6 Return to previous activity level (pre-injury Tegner activity scale level).
1.6.1 At 2 years
Favours conservative 0.1 0.2 0.5 1 2 5 10 Favours surgery
Informed decisions.
Frobell 2010b 27/62 21/59 1.22[0.78,1.91]
1.6.2 At 5 years
Frobell 2010b 14/61 12/59 1.13[0.57,2.23]
Favours conservative 0.1 0.2 0.5 1 2 5 10 Favours surgery

treatment, Outcome 7 Knee stability (KT-1000 test) at 2 years (mm).
Frobell 2010b 62 6.6 (2.4) 59 8.3 (3.1) -1.7[-2.68,-0.72]
Favours surgery -5 -2.5 0 2.5 5 Favours conservative

treatment, Outcome 8 Knee stability (normal pivot shift or Lachman tests).
1.8.1 Knee stability (normal pivot shift test) at 2 years
Frobell 2010b 45/60 27/58 1.61[1.18,2.2]
1.8.2 Knee stability (normal pivot shift test) at 5 years

Frobell 2010b 45/58 23/58 1.96[1.38,2.77]
1.8.3 Knee stability (normal Lachman test) at 2 years

Frobell 2010b 39/60 17/58 2.22[1.43,3.45]
1.8.4 Knee stability (normal Lachman test) at 5 years

Frobell 2010b 45/58 19/58 2.37[1.6,3.51]
Favours conservative 0.2 0.5 1 2 5 Favours surgery
APPENDICES
Appendix 1. Search strategies

CENTRAL (Cochrane Register of Studies Online)
#1 MESH DESCRIPTOR Anterior Cruciate Ligament (688)
#2 MESH DESCRIPTOR Anterior Cruciate Ligament Reconstruction EXPLODE ALL TREES (197)
#3 (((anterior adj2 cruciate* adj2 ligament*) or acl)):TI,AB,KY (1488)
#4 #1 OR #2 OR #3 (1489)
#5 MESH DESCRIPTOR Orthopedics (293)
#6 MESH DESCRIPTOR Surgical Procedures, Operative EXPLODE ALL TREES (93308)
#7 MESH DESCRIPTOR Orthopedic Fixation Devices EXPLODE ALL TREES (1976)
#8 MESH DESCRIPTOR Orthopedic Procedures EXPLODE ALL TREES (8745)
#9 MESH DESCRIPTOR Arthroscopy EXPLODE ALL TREES (1104)
Informed decisions.
#10 MESH DESCRIPTOR Suture Techniques EXPLODE ALL TREES (1590)

#11 (surg* or operat* or reconstruct* or repair* or graft* or arthroscop*):TI,AB,KY (142949)
#12 #5 OR #6 OR #7 OR #8 OR #9 OR #10 OR #11 (172347)
#13 MESH DESCRIPTOR Physical Therapy Modalities EXPLODE ALL TREES (16072)
#14 MESH DESCRIPTOR Braces EXPLODE ALL TREES (319)
#15 (non-surg* or nonsurg* or non-operat* or nonoperat* or conserv* or rehab* or physiotherapy or physical therapy or brace* or exercis*
or cast or casts):TI,AB,KY (76067)
#16 #13 OR #14 OR #15 (80623)
#17 #4 AND #12 AND #16 (431)
MEDLINE (Ovid Online)

1 Anterior Cruciate Ligament/ or Anterior Cruciate Ligament Reconstruction/ (11139)
2 ((anterior adj2 cruciate* adj2 ligament*) or acl).tw. (16152)
3 or/1-2 (17797)
4 Orthopedics/ (17890)
5 exp Surgical Procedures,Operative/ (2599718)
6 exp Orthopedic Fixation Devices/ (64256)
7 Orthopedic Procedures/ (19857)
8 Arthroscopy/ (17948)
9 Suture Techniques/ (37962)
10 su.fs. (1668419)
11 (surg* or operat* or reconstruct* or repair* or graft* or arthroscop*).tw. (2418655)
12 or/4-11 (4264040)
13 exp Physical Therapy Modalities/ (119036)
14 Braces/ (4724)
15 (non-surg* or nonsurg* or non-operat* or nonoperat* or conserv* or rehab* or physiotherapy or physical therapy or brace* or exercis*
or cast*1).tw. (739890)
16 or/13-15 (816600)
17 3 and 12 and 16 (2466)
18 Randomized controlled trial.pt. (403863)
19 Controlled clinical trial.pt. (89971)
20 randomized.ab. (332231)
21 placebo.ab. (165137)
22 Clinical trials as topic/ (174213)
23 randomly.ab. (240016)
24 trial.ti. (143900)
25 18 or 19 or 20 or 21 or 22 or 23 or 24 (986765)
26 exp Animals/ not Humans/ (4171020)
27 25 not 26 (908881)
28 17 and 27 (358)
EMBASE (Ovid Online)

1 Anterior cruciate ligament/ (9112)
2 Anterior cruciate ligament rupture/ (3623)
3 Anterior cruciate ligament reconstruction/ (4521)
4 Anterior cruciate ligament injury/ (2019)
5 ((anterior adj2 cruciate* adj2 ligament*) or acl).tw. (19597)
6 or/1-5 (22467)
7 exp Orthopedic surgery/ (365125)
8 exp Arthroscopy/ (22580)
9 Suturing method/ (28904)
10 su.fs. (1828190)
11 (surg* or operat* or reconstruct* or repair* or graft* or arthroscop*).tw. (3128145)
12 or/7-11 (4070516)
13 exp Physiotherapy/ (65336)
14 exp Kinesiotherapy/ (55643)
15 exp Brace/ (8144)
16 (non-surg* or nonsurg* or non-operat* or nonoperat* or conserv* or rehab* or physiotherapy or physical therapy or brace* or exercis*
or cast*1).tw. (927885)
17 or/13-16 (981076)
18 6 and 12 and 17 (3317)
Informed decisions.
19 exp Randomized Controlled Trial/ or exp Single Blind Procedure/ or exp Double Blind Procedure/ or Crossover Procedure/ (445012)
20 (random* or RCT or placebo or allocat* or crossover* or 'cross over' or trial or (doubl* adj1 blind*) or (singl* adj1 blind*)).ti,ab. (1490757)
21 19 or 20 (1570098)
22 (exp Animal/ or animal.hw. or Nonhuman/) not (exp Human/ or Human cell/ or (human or humans).ti.) (5749662)
23 21 not 22 (1385888)
24 18 and 23 (468)
WHO ICTRP
1. Anterior cruciate ligament (N = 247)
ClinicalTrials.gov
1. Anterior cruciate ligament (N = 186)
CONTRIBUTIONS OF AUTHORS
Andrew P Monk: lead author, writing of manuscript, co-ordination of team.
Sally Hopewell: development of review, writing of the manuscript, 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.
Four review authors (DB, LJD, APM, AP) are member of the trial management group of an ongoing trial (ACL SNNAP); arrangements will be
made for independent review of this trial when completed.
DIFFERENCES BETWEEN PROTOCOL AND REVIEW

In the protocol we had planned to present the first seven outcomes listed in Types of outcome measures in the 'Summary of findings' table.
We adjusted this to accommodate our splitting of treatment failure into two categories. Given data for objectively measured functional
outcomes were unavailable, we replaced this with osteoarthritis at five years, even though this was radiologically assessed rather than
symptomatic.
INDEX TERMS
Medical Subject Headings (MeSH)

*Anterior Cruciate Ligament Injuries; Anterior Cruciate Ligament Reconstruction [*methods]; Joint Instability [etiology] [rehabilitation]
[*therapy]; Randomized Controlled Trials as Topic
MeSH check words

Adult; Humans; Young Adult

Monk_et_al-2016-Cochrane_Database_of_Systematic_Reviews

Uploaded by

Copyright:

Available Formats

Monk_et_al-2016-Cochrane_Database_of_Systematic_Reviews

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Monk_et_al-2016-Cochrane_Database_of_Systematic_Reviews

Uploaded by

Copyright:

Available Formats

Cochrane

Surgical versus conservative interventions for treating anterior

Monk AP, Davies LJ, Hopewell S, Harris K, Beard DJ, Price AJ

Surgical versus conservative interventions for treating anterior cruciate

Editorial group: Cochrane Bone, Joint and Muscle Trauma Group.

Data collection and analysis

PLAIN LANGUAGE SUMMARY

Results of the search

Quality of the evidence

Summary of findings for the main comparison.

Intervention: ACL reconstruction (within 10 weeks of injury) followed by structured rehabilitation

Comparison: Conservative treatment (comprising structured rehabilitation)1

Conservative ACL reconstruction2

Cochrane Database of Systematic Reviews

In terms of applicability, considera-

CI: Confidence interval; MD: Mean difference

GRADE Working Group grades of evidence

Cochrane Database of Systematic Reviews

Cochrane Database of Systematic Reviews

Figure 1. Study flow diagram

Allocation Incomplete outcome data

References to studies included in this review References to ongoing studies

cruciate ligament tears. New England Journal of Medicine Linko 2009

Roos 1998 Tegner 1985

Characteristics of included studies [ordered by study ID]

Dates of enrolment: February 2002 to June 2006

Follow-up: 2 and 5 years

Assigned: 69/72 (ACL reconstruction/conservative treatment)

Assessed: 62/59 (ACL reconstruction/conservative treatment)

Frobell 2010b (Continued)

Post-randomisation exclusion criteria:

• A total collateral ligament rupture;

Bias Authors' judgement Support for judgement

Frobell 2010b (Continued)

ACL: Anterior cruciate ligament

Characteristics of excluded studies [ordered by study ID]

Study Reason for exclusion

ACL: Anterior cruciate ligament

Characteristics of ongoing studies [ordered by study ID]

Methods Randomised controlled trial

Participants Target number of participants: 320

ACL SNNAP (Continued)

1. Less than 4 months since injury;

5. Grade 3 MCL/LCL injury, associated PCL/PLC injury;

Group 2: Surgical management group: ACL reconstruction.

• Knee Injury and Osteoarthritis Outcome Score (KOOS4) at 18 months.

Starting date July 2016

Contact information Professor David Beard

Royal College of Surgeons Surgical Intervention Trials Unit (SITU)

Nuffield Dept of Orthopaedics, Rheumatology and Musculoskeletal Sciences

Methods Randomised controlled trial

Participants Target number of participants: 188

3. Patient agreement with randomisation;

4. Trauma within 2 months of inclusion.

3. Patient is unlikely to complete study through 2-year follow-up;

4. Insufficient command of the English language, spoken, written, or both;

6. Malalignment of the knee-hip-ankle axis that requires intervention.

Group 2: Conservative management group: Rehabilitation training for 3 to 4 months, followed by

• Subjective knee scores, measured as change in International Knee Documentation Committee

Starting date May 2011