1 s2.0 S1440244021000554 Main PDF
1 s2.0 S1440244021000554 Main PDF
1 s2.0 S1440244021000554 Main PDF
Review
a r t i c l e i n f o a b s t r a c t
Article history: Objectives: To investigate whether intensive supervised rehabilitation following ACL reconstruction leads
Received 29 October 2020 to superior self-reported function and sports participation compared to less supervised rehabilitation.
Received in revised form 1 March 2021 Design: Systematic review and meta-analysis.
Accepted 3 March 2021
Methods: We included randomised controlled trials (RCTs) comparing supervised rehabilitation to
Available online xxx
rehabilitation with a similar protocol that used less supervised sessions for athletes following ACL recon-
struction. Two reviewers independently screened studies and extracted data. The Physiotherapy Evidence
Keywords:
Database (PEDro) scale was used to evaluate methodological quality and GRADE to evaluate overall qual-
Rehabilitation
Function
ity of evidence. Self-reported function and sports participation were the primary outcomes. Data were
Remote pooled using random effects meta-analyses.
Home Results: Our search retrieved 4075 articles. Seven articles reporting on six RCTs were included (n = 353).
Supervision Very-low to low-certainty evidence suggests intensive supervised rehabilitation is not superior to less
ACL surgery supervised rehabilitation following ACL reconstruction for improving self-reported function, sports par-
ticipation, knee flexor and extensor strength, range of motion, sagittal plane knee laxity, single leg hop
performance, or quality of life.
Conclusion: Based on uncertain evidence, intensive supervised rehabilitation is not superior to less super-
vised rehabilitation for athletes following ACL reconstruction. Although high-quality RCTs are needed to
provide more certain evidence, clinicians should engage athletes in shared decision making to ensure
athletes’ rehabilitation decisions align with current evidence on supervised rehabilitation as well as their
preferences and values.
© 2021 Sports Medicine Australia. Published by Elsevier Ltd. All rights reserved.
https://doi.org/10.1016/j.jsams.2021.03.003
1440-2440/© 2021 Sports Medicine Australia. Published by Elsevier Ltd. All rights reserved.
Please cite this article as: Gamble AR, et al,Intensive supervised rehabilitation versus less supervised rehabilitation following anterior
cruciate ligament reconstruction? A systematic review and meta-analysis, J Sci Med Sport, https://doi.org/10.1016/j.jsams.2021.03.003
G Model
JSAMS-2483; No. of Pages 9 ARTICLE IN PRESS
A.R. Gamble et al. Journal of Science and Medicine in Sport xxx (xxxx) xxx–xxx
reconstruction surgery followed by rehabilitation is recommended We included trials enrolling recreational and elite athletes aged
for most athletes wishing to return to sport, particularly athletes 15 years or older following ACL reconstruction.
involved in sports that require pivoting and cutting.4 Although The included trials had to compare two post-operative rehabil-
there is evidence that some athletes can return to sport without itation protocols that only differed by the number of supervised
surgery following an ACL rupture,5 the use of ACL reconstruction (or clinic-based) sessions athletes received. The protocol that had
continues to rise.3 Following ACL reconstruction, 65% of athletes fewer number of supervised sessions was defined as ‘less super-
will return to their pre-injury level of sports participation and 1 in vised rehabilitation’. Trials evaluating rehabilitation protocols with
4 of those who return to sport will experience a subsequent ACL the same number of supervised sessions or evaluating rehabili-
rupture.6 tation protocols with different intervention parameters between
Rehabilitation following ACL reconstruction aims to return ath- groups (e.g. type of exercises, intensity, duration) were excluded.
letes to their previous level of function and sports participation Self-reported knee function (e.g. Lysholm scale) and sports
while reducing the risk of re-injury. Maximising strength, func- participation (e.g. Tegner scale) were our primary outcomes. Sec-
tional performance, and addressing psychological factors such as ondary outcomes included knee muscle strength (e.g. assessed
fear of re-injury are key targets of rehabilitation.7 Achieving the isometrically, eccentrically or concentrically using a dynamome-
right balance between training and rest during rehabilitation is also ter), knee flexion and extension range of motion, sagittal knee plane
important because both under- and over-loading may increase the laxity (e.g. assessed using an arthrometer), single leg hop perfor-
risk of injury.8 Guidelines recommend using a battery of tests to mance (e.g. assessed using the LSI for single leg hop for maximum
ensure important deficits in strength and functional performance distance, single leg timed hop for 6 m and single leg vertical hop for
are addressed.9 For example, achieving >90% of quadriceps strength maximum height), and quality of life.
and single leg hop performance compared to the unaffected limb
(i.e. Limb Symmetry Index, LSI > 90%) can reduce the risk of re-injury
by 84%.10
2.3. Data sources and searches
There is currently no consensus on the ‘best’ rehabilitation
protocol. Many aspects of rehabilitation are of unknown value,
The following electronic databases and clinical trial registries
such as the need for intensive supervision by a health profes-
were searched using terms synonymous with “ACL”, “surgery” and
sional. Numerous randomised controlled trials and systematic
“rehabilitation” from the earliest record to April 2020: MEDLINE,
reviews have demonstrated that intensive supervised rehabilita-
Cochrane Central Register of Controlled Trials (CENTRAL), EMBASE,
tion is not superior to less supervised or primarily home-based
CINAHL, SPORTDiscuss, Scopus Web of Science, ClinicalTrials.gov,
rehabilitation for a range of orthopaedic and musculoskeletal pre-
The Australian New Zealand Clinical Trials Registry (ANZCTR)
sentations, including post-lumbar disc surgery,11 post-knee12 and
(www.anzctr.org.au) and The World Health Organisation (WHO)
hip arthroplasty,13 and rotator cuff tendinopathy.14 A systematic
International Clinical Trials Registry Platform (ICTRP) (https://apps.
review by Papalia et al. – the most recent review on this topic –
who.int/trialsearch/). Our search strategy is in Appendix B. Two
found that intensive supervised rehabilitation was not superior to
researchers (AG and JZ) independently performed the selection
less supervised rehabilitation following various knee operations,
of studies by screening titles and abstracts, followed by full-text
including ACL reconstruction.15 However, several methodological
articles according to the inclusion criteria. All disagreements were
issues with this review warranted an update of the evidence on
resolved by discussion. To ensure no eligible trials were missed
this topic (specifically following ACL reconstruction). The review
from the above searches, these authors hand-searched reference
by Papalia et al.15 did not use meta-analysis, did not evaluate the
lists of included trials and relevant reviews,15,18 performed cita-
overall quality of the evidence (e.g. using the GRADE approach), it
tion tracking, and contacted investigators known to be involved in
included non-randomised trials (which is inappropriate for assess-
trials that were yet to be published.
ing the effectiveness of interventions as per Cochrane guidelines16 ),
and only provided a brief description of the rehabilitation protocols.
The aim of this systematic review was to investigate whether
intensive supervised rehabilitation following ACL reconstruction 2.4. Data extraction and quality assessment
leads to superior self-reported function, sports participation and
various secondary outcomes (strength, range of motion, sagittal Using a standardised data collection form, two researchers (AG
plane knee laxity, single leg hop performance and quality of life) and JZ) independently extracted data on trial characteristics (e.g.
compared to less supervised rehabilitation. setting, country, sample size, intervention, comparison) and out-
come measures (i.e. baseline and follow-up time points) for each
trial. Disagreements were resolved by discussion and re-checking
2. Methods the trial report. Intervention details were extracted according to the
TIDieR checklist (see Appendix C).19 We extracted four time points
2.1. Protocol and registration for outcome data in this review: up to 3 months post-operatively
(or time point closest to 3 months), 3–6 months (or time point clos-
This systematic review was prospectively registered on PROS- est to 6 months), 6–12 months (or time point closest to 12 months)
PERO (CRD42020163007) and conducted in accordance with the and >12 months (or last time point beyond 12 months). When rele-
PRISMA statement17 and AMSTAR-2 checklist (Appendix A). Two vant outcome data was missing, we extracted data from published
deviations from the protocol were made. We did not restrict par- plots using WebPlotDigitizer (version 4.1) and contacted authors
ticipants’ age to 16 years and older as we wanted to capture all to clarify if needed.
trials on this topic. We used the Physiotherapy Evidence Database Methodological quality was assessed using the PEDro scale
(PEDro) scale to assess risk of bias instead of the Cochrane Risk of (0−10, where higher scores indicate higher methodological
Bias tool as the authors had more experience using the PEDro scale. quality). Trials indexed in PEDro are independently rated for
methodological quality by two trained evaluators, with a third eval-
2.2. Study selection uator arbitrating any disagreements. We extracted these scores for
each trial report from PEDro (https://www.pedro.org.au/).20 Trials
We only included randomised controlled trials (RCTs). scoring ≥7 were considered to have high methodological quality.
2
G Model
JSAMS-2483; No. of Pages 9 ARTICLE IN PRESS
A.R. Gamble et al. Journal of Science and Medicine in Sport xxx (xxxx) xxx–xxx
2.5. Data synthesis and analysis the percentage of females from 19 to 48%, and the sample size was
26−145.
All outcome data were converted to mean (SD) and entered in
Review Manager (version 5.3) for analysis. Outcome data expressed 3.2. Trial characteristics
using mean (range), median (interquartile range, IQR) and median
(range) were converted to mean (standard deviation, SD) using the The included trials were conducted in the United States
calculator by Wan et al.21 Standard errors (SE) were converted to (n = 2),28,29 Canada (n = 1),25 Australia (n = 1),30 United Kingdom
SD using the Review Manager calculator.22 For studies that did not (n = 1),31 and Sweden (n = 1).32 Three trials reported funding,25,29,31
report a SD and it could not be calculated using the above methods, while three reported no funding28,30,32 (Appendix F).
we used the SD from the trial with the highest PEDro score in that
analysis (in accordance with Cochrane).16 If two or more trials had
3.2.1. ACL reconstruction details
identical PEDro scores, we took the SD from the trial with the largest
Time from ACL rupture to reconstruction was a minimum of
sample size.
6 weeks in two trials,25,28 less than 3 months in two trials,30,32
We used random effects meta-analyses to summarise the mean
and between 12 and 132 months in one trial.31 One trial did
intervention effect of intensive supervised rehabilitation (rela-
not report time since ACL rupture for participants.29 ACL graft
tive to less supervised rehabilitation) on self-reported function,
types included bone-patella tendon-bone autograft (n = 3),25,29,31
sports participation and all secondary outcomes. Two measures
bone-patella tendon-bone, not specified as autograft or allograft
of self-reported function were assessed by trials in our review:
(n = 1),30 bone-patella tendon-bone autograft or allograft (n = 1),28
the Lysholm score and International Knee Documenting Commit-
bone-patella tendon-bone or semitendinosus autograft (n = 1)32
tee (IKDC) score. We only pooled data using the Lysholm score
(Appendix F).
because it was reported in more trials than the IKDC score and no
trial reported the IKDC score without reporting a Lysholm score.
3.2.2. Setting
Based on previous research, we defined the minimally important
Participants saw a physiotherapist at an outpatient orthopaedic
clinical difference (MCID) as 10 on a 0−100 scale for the Lysholm
clinic in two trials,31,32 an outpatient physical therapy department
score.23 Pooled estimates were reported using weighted mean
in one trial,30 a University physical therapy clinic in one trial29
difference (MD) and 95% confidence intervals (CI). Heterogeneity
and a University sports medicine clinic in one trial.25 One trial
was assessed using the I2 statistic and interpreted as follows: (a)
did not provide details on the setting.28 In four trials,29–32 partici-
0–40%, unlikely to be important heterogeneity; (b) 30–60%, mod-
pants in both groups (i.e. intensive vs. less supervised groups) were
erate heterogeneity; (c) 50–90%, substantial heterogeneity; and
seen by the same physiotherapist. In one trial,25 participants in the
(d) 75–100%, considerable heterogeneity.24 Since no meta-analysis
intensive supervised group were able to choose their own physio-
included more than 10 trials, we did not assess publication bias.
therapist, while participants in the less supervised group saw the
We were unable to perform a sensitivity analysis restricted to tri-
same physiotherapist at a University sports medicine clinic. One
als with high methodological quality (i.e. PEDro score ≥7) because
trial did not specify whether participants in the intensive super-
only one trial had high methodological quality.25 Due to a limited
vised and less supervised groups saw the same physiotherapist28
number of trials, we were unable to perform sub-group analy-
(Appendix F).
ses investigating whether trial characteristics influenced our main
findings.
The Grading of Recommendations Assessment, Development 3.2.3. Rehabilitation protocols
and Evaluation (GRADE) approach was used to evaluate the over- Duration of prescribed rehabilitation for both the intensive and
all quality of evidence and strength of recommendations.26 Quality less supervised groups was 6 weeks in one trial, 29 3 months
of evidence was rated as high, moderate, low, or very low. Quality in one trial,25 6 months in three trials,28,31,32 and 9 months in
of evidence started at “high” (due to the inclusion of RCTs only) one trial.30 Rehabilitation began immediately post-operatively in
and was downgraded for each of the following issues encountered: five trials.25,29–32 One trial did not specify when rehabilitation
limitations in the study design, imprecision, inconsistency, indi- commenced.28 Rehabilitation protocols were largely comparable
rectness, and publication bias. Quality of evidence was downgraded across trials. All protocols transitioned participants through range
by one level if the limitation was judged as ‘serious’ or by two levels of motion exercises, aerobic exercises, muscle strengthening exer-
if it was judged as ‘very serious’. Criteria on how we judged each cises (with increasing range of motion and stability demands) and
domain are available in Appendix D. sport-specific activities (Appendix F).
3
G Model
JSAMS-2483; No. of Pages 9 ARTICLE IN PRESS
A.R. Gamble et al. Journal of Science and Medicine in Sport xxx (xxxx) xxx–xxx
6 weeks post-operatively and provided participants in the inten- separately in this section.25,27 All trials used true randomization to
sive supervised group with 13–36 additional supervised sessions.32 allocate participants to groups and reported point measures and
The exercise class ran until 6 months post-operatively in both trials variability measures. Six trials reported between-group statisti-
(Appendix F). cal comparisons,25,27,28,30–32 five blinded outcome assessors25,29–32
and five specified the eligibility criteria.25,28–30,32 Baseline charac-
3.2.5. Outcomes teristics were similar between groups in three trials.25,27,30 Two
To assess self-reported function, five trials used the Lysholm trials had low loss to follow up,28,29 two used intention to treat
Scale28–32 and two used the IKDC.31,32 One trial measured func- analyses,25,31 and one used concealed allocation.25 No trial blinded
tion using the Lysholm Scale showing no significant difference participants or therapists.
between groups at 12 months but did not report the preoper-
ative data or SD.29 Three trials used the Tegner scale to assess 3.4. Effects of intervention
sports participation.30–32 Knee flexion and extension strength
were assessed in four trials.25,30–32 Two trials measured isometric 3.4.1. Primary outcomes
strength,30,32 three measured concentric strength,25,30,31 and one There was low-certainty evidence that intensive supervised
measured eccentric strength30 using the LSI. Flexion and exten- rehabilitation following ACL reconstruction did not improve self-
sion range of motion was assessed in three trials.25,28,32 One trial reported function compared to less supervised rehabilitation at 3
measured flexion in supine (active assisted), extension in prone months (MD −1.24, 95% CI: −6.09 to 3.61; 3 trials, n = 114) and 3–6
(passive) and flexion and extension during walking using video- months (MD 1.80, 95% CI: −0.91 to 4.50; 4 trials n = 149), and very-
analysis. 25 One trial measured active flexion without specifying low-certainty evidence at 6–12 months (MD −0.92, 95% CI: −3.73
the position,28 and one measured active flexion and extension in to 1.88; 3 trials n = 114) (Table 2; Appendix G).
supine.32 Sagittal plane knee laxity was measured in three tri- Similar results were found for sports participation, with low-
als using an arthrometer.25,29,31 One trial did not report mean certainty evidence at 3 months (MD 0.50, 95% CI: −0.35 to 1.35; 1
scores for laxity but reported graft failures (defined as side-to-side trial n = 40), 3–6 months (MD 0.52, 95% CI: −0.01 to 1.04; 3 trials
differences in laxity of >4 mm).29 Three trials measured func- n = 96), and 6–12 months (MD 0.91, 95% CI: −1.04 to 2.86; two trials
tional performance using Limb Symmetry Index of single leg hop n = 77) (Table 2; Appendix H).
tests.28,30,32 One trial reported that it measured hop testing but
their assessment method was not specified and they did not report 3.4.2. Secondary outcomes
their results.28 Two trials measured single leg hop for maximum Compared to less supervised rehabilitation, intensive super-
distance,30,32 and one also measured single leg timed hop for 6 m vised rehabilitation did not improve concentric, eccentric and
and single leg vertical hop for maximum height.30 For quality of isometric knee extension (low-certainty; Appendix I–K) or flex-
life, one trial used an ACL-specific measure of quality of life25 and ion strength (low-certainty; Appendix L–N), flexion (very-low to
one used the Sickness Impact scale29 (Appendix F). low-certainty; Appendix O) and extension range of motion (low-
certainty; Appendix P), sagittal plane knee laxity (low-certainty;
3.3. Methodological quality Appendix Q), single leg hop for maximum distance (low-certainty;
Appendix R), single leg timed hop for 6 m (low-certainty; Appendix
PEDro scores ranged from 4 to 7 across the trials (Table 1). The S), single leg vertical hop for maximum height (low-certainty;
two articles by Grant et al. that report on the same trial had dif- Appendix T) and quality of life (low-certainty; Appendix U) at any
ferent PEDro scores (n = 7 and 4, respectively) and were considered time point, except for eccentric knee extension strength at 3–6
4
G Model
JSAMS-2483; No. of Pages 9 ARTICLE IN PRESS
A.R. Gamble et al. Journal of Science and Medicine in Sport xxx (xxxx) xxx–xxx
129
21
53
88
37
37
40
the data entered in Review Manager (version 5.3) for each meta-
n
analysis is in Appendix V.
Total score
(/10)a 4. Discussion
5
4
7
4
5
4
4
measures and 4.1. Summary of main findings and key implications
Report point
100
Y of motion, sagittal plane knee laxity, single leg hop performance
Y
Y
Y
Y
Y
Y
or quality of life between athletes who have intensive supervised
comparisons
small trials at high risk of bias which may not have implemented
86
N contemporary rehabilitation protocols. Although there is a need
Y
Y
Y
Y
Y
Y
N
N
N
N
Y
N
N
N
N
Y
assessors
outcome
N
Y
Y
Y
Y
fulfils most of the AMSTAR-2 criteria (Appendix A), and two review-
ers independently performed the selection of studies and extracted
n = number of participants included in meta-analysis; N: no; PEDro: Physiotherapy Evidence Database; Y: yes.
data. The main limitation is that our conclusions are based on very
N
N
N
N
N
N
N
0
groups
N
N
Y
Y
Y
patients who had a hamstring graft which is by far the most popu-
Concealed
allocation
N
N
N
N
Y
N
Y
Y
Y
Y
Y
Beard (1998)
Grant (2005)
Grant (2010)
each item
5
G Model
JSAMS-2483; No. of Pages 9 ARTICLE IN PRESS
A.R. Gamble et al. Journal of Science and Medicine in Sport xxx (xxxx) xxx–xxx
Table 2
Summary of findings for comparisons between intensive supervised and less supervised rehabilitation following anterior cruciate ligament reconstruction.
Flexion range of motion (degrees for between-limb difference or surgical limb) SMD (95% CI) GRADE
Up to 3 months 182 (2 RCTs) 0.03 (−1.06 to 1.13) Very-lowa,b,c ⊕
3–6 months 90 (2 RCTs) −0.31 (−0.84 to 0.23) Lowa,b ⊕⊕
6–12 months 37 (1 RCT) 0.00 (−0.67 to 0.67) Lowa,b ⊕⊕
12–24 months 88 (1 RCT) 0.20 (−0.22 to 0.62) Lowa,b ⊕⊕
Extension range of motion (degrees for between-limb difference or surgical limb) MD (95% CI) GRADE
Up to 3 months 129 (1 RCT) −1.00 (−1.81 to −0.19) Lowa,b ⊕⊕
3–6 months 37 (1 RCT) 0.00 (−1.99 to 1.99) Lowa,b ⊕⊕
6–12 months 37 (1 RCT) 0.00 (−2.02 to 2.02) Lowa,b ⊕⊕
12–24 months 88 (1 RCT) 1.00 (−0.09 to 2.09) Lowa,b ⊕⊕
Single leg hop for maximum distance (LSI) MD (95% CI) GRADE
Up to 3 months 40 (1 RCT) −0.40 (−15.34 to 14.54) Lowa,b ⊕⊕
3–6 months 77 (2 RCTs) −0.56 (−7.18 to 6.06) Lowa,b ⊕⊕
6–12 months 77 (2 RCTs) 0.29 (−5.74 to 6.31) Lowa,b ⊕⊕
Single leg vertical hop for maximum height (LSI) MD (95% CI) GRADE
Up to 3 months 40 (1 RCT) 10.20 (−5.50 to 25.90) Lowa,b ⊕⊕
3–6 months 40 (1 RCT) 8.90 (−4.48 to 22.28) Lowa,b ⊕⊕
6
G Model
JSAMS-2483; No. of Pages 9 ARTICLE IN PRESS
A.R. Gamble et al. Journal of Science and Medicine in Sport xxx (xxxx) xxx–xxx
Table 2 (Continued)
ACL: anterior cruciate ligament; CI: confidence interval; LSI: limb symmetry index; MD: mean difference; SMD: standardised mean difference; QoL: quality of life; RCT:
randomised controlled trial; SIP: sickness impact profile.
a
Downgraded by one level because >25% of the trials, weighted by their sample size, were at high risk of bias (i.e. scored <7 on the PEDro scale).
b
Downgraded by one level because there is only one study, or where there is more than one study the total sample size is less than 400 (imprecision).
c
Downgraded by one level because I2 ≥ 80% (inconsistency).
outcome, despite 2–5 times more supervised sessions across the might not be necessary for a variety of musculoskeletal and
six trials. There were also no clinically relevant effects sizes. For orthopaedic presentations.39 For example, intensive supervised
example, between-group differences for self-reported function rehabilitation is not superior to less supervised rehabilitation
(a primary outcome) were well-below the MCID of 10 (range: or primarily home-based rehabilitation following knee and hip
0.9–1.8).23 arthroplasty (systematic reviews that both include 5 RCTs,
Since all groups with less supervision in our review had some n = 524 and n = 234 participants, respectively),12,13 following lum-
level of physiotherapist supervision, we cannot recommend home- bar disc surgery (systematic review of 5 RCTs, n = 272),11 for the
exercise without any physiotherapy oversight. Nevertheless, a key management of rotator cuff tendinopathy (one RCT, n = 86),14
implication of these findings is that large amounts of supervised following arthroscopic rotator cuff repair (RCT, n = 117),40 and post-
rehabilitation may not be necessary following ACL reconstruction, immobilization of ankle fractures (Cochrane review of 4 RCTs,
particularly for recreational athletes (as only one trial included n = 366)41 and upper limb fractures (systematic review of 3 RCTs,
elite athletes). This finding is good news for athletes who cannot n = 167).42
afford supervised rehabilitation, have poor access to physiotherapy Guidelines for rehabilitation following ACL reconstruction vary
or have high motivation to perform their rehabilitation inde- in their recommendations for the use of intensive supervised
pendently. It is also comforting considering the effects of the rehabilitation.43 Some guidelines recommend intensive supervised
COVID-19 pandemic, where individual and social circumstances physiotherapy for some people,44 some recommend home-based
have changed access to attend clinic-based appointments. Since protocols for motivated athletes,45 and others highlight the uncer-
intensive supervised rehabilitation does not demonstrate superior tainty of the evidence on this topic (similar to our study).9 Some
outcomes compared to less supervised rehabilitation, clinicians cohort studies suggest intensive clinic-based rehabilitation might
should involve athletes in shared decision making to ensure ath- be superior to primarily home-based programs.46 However, rig-
letes’ rehabilitation decision align with their preferences and orous RCTs are needed to confirm these findings before strong
values. recommendations supporting intensive supervision are made.
Goal setting, pre-operative assessment and education with peri-
odic assessment of objective measures post-operatively could be
4.5. Implications for research and unanswered questions
vital for athletes undergoing ACL reconstruction, particularly those
who prefer to perform most of their rehabilitation with less super-
Very low- to low-certainty evidence in this review highlights the
vision. Goal setting could help individualise programs and increase
need for large, methodologically rigorous RCTs investigating the
athlete motivation to rehabilitate independently. Pre-operative
effectiveness of intensive supervised rehabilitation compared to
education on the use of crutches, importance of early loading and
less supervised rehabilitation following ACL reconstruction. Since
exercise (e.g. weight bearing, range of motion, quadriceps exer-
trials were published between 1997 and 2011 and rehabilitation
cises) with advice on preventing flares in symptoms or jeopardising
protocols were poorly reported, there is also a need to investi-
the integrity of the graft could improve athletes’ confidence and
gate the value of intensive supervised rehabilitation following ACL
self-efficacy.9 Pre-operative assessment could be used to iden-
reconstruction with rehabilitation protocols that are adequately
tify range of motion or strength deficits that an athlete could try
reported (e.g. adhere to the Consensus on Exercise Reporting Tem-
to address before surgery.9 A systematic review of eight RCTs
plate (CERT) or TIDieR checklist) and reflect contemporary practice
(n = 451 athletes) found that pre-operative education and exercise
(i.e. progressive rehabilitation, goal-based progression). Only one
was superior to usual care for improving self-reported function
trial assessed long-term outcomes (i.e. > 12 months)27 . Long-term
and quadriceps strength following ACL surgery.37 Finally, periodic
data is needed to understand whether intensive supervised reha-
assessment of objective measures such as knee strength and hop
bilitation has any long-term benefits (e.g. reduced injury risk,
performance could be used throughout rehabilitation to ensure
increased sports participation, more optimal progression of exer-
athletes reach performance milestones that are associated with a
cises). Another important gap in the literature is whether intensive
reduced risk of re-injury.38
supervision is needed for elite athletes. Elite athletes may have
more advanced rehabilitation needs compared to recreational ath-
4.4. Comparison to existing literature letes and require intensive supervision from a physiotherapist. On
the other hand, they might have higher motivation to complete
Our review adds to the findings of numerous RCTs and sys- rehabilitation with less supervision or might already be working
tematic reviews which suggest intensive supervised rehabilitation with strength and conditioning coaches.
7
G Model
JSAMS-2483; No. of Pages 9 ARTICLE IN PRESS
A.R. Gamble et al. Journal of Science and Medicine in Sport xxx (xxxx) xxx–xxx
8
G Model
JSAMS-2483; No. of Pages 9 ARTICLE IN PRESS
A.R. Gamble et al. Journal of Science and Medicine in Sport xxx (xxxx) xxx–xxx
26. Guyatt G, Oxman AD, Akl EA et al. GRADE guidelines: 1. Introduction-GRADE 37. Alshewaier S, Yeowell G, Fatoye F. The effectiveness of pre-operative exercise
evidence profiles and summary of findings tables. J Clin Epidemiol 2011; physiotherapy rehabilitation on the outcomes of treatment following anterior
64(4):383–394. cruciate ligament injury: a systematic review. Clin Rehabil 2017; 31(1):34–44.
27. Grant JA, Mohtadi NG. Two- to 4-year follow-up to a comparison of home versus 38. Kyritsis P, Bahr R, Landreau P et al. Likelihood of ACL graft rupture: not meeting
physical therapy-supervised rehabilitation programs after anterior cruciate lig- six clinical discharge criteria before return to sport is associated with a four
ament reconstruction. Am J Sports Med 2010; 38(7):1389–1394. times greater risk of rupture. Br J Sports Med 2016; 50(15):946–951.
28. Fischer DA, Tewes DP, Boyd JL et al. Home based rehabilitation for anterior 39. Zadro JR, Harris IA, Abdelshaheed C et al. Choosing wisely after a sport and
cruciate ligament reconstruction. Clin Orthop Relat Res 1998;(347):194–199. exercise-related injury. Best Pract Res Clin Rheumatol 2019; 33(1):16–32.
29. Schenck RC, Blaschak MJ, Lance ED et al. A prospective outcome study of reha- 40. Kim YS, Chung SW, Kim JY et al. Is early passive motion exercise necessary after
bilitation programs and anterior cruciate ligament reconstruction. Arthroscopy arthroscopic rotator cuff repair? Am J Sports Med 2012; 40(4):815–821.
1997; 13(3):285–290. 41. Goost H, Wimmer MD, Barg A et al. Fractures of the ankle joint: investigation
30. Hohmann E, Tetsworth K, Bryant A. Physiotherapy-guided versus home- and treatment options. Dtsch Arztebl Int 2014; 111(21):377–388.
based, unsupervised rehabilitation in isolated anterior cruciate injuries 42. Bruder AM, Shields N, Dodd KJ et al. Prescribed exercise programs may not be
following surgical reconstruction. Knee Surg Sports Traumatol Arthrosc 2011; effective in reducing impairments and improving activity during upper limb
19(7):1158–1167. fracture rehabilitation: a systematic review. J Physiother 2017; 63(4):205–220.
31. Beard DJ, Dodd CA. Home or supervised rehabilitation following anterior cruciate 43. Andrade R, Pereira R, van Cingel R et al. How should clinicians rehabilitate
ligament reconstruction: a randomized controlled trial. J Orthop Sports Phys Ther patients after ACL reconstruction? A systematic review of clinical practice guide-
1998; 27(2):134–143. lines (CPGs) with a focus on quality appraisal (AGREE II). Br J Sports Med 2020;
32. Revenas Å, Johansson A, Leppert J. A randomized study of two physiothera- 54(9):512–519.
peutic approaches after knee ligament reconstruction. Adv Physiother 2009; 44. Arroll B, Robb G, Sutich E. The Diagnosis and Management of Soft Tissue Knee
11(1):30–41. Injuries: Internal Derangements, New Zealand Guidelines Group, 2003. p. 1–100.
33. Ekeland A, Engebretsen L, Fenstad AM et al. Similar risk of ACL graft revision for 45. Wright RW, Haas AK, Anderson J et al. Anterior cruciate ligament reconstruction
alpine skiers, football and handball players: the graft revision rate is influenced rehabilitation: MOON guidelines. Sports Health 2015; 7(3):239–243.
by age and graft choice. Br J Sports Med 2020; 54(1):33. 46. Grindem H, Granan LP, Risberg MA et al. How does a combined preopera-
34. Holland T, Chitkara R, McNicholas M et al. Current trends in ACL reconstruction tive and postoperative rehabilitation programme influence the outcome of ACL
surgery in the United Kingdom. J Orthop Trauma 2017; 7:1–4. reconstruction 2 years after surgery? A comparison between patients in the
35. Rahardja R, Zhu M, Love H et al. Effect of graft choice on revision and contralateral Delaware-Oslo ACL Cohort and the Norwegian National Knee Ligament Registry.
anterior cruciate ligament reconstruction: results from the New Zealand ACL Br J Sports Med 2015; 49(6):385–389.
Registry. Am J Sports Med 2020; 48(1):63–69. 47. Evans L, Hardy L. Injury rehabilitation: a qualitative follow-up study. Res Q Exerc
36. Snaebjörnsson T, Hamrin-Senorski E, Svantesson E et al. Graft diameter and graft Sport 2002; 73(3):320–329.
type as predictors of anterior cruciate ligament revision: a cohort study includ- 48. Heijne A, Axelsson K, Werner S et al. Rehabilitation and recovery after anterior
ing 18,425 patients from the Swedish and Norwegian National Knee Ligament cruciate ligament reconstruction: patients’ experiences. Scand J Med Sci Sports
Registries. JBJS 2019; 101(20). 2008; 18(3):325–335.