The purpose of this clinical practice guideline document is to describe the evidence of effectiveness for the components of rehabilitation after ACLR. This information can then be used to inform ACLR rehabilitation protocols. This guideline is intended to be used by physiotherapists managing patients after ACLR in outpatient clinics. Physicians, orthopaedic surgeons, athletic trainers, nurse practitioners and other healthcare professionals may also benefit from this guideline. Insurance payers, governmental bodies and health-policy decision-makers may also find this guideline to be useful as an evolving standard of evidence regarding rehabilitation after ACLR. Additional key users of this guideline include researchers since this document may highlight gaps in the literature and grey areas that require future research.

We followed the Cochrane Handbook for Systematic Reviews of Interventions20 and the Prisma in Exercise, Rehabilitation, Sport medicine and SporTs science tool.21 We adhered to the refined Appraisal of Guidelines for REsearch & Evaluation (AGREE II) instrument22 to ensure the methodological rigour and transparency.

A Guideline Development Group (GDG) was established comprising impartial clinical and methodology experts (nine physiotherapists/researchers, RK, VK, OB, DM, MP, AB, JL, JW and RW) from Aspetar, Orthopaedic and Sports Medicine Hospital, Doha, Qatar. The GDG consisted of two women (one as first author) and seven men, junior, mid-career and senior researchers of different ethnicities. A patient after ACLR (also physiotherapist) was part of the guideline’s development group. We did not include patient opinion or other stakeholders via focus groups.

At the first meeting, the GDG reviewed and finalised the scope of the guideline and agreed on the set of population, intervention or exposure, comparator, outcome questions, and critical and important outcomes to be assessed. Selected outcomes included: adverse events, return to activity, pain, laxity, strength, muscle atrophy, range of motion, subjective function using patient-reported outcome measures (PROMs), swelling, functional activities, proprioception and balance. Next, the chair of the GDG coordinated the commissioning of literature searches and systematic evidence reviews and the GDG subworking groups met to review the literature.

The following databases were searched from inception to 27 December 2021: MEDLINE (PubMed), EMBASE (Elsevier), Cochrane Library (Wiley), CINAHL (EBSCO) and SPORTDiscus (EBSCO) (online supplemental file—systematic search strategy). We included peer reviewed, English language, randomised clinical trials (RCTs) in patients after ACLR that compared between physical therapy interventions or against no intervention, placebo or standard care. We excluded randomised trials in patients after ACL treated non-operatively, in patients after completion of their rehabilitation, children (<16 years), studies reporting only biomechanical results, studies reporting only on concomitant injuries such as other knee ligament injuries, meniscal or cartilage injuries, surgical decisions (eg, brace), nutritional and psychological interventions (online supplemental file—study selection and criteria). As it would be unethical to assign patients to return to sport without meeting criteria, it is unlikely there will ever be RCT data on this aspect. Accordingly, for the recommendations regarding return to activities, we included only systematic and scoping reviews.

All eligible articles were first screened by title and abstract independently by three pairs of two GDG members, and subsequently the full texts of trials that were identified as potentially eligible were retrieved and assessed. For each eligible trial, pairs of GDG members extracted data independently using a standardised, pilot tested, data extraction form developed in accordance with the Cochrane Handbook for Systematic Reviews of Interventions.20 GDG members collected information regarding patient characteristics (age, sex, type of graft used) and outcomes of interest (means or medians and measures of variability for continuous outcomes, the number of participants analysed and the number of participants who experienced an event for dichotomous outcomes).

We used a priori-defined rules for data extraction: (1) We did not include manual testing as a valid method to measure strength.23 (2) If data are reported in several ways, we chose to extract results in the following order: difference from baseline>limb symmetry index>raw data. (3) Swelling outcome was extracted if measured at mid-patella (not above or below). (4) Atrophy outcome was extracted if measured>7.5 cm above patella. Data were extracted from figures and graphs when necessary. Continuous data were transformed to mean and SD. Discrepancies were resolved by discussion and, when necessary, with adjudication by the GDG chair.

Extracted data were imported to Review Manager V.5.4 (Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2014) for analysis. We summarised the effect of interventions on continuous outcomes, using the standardised mean difference (random effects) and corresponding 95% confidence interval. For dichotomous outcomes, we used the risk ratio and corresponding 95% confidence interval. When more than one study reported results for the same outcome, data was pooled. Cohen’s criteria were used to interpret pooled standardised mean difference: large effect≥0.8, moderate effect 0.5–0.8 and small effect 0.2–0.5.24

Risk of bias was assessed using a revision of the Cochrane tool for assessing risk of bias in randomised trials (RoB V.2.0).25 Risk of bias for systematic reviews included in the recommendations was assessed using the ROBIS tool.26

We followed the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) approach (https://gdt.gradepro.org/app/handbook/handbook.html), and used the GRADEpro Guideline Development Tool online software (https://www.gradepro.org/) to assess the quality of the body of evidence and develop and report the summary of findings tables (online supplemental file—summary of findings tables). We rated the certainty of evidence for each comparison and outcome as high, moderate, low, or very low, based on considerations of: risk of bias, inconsistency, indirectness and imprecision (online supplemental file—GRADE evidence assessment) (table 1). To assess publication bias, we planned to generate funnel plots for meta-analyses including at least 10 trials.27

Certainty of evidence grades

As an additional step, we summarised the evidence findings, in a clinically meaningful way, following the a priori-defined rules: (1) When available, we prioritised pooled results coming from more than one study, over results from single studies. (2) For muscle strength outcomes, we prioritised concentric assessment over isometric. (3) For isokinetic outcomes, we prioritised results in slower-speed over results in higher-speed (degrees/second). Findings were stratified according to the rehabilitation period in phases: very early (<1 month), early (1–2 months), intermediate (2–4 months) and advanced (>4 months).

For better understanding and interpretation of the evidence findings, the wording of the summary reads as follow: (1) ‘might/can/may’ for suggestive of improvement or relatively consistent beneficial effect; (2) the ‘size of effect’ or ‘no effect’ when consistent significant and clinically relevant findings (either in favour or against); and (3) ‘conflicting’ when findings were both in favour or against. Due to the extensive list of outcomes (Supplementary file—summary of critical and important outcomes), we opted to report those deemed clinically important in the results section and the full report of findings is available as online supplemental (summary of findings tables).