Total and partial knee arthroplasty versus non‐surgical interventions of the knee for moderate to severe osteoarthritis

Objectives

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

To assess the benefits and harms of TKA and PKA for people with moderate to severe knee osteoarthritis compared to placebo or sham (efficacy) and to non‐surgical interventions for the knee (comparative effectiveness).

Background

Description of the condition

Knee osteoarthritis is a leading cause of disability worldwide, impacting health on an individual and population level (Bortoluzzi 2018; O'Donnell 2011). Internationally, over 263 million adults are affected by knee osteoarthritis as of 2017 (GBD 2017). Osteoarthritis is a progressive condition characterised by varying degrees of change in the cartilage, bones, and surrounding joint structures (Zhang 2010). Distinctive symptoms accompanying osteoarthritis are joint pain and stiffness, and a decrease in joint mobility and function. These symptoms result in poor health‐related quality of life in patients (NICE 2020a). Symptoms in osteoarthritis are typically fluctuating but manageable; however, a minority of adults with osteoarthritis can experience worsening symptoms (e.g. persistent knee pain, pain at night), indicating the need for knee arthroplasty (Collins 2014; Riddle 2015). 

Knee osteoarthritis is classified based on its severity: mild, moderate, and severe. Severity is determined using many methods, including imaging, clinical symptoms, and physical assessment (Altman 1991; Kellgren 1957; Luyten 2012; Zhang 2010). Radiographically, knee osteoarthritis shows slow evolution of disease over time (3% to 4% per year), with most patients remaining stable (Felson 1995). Yet, it is estimated that an average of 50% of adults diagnosed with knee osteoarthritis will undergo total/partial knee arthroplasty (TKA/PKA) during their lifetime (Weinstein 2013).

Description of the intervention

A broad range of conservative (pharmacological and non‐pharmacological) treatments are available to manage milder osteoarthritis symptoms and a combination of pharmacological and non‐pharmacological therapies are recommended as the first line of treatment (AAOS 2013). Typically, principal core treatments include education (including self‐efficacy and management skills), physiotherapy (e.g. exercises, stretches, walking aids, thermotherapy, cryotherapy, insoles, braces), weight loss, and pharmacological therapies (topical and oral non‐steroidal anti‐inflammatory drugs) (Bruyère 2019; Kolasinski 2020; NICE 2020a). Additional treatment options include intraarticular injections (e.g. corticosteroids), and opioids (Bruyère 2019; Kolasinski 2020).

For moderate to severe osteoarthritis, when conservative treatment fails to alleviate symptoms, evidence‐based guidelines recommend total knee arthroplasty (TKA) and partial knee arthroplasty (PKA) as surgical treatment options (Nelson 2014; Rillo 2016; Zhang 2007, Zhang 2008; Zhang 2010a). TKA and PKA are not advised for milder osteoarthritis as they result in poor outcomes (Youlden 2020), and will require replacement within 15 to 20 years (CIHI 2019). Some international guidelines recommend TKA for individuals who have osteoarthritis symptoms that negatively impact their quality of life (e.g. pain, functional limitations), radiographic evidence of osteoarthritis, or fail to have symptom control using conservative treatments (McGrory 2016; NICE 2020b). Recent evidence‐based clinical practice guidelines recommend considering the perspectives of osteoarthritis patients and engaging them in the TKA decision‐making process by using shared decision‐making approaches (Kolasinski 2020; NICE 2020b). However, there is no agreement on the key clinical indications for TKA in osteoarthritis, which contributes to global clinical practice variations (Katz 2017).

TKA involves the surgical replacement of the surface of the bones (tibia, femur, and occasionally patella) with synthetic prostheses (metal, plastic, ceramic) and the insertion of a spacer (plastic) between the tibia and femur (Foran 2020). PKA involves the same procedure except that it involves replacing only one area of the knee (medial, lateral, or patellofemoral) (Foran 2016). 

How the intervention might work

By replacing the structure of the knee joint, TKA and PKA result in good pain management and increased physical function following these surgeries, leading to better health‐related quality of life (Ethgen 2004; Hammett 2018; Shan 2015; da Silva 2014). Both surgeries are associated with rare adverse events (e.g. mortality, cardiac events) (Morris 2013; Wilson 2019). However, surgical revision of the prosthesis may be required in the short term (e.g. infection) and long term (e.g. prosthesis survival) (Arirachakaran 2015).

Conservative treatments for osteoarthritis help to maintain overall joint and physical function, support joint unloading, and lessen joint pain (AAOS 2013; National Collaborating Center (UK) 2008). 

Why it is important to do this review

Although TKA and PKA are widely recommended surgical treatments to manage moderate to severe osteoarthritis symptoms, up to 34% of patients are dissatisfied with the surgery results (Gademan 2016; Kahlenberg 2018; Scott 2010). Factors contributing to patients' dissatisfaction with TK and PKA vary. 

There is a poor association between surgeons' judgement for TKA/PKA and patients’ informed preferences for these surgeries (Barlow 2015; Frankel 2012; Postler 2020). However, most decisions to undergo TKA/PKA are based on surgeons’ judgement without considering patients’ perspectives (Frankel 2012; Hudak 2002; Waugh 2016; Webster 2013). The criteria deemed most important by orthopaedic surgeons when recommending TKA/PKA are radiographic severity and patient‐reported pain (Churchill 2016; Frankel 2016; Gademan 2016; Skou 2016; Verra 2016). However, the criteria influencing patients’ preference for the surgery are the impact of osteoarthritis on quality of life and pain (Conner‐Spady 2014; Frankel 2012; Gademan 2016; Hsu 2018; Yeh 2017). 

Patients’ preoperative expectations influence the decision to undergo these surgeries (Hawker 2006). However, patients may have unrealistic expectations of TKA/PKA outcomes, leading to their dissatisfaction with the results of surgery (Frankel 2012; Hamilton 2013; Hoffmann 2015; Rosemann 2006; Suarez‐Almazor 2010). Not all patients will benefit from surgery and approximately 20% report persistent postoperative knee pain following TKA (Beswick 2012; Wylde 2011). Many patients report receiving little information on the possible surgical outcomes and alternative treatment options, which can be considered as an uninformed decision to have TKA/PKA (Conner‐Spady 2014; King 2020; Yeh 2017). 

To be consistent with patient‐centred care (CMA 2010), it is essential that patients are given evidence‐based information about the possible benefits and harms of TKA/PKA and alternative treatment options. PKA and TKA are elective procedures in a pool of many alternative treatment options for knee osteoarthritis with no best choice (National Collaborating Center (UK) 2008). As a result, patients need to be aware of all treatment options and given the opportunity to discuss tradeoffs between benefits and harms across options, and to decide which option is most aligned with their informed values and preferences. Patient decision aids are tools that can support this process by providing information on options, benefits, and harms, and helping patients to clarify their values for the outcomes of the options (Stacey 2017). In a Cochrane systematic review of over 100 trials, patient decision aids have been shown to be effective in increasing patients’ knowledge, supporting realistic expectations, and minimising decisional conflict (Stacey 2017). 

However, current TKA/PKA patient decision aids are not optimal for adults with knee osteoarthritis considering these surgeries because they are outdated, based on low‐quality evidence (e.g. non‐randomised controlled trials), or not publicly available (Pacheco‐Brousseau 2021). High‐quality evidence is therefore needed on the effectiveness of TKA/PKA compared to other treatment options to inform healthcare professionals and patients during the decision‐making process (Pacheco‐Brousseau 2021; Stacey 2016).

A recent Cochrane systematic review assessed the benefits and harms of surgical intervention for the management of mild to moderate knee osteoarthritis (Palmer 2019). The authors acknowledged a priori that TKA and PKA were not indicated for milder knee osteoarthritis, but wanted to demonstrate the evidence. Their review found no randomised controlled trials on the effectiveness of TKA and PKA for adults with mild to moderate knee osteoarthritis (Palmer 2019). Our review is different from Palmer 2019 as we are interested in TKA/PKA and non‐surgical interventions for the knee for moderate to severe osteoarthritis. 

This review will be conducted according to the guidelines recommended by the Cochrane Musculoskeletal Editorial Board (Ghogomu 2014).

Objectives

To assess the benefits and harms of TKA and PKA for people with moderate to severe knee osteoarthritis compared to placebo or sham (efficacy) and to non‐surgical interventions for the knee (comparative effectiveness).

Methods

Criteria for considering studies for this review

Types of studies

We will include randomised controlled trials (RCTs). There will be no language restrictions.

Types of participants

The population of interest is adults (aged 18 or older) with a diagnosis of moderate to severe osteoarthritis in one or both knees who are eligible for surgical treatment. We will exclude participants with other joint conditions (e.g. rheumatoid arthritis) or those requiring TKA/PKA because of trauma (e.g. fracture), infection, or surgical revision. We will include trials with a mixture of participants with different joint conditions if we are able to extract data on the eligible participants. 

Types of interventions

We will include TKA or PKA interventions using any surgical approach compared to placebo or sham, and any non‐surgical interventions for the knee (e.g. rehabilitation programmes, weight loss, braces, insoles, pharmacotherapy, any intraarticular injection (e.g. stem cells, glucocorticoids, platelet‐rich plasma). We will pool non‐surgical interventions for the knee together if we judge them to be similar enough based on their description. 

For the purpose of this review, we define placebo surgery as a "procedure in which there is routine delivery of most of the operation, but the presumed active components, or the essential or critical surgical element, are removed" (Beard 2018).

We define sham surgery as entirely simulated surgery or small superficial incisions only, without breaking the skin or with very minimal incision (Beard 2020). 

Usual care non‐surgical interventions that are also provided equally as a co‐intervention in the TKA or PKA group (e.g. rehabilitation programmes, weight loss, braces, insoles, pharmacotherapy, any intraarticular injection (e.g. stem cells, glucocorticoids, platelet‐rich plasma)) are eligible. Co‐interventions may include but are not limited to pharmacotherapy, alternative therapies, preoperative rehabilitation programmes, and lifestyle changes. 

We will exclude patellofemoral arthroplasties and TKA or PKA revisions. We will also exclude studies comparing different types of surgeries (e.g. different prostheses and/or surgical approach). 

We will stratify TKA and PKA in the primary analysis.

Types of outcome measures

Major outcomes

Based on the OMERACT‐OARSI core domain set for measurement in clinical trials of hip and knee osteoarthritis (Smith 2019), and on the proposed outcomes of Cochrane Musculoskeletal, the major outcomes of interest are the following:

1. Pain, with a hierarchy of 11 levels (when more than one is reported, we will use the highest on the list):

• Pain overall

• Pain on walking

• Western Ontario and McMaster Universities Arthritis (WOMAC) index pain subscale

• Pain on activities other than walking

• WOMAC global scale

• Lequesne osteoarthritis index global score

• Other algofunctional scale (e.g. Knee Injury and Osteoarthritis Outcome (KOOS) score)

• Patient’s global assessment

• Physician’s global assessment

• Other continuous pain outcomes

• Non‐continuous pain outcome (e.g. binary)

2. Physical function, with a hierarchy of eight levels (when more than one is reported, we will use the highest on the list):

• Global disability score

• Walking disability

• WOMAC disability subscore

• Composite disability scores other than WOMAC

• Knee‐specific functional scales (e.g. KOOS)

• Disability other than walking

• WOMAC global scale

• Lequesne osteoarthritis index global score

• Other algofunctional scale

3. Knee surgery (including revision or subsequent operation for the TKA/PKA group or initial TKA/PKA for the non‐surgical group

4. Patient satisfaction with treatment outcomes (e.g. satisfaction with postoperative function)

5. Health‐related quality of life

6. Serious adverse events (as defined by authors of included studies, likely infection, deep vein thrombosis, pulmonary embolism, blood transfusion, cardiac events, death)

7. Withdrawals due to adverse events 

Minor outcomes

1. Performance‐based tests (e.g. Timed Up and Go, 6‐Minute Walk Test)

2. All adverse events (e.g. swelling)

3. Cost and cost‐effectiveness

Timing of outcome assessments

We will measure outcomes at short‐term (≤ 6 months), intermediate‐term (6 to 12 months), and long‐term (≥ one year) follow‐up. For efficacy outcomes, the primary analysis will be at long‐term follow‐up. For safety outcomes, we will report and analyse adverse events at all time points.

Minimally important differences

Following the guidance in Chapter 15 of the Cochrane Handbook for Systematic Reviews of Interventions (Schünemann 2021a), we will use the following minimal clinically important differences (MCIDs) for interpreting our results.

We will assume a MCID of 1.5 points in a 10‐point scale for pain, and 10 points on a 100‐point scale for function or disability, for input into the calculator (Tubach 2012). 

Search methods for identification of studies

Electronic searches

We will search the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, and Embase.

We will also conduct a search of ClinicalTrials.gov (www.ClinicalTrials.gov) and the WHO trials portal (www.who.int/ictrp/en/).

We will search all databases from 2010 to the present because of major practice changes in TKA surgical techniques (Blasier 2009). Additionally, the Cochrane review of surgical interventions for mild to moderate knee osteoarthritis, which searched databases from inception to 2018, did not find any trials of TKA or PKA (Palmer 2019). 

We will impose no restriction on language of publication.

See Appendix 1 for the MEDLINE search strategy.

Searching other resources

We will check the reference lists of all primary trials and review articles for additional references. We will search for errata or retractions from included trials published in full text on PubMed (www.ncbi.nlm.nih.gov/pubmed) and the Retraction Watch database for the first author of trials, and we will report the date this was done within the review. We will track citations of included trials using the Bond University Systematic Review Accelerator (https://sr-accelerator.com/#/). 

Data collection and analysis

Selection of studies

Four review authors (LPB, DS, SA, NN) will independently screen titles and abstracts against the eligibility criteria for inclusion of all the potentially relevant trials we identify as a result of the search, and code them as 'retrieve' (eligible or potentially eligible/unclear) or 'do not retrieve'. We will retrieve the full‐text study reports. Two review authors (LPB, SA) will independently screen the full text against the eligibility criteria and identify trials for inclusion; we will record the reasons for exclusion of the ineligible trials. We will resolve any disagreement through discussion with all review authors. We will identify and exclude duplicates and collate multiple reports of the same study so that each study, rather than each report, is the unit of interest in the review. We will record the selection process in sufficient detail to complete a Preferred Reporting Items for Systematic Reviews and Meta‐Analyses (PRISMA) flow diagram (Page 2021a; Page 2021b) and 'Characteristics of excluded studies' table.

Data extraction and management

We will use a pre‐developed data collection form for study characteristics, interventions, and outcome data, which has been piloted on five trials in the review. One review author (LPB) will extract study characteristics from included trials. A second review author (NN) will spot‐check study characteristics for accuracy against the study report. We will extract the following study characteristics:

1. Methods: study design, total duration of study, details of any 'run‐in' period, number of study centres and location, study setting, withdrawals, and date of study.

2. Participants: N, mean age, age range, sex, disease duration, severity of condition, comorbidities, sociodemographics, ethnicity, diagnostic criteria, important baseline data, inclusion criteria, and exclusion criteria.

3. Interventions: total number of intervention arms, intervention details, comparison details, co‐interventions details.

4. Outcomes: all major and minor outcomes specified and collected, and time points reported.

5. Characteristics of the design of the study as outlined below in the Assessment of risk of bias in included studies section.

6. Notes: funding for study, and notable declarations of interest of study authors.

Two review authors (LPB, SA) will independently extract outcome data from included trials. We will extract the number of events and number of participants per treatment group for dichotomous outcomes, and the means and standard deviations (or statistics in which standard deviations can be derived) and number of participants per treatment group for continuous outcomes. We will note in the ‘Characteristics of included studies’ table if outcome data were not reported in a usable way and when data were transformed or estimated from a graph. We will resolve disagreements by consensus within the review author group. We will transfer data into the Review Manager file (RevMan Web 2022). We will double‐check that data are entered correctly by comparing the data presented in the systematic review with the study reports.

We will use PlotDigitizer to extract data from graphs or figures (Plotdigitizer). We will also extract these data in duplicate.

If more than one measure for an outcome is reported, we will prioritise them based on the hierarchy of major outcomes. If data for more than one time point are provided, we will use the longest follow‐up for each time point (e.g. closest to six months for short‐term) for the meta‐analysis.

If both unadjusted and adjusted values for the same outcome are reported, we will extract unadjusted values. If data are analysed based on an intention‐to‐treat (ITT) sample and another sample (e.g. per‐protocol, as‐treated), we will extract ITT. For adverse effects, we will extract as‐treated. If continuous outcome measures are reported in multiple ways, we will follow the following hierarchy: 

1. Final values adjusted for baseline characteristics

2. Final values

3. Change from baseline

Assessment of risk of bias in included studies

Three review authors (LPB, SA, NN) will independently assess risk of bias for each study using the RoB 1 criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2021). We will resolve any disagreements by discussion or within the review author group. We will assess the risk of bias according to the following domains.

1. Random sequence generation.

2. Allocation concealment.

3. Blinding of participants and personnel.

4. Blinding of outcome assessment.

5. Incomplete outcome data.

6. Selective outcome reporting.

7. Other bias (e.g. potential sources of bias reported by authors, participant selection, study design).

If cluster trials are found, we will follow appropriate methods as outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2020a).

We will rate each potential source of bias as high, low, or unclear risk, and provide a quote from the study report together with a justification for our judgement in the risk of bias table. We will summarise the risk of bias judgements across different trials for each of the domains listed. Due to the nature of the trials, we will consider bias due to lack of blinding for all outcomes. In addition, we will consider the impact of missing data by key outcomes.

Where information on risk of bias relates to unpublished data or correspondence with a trialist, we will note this in the risk of bias table.

When considering treatment effects, we will take into account the risk of bias for the trials that contribute to that outcome.

We will present the figures generated by the risk of bias tool to provide summary assessments of the risk of bias.

Assessment of bias in conducting the systematic review

We will conduct the review according to this published protocol and report any deviations from it in the 'Differences between protocol and review' section of the systematic review.

Measures of treatment effect

We will analyse dichotomous data as risk ratios (RR) or Peto odds ratios (OR) when the outcome is a rare event (approximately less than 10%), and use 95% confidence intervals (CIs). We will analyse continuous data as mean difference (MD) or standardised mean difference (SMD), depending on whether the same scale is used to measure an outcome, with 95% CIs. We will enter data presented as a scale with a consistent direction of effect across trials.

When different scales are used to measure the same conceptual outcome (e.g. disability), we will calculate SMDs instead, with corresponding 95% CIs. We will back‐translate SMDs to a typical scale (e.g. 0 to 10 for pain) by multiplying the SMD by a typical among‐person standard deviation (e.g. the standard deviation of the control group at baseline from the most representative study) (Higgins 2020a; Higgins 2020b).

For dichotomous outcomes, we will calculate the number needed to treat for an additional beneficial outcome (NNTB), or the number needed to treat for an additional harmful outcome (NNTH), from the control group event rate and the relative risk using the Visual Rx NNT calculator (Cates 2008). We will calculate the NNTB or NNTH for continuous measures using the Wells calculator (available at the Cochrane Musculoskeletal editorial office, musculoskeletal.cochrane.org/). We will use the minimal clinically important difference (MCID) in the calculation of NNTB or NNTH; we will assume a MCID of 1.5 points in a 10‐point scale for pain, and 10 points on a 100‐point scale for function or disability, for input into the calculator (Tubach 2012). 

For dichotomous outcomes, we will calculate the absolute percent change from the difference in the risks between the intervention and control group using GRADEpro (GRADEpro GDT), and express as a percentage. We will calculate the relative percent change as the absolute value of the risk ratio ‐ 1, and express as a percentage.

For continuous outcomes, we will calculate the absolute percent change by dividing the mean difference by the scale of the measure, and express as a percentage. We will calculate the relative difference as the absolute benefit (mean difference) divided by the baseline mean of the control group, and express as a percentage.

In the 'Effects of interventions' results section and the 'What happens' column of the summary of findings table, we will provide the absolute percent change, the relative percent change from baseline, and the NNTB or NNTH (the NNTB or NNTH will be provided only when the outcome shows a clinically significant difference).

Unit of analysis issues

The unit of analysis will be the participant. However, we will analyse RCTs with knees as the unit of analysis as such, but will specify this in the results section. We do not expect to find RCTs with non‐standard designs (e.g. cross‐over, multiple treatment arms) since the intervention is at the individual level. In the situation that multiple study arms are reported in a single study, we will include only the relevant arms. When possible, we will pool arms together. If two comparisons (e.g. physiotherapy versus TKA and weight loss versus TKA) are combined in the same meta‐analysis, we will halve the control group to avoid double‐counting. For continuous outcomes, we will halve the control group and the mean and standard deviation will remain the same. For binary outcomes, we will halve the number of events in the numerator and denominator in the control group. 

Dealing with missing data

We will contact investigators or study sponsors in order to verify key study characteristics and obtain missing numerical outcome data where possible (e.g. when a study is identified as abstract only or when data are not available for all participants). Where this is not possible, and the missing data are thought to introduce serious bias, we will explore the impact of including such trials in the overall assessment of results by a sensitivity analysis. We will clearly describe any assumptions and imputations to handle missing data and explore the effect of imputation by sensitivity analyses.

For dichotomous outcomes (e.g. number of withdrawals due to adverse events), we will calculate the withdrawal rate using the number of patients randomised in the group as the denominator.

For continuous outcomes (e.g. mean change in pain score), we will calculate the MD or SMD based on the number of patients analysed at that time point. If the number of patients analysed is not presented for each time point, we will use the number of randomised patients in each group at baseline.

Where possible, we will compute missing standard deviations from other statistics such as standard errors, CIs or P values, according to the methods recommended in the Cochrane Handbook for Systematic Reviews of Interventions (Deeks 2020). If standard deviations cannot be calculated, we will impute them (e.g. from other trials in the meta‐analysis).

Assessment of heterogeneity

We will assess clinical and methodological diversity in terms of participants, interventions, outcomes, and study characteristics for the included trials to determine whether a meta‐analysis is appropriate. We will conduct this by observing these data from the data extraction tables. We will assess statistical heterogeneity by visual inspection of the forest plot to assess for obvious differences in results between the trials, and using the I² and Chi² statistical tests.

As recommended in the Cochrane Handbook for Systematic Reviews of Interventions (Deeks 2020), the interpretation of an I² value of 0% to 40% 'might not be important'; 30% to 60% may represent 'moderate' heterogeneity; 50% to 90% may represent 'substantial' heterogeneity; and 75% to 100% represents 'considerable' heterogeneity. As noted in the Cochrane Handbook for Systematic Reviews of Interventions (Deeks 2020), we will keep in mind that the importance of I² depends on: (i) the magnitude and direction of effects and (ii) the strength of evidence for heterogeneity.

We will interpret the Chi² test as a P value ≤ 0.10 indicating evidence of statistical heterogeneity.

If we identify substantial heterogeneity, we will report it and investigate possible causes by following the recommendations in section 10.10 of the Cochrane Handbook for Systematic Reviews of Interventions (Deeks 2020). 

Assessment of reporting biases

We will create and examine a funnel plot to explore possible small study biases if the systematic review includes more than 10 trials (Sterne 2011). In interpreting funnel plots, we will examine the different possible reasons for funnel plot asymmetry as outlined in section 13 of the Cochrane Handbook for Systematic Reviews of Interventions and relate this to the results of the review. If we are able to pool more than 10 trials, we will undertake formal statistical tests to investigate funnel plot asymmetry, and will follow the recommendations in section 13.3 of the Cochrane Handbook for Systematic Reviews of Interventions (Page 2020).

To assess outcome reporting bias, we will check study protocols against published reports. For trials published after 1 July 2005, we will screen the Clinical Trial Register at the International Clinical Trials Registry Platform of the World Health Organization (http://apps.who.int/trialssearch) for the a priori study protocol. We will evaluate whether selective reporting of outcomes is present.

Data synthesis

Main planned comparisons

We will undertake meta‐analyses for the following comparisons using Review Manager Web (RevMan Web 2022).

1. Comparison 1: TKA and PKA versus sham/placebo.

2. Comparison 2: TKA and PKA versus usual care non‐surgical interventions, including but not limited to rehabilitation programmes, weight loss, braces, insoles, pharmacotherapy, any intraarticular injection (e.g. stem cells, glucocorticoids, platelet‐rich plasma), when usual care is not a co‐intervention of TKA.

3. Comparison 3: TKA and PKA plus usual care non‐surgical interventions versus usual care non‐surgical interventions alone.

We will use a random‐effects model and perform a sensitivity analysis with a fixed‐effect model. The primary analysis will include all trials regardless of their risk of bias. 

If a meta‐analysis is not possible due to limited availability of data, we will present the results for included studies in this review in a narrative synthesis, according to the Synthesis Without Meta‐analysis (SWiM) reporting guideline, as outlined in section 12 of the Cochrane Handbook for Systematic Reviews of Interventions (McKenzie 2021).

Subgroup analysis and investigation of heterogeneity

We plan to carry out the following subgroup analyses if sufficient data are available.

1. Age: < 55, 55 to 75, < 75.

2. Sex/gender.

3. Ethnicity – when possible, Caucasian, Aboriginal, South Asian, Chinese, Black, Latin American, Arab, Southeast Asian, west Asian, multiple ethnicities, other ethnicities; or authors reported grouping (Government Canada 2015).

4. Non‐surgical intervention types (e.g. education, physiotherapy, pharmacotherapy).

5. Implant type (e.g. fixed, mobile).

6. Placebo/sham.

We chose these subgroups because: 1) knee osteoarthritis is more prevalent in Caucasian women (Deshpande 2016); 2) there are disparities in demand and recommendation for TKA/PKA based on age, sex/gender, and ethnicity (Borkhoff 2008; Mota 2012; Singh 2014); 3) there are differences in preoperative expectations and postoperative outcomes based on sex and ethnicity (Edwards 2018; Perez 2020; Winemaker 2015); 4) there are different types of non‐surgical interventions for knee osteoarthritis (e.g. education, pharmacotherapy) with a different effect on management (Kolasinski 2020); 5) implant types can potentially influence outcomes (Fransen 2017); 6) placebo and sham surgery are not the same concept (Beard 2020), and it is unclear whether this would make a difference in this setting. 

We will use major outcomes (pain and function) in the subgroup analyses if sufficient data are available.

We will use the formal test for subgroup interactions in Review Manager (RevMan Web 2022), and we will use caution in the interpretation of subgroup analyses as advised in section 10 of the Cochrane Handbook for Systematic Reviews of Interventions (Deeks 2020). We will compare the magnitude of the effects between the subgroups by assessing the overlap of the CIs of the summary estimate. Non‐overlap of the CIs indicates statistical significance.

Sensitivity analysis

We plan to carry out the following sensitivity analyses to investigate the robustness of the treatment effect on pain and function:

1. Unclear or high risk of selection bias (random sequence generation and allocation concealment).

2. Unclear or high risk of detection bias (blinding of outcome assessment).

3. Unclear or high risk of attrition bias (incomplete outcome data).

Summary of findings and assessment of the certainty of the evidence

We will follow the guidelines in the Cochrane Handbook for Systematic Reviews of Interventions, chapter 15 (Schünemann 2021a), for interpreting results, and will be aware of distinguishing a lack of evidence of effect from a lack of effect. We will base our conclusions only on findings from the quantitative or narrative synthesis of included trials for this review. We will avoid making recommendations for practice, and our implications for research will suggest priorities for future research and outline what the remaining uncertainties are in the area.

We will create a summary of findings table for the major outcomes of pain, physical function, knee surgery, patient satisfaction, health‐related quality of life, serious adverse events, and withdrawals due to adverse events. The first summary of findings table will compare TKA and PKA versus sham/placebo. A second summary of findings table will compare TKA and PKA versus usual care non‐surgical interventions. 

Two review authors (LPB, SA) will independently assess the certainty of the evidence with disagreements resolved by discussion or involving a third review author (NN, DS, or GW). We will use the five GRADE considerations (study limitations, consistency of effect, imprecision, indirectness, and publication bias) to assess the certainty of the body of evidence as it relates to the trials that contribute data to the meta‐analyses for the prespecified outcomes, and report the certainty of evidence as high, moderate, low, or very low. We will justify, document, and incorporate judgements into the reporting of results for each outcome. We will use the methods and recommendations described in chapter 14 of the Cochrane Handbook for Systematic Reviews of Interventions (Schünemann 2021).

We will use GRADEpro software to prepare the summary of findings tables (GRADEpro GDT). We will use version 3 of the GRADEPro view to display our summary of findings tables. We will justify all decisions to downgrade the certainty of evidence for each outcome using footnotes, and we will make comments to aid the reader's understanding of the review where necessary. We will provide the NNTB or NNTH, and absolute and relative percent change in the 'What happens' column of the summary of findings table as described in the 'Measures of treatment effect' section above, with the exception of the absolute difference for dichotomous outcomes, which is displayed by default in version 3 of the GRADEPro view.

Appendices

Appendix 1. Preliminary MEDLINE (Ovid) search strategy

C1 ‐ Database: Ovid MEDLINE(R)

‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐

1 osteoarthritis/

2 osteoarthrit$.tw.

3 osteoarthro$.tw.

4 oa.tw.

5 degenerative joint disease.tw.

6 degenerative arthritis.tw.

7 djd.tw.

8 gonarthro$.tw.

9 or/1‐8

10 exp KNEE/

11 knee$.tw.

12 femorotibia$.tw.

13 or/10‐12

14 9 and 13

15 Knee Osteoarthritis/

16 14 or 15

17 exp Arthroplasty, Replacement, Knee/

18 (knee$ adj4 (replace$ or arthroplast$ or prosthe$ or endoprosthe$ or implant$)).tw.

19 (tka or tkr or ukr).tw.

20 Knee Prosthesis/

21 exp Arthroplasty, Replacement/ or exp Arthroplasty/

22 Knee Prosthesis/ or "Prostheses and Implants"/

23 or/17‐22

24 randomized controlled trial.pt.

25 controlled clinical trial.pt.

26 randomized.ab.

27 placebo.ab.

28 clinical trials as topic.sh.

29 randomly.ab.

30 trial.ti.

31 or/24‐30

32 exp animals/ not humans.sh.

33 31 not 32

34 16 and 23 and 33

35 limit 34 to yr="2010 ‐Current"

Contributions of authors

LPB wrote the protocol. JPP developed the search strategy. LPB, JPP, DS, SA, NN, GD, SK, and GW contributed to the decisions and development of the protocol and approved the final protocol.

Sources of support

Internal sources

• NA, Other

  No support was provided.

External sources

• First author's PhD scholarships, Canada

  LPB is supported by the Arthritis Society PhD Salary Award (#21‐0000000085) and matched funding from the University of Ottawa and LPB co‐supervisors research funds, the Ontario Graduate Scholarship, and the University of Ottawa Admission Scholarship. In the last three years, LPB received support from the Hans K. Uhthoff MD FRCSC Graduate Fellowship (#712240301930), the Queen Elizabeth II Graduate Scholarships in Science and Technology, the University of Ottawa Excellence Scholarship, the Eastern District of the Ontario Physiotherapy Association (#712140302327 and #712200305332), and l’Ordre Professionnel de la Physiothérapie du Québec. The funding bodies had no role in the design, preparation, review, approval, or any other decisions pertaining to the study. 

Declarations of interest

LPB, SA, SK, JPP, NN, DS, and GW have no conflicts of interests to declare. GD is a consultant with Microport Orthopedics and has previously received research funding from them. GD had a speaking engagement with Zimmer Biomet Canada in the last 12 months.

New