Total joint replacement surgery for knee osteoarthritis and other non‐traumatic diseases: a network meta‐analysis

Abstract

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

To assess the benefits and harms of TKA compared to other surgical therapies and medical therapies and comparison of various surgical approaches, techniques, and implants for TKA in patients with knee osteoarthritis and other non‐traumatic diseases.

Background

Description of the condition

Arthritis is a major public health problem in the world (Helmick 2008; Bone and Joint Decade).  In 2003, the total cost burden of arthritis was $128 billion in the U.S. (CDC 2007) and led to 992,100 annual hospitalizations (Helmick 2008) and 44 million annual outpatient visits in the U.S. (American Academy of Orthopaedic Surgeons).  Of the $128 million expense, $47 billion was in indirect costs (primarily lost earnings).

Osteoarthritis is the most common form of arthritis in adults. Osteoarthritis alone was associated with indirect costs between $3.4 and $13.2 billion each year in the U.S. (Buckwalter 2004; CDC 2010).  Osteoarthritis commonly affects hands, knees, hips and other joints. Osteoarthritis, the leading cause of primary total knee arthroplasty (TKA), was the 9^th top cause of years of life lived with disability (YLD) in the U.S. in 2010 (Murray 2013) and the 11^th leading cause of YLDs globally (Vos 2012). Knee osteoarthritis is treated with multiple treatments including non‐steroidal anti‐inflammatory drugs (NSAIDs), analgesics such as acetaminophen, narcotic pain medications etc., but they have side effects and may not be effective in advanced osteoarthritis (Zhang 2007; Singh 1998). TKA is an option for such patients with knee osteoarthritis with refractory disease.

Although the commonest underlying diagnosis for TKA is knee osteoarthritis, other non‐traumatic arthritic conditions, including rheumatoid arthritis, psoriatic arthritis, ankylosing spondylitis and avascular necrosis, can also lead to TKA. TKA is associated with significant improvement in pain, function and quality of life of patients (Ethgen 2004; Jones 2003; Kane 2005; Wright 2004).

Description of the interventions

TKA is undoubtedly one of the most successful treatments for patients with end‐stage knee arthritis (AAOS).  The improvements in pain, function, patient satisfaction and quality of life are impressive (Kane 2005; Ethgen 2004; Jonsson 1991; Hawker 2004).  Since its introduction almost four decades ago, it has become an increasingly common surgery.  Recent estimates project a 6‐fold increase in yearly TKA volume in the U.S. by 2030, and reported the volume of TKA at 0.5 million per year in 2005 in the U.S. (Kurtz 2007).  Several Cochrane systematic reviews have been performed, addressing various surgical approaches (patellar resurfacing versus not; posterior cruciate ligament preserving versus not) and implant types for TKA (Harvey 2010; Jacobs 2004; Jacobs 2005; Khan 2008; McDonald 2004). 

TKA includes replacement of both surfaces of a knee joint, the femoral and tibia. In some cases, the patellar surface is also resurfaced. TKA surgery includes the removal of the distal end of the femur bone and replacement with a metal shell, removal of the proximal end of tibia and replacement with a metal plate and stem with a polyethylene insert at the surface to restore mobility and balancing of ligaments to provide stability of knee joint. Cement may be used to fix the components to the bone.  Additional procedures can be performed on soft tissues of the knee, such as sparing or sacrificing the posterior cruciate ligament, lengthening or shortening of ligaments and tendons to achieve adequate balancing, debridement and denervation of the patella if not resurfaced.

How the intervention might work

TKA provides symptom relief in patients with severe arthritis by removal of pathologic joint tissue and restoration of the joint anatomy and function by the use of metal or plastic components leading to stable biomechanics. Although the exact mechanism of symptom relief with TKA is unclear, removal of diseased tissue and restoration of joint anatomy and biomechanics leads to significant pain relief, improvement of function and quality of life in patients who undergo these procedures.

Why it is important to do this overview

A lot of progress has been made in the last two decades in improving the outcomes of TKA in various aspects of care of patients undergoing TKA (pre‐, intra‐ and post‐operative).  As described above, while Cochrane systematic reviews exist for some aspects of TKA, many have not been updated.  In addition, systematic evaluation of several surgical aspects of TKA has not been performed. This network meta‐analysis (NMA) attempts to put all evidence in one place for patients, surgeons and policy makers. An overview of literature in the area of total knee replacement is needed to help practitioners, surgeons and patients to have a clearer understanding of benefits and harms of various surgical interventions used in patients undergoing TKA. 

Objectives

To assess the benefits and harms of TKA compared to other surgical therapies and medical therapies and comparison of various surgical approaches, techniques, and implants for TKA in patients with knee osteoarthritis and other non‐traumatic diseases.

Methods

Criteria for considering reviews for inclusion

Note this section refers to studies for inclusion:

Types of studies

We will include all randomized controlled trials (RCTs) and controlled clinical trials (CCTs) that evaluate any aspect of TKA (e.g. choice of implant, modality of fixation). 

Types of participants

Adults (aged 16 or higher) with knee osteoarthririts and other non‐traumatic diseases will be considered for inclusion. Trials that include traumatic or post‐traumatic arthritis will be included if osteoarthritis and non‐traumatic arthritis constitutes more than 80% of patients.

Types of interventions

We plan to make the following specific comparisons. Additional comparisons will be made as applicable, since the network meta‐analysis covers all surgical aspects of TKA care.

1. Total knee replacement surgery compared to other surgical therapies

2. Total knee replacement surgery compared to medical therapies

3. Different surgical techniques for TKA compared to each other (PCL retaining versus PCL sacrificing; patellar resurfacing versus not resurfacing; comparison of other surgical techniques)

4. Different types of TKA implants compared to each other (metal‐backed versus all‐polyethylene tibial component; rotating platform versus fixed‐bearing; other comparisons of implant designs and types)

5. Different methods of performing TKA (Computer assisted navigation vs. manual instrumentation; different surgical approaches)

Types of outcome measures

Major Outcomes: We will have following seven major outcomes, which will be presented in the SoF tables.

1. Pain: measured with pain severity scales, semi‐quantitative or quantitative such as numeric rating (NRS), visual analog scale (VAS) and other similar continuous or nominal scales

2. Function: as assessed by Western Ontario McMaster Osteoarthritis Index (WOMAC), OXFORD knee, Knee Osteoarthritis and Outcome Score (KOOS), American Knee Society Score (AKS) etc.

3. Health‐related Quality of life (HRQoL), as assessed by Short form‐36 (SF‐36), Short Form‐12 (SF‐12) or other instruments

4. Revision rate

5. Any adverse event: defined as any undesirable experience associated with the use of a medical product in a patient, in this case, TKA (FDA 2013).

6. Serious adverse events (SAEs): as defined in each study; usually these include an AE that is life‐threatening or leads to Hospitalization (initial or prolonged), Disability or Permanent Damage, Congenital Anomaly/Birth Defect or death (FDA 2013).

7. Death

Minor Outcomes will include the following:

1. Cardiac adverse events: myocardial infarction, cardiac arrest, arrhythmia, malignant hypertension, congestive heart failure and others

2. Pulmonary adverse events: pneumonia, pneumothorax, venous thromboembolism and others

3. Local complications: wound Infection, Septic arthritis, Bleeding

4. Cost: inpatient costs, outpatient costs, total costs

5. Inpatient and outpatient health care utilization

6. Patient satisfaction

7. Reoperation rate

8. X‐ray measured with valid quantitative techniques (such as radiostereometric assessment [RSA])

9. Range of motion in degrees, including flexion, extension, rotation etc.

Search methods for identification of reviews

Note this section refers to identification of studies:

We will search the following databases to find published RCTs or CCTs: MEDLINE, EMBASE, The Cochrane Library including the Cochrane Central Register of Controlled Trials (CENTRAL), CINAHL, and ISI Web of Knowledge, as described respectively in Appendix 1, Appendix 2, Appendix 3, Appendix 4, Appendix 5 and Appendix 6.

We will search the WHO International Clinical Trials Registry Platform (www.who.int/ictrp/en/) and ClinicalTrials.gov (www.ClinicalTrials.gov).

We will limit to studies published in English, due to feasibility issues.

We will also screen relevant Cochrane systematic reviews and reference lists of included studies to identify additional RCTs or CCTs.

Data collection and analysis

Selection of reviews

Note this section refers to selection of studies:

Due to lack of availability of reviews in all the areas included in our overview, lack of availability of updated reviews for those with existing Cochrane reviews and challenges with presentation of different outcomes in various Cochrane reviews, similar to our previous overview (Singh 2011), we will obtain data directly from all eligible RCTs or CCTs.

Two reviewers will independently review titles and abstracts to identify relevant studies for inclusion. 

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 PRISMA flow diagram and 'Characteristics of excluded studies' table.

Data extraction and management

Two reviewers will extract the data independently in duplicate by using a structured Excel sheet (Microsoft Inc.).  This will be piloted first on ten articles. Any disagreements on extractions will be resolved by discussion. 

We will use software for extraction for graphs, e.g. PlotDigitizer to extract data from graphs or figures. This data will also be extracted in duplicate.

The number of events and number of participants per treatment group for dichotomous outcomes, and means and standard deviations and number of participants per treatment group for continuous outcomes will be extracted. We will note in the 'Characteristics of included studies' table if outcome data was not reported in a usable way and when data was transformed or estimated from a graph.

We will extract data at the last follow‐up time‐point for each outcome.

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 is 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 studies in the overall assessment of results by a sensitivity analysis. Any assumptions and imputations to handle missing data will be clearly described and the effect of imputation will be explored by sensitivity analyses.Missing data will be imputed as follows: When SD is missing, and if standard error or CI are reported instead of standard deviation, we will calculate SD, as recommended, i.e., SD = SE X sqrt (n); SD = (upper CI ‐lower CI) X sqrt (n)/1.96.

Where possible data presented by at patient‐level will be used. Where only joint‐level data are presented, we will extract and use these data. We are aware of potential unit of analysis errors if the level at which randomization occurred does not match the number of ‘units’ that were randomized. In this case, we will consult the Cochrane Handbook guidance and with a statistician. Where multiple trial arms are reported in a single trial, we will include only the relevant arms. If two comparisons (e.g. drug A versus placebo and drug B versus placebo) are combined in the same meta‐analysis, we will halve the control group to avoid double‐counting and note this in the Table of Included Studies.

For pain and function, we will use the following hierarchy (Ghogomu 2014).

If trials report more than one measure for an outcome :

• for pain, we will use a hierarchy of 11 levels:

  • pain overall (global pain);

  • pain on walking;

  • WOMAC pain subscale;

  • pain on activities other than walking;

  • WOMAC global scale;

  • Lequesne osteoarthritis index global score;

  • other algofunctional scale;

  • patient’s global assessment;

  • physician’s global assessment;

  • other outcome;

  • no continuous outcome reported.

When more than one is reported, we will take the highest on the list – for example choose global pain over WOMAC pain subscale.

• for physical function, we will choose a continuous scale over an ordinal scale, and use a hierarchy of eight levels;

• • global disability score;

  • walking disability;

  • WOMAC disability subscore;

  • composite disability scores other than WOMAC;

  • disability other than walking;

  • WOMAC global scale;

  • Lequesne osteoarthritis index global score;

  • other algofunctional scale.

When more than one is reported, we will take the highest on the list.

• if both final values and change from baseline values are reported for the same outcome, we will extract the final values;

• if both unadjusted and adjusted values for the same outcome are reported, we plan to extract the adjusted values; and,

• if data are analyzed based on an intention‐to‐treat (ITT) sample and another sample (e.g. per‐protocol, as‐treated), we plan to extract ITT sample values for both benefits and harms.

• if there are multiple time points, we will abstract the data for three time‐frames: (a) short‐term: less than three months after the surgery; (b) intermediate‐term: three to six months after the surgery; and (c) long‐term: more than six months after the surgery.

Assessment of methodological quality of included reviews

Note this section refers to included studies:

The assessment of risk of bias in each of the included studies will be done using the Cochrane Collaboration recommendations. We will assess the risk of bias for each criteria as “high risk”, “low risk” or “uncertain risk” as per the definitions and examples in the Cochrane Handbook: allocation concealment, random sequence generation, presence of blinding in the studies for both patients and personnel (performance bias) and assessors (detection bias), incomplete outcome data, selective outcome reporting and evidence of major baseline imbalance (Higgins 2011a).

Data synthesis

Our plan is to do the main analysis on all studies that satisfy PICO, for the pre‐specified major outcomes. Pre‐specified subgroup and sensitivity analysis will be performed when data are available. We will used risk ratio (RR) and 95% confidence intervals (CI) for categorical measures (complications, AEs, SAEs, death, specific complications) and mean difference and 95% CI for continuous outcomes (pain, function, HRQOL, utilization, costs etc.), or standardized mean difference (SMD) if continuous outcomes are reported using different scales.

For outcomes that are rare, we will use Peto’s odds ratio (OR). We will use the random‐effects model, a more conservative method than fixed‐effects model, as the primary method for the meta‐analysis. 

For dichotomous data, we will calculate the absolute risk difference by using RevMan to generate the Risk Difference analysis and then reporting the result as a percentage. The relative percent change will be calculated by finding the relative risk (RR) from RevMan and then applying the formula RR‐1 equals the relative percent change. We will calculate the number needed to treat to benefit (NNTB) or harm (NNTH), using the Cates calculator Visual Rx (available at www.nntonline.net) (Cates 2008). The Cates calculator provides the NNT by using the pooled RR (or OR) and the control event rate (if not available the control group event rate will be used).

For continuous data, we will calculate the absolute change (benefit or harm) from the mean difference (divided by the number of points on the scale and expressed a per cent). The relative difference will be calculated as the absolute benefit from baseline to follow‐up divided by the baseline mean of the most representative control group from a trial). For statistically significant outcomes for continuous measures, we will calculate NNT using the Wells calculator software available at the Cochrane Musculoskeletal Group editorial office, using the well‐established published minimal clinical important difference (MCID's) thresholds for pain, function and HRQOL scales (for example the 5‐point threshold for Short‐Form 36 subscales and 2.5‐point threshold on physical and mental component summary scales, 0.22 units on health assessment questionnaire etc.).

If we use SMD, for absolute change, we will multiply the SMD by the baseline standard deviation of the control group (from the most representative trial in the meta‐analysis), divide by the scale, express as a percent. For relative change we will have the absolute change divided by baseline mean of control group (of the most representative trial), and express as a percent.

The major outcomes listed above will be presented in "Summary of Findings" table and we will use the five GRADE considerations ‐study limitations, consistency of effect, imprecision, indirectness and publication bias‐ for assessing the overall quality of the body of evidence for key outcomes (Atkins 2004; Atkins 2004a; Schünemann 2008; Schünemann 2008a). For major outcomes 1‐4, we will present the long‐term outcome; for major outcomes 5‐7, we will select short‐term outcomes, since these complications related to surgery are most relevant in the short‐term postoperative period.

We will follow the general methods of meta‐analyses, as described in the Cochrane Handbook (Higgins 2011). Additional specific details for network meta‐analyses are provided below and are similar to methods of our recent Cochrane overview and network meta‐analysis (Singh 2011).

Methods for the Bayesian mixed treatment comparison:

Bayesian NMAs will be conducted for pre‐specified outcomes. The effect estimate chosen (Risk Ratio, Mean Difference etc.) will depend on the pre‐specified outcome of interest and availability of data. We will conduct a Bayesian network meta‐analysis using a model which accommodates complex interventions, such as surgeries incorporating various specific surgical procedures (Welton 2009).  Both fixed and random‐effects network meta‐analyses will be conducted; model fit for Bayesian analyses will be based on the deviance information criterion (DIC) and comparison of residual deviance to number of unconstrained data points (Dias 2011; Dias 2011a; Dias 2011b; Dias 2011c).  R (R Foundation for Statistical Computing, Vienna, Austria) and WinBUGS (MRC Biostatistics Unit, Cambridge, UK) will be used for Bayesian network meta‐analyses according to the routine which accommodates evidence structures which may consist of multi‐arm trials as developed at the Universities of Bristol and Leicester (www.bris.ac.uk/cobm/research/mpes/).

Heterogeneity across trials in terms of patient characteristics, trial methodologies, and treatment protocols will be carefully assessed. Provided sufficient data is available to inform the evidence network, to further investigate heterogeneity, subgroup analyses and meta‐regressions (Dias 2011b; Dias 2011c) will be conducted exploring the effect of various characteristics (key demographic, medical, and study design characteristics).  We will also perform analyses including removal of studies from the network of therapies that were not scored as being of high quality.

We will formally assess consistency using back‐calculation and node splitting techniques and informally assess consistency between direct and indirect evidence by comparing direct estimates obtained from pairwise meta‐analysis with estimates from the Bayesian network meta‐analysis (Dias 2010). We will also use the Bucher method (as applicable) to compare the estimate from direct vs. indirect comparison using a simple Z‐test and examining for statistical significance. If there is no statistically significant difference noted, then we will pool direct and indirect estimates. If the direct and indirect estimates are different, we will consider the network as inconsistent and explore and explain the reasons for it, by examining the clinical characteristics in various trials included in the analyses.

Model diagnostics including trace plots and the Brooks‐Gelman‐Rubin statistic will be assessed to ensure model convergence.

At least two chains will be fit in WinBUGS for each analysis, each employing at least 40,000 iterations, with a burn‐in of at least 20,000 iterations (Spiegelhalter 2003; Ntzoufras 2009; Dias 2011a).  Posterior densities for unknown parameters will be estimated using Markov Chain Monte Carlo (MCMC) methods.  Basic parameters will be assigned non‐informative or vague prior distributions; more informative priors will be considered after evaluation of the information base, clinical expert advice and empirical expert evidence (Turner 2012).

Point estimates and 95% credible intervals will be used to summarize findings. The probability of a comparator being optimal will be estimated for each outcome based on the proportion of MCMC simulations in which its relative measure of effect was best.  The mean rank for each comparator will also be calculated.  Graphical methods and numerical summaries will be developed for presenting results from network meta‐analysis (Salanti 2011).

We will test the transitivity assumption, which implies that for indirect comparisons the comparison standard is similar between two trials testing two different treatments of interest (TKA vs. another surgery), and that the distribution of effect modifiers of the relative treatment effects is similar between two trials (Salanti 2012). Some examples of effect modifiers in our study are underlying diagnosis (osteoarthritis vs. non‐osteoarthritis) and elective vs. non‐elective surgery. We will examine the distributions of these variables across comparisons.

Due to a large number of comparisons (due to multiple important outcomes), it is possible that some findings may be due to chance. We will therefore, interpret our findings in light of this limitation. A more common scenario is few data for each outcome and missing a significant effect, when one truly could have been there, had there been enough data points. We are also aware of this limitation.

Sensitivity analysis: Small‐study bias

We will compare the fixed‐ and random‐effects estimates of the intervention effect, when there is between‐study heterogeneity (I‐squared >0). If the estimates are similar, it implies that any small‐study effects had little effect on the intervention effect estimate. In presence of differences between the random and fixed effects, if the larger studies are typical of the use of intervention in practice, then we will report the results of meta‐analyses limited to the larger studies only.

We plan to investigate the robustness of the treatment effect by restricting the analysis to studies with low overall risk of bias, defined based on blinding and on allocation concealment, for the major outcomes of pain and function.

Sub‐group Analyses and Planned Comparisons

We plan the following subgroup analyses for the pairwise comparisons of two main beneficial outcomes, pain and function. Most trials provide data for one subgroup or another, therefore these comparisons will be done by trial characteristics. If a trial includes both subgroups and provides data for each subgroup, then these data will be abstracted during data abstraction from each trial, if provided. We will compare the estimates informally between the subgroups. If enough data are available, we will attempt meta‐regression analyses, as described in the Cochrane Handbook (Higgins 2011).

1. By the length of follow‐up: (a) short‐term: less than three months after the surgery; (b) intermediate‐term: three to six months after the surgery; and (c) long‐term: more than six months after the surgery.

Appendices

Appendix 1. Appendix 1. Medline Search Strategy

MEDLINE

1. exp arthritis/

2. ((rheumatoid or reumatoid or revmatoid or rheumatic or reumatic or revmatic or rheumat$ or reumat$ or revmarthrit$) adj3 (arthrit$ or artrit$ or diseas$ or condition$ or nodule$)).tw.

3. Spondylitis, Ankylosing/

4. (ankylosing or spondyl$).tw.

5. osteoarthr$.tw.

6. (arthrit$ or arthros$).tw.

7. jia.tw.

8. avascular necrosis.tw.

9. or/1‐8

10. Arthroplasty/

11. Joint Prosthesis/

12. "Prostheses and Implants"/

13. Knee/

14. exp Knee Joint/

15. or/10‐12

16. 13 or 14

17. 15 and 16

18. Arthroplasty, Replacement, Knee/

19. Knee Prosthesis/

20. (tka or tkr).tw.

21. (knee$ adj (replace$ or arthroplast$ or prosthe$ or endoprosthe$ or implant$)).tw.

22. or/17‐21

23. 9 and 22

24. randomized controlled trial.pt.

25. controlled clinical trial.pt.

26. randomized.ab.

27. placebo.ab.

28. drug therapy.fs.

29. randomly.ab.

30. trial.ab.

31. groups.ab.

32. or/24‐31

33. (animals not (humans and animals)).sh.

34. 32 not 33

35. 23 and 34

35. (2005$ or 2006$ or 2007$ or 2008$ or 2009$ or 2010$).ed.

36. 34 and 36

 

Appendix 2. EMBASE search strategy

EMBASE

 

1. exp arthritis/

2. ((rheumatoid or reumatoid or revmatoid or rheumatic or reumatic or revmatic or rheumat$ or reumat$ or revmarthrit$) adj3 (arthrit$ or artrit$ or diseas$ or condition$ or nodule$)).tw.

3. ankylosing spondylitis/

4. (ankylosing or spondyl$).tw.

5. osteoarthr$.tw.

6. (arthrit$ or arthros$).tw.

7. jia.tw.

8. avascular necrosis.tw.

9. or/1‐8

10. exp Arthroplasty/

11. exp Joint Prosthesis/

12. exp "Prostheses and Implants"/

13. exp KNEE/

14. exp Knee Joint/

15. or/10‐12

16. 13 or 14

17. 15 and 16

18. exp Arthroplasty, Replacement, Knee/

19. exp Knee Prosthesis/

20. tka.tw.

21. (knee$ and (replace$ or arthroplast$ or prosthe$)).tw.

22. or/17‐21

23. 9 and 22

24. (random$ or placebo$).ti,ab.

25. ((single$ or double$ or triple$ or treble$) and (blind$ or mask$)).ti,ab.

26. controlled clinical trial$.ti,ab.

27. RETRACTED ARTICLE/

28. or/24‐27

29. (animal$ not human$).sh,hw.

30. 28 not 29

31. 23 and 30

32. (2010$ or 2011$).em.

33. 31 and 32

34. 2010$.em

35. 33 and 34

Appendix 3. The Cochrane Library search strategy

 

The Cochrane Library

 

#1           MeSH descriptor Arthritis explode all trees

#2           ((rheumatoid or reumatoid or revmatoid or rheumatic or reumatic or revmatic or rheumat* or reumat* or revmarthrit*) near/3 (arthrit* or artrit* or diseas* or condition* or nodule*)):ti,ab

#3           MeSH descriptor Spondylitis, Ankylosing, this term only

#4           (ankylosing or spondyl*):ti,ab

#5           osteoarthr*:ti,ab

#6           (arthrit* or arthros*):ti,ab

#7           jia:ti,ab

#8           (avascular next necrosis):ti,ab

#9           (#1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7 OR #8)

#10         MeSH descriptor Arthroplasty explode all trees

#11         MeSH descriptor Joint Prosthesis, this term only

#12         MeSH descriptor Prostheses and Implants explode all trees

#13         MeSH descriptor Knee, this term only

#14         MeSH descriptor Knee Joint explode all trees

#15         (#10 OR #11 OR #12)

#16         (#13 OR #14)

#17         (#15 AND #16)

#18         MeSH descriptor Arthroplasty, Replacement, Knee, this term only

#19         MeSH descriptor Knee Prosthesis, this term only

#20         (tka or tkr):ti,ab

#21         (knee* next (replace* or arthroplast* or prosthe* or endoprosthe* or implant*)):ti,ab

#22         (#17 OR #18 OR #19 OR #20 OR #21)

#23         (#9 AND #22)

Appendix 4. CINAHL Search Strategy

 

CINAHL

 

S46 S16 and S32 and S45

S45 S33 or S34 or S35 or S36 or S37 or S38 or S39 or S40 or S41 or S42 or S43 or S44

S44 TI Allocat* random* or AB Allocat* random*

S43 (MH "Quantitative Studies")

S42 (MH "Placebos")

S41 TI Placebo* or AB Placebo*

S40 TI Random* allocat* or AB Random* allocat*

S39 (MH "Random Assignment")

S38 TI Randomi?ed control* trial* or AB Randomi?ed control* trial*

S37 AB singl* blind* or AB singl* mask* or AB doub* blind* or AB doubl* mask* or AB trebl* blind* or AB trebl* mask* or AB tripl* blind* or AB tripl* mask*

S36 TI singl* blind* or TI singl* mask* or TI doub* blind* or TI doubl* mask* or TI trebl* blind* or TI trebl* mask* or TI tripl* blind* or TI tripl* mask*

S35 TI clinical* trial* or AB clinical* trial*

S34 PT clinical trial

S33 (MH "Clinical Trials+")

S32 S24 or S25 or S26 or S27 or S28 or S29 or S30 or S31

S31 TI knee* implant* OR AB knee* implant*

S30 TI knee* endoprosthe* OR AB knee* endoprosthe*

S29 TI knee* prosthe* OR AB knee* prosthe*

S28 TI knee* arthroplast* OR AB knee* arthroplast*

S27 TI knee* replace* OR AB knee* replace*

S26 TI tkr OR AB tkr OR TI tka OR AB tka

S25 (MH "Arthroplasty, Replacement, Knee")

S24 S22 and S23

S23 S20 or S21

S22 S17 or S18 or S19

S21 (MH "Knee Joint+")

S20 (MH "Knee")

S19 (MH "Prostheses and Implants")

S18 (MH "Joint Prosthesis")

S17 (MH "Arthroplasty")

S16 S1 or S2 or S3 or S4 or S5 or S6 or S7 or S8 or S9 or S10 or S11 or S12 or S13 or S14 or S15

S15 TI avascular necrosis OR AB avascular necrosis

S14 TI jia OR AB jia

S13 TI arthrit* OR TI arthros* OR AB arthrit* OR AB arthros*

S12 TI osteoarthr* OR AB osteoarthr*

S11 TI ankylosing OR TI spondyl* OR AB ankylosing or AB spondyl*

S10 (MH "Spondylitis, Ankylosing")

S9 TI reumat* N3 arthrit* or AB reumat* N3 arthrit* or TI reumat* N3 artrit* or AB reumat* N3 artrit* or TI reumat* N3 diseas* or AB reumat* N3 diseas* or TI reumat* N3 condition* or AB reumat* N3 condition* or TI reumat* N3 nodule* or AB reumat* N3 nodule*

S8 TI revmarthrit* N3 arthrit* or AB revmarthrit* N3 arthrit* or TI revmarthrit* N3 artrit* or AB revmarthrit* N3 artrit* or TI revmarthrit* N3 diseas* or AB revmarthrit* N3 diseas* or TI revmarthrit* N3 condition* or AB revmarthrit* N3 condition* or TI revmarthrit* N3 nodule* or AB revmarthrit* N3

S7 TI rheumat* N3 arthrit* or AB rheumat* N3 arthrit* or TI rheumat* N3 artrit* or AB rheumat* N3 artrit* or TI rheumat* N3 diseas* or AB rheumat* N3 diseas* or TI rheumat* N3 condition* or AB rheumat* N3 condition* or TI rheumat* N3 nodule* or AB rheumat* N3 nodule*

S6 TI revmatic N3 arthrit* or AB revmatic N3 arthrit* or TI revmaticN3 artrit* or AB revmatic N3 artrit* or TI revmatic N3 diseas* or AB revmatic N3 diseas* or TI revmatic N3 condition* or AB revmatic N3 condition* or TI revmatic N3 nodule* or AB revmatic N3 nodule*

S5 TI rheumatic N3 arthrit* or AB rheumatic N3 arthrit* or TI rheumatic N3 artrit* or AB rheumatic N3 artrit* or TI rheumatic N3 diseas* or AB rheumatic N3 diseas* or TI rheumatic N3 condition* or AB rheumatic N3 condition* or TI rheumatic N3 nodule* or AB rheumatic N3 nodule*

S4 TI revmatoid N3 arthrit* or AB revmatoid N3 arthrit* or TI revmatoid N3 artrit* or AB revmatoid N3 artrit* or TI revmatoid N3 diseas* or AB revmatoid N3 diseas* or TI revmatoid N3 condition* or AB revmatoid N3 condition* or TI revmatoid N3 nodule* or AB revmatoid N3 nodule*

S3 TI reumatoid N3 arthrit* or AB reumatoid N3 arthrit* or TI reumatoid N3 artrit* or AB reumatoid N3 artrit* or TI reumatoid N3 diseas* or AB reumatoid N3 diseas* or TI reumatoid N3 condition* or AB reumatoid N3 condition* or TI reumatoid N3 nodule* or AB reumatoid N3 nodule*

S2 TI rheumatoid N3 arthrit* or AB rheumatoid N3 arthrit* or TI rheumatoid N3 artrit* or AB rheumatoid N3 artrit* or TI rheumatoid N3 diseas* or AB rheumatoid N3 diseas* or TI rheumatoid N3 condition* or AB rheumatoid N3 condition* or TI rheumatoid N3 nodule* or AB rheumatoid N3 nodule*

S1 (MH "Arthritis+")

Appendix 5. ISI Web of Knowledge Search Strategy

 

ISI Web of Knowledge

 

# 5 #3 AND #2 AND #1  Refined by: Document Type=( ABSTRACT OR MEETING OR CLINICAL TRIAL )

# 4 #3 AND #2 AND #1

#3 Topic=((trial* or random* or placebo* or control* or double or treble or triple or blind* or mask* or allocat* or prospective* or volunteer*or comparative or evaluation or follow‐up or followup))

#2 Topic=((knee* and (replace* or arthroplast* or prosthe* or endoprosthe* or implant*)))

#1 Topic=(((rheumatoid or reumatoid or revmatoid or rheumatic or reumatic or revmatic or rheumat* or reumat* or revmarthrit*) and (arthrit* or artrit* or diseas* or condition* or nodule*))) OR Topic=(ankylosing or spondyl* or osteoarthr* or arthrit* or arthros* or jia or avascular necrosis)

 

Appendix 6. WHO Clinical Trial International Clinical Trials Registry Platform and ClinicalTrials.gov Search Strategy

World Health Organization International Clinical Trials Registry Platform and ClinicalTrials.gov

 

Rheumatoid OR ankylosing OR spondyl* OR osteoarthr* OR arthrit* OR arthros* OR jia OR avascular necrosis In Intervention

AND

Total knee In Condition

What's new

Date	Event	Description
3 July 2015	Amended	Amended COI

History

Protocol first published: Issue 6, 2015

Date	Event	Description
29 June 2015	Amended	COI amended
24 June 2015	Amended	COI amended

Contributions of authors

JS: title registration

JS: protocol development

JS, GW, PT, GZ: editing protocol

Sources of support

Internal sources

• Birmingham VA Medical Center, Other.

  Time protection for research

External sources

• No sources of support supplied

Declarations of interest

JAS: JAS has received research and travel grants from Takeda, Savient, and consultant fees from Allergan, Renegeron, Takeda and Savient;  he is a member of the executive of OMERACT, an organization that develops outcome measures in rheumatology and receives arms‐length funding from 36 companies. The COI declared do not apply to this specific review topic.

GZ‐In the last 3 years he has received payment as a speaker or advisory panel member from the following companies: Johnson & Johnson (Depuy). In the last 3 years he has received travel funding form the following companies: B‐Braun, Thornier. His main income derives from clinical practice as an orthopedic surgeon (performing mostly total joint replacement), either directly from private patients or indirectly from the health system or insurances acting as a private consultant.

PT‐ PT has received honoraria for quality of life consultancy with BMS; chairs Data Safety Monitoring Boards for Chelsea and UCB, mandated for the FDA; is a member of the executive of OMERACT, an organization that develops outcome measures in rheumatology and receives arms‐length funding from 36 companies;  co‐chairs an Industry Board of 6 companies [Abbott, BMS, Merck, Roche, Schering Plough, UCB] for a biologics registry of the Ontario Rheumatology Association, the Ontario Biologics Research Initiative (an arms‐length registry which independently collects information, supported by a grant from the Canadian Institutes of Health Research (CIHR) matched by funding from these pharmaceutical companies); and receives an honorarium from the CIHR grant. The COI declared do not apply to this specific review topic.

GW‐ GW has received honoraria for consultancy with Novartis, Bristol‐Myers Squibb, and Abbott, a grant from Bristol‐Myers Squibb, and speaker honoraria from Abbott; he is a member of the executive of OMERACT and of the Scientific committee for the Ontario Biologics Research Initiative.The COI declared do not apply to this specific review topic.

Notes

This protocol for a Cochrane Review has been withdrawn by a joint decision of authors and editorial team. A new review team will be taking this topic on, and prepare a new protocol.

Withdrawn from publication for reasons stated in the review