Table 1 OARSI clinical trials recommendations [10]

Randomization

1. Trial methodology includes effective randomization procedures that ensure that members of the study team and participants remain unable to predict or influence their treatment assignment.

Blocking and stratification

2. Any stratification and/or subset analyses are specified prior to study development.

Blinding

3. Adequate blinding procedures are used to prevent disclosure of allocation to participants and study staff.

4. If adequate blinding is not possible, an independent staff member, ideally not aware of the study hypotheses, performs the assessments or procedures that may lead to disclosure of allocation assignment (e.g., injections or exercise intervention).

5. It is clearly indicated who was blinded and the mechanisms by which this was accomplished are described.

Expectations

6. Research staff is trained about the importance of equipoise when discussing the study interventions with the participants.

Washout periods and concomitant pain medications

7. Design strategies are adopted to manage confounding by concomitant medications.

Outcome reporting training

8. Steps are taken to ensure consistency and accuracy of outcome reporting by participants.

Comorbidities and subphenotypes

9. Study design is explicit about inclusion or exclusion of comorbidities.

10. Planning is conducted for recruiting or analyzing sub-phenotypes.

Characterizing baseline disease

11. The severity of the disease, the pathological sub-phenotype (e.g., synovitic, bone marrow lesion, meniscal), structural sub-phenotype (e.g., knee compartment), and pain sub-phenotype (e.g., neuropathic, nociceptive) are characterized at baseline

Selecting an index knee

12. If selection of an index knee is required, selection strategy is defined in advance.

Symptom-modifying interventions

13. Symptomatic cut-points are selected to avoid ceiling or floor effects and permit analysis of the minimally clinically important differences.

Structure-modifying interventions

14. Structural severity cut-off points are selected to avoid ceiling or floor effects and permit analysis of the minimally clinically important differences.

Trial interventions

15. Interventions (active and placebo) are described in sufficient detail to allow others to replicate them.

Trial outcome measures

16. Trials use patient-reported and objective outcome measures that are valid, reliable, and responsive to change.

17. Primary and secondary outcome measures are defined a priori and indicated when registering a trial.

Patient-reported outcome measures

18. Symptomatic outcomes are assessed using the three core clinical measures: pain, physical function, and patient global assessment.

Objective outcome measures (e.g., physical function)

19. A set of physical performance measures for knee OA are used.

Structural outcome measures

20. Radiography or MRI are used for demonstration of structure modification. The choice of imaging technique and outcome measures (primary and secondary) should be predicated on the expected mechanism of the intervention (e.g., synovitis/effusion volume for anti-inflammatory agents)

21. Reliability and other metrics of measurement error and sensitivity, including scan-rescan reproducibility, are assessed at each study site.

22. Disease modification is defined as an improvement in KOA-related symptoms (e.g., joint pain) and one of the following structural outcomes: reduction or reversal of joint space narrowing (continuous outcome); or reducing the progression of cartilage damage or reversal of cartilage damage on MRI (e.g., thickness, denudation).

Biochemical biomarkers

23. Biological fluids are collected and stored to assess the metabolic effect of a treatment on joint tissues.

Reporting

24. The clinical trial is registered in the appropriate registry prior to enrolling participants (e.g., Clinicaltrials.gov).

25. The clinical trial methodology and results are reported in a format that allows for their inclusion in pooled analyses.