Table V.

Recommendations for points to consider: outcome measures

Consideration	Recommendation
1. General	□Measures of symptoms and structure are both important and should be recorded □The primary outcome measure(s) are likely to be required after 1–2 years after intervention but should relate to the study question □Short, medium and long term outcomes should be collected □Frequent outcomes should be considered in the first year, particularly for efficacy and biomarker-related questions
2. Patient reported outcome measures (PROMs)	□PROMs which have been validated within appropriate populations and which examine pain, function, performance and quality of life were recommended □The choice of tool should depend on its extent of validation and reliability as well as feasibility including cost □Early assessment of the cost effectiveness of any given intervention, or interventions should be considered
3. Imaging	□Imaging should be used a) to categorize and phenotype, and b) as an important outcome measure □MRI and X-ray are both important outcome measures, but MRI may have increased sensitivity at earlier times after injury □The patello-femoral joint and tibio-femoral joints should both be included in imaging assessments □An index/signal knee should be defined (given that the opposite side may subsequently be affected) □The contralateral knee may be a useful imaging control or comparator for the index/signal knee □The index/signal knee, and ideally both knees, should be imaged at 0 (baseline), 12 months and 24 months for structural changes after intervention; inclusion of a later time point, such as 5 years was also recommended □Morphology and change in all joint tissues should be captured, using validated semi-quantitative and/or quantitative measures □Compositional assessment at 6 months for cartilage (MRI) or bone changes (MRI, PET, CT) is more experimental but should be considered in addition to structural assessments
4. Molecular biomarkers	□ No specific biomarker(s) can be recommended for routine use in interventional studies ○ Biomarkers cannot yet act as independent surrogate endpoints for early OA diagnosis ○ Biomarkers have not been validated for aiding selection of patients for interventional studies □ Molecular biomarkers should be considered as exploratory outcome measures in interventional studies ○ Choice(s) will depend on the target and outcomes under study □ Bio-samples (including synovial fluid, in addition to serum/plasma and urine) should be collected in all future studies where possible ○ Serum and urine should be collected at all available time points ○ Sampling should include DNA storage where appropriate consent is given ○ Synovial fluid can be accessed at the time of surgery or clinical aspiration, or at the time of drug delivery into the index/ signal knee ○ Timing and method of sample collection must be consistent and standardized across all studied patients
5. Functional outcomes	□Stability of the knee and muscle strength are important to patients, and potentially important outcome measures □Symptoms of instability may have value in addition to examination-based measures of mechanical instability/laxity □Other potential functional biomarkers include kinematics, hop or stair climbing tests and muscle co-contraction testing