Construct validity of the OCTOPuS stratification algorithm for allocating patients with knee osteoarthritis into...

Area of Science:

• Orthopedics and Rehabilitation
• Biomedical Engineering
• Clinical Trial Design

Background:

• Knee osteoarthritis (OA) management can be improved with stratified exercise therapy.
• A novel algorithm categorizes knee OA patients into high muscle strength, low muscle strength, or obesity subgroups.
• Each subgroup is designed to receive tailored exercise interventions based on their phenotype.

Purpose of the Study:

• To evaluate the construct validity of a newly developed stratification algorithm for knee osteoarthritis (OA).
• To determine if patient proportions, subgroup characteristics, and exercise therapy effects align with proposed phenotypes.

Main Methods:

• Utilized data from five studies (four exercise therapy trials, one cross-sectional cohort) involving 1211 knee OA patients.
• Tested 63 a priori hypotheses concerning subgroup proportions, characteristics, and treatment effects.
• Assessed construct validity by comparing observed data against expected outcomes for each research question.

Main Results:

• The algorithm demonstrated good construct validity, with 73% of hypotheses confirmed.
• Patient subgroup proportions were consistent across cohorts, and subgroup characteristics matched proposed phenotypes.
• Usual exercise therapy yielded similar medium to large effects on knee pain and function across all subgroups, contrary to hypotheses.

Conclusions:

• The stratification algorithm reliably assigns knee OA patients into hypothesized subgroups.
• However, the study found no substantial differences in the effects of usual exercise therapy among the subgroups.
• Further research is ongoing to determine if subgroup-specific therapy improves clinical and economic outcomes.