Objective: Treatment of OA by stratifying for commonly used and novel therapies will likely improve the range of effective therapy options and their rational deployment in this undertreated, chronic disease. In order to develop appropriate datasets for conducting post hoc analyses to inform approaches to stratification for OA, our aim was to develop recommendations on the minimum data that should be recorded at baseline in all future OA interventional and observational studies.
Methods: An Arthritis Research UK study group comprised of 32 experts used a Delphi-style approach supported by a literature review of systematic reviews to come to a consensus on core data collection for OA studies.
Results: Thirty-five systematic reviews were used as the basis for the consensus group discussion. For studies with a primary structural endpoint, core domains for collection were defined as BMI, age, gender, racial origin, comorbidities, baseline OA pain, pain in other joints and occupation. In addition to the items generalizable to all anatomical sites, joint-specific domains included radiographic measures, surgical history and anatomical factors, including alignment. To demonstrate clinical relevance for symptom studies, the collection of mental health score, self-efficacy and depression scales were advised in addition to the above.
Conclusions: Currently it is not possible to stratify patients with OA into therapeutic groups. A list of core and optional data to be collected in all OA interventional and observational studies was developed, providing a basis for future analyses to identify predictors of progression or response to treatment.
Keywords: clinical trials; osteoarthritis; personalized medicine; prognosis; stratification.
© The Author 2016. Published by Oxford University Press on behalf of the British Society for Rheumatology.