AB0120 IDENTIFYING CORE VARIABLES TO DEVELOP A SEVERITY INDEX IN RHEUMATOID ARTHRITIS: A NATIONWIDE DELPHI CONSENSUS

Background: Early interventions during the “window of opportunity” have been shown to improve clinical outcomes in rheumatoid arthritis (RA). However, intensive treatment can induce toxicity so identifying patients most likely to benefit from it is of great importance. Hence, tools for guiding therapeutic decisions in early disease stages are needed. Objectives: To identify core variables to develop a RA severity index according to an expert panel. Methods: An expert panel was prompted to analyze relevant variables to define a “severity” construct, specified as “early severe disease”, able to classify patients with data collected on the first 2 years of the disease. They were also asked to identify potential modifying factors and external criteria to evaluate criterion validity. An anonymous nationwide 2-round Delphi survey was applied to look for consensus about: (1) the priority of inclusion of each variable, (2) the feasibility to obtain them from usual sources (e.g., medical records) and (3) their definition. Each item was rated on a 10-point (priority) or 5-point (feasibility and definition) Likert scales were 0=complete disagreement and 5-10=complete agreement. After the 1st round, any item rated from 5 to 10 in priority and 3 to 5 for feasibility by at least 70% of responders was included in the final variable list. Items not reaching at least 20% consensus were discarded. The remaining ones would be voted again in the 2nd round and adopted if they reached at least 50% consensus or else discarded. Results: The task force identified 17 variables to define the “severity” construct (Table 1). Socio-economic status, knowledge of their own disease, type or work adherence to treatment, among others were proposed modifying factors. Rheumatoid factor or anti-citrullinated peptide antibody seropositivity as predictive factors The physician global assessment and the “burden of treatment” (lines of treatment, number and dose of DMARDs and cumulative steroid dose received) were proposed for evaluating criterion validity. A total of 61 stakeholders from across Spain took the survey, 56% were female and had 19.8 years of average experience after training. All variables were included after 1st round. Nonetheless, definitions were submitted for a 2nd round after rephrasing, including comments received on the survey. The final list was reduced to 15 items after merging 3 of the initial variables into a new one called “refractoriness”(Table 1). Conclusion: The consensus process resulted in a list of variables and modifying factors deemed of relevance for the severity construct as we defined it. These items will be used to develop a severity index to guide treatment decisions in early disease stages. Disclosure of Interests: Sebastian C Rodriguez-Garcia Speakers bureau: Roche, Sanofi, MSD, UCB-Pharma, Bristol-Myers-Squibb, Novartis, janssen, Consultant of: Bristol-Myers-Squibb, Galapagos, ESTHER TOLEDANO MARTINEZ: None declared, Maria Jesus Garcia de Yebenes: None declared, Loreto Carmona Grant/research support from: Novartis Farmaceutica, SA, Pfizer, S.L.U., Merck Sharp & Dohme Espana, S.A., Roche Farma, S.A, Sanofi Aventis, AbbVie Spain, S.L.U., and Laboratorios Gebro Pharma, SA, Isidoro Gonzalez-Alvaro Speakers bureau: Abbvie, MSD, Roche, Lilly, Paid instructor for: Lilly, Consultant of: Lilly, Sanofi, Grant/research support from: Roche