23

Objectives. – Patient’s perception analysis appears recently in numerous studies. Conjoint analysis has been used extensively by market researchers for studying how people value the characteristics of products and services. This technique was used in a clinical study to describe perceptions and preferences of anaemic patients suffering from chronic kidney disease not on dialysis (CKDnd), regarding erythropoietin stimulating agents (ESA). Methods. – PERCEPOLIS was a French multicenter prospective non-interventional study designed to describe the relative importance of ESA attributes according to CKDnD patients. Patients fulfilled questionnaires using an electronic device (digital tablet) at baseline and after 6 months under continuous erythropoietin receptor activator (CERA) treatment. Choice-based conjoint (CBC) question- naires were developed with multiple components: 7 ESA attributes (2 or 3 levels per attribute), 2 partial profiles per task (2 out of the 7 attributes), and 7 tasks per questionnaire. Analyses were performed according to previous ESA treatment or not. Results. – From 789 analyzed patients, 433 non ESA-naive patients were more than 80% to declare treatment efficacy as the most important expectative in ESA choice process (direct question) but CBC analyses revealed that frequency of injections was more crucial (relative mean weight: ~30% versus ~20% for efficacy). Pain at injection site and haemoglobin not exceeding the recommended target were confirmed as important criteria for patients (relative mean weights: ~15%). No new or unexplained safety signals were noted. Conclusions. – Using CBC design for the first time in a non-interventional ESA study with an electronic Patient Reported Outcome (ePRO) in an elderly population, these data showed that monthly injections and treatment efficacy were key patients’ expectations relative to ESAs. CERA efficacy to maintain stable haemoglobin within the recommended range was confirmed in real-life conditions.