Multiple Sclerosis is a neurodegenerative disease which shows different phenotypes making difficult for clinicians to make short-term decisions related with treatment and prognosis. Diagnosis is usually retrospective. Learning Healthcare Systems (LHS) can support clinical practice as they are devised as constantly improving modules. LHS can identify insights which allow evidence-based clinical decisions and more accurate prognosis. We are developing a LHS with the aim of reducing uncertainty. We are using ReDCAP to collect patients’ data, both from Clinical Reported Outcomes (CRO) and from Patients Reported Outcomes (PRO). Once analyzed, this data will serve as a foundation to our LHS. We conducted bibliographical research to select those CRO and PRO collected in clinical practice or identified as possible risk factors. We designed a data collection and management protocol based on using ReDCAP. We are following a cohort of 300 patients for 18 months. At the moment, we have included 93 patients and received 64 complete responses and 1 partial response. This data will be used to develop a LHS, able to accurate prognosis as well as to automatically include new data and improve its algorithm.