Background:Treatment of rheumatoid arthritis (RA) has changed in the past 20 years due to the arrival of biologic disease-modifying antirheumatic drugs (bDMARDs) and the application of treat to target strategies. Many different combinations of conventional synthetic (cs)DMARDS and bDMARDS are being applied in daily practice. It is difficult to visualize and understand all the different treatment pathways that are sequentially being used over longer periods of time in a real-world setting. We therefore investigated whether a Sankey diagram is a suitable tool to study and visualize which treatment pathways exist and to what extent a RA treatment protocol is being followed.Objectives:To illustrate the treatment pathways over longer periods of time in a cohort of early RA patients and to assess adherence to the treatment protocol using Sankey diagrams.Methods:This single-centre retrospective, observational cohort study used data from the DREAM-RA registry. Patients were treated according to a treat to target strategy with a step-up DMARD treatment protocol when remission was not achieved, initial steps were starting with MTX 15 mg/wk, week 8 MTX 25 mg/wk, week 12 MTX+SSZ 2dd1000mg, week 20 MTX+SSZ 3dd1000mg, week 24 MTX+adalimumab 40mg/2wk, week 36 MTX+adalimumab 40mg/wk. Although the protocol met full consensus, adherence to the protocol in individual patients was at the discretion of the treating rheumatologist. In this study, patients were included if they received a continuous treatment with a conventional synthetic or biologic DMARD between 1 January 2002 and 30 April 2020. During treatment, corticosteroids per protocol were allowed but not considered as an individual treatment. Evaluated outcomes included: the consecutive treatments that patients followed including start- and stop-date of treatments, the proportion of patients that received bDMARDs, the number of switches until first bDMARD, and time to first bDMARD. Furthermore, the lower limit of adherence to the protocol was estimated by considering all patients (% of total) treated according to the protocol. This information was determined by verifying whether the patient was being treated according to the protocol after each switch.Results:A total of 372 patients were included in this study (Table 1). The mean overall follow-up time of the cohort was 8.83 (± 3.59) years. The follow-up time for the first 4 treatments, depicted in the Sankey diagram was 6.28 (± 3.31) years. At least 45 (12%) patients started with a bDMARD before all previous protocol steps were followed. At the start of treatment, 81% of the patients were treated according to the protocol, this was reduced to 28% after one switch. The lower limit of adherence to the predefined protocol after 3 switches were roughly 5% of all patients.Table 1.Absolute counts of T-cell subpopulations at baseline, after 6 and 12 m of TCZ therapyAll patients(n=372)Gender, female, n (%)246 (66.1)Age, year, mean ± SD67.9 ± 13.61Overall follow-up, years ± SD8.83 ± 3.59Follow-up first 4 treatments, years ± SD6.28 ± 3.31Baseline DAS-28 score, mean ± SD3.60 ± 1.41Rheumatoid factor positive, n (%)336 (90.3)Patients who eventually received a bDMARD, n (%)108 (29.0)Number of switches until first bDMARD, mean ± SD2.7 ± 1.41Time to first bDMARD, years ± SD3.66 ± 3.00Figure 1.Sankey diagram of the treatment pathway of the first 3 switches of RA patients. The average duration of treatment of a flow is displayed in years if the flow included more than 20 patients. (MTX= methotrexate; SSZ= sulfasalazine; HCQ = hydroxychloroquine; LEF= leflunomide; bDMARD= monotherapy bDMARD; Combi csDMARDs= combination therapy of csDMARDS; csDMARD(s)+bDMARD= combination therapy of one or two csDMARD(s) + a bDMARD; No therapy= no treatment received >3 months; Other = medication that is not a (cs)(b)DMARD)Conclusion:Sankey diagramming can be used to illustrate complex real-world treatment data of a treat to target cohort of RA patients. Treatment protocol adherence can be assessed with the help of a Sankey diagram. After 3 switches, the lower limit of adherence to the protocol was roughly 5%.Disclosure of Interests:Tristan Coppes: None declared, Naomi Jessurun: None declared, Jurriaan Jansen: None declared, Kimberly Velthuis: None declared, Peter ten Klooster: None declared, Harald Vonkeman Consultant of: BMS, Celgene, Celltrion, Galapagos, Gilead, Janssen-Cilag, Lilly, Novartis, Pfizer, Sanofi-Genzyme, Grant/research support from: Abbvie
