Diagnostic pathway and treatment preferences for systemic lupus erythematosus: a physician-based discrete choice experiment

Clin Exp Rheumatol. 2024 Sep 10. doi: 10.55563/clinexprheumatol/25a2qk. Online ahead of print.

Abstract

Objectives: To assess physicians' preferences on diagnostic pathways and treatment priorities for systemic lupus erythematosus (SLE) using a discrete choice experiment (DCE).

Methods: A board of 11 SLE experts and a DCE expert statistician defined informative profiles of diagnostic pathways, pharmacological therapies, and two distinct profiles of mild-moderate and severe SLE. An independent panel of 115 clinicians involved in SLE management was invited to participate. Parameter estimates from the model were interpreted as relative preference weights (PWs). The mean PWs were used to calculate each attribute's relative importance (RI).

Results: 95 clinicians (57% females, 71% rheumatologists) completed the DCEs. The DCEs could not identify a hierarchy of importance among diagnostic pathway attributes. Nevertheless, "referral time to a rheumatologist" was considered more important for mild-moderate (RI=25%) and severe (RI=20%) SLE. Among the therapeutic attributes, the effect on organ damage progression after 12 months showed the highest preference for mild-moderate (RI=35%) and severe (RI=41%) SLE patients, followed by reduction in disease activity levels (max RI=19%) and glucocorticoid dose (max RI=13%) after six months. Reducing prednisone dose below 5 mg/day scored higher utility levels for mild-moderate (PW=66.1) than severe (PW=14.2) SLE. Administration route, action rapidity, patient-global assessment, and serious infection risk showed lesser relevance (RI 7-8%). No distinctions were found among subgroups categorised by the clinicians' areas of expertise.

Conclusions: These DCEs highlight a high degree of awareness among lupus-treating physicians, with no differences across medical specialties, of the unmet need for early diagnosis and prevention of damage accrual in SLE management.