Early diagnosis and clinical application of systemic lupus erythematosus based on a nomogram model

Abstract

Background: Systemic lupus erythematosus (SLE) is a chronic autoimmune disease involving multi-system and multi-organ dysfunction, and is easily misdiagnosed early in the disease course. We aimed to accurately predict early SLE nomogram to provide a reference basis for the early clinical diagnosis of SLE. Methods: We retrospectively analyzed 167 patients who were first diagnosed with SLE at Fengxian District Central Hospital, Shanghai, between March 2017 and October 2022. Three groups of 129 physically healthy subjects, 67 patients with rheumatoid arthritis, and 40 patients with rashes were selected as controls during the same period. Patients with SLE and control group were randomly divided into training (n = 217) and validation (n = 141) group. Univariate and multivariate analyses were used to identify independent risk factors for early SLE diagnosis. The independent risk factors for diagnosis were used to construct a nomogram to predict early SLE. Results: Based on the training group, three variables were identified as independently influencing early SLE: platelets (odds ratio OR = 0.993, P = 0.047), albumin (OR = 0.833, P = 0.007), and complement component 1q (OR = 0.956, P = 0.000). The precision of the nomogram was assessed using C-index values and calibration plot diagrams. The C-index values were 0.929 for training group and 0.898 for validation group. Both the training group and validation group calibration curves showed good predicted outcomes. Conclusion: The construction of a nomogram can accurately predict the risk of early SLE. The model showed good discriminatory power and calibration for use in the diagnosis of SLE, providing a visual tool and reference basis for the early diagnosis of SLE.

Keywords

• Systemic lupus erythematosus
• Nomogram
• Early diagnosis
• Predictive model