Nomogram-Based model accurately predicts disease flare-ups in axial spondyloarthritis

Abstract

Background

The accuracy of predicting axial spondyloarthritis (axSpA) flares based on clinical experience is limited.

Objective

The aim of this study was to evaluate the efficacy of the previously designed nomogram prediction model in forecasting disease flares among rheumatologists and medical students.

Methods

Patients who met the classification criteria for axSpA were enrolled in the study. Once a low ankylosing spondylitis disease activity score (ASDAS ≤ 2.1) was achieved, patients were monitored for 12 months to observe any disease flare-ups. Investigators assessed the likelihood of axSpA recurrence using the nomogram prediction model and their clinical experience, respectively. This allowed for a comparison of the predictive efficacy of both methods among the specialists and students.

Results

The accuracy, sensitivity, specificity, and Youden index in which disease flare-ups were predicted by the rheumatologist using clinical experience were slightly lower than those obtained using the nomogram prediction model, but the difference was not statistically significant (P > 0.05). In contrast, the indicators above by medical students using clinical experience were significantly lower compared to those predicted by the nomogram prediction model (P < 0.05).

Conclusion

The nomogram prediction model is effective in predicting the probability of disease remission and flare-ups in axSpA patients with low disease activity, demonstrating good clinical practicality and usability. Medical students can also use this model to significantly enhance the accuracy of predicting axSpA flares.