Presentation (CTI)

OBJECTIVES: This study leverages advanced analytical methodologies to identify patients at high risk of transitioning from Psoriasis (PsO) to Psoriatic Arthritis (PsA). We aim to improve early detection and intervention, enabling tailored treatment strategies and better resource allocation.
METHODS: Merative MarketScan payer sourced claims data from 2019-2022 was leveraged and patients with advanced PsO treatments were identified. Survival analysis using Kaplan-Meier curves and predictive modeling via Cox Regression and Random Survival Forest (RSF) were conducted. Chi-Square tests of independence identified key predictors, while RSF provided risk scores for patients likely to progress to PsA. Predictors encompassed demographic factors, treatment pathways (e.g., Interleukin inhibitors), comorbidities, and ICD code groupings. Model validation assessed proportional hazard assumptions and predictive accuracy.
RESULTS: Findings revealed that most PsA cases emerge within the first year of advanced PsO treatment initiation, with significant predictors including Anti-TNF and Interleukin therapies. Brand switching peaked in the initial two years, underscoring critical early intervention opportunities. RSF outperformed other models, accurately stratifying patients by progression risk. Additionally, insights from survival analysis highlighted the role of comorbidities and procedural codes in accelerating PsA development.
CONCLUSIONS: This study demonstrates the value of predictive modeling in identifying patients at high risk of transitioning from PsO to PsA. Such models can drive targeted drug development, optimize clinical trial design, and facilitate precision medicine approaches. Future work should integrate broader datasets to validate findings and explore cost-benefit analyses for early intervention strategies.