Severe Disease in Patients With Recent-Onset Psoriatic Arthritis. Prediction Model Based on Machine Learning

Abstract

Objectives: To identify patient- and disease-related characteristics that make it possible to predict higher disease severity in recent-onset PsA.

Methods: We performed a multicenter observational prospective study (2-year follow-up, regular annual visits). The study population comprised patients aged ≥ 18 years who fulfilled the CASPAR criteria and less than 2 years since the onset of symptoms. Severe disease was defined at each visit as fulfillment of at least 1 of the following criteria: need for systemic treatment, Health Assessment Questionnaire (HAQ) > 0.5, polyarthritis. The dataset contained data for the independent variables from the baseline visit and follow-up visit number 1. These were matched with the outcome measures from follow-up visits 1 and 2, respectively. We trained a logistic regression model and random forest-type and XGBoost machine learning algorithms to analyze the association between the outcome measure and the variables selected in the bivariate analysis.

Results: The sample comprised 158 patients. At the first follow-up visit, 78.2% of the patients who attended the clinic had severe disease. This percentage decreased to 76.4% at the second visit. The variables predicting severe disease were patient global pain, treatment with synthetic DMARDs, clinical form at diagnosis, high CRP, arterial hypertension, and psoriasis affecting the gluteal cleft and/or perianal area. The mean values of the measures of validity of the machine learning algorithms were all ≥ 80%.

Conclusion: Our prediction model of severe disease advocates rigorous control of pain and inflammation, also addressing cardiometabolic comorbidities, in addition to actively searching for hidden psoriasis.

Keywords: global pain; machine learning; perianal psoriasis; prediction model; recent-onset psoriatic arthritis; severe disease.