Still's Disease Patient Data for Predicting Macrophage Activation Syndrome

A multicenter observational study of 737 Still's disease patients from the GIRRCS AOSD Study Group and AIDA Network Still’s Disease Registry, with 11.4% affected by macrophage activation syndrome. The dataset was used to explore machine learning techniques for MAS prediction, highlighting the role of ferritin, age, CRP, and systemic score. The supplementary file, authored by Piero Ruscitti and last updated in April 2026, contains the results of this analysis.

Use Cases

• Train logistic regression models to predict MAS risk based on clinical variables like ferritin and CRP.
• Build decision tree or random forest classifiers for patient stratification using age and systemic score.
• Analyze feature importance for MAS occurrence in Still's disease as described in the study results.
• Estimate probability of MAS across different clinical scenarios combining the highlighted variables.

Strengths

• Includes 737 patient records with specific demographics (age: 35.5 ± 17.8, male sex: 44.7%).
• Clear clinical outcome metrics: 11.4% MAS prevalence and 3% poor prognosis.
• Study details a specific machine learning pipeline including random forest imputation and model validation.
• Released under a permissive CC-BY-4.0 license for reuse.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count for the underlying data is unknown, which may limit suitability assessment.
• The 38.6 KB file size suggests the dataset is a supplementary document, not the primary raw data.

Provenance

Source

Gruppo Italiano di Ricerca in Reumatologia Clinica e Sperimentale (GIRRCS) AOSD Study Group and the AutoInflammatory Disease Alliance (AIDA) Network Still’s Disease Registry.

Collection Method

Multicenter, observational, prospective study.

Time Range

null

Freshness

Last updated 2026-04-14 05:39:23; freshness should be verified.

Geography

null