Fig. 2 (abstract P149).From: 12th European Headache Federation Congress jointly with 32nd National Congress of the Italian Society for the Study of HeadachesMultivariate analysis using random forest showing sorted variable importance (VI) mean VI (on the X-axis) of all variables recorded , in descending order (on the y-axis). The red line is the threshold value. Only five items were above the line but with minor significance, of no clinical relevance (VI mean < 0.002). Variables legend: V2: Current age, V3: Gender, V5: Age of Onset, V6: Episodic CH, V7: Chronic CH, V8: Probable CH , V9: Strictly Unilateral, V10: Side Variable, V11: Bilateral, V12: Site RetroOrbital,V13: Site Orbital, V14: Site Frontal, V15: Site Temporal, V16: Site Parietal, V17: Site Vertex, V18: Site Occiput, V19: Site Nasal, V20: Absence of dysautonomic feature, V21: Ptosis, V22: Eye Oedema, V23: Conjonctival injection, V24: Miosis, V25: Lacrimation, V26: Nasal Blockage, V27: Rhinorrea, V28: Sweating, V29: Facial Flush, V30: Aural Fullness, V31: Restlessness, V32: Average attack frequency per day, V33: Minimum attack frequency per day V34: Maximum attack frequency per day, V35: Average Attack Duration, V36: Minimum Attack Duration, V37: Maximum Attack Duration, V39: Intractable to Acute treatment, V40: Intractable to Preventive treatment, V41: Follow-up durationBack to article page