Characterization and classification of Rare Disease Registries by using exploratory data analyses

Table 1 Main characteristics of Clinical-Genetic Research, Treatment, and Public Health Registries according to the most informative variables emerged after the random forest method. Variables reported in the table characterize most of the registries of each class.

Variables	Public Health Registries	Treatment Registries	Clinical-Genetic research Registries
Aims	- epidemiologic research - disease surveillance - healthcare services planning	- treatment evaluation - treatment monitoring	- clinical research - genetic -natural history of the disease
Collected data	socio-demographic	- clinical - medications, devices and health services - genetic - family history - date of the patient death - patient-reported outcomes - anthropometric info	- clinical - genetic - family history
Coding system	ICD	No coding system or own code	No coding system or own code
Services requested to a EU platform	“Quality control systems”	“Facilitated access to useful data sources”	“Model documents”