An ontology-guided semantic data integration framework to support integrative data analysis of cancer survival

Table 1 Summary of the risk factors extracted from different datasets

Risk factor		Data source	Reference ontology
Individual level	Race Gender Ethnicity Marital status Smoking status Insurance payer Residency: county and census tract^a Age at diagnosis Year of diagnosis Tumor stage Tumor type Treatment procedure	Florida Cancer Data System (FCDS)	NCIt TEO OCRV
Contextual level	Census tract SVI^b household composition and disability Census tract SVI minority status and language Census tract SVI housing and transportation Census tract SVI socioeconomic status	Agency for Toxic Substances & Disease Registry (ATSDR)	OCRV
	Census tract high school completion rates Census tract family poverty rates^c	United States Census Bureau	OCRV
	Census tract rurality status^d		OCRV NCIt
	County adult mental and physical health status^e County density of primary care physicians^f	County Health Ranking & Roadmaps	OCRV
	County smoking rate County alcohol consumption rate	Behavioral Risk Surveillance System (BRFSS)	OCRV NCIt

^aThe residency of the individual at the county- and census tract-level (i.e., which county and census tract the individual lives in), which are the linkage variables used to connect the individuals with contextual-level risk factors
^bSocial Vulnerability Index (SVI) refers to the resilience of communities when confronted by external stresses on human health, such as when facing disasters or disease outbreaks
^cThe percentage of all families whose income in the past 12 months is below the poverty level
^dThe rurality status for each census tract is based on the RUCA code
^eThe average number of days a county’s adult respondents report that their mental/physical health was not good during past 30 days
^fThe ratio of the population to total primary care physicians