Table 2.

Results of the bivariate logistic regression models- looking at the association between hepatitis C and liver cancer and hepatitis C and hepatocellular carcinoma and results of the sensitivity analyses^a, Georgia, 2015–2019

Bivariate Logistic regression for Liver Cancer Analysis and Hepatocellular Cancer Sub-analysis
	General liver cancer analysis bivariate regression model (All ICD-10 codes for hepatobiliary cancer diagnoses included): AICb 2303		Hepatocellular sub-analysis bivariate regression model (Only hepatocellular carcinoma-specific ICD-10 code—C22.0 -included): AIC 1066
	Frequency (n)	ORc (95% CId)	Frequency (n)	OR (95% CI)
Hepatitis C negative	3429	refe	3429	ref
Hepatitis C positive	445	20.01 (15.97 –25.37)	445	16.84 (12.01–23.83)
Sensitivity Analyses (Based on General Liver Cancer Analysis)
	Bivariate logistic Regression- All hepatitis C screened individuals with positive test results without a viremia test considered hepatitis C positive (AIC 2400)		Bivariate logistic Regression- All hepatitis C screened individuals with positive test results without a viremia test considered hepatitis C positive (AIC 2502)
	Frequency (n)	OR (95% CI)	Frequency (n)	OR (95% CI)
Hepatitis C negative	3429	ref	3499	ref
Hepatitis C positive	515	19.38 (15.57–24.19)	445	17.51 (13.98–21.99)

^aBivariate logistic regression models looking at the association between hepatitis C and general liver cancer in the years 2015–2019 in which either all individuals positive for hepatitis C antibodies without a confirmatory test are considered to be hepatitis C positive or all are considered to be hepatitis C negative    

^bAkaike information criterion

^cOdds Ratio

^d95% Confidence Interval

^eReference category