Table 3 |.

Univariate and multivariable logistic regression model predicting eGFR improvement after DAA therapy

Baseline predictors of eGFR improvement after DAAs	Univariate model		Multivariable model
	Odds ratio (95% CI)	P	Odds ratio (95% CI)	P
Age, per 10 yr	0.95 (0.80–1.1)	0.60
Sex (female vs. male)	1.30 (0.87–1.9)	0.19	1.16 (0.77–1.8)	0.47
Race (white vs. nonwhite)	1.30 (0.83–2.1)	0.24	1.31 (0.81–2.1)	0.27
Nondiabetic patients vs. diabetic	1.78 (1.07–2.9)	0.025	1.76 (1.01–3.0)	0.043
Cirrhosis (vs. noncirrhotic)	0.75 (0.51–1.1)	0.16	0.89 (0.58–1.4)	0.58
Hypertension (vs. nonhypertensive)	0.75 (0.51–1.1)	0.14	0.74 (0.48–1.1)	0.17
HIV co-infection	1.05 (0.57–1.9)	0.86	—	—
Ribavirin use	0.91 (0.59–1.4)	0.67	—	—
DAA course length (>12 wk vs. ≤12 wk)	0.62 (0.35–1.1)	0.11	0.61 (0.33–1.1)	0.12
Baseline eGFR < 60 ml/min per 1.73 m2	3.49 (2.2–5.7)	<0.001	4.29 (2.6–7.2)	<0.001

CI, confidence interval; DAA, direct-acting antiviral; eGFR, estimated glomerular filtration rate.

eGFR improvement was defined by ≥10% increase in eGFR from baseline to post-treatment average in the 12 months after completing DAA therapy. The multivariable model includes demographics and predictors with P < 0.1 in the univariate model. Only nondiabetic status and baseline eGFR < 60 ml/min per 1.73 m² predicted improvement in eGFR after DAAs in multivariable models.