. 2021 Apr 6;4(4):e214132. doi: 10.1001/jamanetworkopen.2021.4132

Table 2. Association Between Bcr-Abl TKI Use and Incident Hepatitis B Flare.

TKI use prior to index date, d	Rate ratio (95% CI)^a
TKI use prior to index date, d	Primary^b	Sensitivity^c
60	1.41 (0.99-2.01)	1.49 (1.04-2.14)
90	1.56 (1.11-2.20)	1.62 (1.14-2.30)
120	1.64 (1.17-2.30)	1.69 (1.20-2.38)
180	1.60 (1.15-2.23)	1.67 (1.19-2.35)
365	1.66 (1.20-2.28)	1.69 (1.22-2.34)

Abbreviation: TKI, tyrosine kinase inhibitors.

^{^a}

Rate ratios were estimated by the conditional logistic regression. Covariates considered in the models were Bcr-Abl TKI use, cancer history in 6 months prior to the index date, cirrhosis, transplantation for hematologic malignant neoplasms, and other medications that are well known for their hepatitis B flare HBV reactivation risk (ie, immunosuppressants, cytotoxic chemotherapy, and rituximab).

^{^b}

In the primary analysis, case patients were those who received their first antiviral agents for more than 28 days with no evidence of receiving chemotherapy within 30 days after index date.

^{^c}

In the sensitivity analysis, case patients were those who received their first antiviral agents for more than 28 days with no evidence of receiving chemotherapy within 30 days before and after index date.