Incidence of hepatocellular carcinoma among individuals with hepatitis B virus infection identified using an...

Area of Science:

• Hepatology
• Medical Informatics
• Oncology

Background:

• Chronic hepatitis B virus (HBV) infection is a major risk factor for hepatocellular carcinoma (HCC).
• Accurate identification of patients with chronic HBV is crucial for timely management and surveillance.
• Automated data sources offer potential for large-scale patient identification but require robust algorithms.

Purpose of the Study:

• To develop and evaluate an algorithm for identifying patients with chronic HBV infection using automated data from US health systems.
• To quantify the incidence of HCC among patients identified with chronic HBV using the developed algorithm.
• To compare HCC incidence across different definitions of chronic HBV infection.

Main Methods:

• Developed an algorithm using automated data from two US health systems to identify chronic HBV patients.
• Created three overlapping cohorts based on varying definitions of chronic HBV infection.
• Compared HCC incidence in chronic HBV cohorts against a matched general population cohort without HBV.

Main Results:

• Patients meeting stringent chronic HBV criteria (6+ months infection) had a 146-fold increased HCC risk (aHR=146.5).
• Patients with at least one positive hepatitis B surface antigen test showed a 30-fold increased HCC risk (aHR=29.8).
• Patients with at least one HBV diagnosis had a 38-fold increased HCC risk (aHR=37.8).

Conclusions:

• Automated data algorithms can effectively identify patients with chronic HBV infection.
• Different criteria for defining chronic HBV yield varying but significant increases in HCC risk.
• The developed algorithm demonstrates the utility of automated data for identifying high-risk populations for HCC surveillance.