Data from the Korean National Health Insurance Service (NHIS) have been widely used to provide real-world evidence. Due to the nature of claims data, researchers use operational definitions to define patients with specific diseases. This study aimed to conduct a systematic review of the operational definitions of liver cancer used in studies based on the NHIS database and to suggest the most appropriate operational definition. Literature search was completed on January 6, 2021, using PubMed and KoreaMed. We applied the most frequently used operational definitions of liver cancer to the NHIS–National Sample Cohort and calculated age-standardized incidence rates (ASRs) of liver cancer by year. The ASRs using each operational definition were compared with the ASR from the Korea Central Cancer (KCCR) data. Among 236 articles, 90 were selected for review, covering histologically various kinds of liver cancer and varied by study subjects. Most studies (n = 79) did not mention whether the codes for the operational definition were from only the main diagnosis or from both the main and sub-diagnosis. The most frequently used operational definition was C22 (n = 39); however, the most similar operational definition was the ASR using “C22.0 or C22.9” for men and “C22.0” for women as the main diagnosis to the ASR from the KCCR. Based on the comparison with KCCR data, we suggest using “C22.0 or C22.9” for men and “C22.0” for women as the main diagnosis for the operational definition of liver cancer when using the NHIS data.

Keywords: Systematic review, Liver neoplasms, Incidence, National health programs