Validation of a coding algorithm to identify patients with hepatocellular carcinoma in an administrative database

Area of Science:

• Hepatobiliary Medicine
• Medical Informatics
• Epidemiology

Background:

• International Classification of Disease, Ninth Revision, Clinical Modification (ICD-9-CM)-based algorithms for identifying hepatocellular carcinoma (HCC) are lacking outside specific healthcare systems.
• Developing and validating such algorithms is essential for population-based research on HCC epidemiology, treatment efficacy, and safety.

Purpose of the Study:

• To develop and validate an ICD-9-CM-based algorithm for identifying patients with hepatocellular carcinoma (HCC).
• To enable population-based studies on HCC epidemiology and comparative effectiveness of therapies.

Main Methods:

• Electronic medical records from two tertiary care hospitals were queried.
• The algorithm identified patients with at least two ICD-9-CM codes for chronic liver disease/cirrhosis and two ICD-9-CM codes for HCC.
• Positive predictive value (PPV) was determined by expert medical record review.

Main Results:

• The algorithm identified 101 patients, with 88 confirmed cases of HCC (PPV: 87.1%).
• Sensitivity was 91.7% for confirmed HCC cases, and specificity was 98.7% for non-HCC chronic liver disease patients.
• Excluding patients receiving recent chemotherapy improved the PPV to 91.6%.

Conclusions:

• An algorithm requiring at least two ICD-9-CM codes for chronic liver disease/cirrhosis and two for HCC demonstrates high PPV.
• This claims-based algorithm is suitable for future epidemiologic studies of HCC.
• It can also be used to evaluate outcomes and adverse events of HCC therapies.