The mortality burden of chronic liver disease may be substantially underestimated in the United States.

PURPOSE: The United States National Center for Health Statistics (NCHS) uses death certificate data to estimate the burden of serious disease. This study aimed to determine the accuracy of the NCHS method for estimating the burden of chronic liver disease (CLD).

METHOD: The authors identified death certificates of New Haven County residents who died from October 1999-September 2000 that were assigned one of 115 ICD-10 codes that might indicate CLD. They reviewed medical charts, medical examiner records and a certifier questionnaire to determine whether CLD was the cause of death.

RESULT: Using the authors' determination of CLD status as the gold standard, the specificity of the NCHS classification was high (86%), but the sensitivity was low (36%). The authors found that adding selected ICD-10 codes to those considered by the NCHS to be CLD (certain CLD malignancies and viral hepatitis) could improve sensitivity. Ensuring that deaths attributed by certifiers to "End Stage Liver Disease" were coded as CLD could also improve completeness. These modifications could increase sensitivity substantially with little effect on specificity.

CONCLUSION: The NCHS method may understate the CLD burden substantially which could have a detrimental effect on planning for and evaluating prevention and treatment. Modifications could improve completeness.