摘要：Objectives. We compared chronic liver disease (CLD) mortality from 1999 to 2009 between American Indians and Alaska Natives (AI/ANs) and Whites in the United States after improving CLD case ascertainment and AI/AN race classification. Methods. We defined CLD deaths and causes by comprehensive death certificate-based diagnostic codes. To improve race classification, we linked US mortality data to Indian Health Service enrollment records, and we restricted analyses to Contract Health Service Delivery Areas and to non-Hispanic populations. We calculated CLD death rates (per 100 000) in 6 geographic regions. We then described trends using linear modeling. Results. CLD mortality increased from 1999 to 2009 in AI/AN persons and Whites. Overall, the CLD death rate ratio (RR) of AI/AN individuals to Whites was 3.7 and varied by region. The RR was higher in women (4.7), those aged 25 to 44 years (7.4), persons residing in the Northern Plains (6.4), and persons dying of cirrhosis (4.0) versus hepatocellular carcinoma (2.5), particularly those aged 25 to 44 years (7.7). Conclusions. AI/AN persons had greater CLD mortality, particularly from premature cirrhosis, than Whites, with variable mortality by region. Comprehensive prevention and care strategies are urgently needed to stem the CLD epidemic among AI/AN individuals. In 2009, chronic liver disease (CLD) was classified as the fifth leading cause of death among American Indians and Alaska Natives (AI/ANs), accounting for 4.8% of AI/AN deaths. 1 A 2004 mortality study used a broader set of disease codes from the International Classification of Diseases, Ninth Revision ( ICD-9 ) than was used in the National Vital Statistics System (NVSS) to define a CLD death, and reported an AI/AN age-adjusted death rate more than twice that of other races. 2 Few US population-based studies of CLD have described its prevalence among AI/AN persons. 3 Two studies found that 4.9% to 7.2% of patients in 3 facilities serving AI/AN populations had clinically confirmed CLD. 4,5 Both identified alcoholic liver disease (ALD), chronic HCV infection, and nonalcoholic fatty liver disease (NAFLD) as the most common contributing causes. Multiple factors limit the precision of AI/AN mortality estimates. First, racial misclassification of AI/AN individuals can affect mortality data and subsequent surveillance estimates. 6,7 In cancer incidence studies, the Indian Health Service (IHS) has improved race classification by linking surveillance data with IHS patient enrollment records and restricting analyses to counties where AI/AN race is more accurately classified. 6 In addition, previously used ICD -based CLD definitions have underestimated CLD cases in death records. 8 A 2004 mortality study provided improved estimates of CLD deaths from 1990 to 1998 with inclusion of viral hepatitis disease codes. 2 Since then, an even more comprehensive ICD -based definition of CLD has provided more accurate estimates of clinically confirmed CLD deaths, 8,9 but has not yet been used in national mortality studies. In this study, we employed a comprehensive ICD -based definition of CLD deaths and used established techniques to improve race classification to accurately describe disparities and compare trends in CLD mortality from 1999 to 2009 between AI/AN persons and Whites in the United States.
