#4 Using hierarchial linear modeling to examine geographic variation in hepatocellular carcinoma

PURPOSE: Geographic variation in hepatocellular carcinoma (HCC) has not been previously studied in the US. The purpose of this study was to determine whether geographic variation exists in the incidence of HCC and to assess the effects of individual and population-level risk factors. METHODS: Using NCI SEER program data, we identified cases of HCC from 5 states and 4 metropolitan areas. The 1990 BRFSS survey provided regional prevalence data for several HCC risk factors. Age-adjusted incidence rates were calculated for each geographic region. Hierarchical linear modeling was used to assess the association between HCC incidence and geographic region, adjusting for age, gender, race, year of diagnosis, and the population prevalence of current smoking (smoke everyday/some days and ≥100 cigarettes in lifetime), chronic alcohol use (greater-or-equal, slanted2 drinks/day or greater-or-equal, slanted60 drinks/month), heavier body weight (BMI:25.0–29.9), and diabetes. RESULTS: Between 1975–98, 11,547 cases were identified. Hawaii had the highest age-adjusted incidence rate (4.6), followed by San Francisco-Oakland (3.2), New Mexico (2.0), Detroit (1.9), Seattle-Puget Sound (1.8), Atlanta-Metropolitan (1.7), Connecticut (1.6), Iowa (1.1), and Utah (1.0); all rates per 100,000. Overall, men had a 2–3 times higher age-adjusted incidence rate than women. Whites had an age-adjusted incidence rate of 1.5, blacks 3.2, and other races 7.0. However, variation in HCC incidence rates persisted among geographic areas irrespective of age, gender, or race. These geographic differences were confirmed by hierarchical linear modeling, adjusting for demographic factors; these analyses also showed that the risk of HCC was greater in regions with a higher proportion of alcohol users (p = 0.02) and heavier body weights (p = 0.06). CONCLUSION: Significant geographic variation in HCC incidence exists in the US. These variations are only partly explained by differences in age, gender, race, year of diagnosis, and prevalence of chronic alcohol use and heavier body weight in the underlying population.