Risk Factors for Hepatocellular Carcinoma in the US
Risk Factors for Hepatocellular Carcinoma in the US
Data were obtained from the SEER-Medicare database, which is the linkage of cancer registry data from the National Cancer Institute's Surveillance, Epidemiology and End Results (SEER) program with Medicare medical claims data. The SEER program began collating data on cancer cases in 1973, in the states of Connecticut, Iowa, New Mexico, Utah, and Hawaii and the metropolitan areas of Detroit and San Francisco-Oakland. In subsequent years, SEER was expanded to include registries in the Atlanta and Seattle-Puget Sound areas (1974–1975), rural counties in Georgia (1978), American Indians in Arizona (1980), New Orleans (1974–1977, rejoined 2001); New Jersey (1979–1989, rejoined 2001), Los Angeles County and the San Jose-Monterey area (1992), and Kentucky and the remaining counties in California (2001). SEER also provides assistance to the Alaska Native Tumor Registry. The data quality and completeness are ascertained regularly.
Medicare is the primary health insurer for ~97% of individuals aged 65 years and older in the United States. Approximately 99% of Medicare beneficiaries are covered by Medicare part A benefits, which cover inpatient hospitalizations, and 95% by part B benefits, which cover outpatient visits and supplier services. The linked SEER-Medicare data set contains part A and part B claims data beginning in 1991 for all Medicare-enrolled patients identified by SEER registries. The linkage captures ~93% of patients in the SEER database aged 65 years and older.
All persons diagnosed with HCC (ICD-O topography code=C22 and morphology codes=8170–8175) in a SEER registry who met the following criteria were eligible for inclusion as cases: diagnostic confirmation of HCC, age 65 years or older at diagnosis, and enrolled in Medicare between 1991 and 2007. Diagnostic confirmation was defined as having positive histology, cytology, laboratory test/marker study, direct visualization, or positive radiology tests. Individuals with clinical diagnosis only, unknown method of confirmation, or HCC reported exclusively by death certificate or at autopsy were not eligible for inclusion. To include persons with equal exposure to risk factor information, all individuals had to have continuous enrollment in Medicare parts A and B for at least 3 years before HCC diagnosis. This criterion resulted in a minimum age of 68 years for all participants.
Persons enrolled in a health maintenance organization (HMO) were excluded because Medicare HMO plans have not been required to submit individual claims to CMS for specific services. To avoid the inclusion of metastatic liver disease, all persons with prior diagnoses of stomach, colon, lung, pancreatic, breast, prostate or rectal cancers were excluded.
A 5% random sample of Medicare beneficiaries with no prior cancer diagnosis of HCC who resided in the geographic regions of the SEER-13 registries were selected as controls. The same inclusion/exclusion criteria as used for the cases were applied. A pseudo-diagnosis date using a random number generator was assigned to controls. Controls were matched to cases on the year of search for risk factors to minimize possible diagnostic trends.
Risk factors were identified for a minimum of 3 years prior to HCC diagnosis or reference data for the comparison group. The ICD-9 codes used to identify the risk factors are shown in Supplementary Table 1 online. As presence of diabetes and obesity are highly correlated, these conditions were combined in the data analyses.
The study protocol was approved by the Office of Human Subjects Research of the National Institutes of Health, Bethesda, MD.
Age, race/ethnicity (white, black, Hispanic, Asian, other), geographic region (SEER-13 registry region), and state buy-in status were included as covariates. The state buy-in variable indicates whether a third-party pays a beneficiary's Medicare premiums, and was thus used as an indicator of lower socioeconomic status. Demographic features and pre-existing medical conditions were compared between cases and controls using t-tests for continuous variables and Chi-square or Fisher's exact tests for categorical variables. Logistic regression was used to calculate ORs and 95% confidence intervals (95% CI). Wald χ tests determined the significance of variables in the logistic regressions. Adjusted PAFs for a specific risk factor or set of risk factors were computed using a method for estimating PAFs from population-based case–controls studies with odd ratios for the risk factor set of risk factors adjusted for confounders and other risk factors. 95% confidence intervals were computed using s.e. estimates of adjusted PAFs.
All tests of statistical significance were two-sided and a P value < 0.05 was considered statistically significant. Statistical analyses were performed using SAS Version 9.1 (SAS Institute, Cary, NC). In compliance with SEER-Medicare privacy standards, all cells with fewer than 11 individuals in any table are displayed as "<11".
Methods
Data were obtained from the SEER-Medicare database, which is the linkage of cancer registry data from the National Cancer Institute's Surveillance, Epidemiology and End Results (SEER) program with Medicare medical claims data. The SEER program began collating data on cancer cases in 1973, in the states of Connecticut, Iowa, New Mexico, Utah, and Hawaii and the metropolitan areas of Detroit and San Francisco-Oakland. In subsequent years, SEER was expanded to include registries in the Atlanta and Seattle-Puget Sound areas (1974–1975), rural counties in Georgia (1978), American Indians in Arizona (1980), New Orleans (1974–1977, rejoined 2001); New Jersey (1979–1989, rejoined 2001), Los Angeles County and the San Jose-Monterey area (1992), and Kentucky and the remaining counties in California (2001). SEER also provides assistance to the Alaska Native Tumor Registry. The data quality and completeness are ascertained regularly.
Medicare is the primary health insurer for ~97% of individuals aged 65 years and older in the United States. Approximately 99% of Medicare beneficiaries are covered by Medicare part A benefits, which cover inpatient hospitalizations, and 95% by part B benefits, which cover outpatient visits and supplier services. The linked SEER-Medicare data set contains part A and part B claims data beginning in 1991 for all Medicare-enrolled patients identified by SEER registries. The linkage captures ~93% of patients in the SEER database aged 65 years and older.
Cases
All persons diagnosed with HCC (ICD-O topography code=C22 and morphology codes=8170–8175) in a SEER registry who met the following criteria were eligible for inclusion as cases: diagnostic confirmation of HCC, age 65 years or older at diagnosis, and enrolled in Medicare between 1991 and 2007. Diagnostic confirmation was defined as having positive histology, cytology, laboratory test/marker study, direct visualization, or positive radiology tests. Individuals with clinical diagnosis only, unknown method of confirmation, or HCC reported exclusively by death certificate or at autopsy were not eligible for inclusion. To include persons with equal exposure to risk factor information, all individuals had to have continuous enrollment in Medicare parts A and B for at least 3 years before HCC diagnosis. This criterion resulted in a minimum age of 68 years for all participants.
Persons enrolled in a health maintenance organization (HMO) were excluded because Medicare HMO plans have not been required to submit individual claims to CMS for specific services. To avoid the inclusion of metastatic liver disease, all persons with prior diagnoses of stomach, colon, lung, pancreatic, breast, prostate or rectal cancers were excluded.
Controls
A 5% random sample of Medicare beneficiaries with no prior cancer diagnosis of HCC who resided in the geographic regions of the SEER-13 registries were selected as controls. The same inclusion/exclusion criteria as used for the cases were applied. A pseudo-diagnosis date using a random number generator was assigned to controls. Controls were matched to cases on the year of search for risk factors to minimize possible diagnostic trends.
Risk Factors of Interest
Risk factors were identified for a minimum of 3 years prior to HCC diagnosis or reference data for the comparison group. The ICD-9 codes used to identify the risk factors are shown in Supplementary Table 1 online. As presence of diabetes and obesity are highly correlated, these conditions were combined in the data analyses.
The study protocol was approved by the Office of Human Subjects Research of the National Institutes of Health, Bethesda, MD.
Statistical Analyses
Age, race/ethnicity (white, black, Hispanic, Asian, other), geographic region (SEER-13 registry region), and state buy-in status were included as covariates. The state buy-in variable indicates whether a third-party pays a beneficiary's Medicare premiums, and was thus used as an indicator of lower socioeconomic status. Demographic features and pre-existing medical conditions were compared between cases and controls using t-tests for continuous variables and Chi-square or Fisher's exact tests for categorical variables. Logistic regression was used to calculate ORs and 95% confidence intervals (95% CI). Wald χ tests determined the significance of variables in the logistic regressions. Adjusted PAFs for a specific risk factor or set of risk factors were computed using a method for estimating PAFs from population-based case–controls studies with odd ratios for the risk factor set of risk factors adjusted for confounders and other risk factors. 95% confidence intervals were computed using s.e. estimates of adjusted PAFs.
All tests of statistical significance were two-sided and a P value < 0.05 was considered statistically significant. Statistical analyses were performed using SAS Version 9.1 (SAS Institute, Cary, NC). In compliance with SEER-Medicare privacy standards, all cells with fewer than 11 individuals in any table are displayed as "<11".
Source...