Risk of HCC in Chronic HCV Patients With New Onset Diabetes
Risk of HCC in Chronic HCV Patients With New Onset Diabetes
The procedures followed were in accordance with the ethical standards of the Institutional Review Board of the Cathay General Hospital (CGH-P101056) and with the Helsinki Declaration (1964, amended most recently in 2008) of the World Medical Association. Informed consent was not obtained as the data were analysed anonymously.
National Health Insurance in Taiwan is a nation-wide health programme introduced in March 1995. The NHIRD comprises of health care claim data from more than 99% of the 24 million inhabitants of Taiwan. Enormous amount of researches using NHIRD have been published (http://w3.nhri.org.tw/nhird/talk_07.htm). NHIRD includes information on ambulatory and in-patient care, prescription drugs and medical institutions. Diagnoses are coded by physician according to the 9th International Classification of Diseases (ICD-9; Table S1 http://onlinelibrary.wiley.com/doi/10.1111/apt.13341/suppinfo). For the precision of the claim data, the Bureau of National Health Insurance (BNHI) performs expert reviews on a random sample of every 50–100 ambulatory and in-patient claims in each hospital and clinic quarterly. False reports of diagnosis yield a severe penalty from BNHI.
The study population included one million enrollees who were randomly sampled by BNHI from the original claim data of NHIRD. There was no significant difference in the gender (P = 0.613; http://w3.nhri.org.tw/nhird/en/Data_Subsets.htmlS2_Subsets.htmlS2) and age distribution (http://nhird.nhri.org.tw/date_cohort.htm) between the patients in the randomly sampled data and the original NHIRD. Figure 1 depicts flow chart of enrolment. We identified CHC patients by ICD-9 codes (Table S1 http://onlinelibrary.wiley.com/doi/10.1111/apt.13341/suppinfo) from 1997 to 2009. New onset diabetes was defined as a diagnosis of diabetes in the years 1999–2009 but not in the year 1997–1998. The nondiabetes cohort was comprised of CHC patients without an assigned diagnosis of diabetes from 1997 to 2009.
(Enlarge Image)
Figure 1.
Flow chart of enrolment. Chronic hepatitis C patients in new onset diabetes and nondiabetes cohort.
Hepatocellular carcinoma was defined as a diagnosis of HCC and inclusion in the Registry for Catastrophic Illness Patient Database (RCIPD), a subpart of the NHIRD. To be registered in the RCIPD, a patient should provide certificate of diagnosis from his or her physician and pathological or radiological report confirming the diagnosis of HCC, which are verified by an expert review committee.
To increase the probability of capturing new (incident) cases of HCC, we excluded patients with the diagnosis of HCC before the inception point for follow-up. We also excluded patients with the diagnosis of alcoholic cirrhosis or biliary cirrhosis. To increase the homogeneity of the study population, we excluded patients with alcoholic liver disease or hepatitis B infection. The new onset diabetes cohort and 1:1 ratio age, gender and inception point-matched nondiabetes cohort were followed up from the inception point until the development of HCC, withdrawal from insurance, or December 2009.
We searched CHC patients receiving treatment by the Anatomical Therapeutic Chemical (ATC) Classification System (Table S1 http://onlinelibrary.wiley.com/doi/10.1111/apt.13341/suppinfo): interferons and ribavirin. To adjust for the effect of comorbidities, we searched comorbid illnesses in the NHIRD to calculate the Deyo comorbidity index as a modification of the Charlson disease severity index. The index comprises 12 diagnostic categories: myocardial infarction; congestive heart failure; peripheral vascular disease; cerebrovascular disease; dementia; chronic pulmonary disease; rheumatic disease; peptic ulcer disease; hemiplegia or paraplegia; renal disease; any malignancy, including lymphoma and leukaemia, except malignant neoplasm of skin; metastatic solid tumour; and human immunodeficiency virus/AIDS. In constructing the index, we did not include mild liver disease, moderate or severe liver disease, diabetes without chronic complication or diabetes with chronic complication. The composite score for the modified Deyo comorbidity index was compared between the cohorts of CHC patients with new onset diabetes and those without diabetes, which was also included as a covariate in the Cox regression analysis.
Demographics and clinical characteristics between CHC patients with new onset diabetes and those without diabetes were analysed by chi-square test for univariate comparisons between dichotomous variables, and Student's t-tests for continuous variables. All of the tests of significance were 2-tailed, and a p-value of less than 0.05 was considered statistically significant. The cumulative incidence of the study outcome was estimated in modified Kaplan–Meier survival analyses and Grey method, after adjustment for competing mortality, and the modified log-rank test was used to test the statistical significance for difference in the incidence rates of HCC between patients with new onset diabetes and those without diabetes. The risk of HCC was modelled in a Cox proportional hazards survival analysis that examined the risk associated with diabetes after adjustment for age in each incremental year, gender, cirrhosis, hyperlipidaemia, CHC treatment, diabetes treatment, statins therapy, obesity and comorbidity index. The sas and spss software were used for the modified Kaplan–Meier survival analysis and the Cox proportional hazards model. All analyses were performed using sas software version 9.2 (SAS Institute, Cary, NC, USA) and spss program for Windows 10.0 (SPSS Inc., Chicago, IL, USA).
Methods
Ethics Statement
The procedures followed were in accordance with the ethical standards of the Institutional Review Board of the Cathay General Hospital (CGH-P101056) and with the Helsinki Declaration (1964, amended most recently in 2008) of the World Medical Association. Informed consent was not obtained as the data were analysed anonymously.
Databases
National Health Insurance in Taiwan is a nation-wide health programme introduced in March 1995. The NHIRD comprises of health care claim data from more than 99% of the 24 million inhabitants of Taiwan. Enormous amount of researches using NHIRD have been published (http://w3.nhri.org.tw/nhird/talk_07.htm). NHIRD includes information on ambulatory and in-patient care, prescription drugs and medical institutions. Diagnoses are coded by physician according to the 9th International Classification of Diseases (ICD-9; Table S1 http://onlinelibrary.wiley.com/doi/10.1111/apt.13341/suppinfo). For the precision of the claim data, the Bureau of National Health Insurance (BNHI) performs expert reviews on a random sample of every 50–100 ambulatory and in-patient claims in each hospital and clinic quarterly. False reports of diagnosis yield a severe penalty from BNHI.
Study Population
The study population included one million enrollees who were randomly sampled by BNHI from the original claim data of NHIRD. There was no significant difference in the gender (P = 0.613; http://w3.nhri.org.tw/nhird/en/Data_Subsets.htmlS2_Subsets.htmlS2) and age distribution (http://nhird.nhri.org.tw/date_cohort.htm) between the patients in the randomly sampled data and the original NHIRD. Figure 1 depicts flow chart of enrolment. We identified CHC patients by ICD-9 codes (Table S1 http://onlinelibrary.wiley.com/doi/10.1111/apt.13341/suppinfo) from 1997 to 2009. New onset diabetes was defined as a diagnosis of diabetes in the years 1999–2009 but not in the year 1997–1998. The nondiabetes cohort was comprised of CHC patients without an assigned diagnosis of diabetes from 1997 to 2009.
(Enlarge Image)
Figure 1.
Flow chart of enrolment. Chronic hepatitis C patients in new onset diabetes and nondiabetes cohort.
Hepatocellular carcinoma was defined as a diagnosis of HCC and inclusion in the Registry for Catastrophic Illness Patient Database (RCIPD), a subpart of the NHIRD. To be registered in the RCIPD, a patient should provide certificate of diagnosis from his or her physician and pathological or radiological report confirming the diagnosis of HCC, which are verified by an expert review committee.
To increase the probability of capturing new (incident) cases of HCC, we excluded patients with the diagnosis of HCC before the inception point for follow-up. We also excluded patients with the diagnosis of alcoholic cirrhosis or biliary cirrhosis. To increase the homogeneity of the study population, we excluded patients with alcoholic liver disease or hepatitis B infection. The new onset diabetes cohort and 1:1 ratio age, gender and inception point-matched nondiabetes cohort were followed up from the inception point until the development of HCC, withdrawal from insurance, or December 2009.
We searched CHC patients receiving treatment by the Anatomical Therapeutic Chemical (ATC) Classification System (Table S1 http://onlinelibrary.wiley.com/doi/10.1111/apt.13341/suppinfo): interferons and ribavirin. To adjust for the effect of comorbidities, we searched comorbid illnesses in the NHIRD to calculate the Deyo comorbidity index as a modification of the Charlson disease severity index. The index comprises 12 diagnostic categories: myocardial infarction; congestive heart failure; peripheral vascular disease; cerebrovascular disease; dementia; chronic pulmonary disease; rheumatic disease; peptic ulcer disease; hemiplegia or paraplegia; renal disease; any malignancy, including lymphoma and leukaemia, except malignant neoplasm of skin; metastatic solid tumour; and human immunodeficiency virus/AIDS. In constructing the index, we did not include mild liver disease, moderate or severe liver disease, diabetes without chronic complication or diabetes with chronic complication. The composite score for the modified Deyo comorbidity index was compared between the cohorts of CHC patients with new onset diabetes and those without diabetes, which was also included as a covariate in the Cox regression analysis.
Statistical Analyses
Demographics and clinical characteristics between CHC patients with new onset diabetes and those without diabetes were analysed by chi-square test for univariate comparisons between dichotomous variables, and Student's t-tests for continuous variables. All of the tests of significance were 2-tailed, and a p-value of less than 0.05 was considered statistically significant. The cumulative incidence of the study outcome was estimated in modified Kaplan–Meier survival analyses and Grey method, after adjustment for competing mortality, and the modified log-rank test was used to test the statistical significance for difference in the incidence rates of HCC between patients with new onset diabetes and those without diabetes. The risk of HCC was modelled in a Cox proportional hazards survival analysis that examined the risk associated with diabetes after adjustment for age in each incremental year, gender, cirrhosis, hyperlipidaemia, CHC treatment, diabetes treatment, statins therapy, obesity and comorbidity index. The sas and spss software were used for the modified Kaplan–Meier survival analysis and the Cox proportional hazards model. All analyses were performed using sas software version 9.2 (SAS Institute, Cary, NC, USA) and spss program for Windows 10.0 (SPSS Inc., Chicago, IL, USA).
Source...