ABSTRACT
Because of the cancer incidence increase and population aging in Taiwan, we aimed to assess the cancer prevalence, to summarize the comorbidities of older patients with the five most common cancers (i.e., breast, colorectal, liver, lung, and oral), and to develop a Taiwan cancer comorbidity index (TCCI) for studying their actual prognosis. The linkage of the Taiwan Cancer Registry, Cause of Death Database, and National Health Insurance Research Database was used. We followed the standard statistical learning steps to obtain a survival model with good discriminatory accuracy in predicting death due to noncancer causes, from which we obtained the TCCI and defined comorbidity levels. We reported the actual prognosis by age, stage, and comorbidity level. In Taiwan, cancer prevalence nearly doubled in 2004-2014, and comorbidities were common among older patients. Stage was the major predictor of patients' actual prognoses. For localized and regional breast, colorectal, and oral cancers, comorbidities correlated with noncancer-related deaths. Compared with the US, the chances of dying from comorbidities in Taiwan were lower and the chances of dying from cancer were higher for breast, colorectal, and male lung cancers. These actual prognoses could help clinicians and patients in treatment decision-making and help policymakers in resource planning.
Subject(s)
Colorectal Neoplasms , Lung Neoplasms , Humans , Male , Aged , Prevalence , Taiwan/epidemiology , Comorbidity , Lung Neoplasms/epidemiology , Colorectal Neoplasms/epidemiologyABSTRACT
INTRODUCTION: In high-income countries, advances in early diagnosis and treatment have improved cancer survival. However, socioeconomic inequalities in survival have persisted or increased for some adult cancers. MATERIALS AND METHODS: We assessed net survival for the 20 most common adult cancers in Taiwan. They were stratified into six age groups and three socioeconomic groups. RESULTS: Out of 120 cancer site and age group combinations, 49 showed improvements in 5-year net survival from 2000-2004 to 2005-2010. Only cervix uteri cancer in the 35-49-year age group showed a deterioration. During 2000-2010, 13 of the 20 cancer cases experienced socioeconomic inequalities for all age groups combined, and the deprivation gaps varied with cancer site and age at diagnosis. For the five most common cancers - liver, colon and rectum, lung, breast, and oral - there were socioeconomic inequalities, and 5-year net survival improved for most or all of the six age groups from 2000-2004 to 2005-2010. CONCLUSION: Reducing socioeconomic inequality in survival may lead to improvements in survival overall. We should focus on the age groups with large deprivation gaps. Our results are useful for prioritizing cancer sites and age groups for in-depth socioeconomic disparity studies and for proposing interventions for health disparity reductions and net cancer survival improvements.
Subject(s)
Neoplasms/mortality , Socioeconomic Factors , Adult , Aged , Female , Humans , Income , Male , Middle Aged , Taiwan/epidemiologyABSTRACT
Recent studies compared the age effects and birth cohort effects on female invasive breast cancer (FIBC) incidence in Asian populations with those in the US white population. They were based on age-period-cohort model extrapolation and estimated annual percentage change (EAPC) in the age-standardized incidence rates (ASR). It is of interest to examine these results based on cohort-specific annual percentage change in rate (APCR) by age and without age-period-cohort model extrapolation. FIBC data (1991-2010) were obtained from the Taiwan Cancer Registry and the U.S. SEER 9 registries. APCR based on smoothed Lexis diagrams were constructed to study the age, period, and cohort effects on FIBC incidence. The patterns of age-specific rates by birth cohort are similar between Taiwan and the US. Given any age-at-diagnosis group, cohort-specific rates increased overtime in Taiwan but not in the US; cohort-specific APCR by age decreased with birth year in both Taiwan and the US but was always positive and large in Taiwan. Given a diagnosis year, APCR decreased as birth year increased in Taiwan but not in the US. In Taiwan, the proportion of APCR attributable to cohort effect was substantial and that due to case ascertainment was becoming smaller. Although our study shows that incidence rates of FIBC have increased rapidly in Taiwan, thereby confirming previous results, the rate of increase over time is slowing. Continued monitoring of APCR and further investigation of the cause of the APCR decrease in Taiwan are warranted.