According to registered statistics from the US Surveillance, Epidemiology, and End Results Program (SEER) from 1975 to 2011, the incidence rate of thyroid cancer increased threefold over this period. On the other hand, there were no changes in death rates (123). This increase in incidence rates is considered to have been due to a growth in diagnoses brought about by greater ease of access to medical services; thus, an increase in health insurance coverage is thought to have led to an more widespread in the diagnosis rates of small thyroid cancers (456789). Socioeconomic status (SES) is a key determinant regarding the health status of individuals as well as entire populations. In the US, it has been reported that lower SES is associated with higher incidence rates of uterine cervical cancer, gastric cancer, and lung cancer, and also that higher SES is associated with higher incidence rates of breast cancer and melanoma (10). In the case of thyroid cancer, despite some reports indicating higher incidence rates among those of higher SES, other reports indicate no relations between SES and thyroid cancer incidence rates (1112131415). In the case of the Republic of Korea, according to the National Cancer Registration & Statistics Program, the incidence rate of thyroid cancer per 100,000 females for the year 2012 was 120.4, with the incidence rate having increased by approximately 24% every year from 1999 to 2012 (16). In addition, upon comparing the results of GLOBOCAN 2012, an international cancer statistics estimation project in which age-standardized incidence rates have been proposed, it was found that the incidence of thyroid cancer in the Republic of Korea was 4 times higher than in the US and 10 times higher than in Japan and China (17). This steep rise in the incidence of thyroid cancer is difficult to consider a natural phenomenon, and it has been reported that this rise has been caused by overdiagnosis (18). It is thought that in the Republic of Korea, the enactment of the Medical Insurance Act in 1963 and the completion of national coverage of medical insurance by 1989, which resulted in easier access to medical services compared to other countries, are key factors that have contributed to the overdiagnosis. It is also thought that there are differences in the use of medical services according to socioeconomic status. In light of this, the authors of this study undertook analyses to understand the correlations between SES and incidence rates of thyroid cancer using the National Health Insurance Service sample cohort database, from which big data for research purposes can be acquired.

This study was undertaken using a big data sample cohort database for researchers that were acquired from the National Health Insurance Service with approval of Institutional Review Board of National Health Insurance Service Ilsan Hospital (2016-03-035).

Samples accounting for approximately 2.2% of the 2002 National Health Insurance Recipient Qualifications Database (1,025,340 recipients) were first collected and then used in conjunction with corresponding sample treatment history, medical institution, health exam, and the National Statistical Office of Korea Cause-of-Death databases to undertake retrospective analysis of the data. Samples were collected through a systematic sampling process in which annual total medical expenses for 1,476 strata associated with different combinations of age (18 groups), sex, qualification (3 groups; self-employed members, employed members, and qualifying recipients of medical care), and income bracket (21 groups) were taken into account. Natural reductions (due to death or immigration) of the cohort data due to the passing of time were accounted for by supplementing the data with samples of newborn infants.

Although the cohort data, excluding the year 2002 in which the cohort was formed, was not considered to have sufficiently maintained the representativeness of its sample base for the years after 2002, it was considered possible to address such shortcomings through the characteristics of the representative sample.

The number of samples by year of the sample cohort database was as shown in Table 1. Although the number of individuals fluctuated by year, the total number of samples was maintained at levels accounting for approximately 2% of the total population. 11-year thyroid cancer patients from January 1, 2002 to December 31, 2013 are those patients that have insurance coverage histories in which their main illness codes begin with C73 according to their claims documents. All claims were considered, regardless of in-patient or out-patient status, or insurance qualifications.

The number of claimed cases of thyroid cancer by year and the number of overlapping patients within each given year were as shown in Table 2. Over the 11 years, the number of claimed cases of thyroid cancer and the number of patients were found to have risen. For new thyroid cancer patients, the patients were sorted by their initial date of treatment; the date at which claims associated with the illness codes concerned were first made was considered to be the initial date of diagnosis (Table 3).

Monthly contributions made according to the health insurance duty system were considered indirect parameters of evaluating SES, and were classified into 10 levels to study the correlations between each level and their associated incidence rates of thyroid cancer (Tables 4 and 5). In addition, general health exams, cancer screenings, and turning-point-of-life health exams provided for free by the government were studied to indirectly analyze access to medical services. To account for the possibility of error regarding the initial diagnosis date of thyroid cancer patients during the earlier years, new thyroid patients for the years 2002 and 2003 were excluded.

All analyses were performed using the SAS v9.4 software package (SAS Institute, Cary, NC, USA).

Thyroid cancer is reported to occur approximately 3 times more in females than in males (1920). In the current study, the largest gender gap occurred in 2004 at a male: female ratio of 1: 7.2. After this time point, the gap gradually decreased (Table 7) to around 1;3.6 by 2013.

Of the thyroid cancers, the highest prevalence of papillary cancer was found in those in their thirties and forties, while the highest prevalence of follicular cancer was found in those in their fifties. However, according to recent reports, the age at which follicular thyroid cancer occurs is becoming increasingly younger, with approximately half of all cases typically being discovered before the age of 40 while the progression of the illness is still at its initial stages. In this study, all thyroid cancers were analyzed without further classification of papillary and follicular cancers. Between the years 2004 and 2010, incidence rates among those in their forties were found to be the highest, whereas from 2011 onward the incidence rates among those in their fifties were found to be highest. After those in their 40s and 50s, those in their 30s presented the next highest incidence rates. Looking at rates of thyroid cancer incidence per 100,000 members of the population reported by the Korea Central Cancer Registry for the year 2011, males between the ages of 45 and 49 presented the highest incidence rates among their sex at 24.5 cases, whereas females between the ages of 50 and 54 presented the highest incidence rates among their sex at 84.8 cases. These statistics present a similar result to that found in this study (Table 8).

The correlations between socioeconomic status and thyroid cancer have been discussed in many overseas reports (212223). Due to the high cost of insurance in the US, despite the relatively smaller number of health insurance holders, those with insurance are typically reported to have higher incidence rates of thyroid cancer due to their ease of access to medical services compared to those without insurance. Due to difficulties in accessing medical services, those without insurance typically only visit hospitals upon the onset of symptoms, leading to late diagnoses and higher incidence rates of progressive thyroid cancers (2425). In Canada, due to its universal health care system, socioeconomic factors and thyroid cancer incidence rates are reported to have no correlation (11). In Switzerland, however, which has a similar universal health care system, differences in death rates according to socioeconomic factors have been reported (26).

In the Republic of Korea, the Medical Insurance Act was first enacted in 1963 and a system requiring the compulsory application of health insurance was first implemented in July 1977. The system was expanded to include public servants and private school employees in 1979, rural agriculture and fisheries workers in 1988, and self-employed workers within cities in 1989. By requiring self-employed workers within cities who were not salaried laborers to be covered by health insurance, the Republic of Korea established a health care system covering all of its citizens.

This study classified income into 10 brackets using data related to collected insurance contributions, and further analyses were undertaken based on an assumption that inclusion in the higher brackets entailed higher income (Table 4). Through the analysis of the 10-year sample cohort data taken from 2004 to 2013, it was found that thyroid cancer incidence rates were higher as the income brackets became higher (Table 5). In addition, the rate of receiving health exams provided free of charge by the government was analyzed according to income bracket. In consideration of the fact that such exams are typically received every 2 years, the analysis was carried out in 2-year intervals. The rate of receiving health exams was also found to be higher for those in higher income brackets. Considering that health exams paid for out-of-pocket by individuals could not be included in this study, the rate of receiving health exams was not considered to be completely accurate. An attempt was made through this study to analyze correlations between socioeconomic factors and thyroid cancer incidence rates. In order to carry out an accurate analysis of socioeconomic factors, the residential areas, level of education, income, status of health insurance subscriptions of each subject needs to be collected. As such, this study is considered limited, in that the data regarding income used in this study was derived indirectly by tracking only health insurance contributions (27282930).

This study indicated an increase in thyroid cancer incidence rates and rates of receiving health exams among individuals with a higher SES. Despite this being considered to be related to a greater access to medical services among those with higher SES, which in turn gives them a greater chance at early diagnosis of thyroid cancer, there still are several other factors that may affect the incidence rates of thyroid cancer that should be studied, which can be an area for further investigation (31).