The incidence and mortality of cancer are rapidly increasing worldwide. According to the International Agency for Research on Cancer (IARC), there were 18.1 million new cases and 9.6 million deaths due to cancer worldwide in 2018 [1]. In the case of South Korea, the incidence of cancer has been reported since 1999, and the annual average increase in age-adjusted cancer incidence from 1999 to 2011 was 3.8%. Since 2011, although the incidence of cancer has decreased by 3.0% annually (except for a 3.1% increase between 2015 and 2016 [2]), the death rate from cancer has been steadily increasing. Based on 2017 data, cancer was still the leading cause of death in South Korea, accounting for 27.8% of all deaths. Also, the cancer mortality rate was 153.0 per 100,000 population, with a 1.4% increase from the previous year [3].

As Korea became an aged society in 2017 [4], the increase in the elderly population will lead to higher numbers of the elderly living with cancer [5]. Although new developments in cancer diagnostics and treatment have led to increases in cancer survival rates, the prevalence of cancer is still increasing [6]. The economic burden on cancer patients has been growing due to aging, increased survival rates, and increased use of medical care after diagnosis [7]. However, little is known about which type of cancer is the costliest in Korea. Likewise, it is not known what the year-to-year trends in cancer costs are in Korea.

According to previous studies, the economic burden of all cancers in Korea was approximately $9.4 billion in 2002 [8], about $12.1 billion (14.1 trillion won) in 2005 [9], and $20.9 billion in 2010 [10]. However, it is difficult to directly compare estimates of economic burden across studies because of differences in methodology, particularly differences related to calculations used and variables examined. For these reasons, there are very few studies comparing the economic burden of cancer year-to-year using recent data.

Data on the economic burden of disease are significant, as they can be used for a variety of purposes, such as formation of medical, health, and economic policies and the generation of reports on the general health of the population [11]. Furthermore, data regarding the performance of the national health system could be utilized for the post-evaluation of policies and comparisons with other countries [12,13], as well as evaluations of the effectiveness of medical and health research resource distribution [14]. As such, some countries routinely assess the economic burden of all diseases at the national level [15,16].

The present study was carried out to confirm and compare the prevalence and economic burden of cancer by cancer type, age, and sex utilizing claims data of the National Health Insurance Service (NHIS) of Korea.

Yearly economic cost was estimated using claims data of the NHIS for the period January 1, 2011, to December 31, 2015, which represented the latest available economic and prevalence data.

The selection criteria for subjects used to calculate disease cost were as follows. Subjects were defined as patients who filed claims with the NHIS from 2011 to 2015 for International Classification of Diseases, 10th revision, Clinical Modification (ICD-10-CM) codes indicating cancer (C00-C97) and (1) visited the outpatient department three or more times or (2) were hospitalized one or more times due to same cancer within 1 year of the diagnosis [8]. Cancer (C00-C97) was classified into the following 24 types: lip, oral cavity, and pharynx (C00-C14); esophagus (C15); stomach (C16); colon and rectum (C18-C20); liver (C22); gallbladder (C23-C24); pancreas (C25); larynx (C32); lung (C33-C34); breast (C50); cervix uteri (C53); corpus uteri (C54); ovary (C56); prostate (C61); testis (C62); kidney (C64); bladder (C67); brain and CNS (C70-C72); thyroid (C73); Hodgkin’s lymphoma (C81); non-Hodgkin’s lymphoma (C82-85, C96); multiple myeloma (C90); leukemia (C91-C95); and other unspecified cancer [17].

The economic burden of disease can be divided into direct and indirect costs. Direct costs consisted of direct medical and non-medical costs. NHIS claims data were used to estimate direct medical costs incurred by cancer patients stemming from hospitalization and outpatient care. The non-benefit service cost was defined as the coinsurance rate of the non-benefit service of cancer patients, as presented in the “Survey on the Benefit Coverage Rate of National Health Insurance” [18].

Direct non-medical costs included transportation and nursing costs. Transportation costs were defined as the expenses incurred by cancer patients visiting medical institutions for hospitalization and outpatient treatment. The total annual transportation cost for cancer patients was calculated by multiplying the number of annual visits to the outpatient department and the number of hospitalizations by the average round-trip transportation cost for one visit to the hospital. For this, the annual hospitalization and outpatient transportation costs of cancer patients, per the Korea Health Panel Survey, were used, and the formula used for the calculations was as follows [19].

The number of patients accessing medical services for cancer was 821,525 in 2011, 955,703 in 2012, 1,080,199 in 2013, 1,184,681 in 2014, and 1,237,739 in 2015, showing a continuous, upward trend of approximately 50.7% when comparing 2011 to 2015. In terms of cancer type, thyroid cancer exhibited the highest prevalence (n=189,711) in 2011, followed by stomach cancer (n=102,836), colorectal cancer (n=88,730), and breast cancer (n=88,594). In terms of prevalence by sex, in 2011, the number of female and male cancer patients was 451,427 (54.9%) and 370,098 (45.1%), respectively. In 2015, the number of female and male patients was 695,293 (56.2%) and 542,446 (43.8%), respectively. Overall, the number of female patients exceeded that of male patients each year (Table 1).

The number of deaths from cancer was 71,944 in 2011, 74,214 in 2012, 75,848 in 2013, 77,137 in 2014, and 76,854 in 2015. In terms of cancer type, the mortality rate of lung cancer was the highest, at 31.8 deaths per 100,000 persons in 2011, followed by liver cancer (22.0 per 100,000), stomach cancer (19.5 per 100,000), and colorectal cancer (15.4 per 100,000). In 2015, the mortality rate was higher for lung cancer (34.1 per 100,000), unchanged for liver cancer (22.2 per 100,000), lower for stomach cancer (16.7 per 100,000), and higher for colorectal cancer (16.3 per 100,000) (Table 2).

The cost for all cancer patients was $11,894.9 million in 2011, $12,735.8 million in 2012, $13,426.6 million in 2013, $13,719.0 million in 2014, and $13,945.3 million in 2015, exhibiting a steady increase from 2011-2015 of ~17% (Table 3). In 2011, the direct cost of disease was $5,054.9 million, and the indirect cost was $6,839.9 million. In 2015, the direct cost was $5,955.1 million, and the indirect cost was $7,990.2 million. Thus, the indirect cost was relatively higher than the direct cost (Fig. 1).

Broken down by sex, the cost for male and female patients was $8,268.4 million and $3,626.5 million, respectively, in 2011. In 2015, the cost for men was $9,469.7 million, roughly 2.12 times the cost for women ($4,475.6 million). From 2011 to 2015, the economic burden continued to increase for both sexes (Table 3).

In 2011, the most expensive cancers for men were liver cancer ($1,881.1 million), lung cancer ($1,310.2 million), stomach cancer ($1,191.8 million), and colorectal cancer ($918.4 million). This was also true in 2015: the cost of liver cancer was the highest ($2,020.3 million), followed by lung cancer ($1,538.6 million), stomach cancer ($1,167.6 million), and colorectal cancer ($1,051.9 million). The cost of colorectal cancer continued to increase from 2011 to 2015; the cost of lung cancer increased until 2013, decreased in 2014, and increased in 2015; and the cost of liver cancer increased until 2014, when it declined modestly. The cost of stomach cancer increased until 2013 and began decreasing in 2014.

For women, in 2011, the cost was highest for breast cancer ($780.1 million), followed by colorectal cancer ($395.2 million), stomach cancer($359.6 million), and lung cancer($310.8 million). In 2015, the cost was highest for breast cancer ($1,044.3 million), followed by colorectal cancer ($459.0 million), lung cancer($390.1 million), and stomach cancer($379.7 million). Since 2014, lung cancer has been more economically burdensome than stomach cancer (Table 3).

When costs were ranked based on age, the highest costs were incurred by those 50-59 years old ($36,464.7 million, 30.7% of all cancer costs), followed by those 40-49 years old ($2,491.1 million, 20.9%), 60-69 years old ($2,365.9 million, 19.9%), and ≥ 70 years old ($1,969.0 million, 16.6%). In 2015, those 50-59 years old accounted for the highest costs ($4,309.0 million, 30.9%), followed by those 60-69 years old ($2,853.2 million, 20.5%), ≥ 70 years old ($2,731.0 million, 19.6%), and 40-49 years old ($2,619.0 million, 18.8%) (Fig. 2).

When a discount rate of 3% was applied (the standard for this study) to the indirect cost calculation, the productivity loss due to medical treatment in 2015 was $474.7 million for the economically active population of 60 years old and younger, $539.6 million for the population 64 years old and younger, and $622.2 million for the population ≥ 65 years old. The premature death cost for these age groups was $3,530.7 million, $4,671.9 million, and $7,367.9 million, respectively. For the sensitivity analysis, discount rates of 5%, 3%, and 0% were applied to calculate the economic loss due to premature death. When a discount rate of 5% was applied to each of the three age groups, the premature death cost was $3,140.6 million, $4,155.8 million, and $6,447.4 million, respectively. When no discount rate was applied (0%), the premature death cost was $4,392.1 million, $5,806.6 million, and $9,442.9 million, respectively. Thus, the premature death cost difference was $6,302.3 million, depending on to the methodology used. When a friction cost approach rather than human capital approach was used to calculate indirect costs, the premature death cost for the three age groups in 2015 was $176.0 million, $210.9 million, and $278.0 million, respectively, for the economically active populations of 60 years and younger, 64 years and younger, and 15 years and older (Table 4).

This study estimated the socioeconomic costs for cancer patients from 2011 to 2015 using NHIS claims data. The number of patients who received medical treatment for cancer was 821,525 in 2011 and 1,237,739 in 2015, an increase of approximately 50%. From 2011 to 2015, the total annual costs for cancer patients increased by 17%, from $11,894.9 million to $13,945.3 million. Ongoing studies of the socioeconomic burden of cancer indicate that in countries that rank highly on the Human Development Index, cancer is already a leading cause of death and will soon become a major cause of death worldwide. By 2030, global cancer incidence is expected to increase by 75%, leading to an enormous increase in the economic burden of cancer [22].

The prevalence rate of cancer in Korea increased from 1,639.4 to 2,429.2 cases per 100,000 persons between 2011 and 2015, with thyroid, stomach, breast, and colorectal cancers being the four most common types, in that order. However, the economic burden of disease in 2015 was the highest for liver, lung, stomach, colorectal, and breast cancers, in that order. Thus, the most common cancers do not necessarily incur the highest costs. Indeed, from 2011 to 2015, liver cancer was the most burdensome, but it was not among the most common cancers overall. The economic burden of liver and stomach cancers has decreased since 2013, whereas the economic burden of colorectal and breast cancers has steadily increased since 2014. Although liver cancer was not one of the most common cancers between 2011 and 2015, it had the highest economic burden for men, followed by lung cancer, stomach cancer, and colorectal cancer. For women, breast cancer imparted the highest economic burden, followed by colorectal cancer, liver cancer, and stomach cancer. Thus, the economic burden of various cancers differs according to sex.

It is therefore necessary to consider sex when establishing relevant policies and identify risk factors for particular cancer types in order to develop effective intervention and prevention strategies. According to Globocan 2018, lung cancer has the highest incidence and mortality rate of all cancers worldwide, and one in four women is affected by breast cancer [1]. In light of these data, the government of South Korea included lung cancer on the National Cancer Screening Program form for 2019.

The economic burden of cancer among patients over 60 years old has continued to grow over time. The proportion of the total cost of cancer for patients aged 60 years and older increased from 36.4% in 2011 to 40.0% in 2015, an increase of 3.6%. This trend is thought to result from increases in life expectancy and the overall aging of the Korean population [23], combined with higher cancer survival rates due to advances in treatment [2]. Regardless of the reason, however, this finding reveals that policy interventions may be required to mitigate rising cancer costs.

The indirect costs associated with cancer increased steadily from $6,839.9 million in 2011 to $7,990.2 million in 2015, an increase of approximately 16.9%. Although the socioeconomic cost of the disease varies according to the applied variable, indirect costs could also vary depending upon how the age of the productive population is defined. In the case of premature death costs, a time discount rate can be applied to calculate future costs; however, the results could vary depending on whether a human capital approach or a friction cost approach is applied.

When calculating the socioeconomic burden of disease, there is considerable debate surrounding methodologies for calculating the wages of the elderly. Most studies consider those over 65 years of age as being economically inactive [10,20]. However, when recent wage and employment rate data for this group were included, the economic costs increased by 22% to 28% for males and 52% to 54% for females. Furthermore, when the elderly population was included, the productive population among all age groups increased sharply [24]; consequently, the assumption that those over 65 years of age earn no income results in underestimation of the economic burden of disease. Therefore, we used the Ministry of Employment and Labor’s raw survey data to determine the actual work status of and wages earned by men and women aged 65 years and older. The employment rate was calculated by applying the results of the survey to the actual work status of the elderly.

The retirement age for public servants in Korea is 60 years [25], but not all individuals retire at that age. Therefore, in this study, three productive population groups were considered: those 60 years old and younger, those 64 years old and younger, and those 15 years old and older. Additionally, we compared the results obtained by applying time discount rates of 0%, 3%, and 5% for our sensitivity analyses. In this study, we applied the basic discount rate of 3% to the income of those aged ≥ 15 years ($13,945.3 million). The socioeconomic burden of cancer ranged from $9,570.4 million (5%, 60 years old and younger) to $16,020.3 million (0%, 15 years old and older) in 2015, a difference of 60%, depending on the methodology used.

The friction cost approach can also be applied to the indirect cost calculation, in addition to the human capital approach [26] applied in this study. The human capital approach estimates the expected income lost due to disease morbidity or premature death from the patient’s perspective, whereas the friction cost approach calculates costs from an employer’s perspective. Therefore, the friction cost approach calculates the losses incurred until the patient’s workload is reassigned or a replacement is hired [27], and it only reflects the value of the substitute employee during the friction period. The human capital approach has the limitation that the loss could be overestimated because it measures potential loss without considering the possibility of labor substitution [28]. However, the friction period can vary from 3-4 months to 22 weeks, or to 58 days based on country, occupation, and specific disease [15].

Canada, which periodically calculates the economic burden of all diseases, recently reported these costs for 2005-2008 and 2010 calculated using the friction cost approach [15]. In Canada, the cost of premature death due to cancer was $278 million in 2010. Unfortunately, there were not enough Korean data available to accurately determine a friction period in the present study; therefore, the Canadian friction period was used to calculate a friction cost of $278 million in 2015 for those ≥ 15 years of age in Korea. When defining the age of the economically active population as those under 60 years old, the friction cost was $176 million. A comparison of the friction cost for those under 60 years old to the highest cost of premature death in our study ($9,442.9 million) showed that estimated costs can vary by up to 192.8%, depending on the applied variables.

This study has several limitations. First, the data used in this study were limited. Only NHIS insurance claims data were used to calculate the cost of cancer to patients, and these data are limited because the system was created for the purpose of claims processing [29]; thus, the scope and scale of any non-benefit service items could not be determined [30]. To compensate for this limitation, non-benefit service costs were estimated using the coinsurance rate of non-benefit services based on the “Report on the Actual Status of Medical Service Cost of Health Insurance Patients,” published by the NHIS. Moreover, this study failed to consider differences in economic activities and wages earned between women and men. In the case of women, if a wider range of costs is applied by including household activities (such as childrearing) in addition to economic activities, the socioeconomic burden of cancer is expected to be higher. There are also limitations related to cost items. In this study, all other costs, including informal medical costs, were excluded from the calculations. Therefore, there were limitations in estimating the costs of purchasing assistive devices and medicines, costs of exercise and diet management, and the use of medical services, such as complementary, alternative, and traditional medicine. Follow-up studies should determine a more precise scale of costs.

Despite these limitations, this study estimated the socioeconomic burden on cancer patients from 2011 to 2015 using the latest Korean data—the NHIS claims data and Statistics Korea’s death data. The economic burden for each of 24 cancer types was also determined based on indirect cost calculations. During the period 2011 to 2015, the economic burden of cancer increased and is expected to increase further. Follow-up studies aimed at predicting the future economic burden of cancer in Korea using trend analyses could be designed based on the results of this study. The results of these follow-up studies could then be used as the basis for allocating limited health care resources and guiding the decisions of policymakers. To reduce the disease burden of cancer, it will be necessary to emphasize changes in lifestyle and promote early cancer detection and treatment [21]. Social and national efforts should continue to establish measures and policies for improving the quality of life for patients and reducing their psychological pain and economic burden.