Introduction

Despite improvements in early detection and treatment, the incidence and mortality rates from gastric cancer remain high both worldwide and in Korea. Gastric cancer is the third leading cause of death from cancer worldwide, with 951,600 new cases and 723,100 deaths in 2012 [1]. The burden of disease from gastric cancer is highest in the east Asian region, particularly in Korea, Japan, and China, with about three quarters of all new cases worldwide occurring in Asian countries [1, 2]. In Korea, gastric cancer remains the most commonest cancer in males; 19,545 new cases were detected in men in 2015. Furthermore, gastric cancer is the fourth most common cancer in women, with 9662 new cases detected in women in 2015 [3].

Gastric cancer screening enables the detection of early stage cancer, contributing to dramatic decreases in mortality from the disease in Korea [4]. Due to the lack of symptoms or signs in the early stage of gastric cancer, it is easy to miss the window for early treatment without regular screening [4]. In Korea, the government started the National Cancer Screening Program (NCSP) in 1999 to provide Korean people aged 40 or older with gastroscopy or upper gastrointestinal series (UGIS) every other year for free, including a biopsy if required [5]. A previous study of the effectiveness of gastric cancer screening in Korea showed that ever-screened subjects had a 21% reduction in mortality from gastric cancer, and as the number of screenings increased, the reduction in the mortality rate from gastric cancer became greater [4]. Partly as a result of the national gastric cancer screening program, the age-standardized gastric cancer mortality rate decreased from 23.8 to 8.9 per 100,000 persons from 1999 to 2015 [3]. Cost-effectiveness of gastric cancer screening in Korea is also well established.[6] Although there are some potential risks of endoscopic screening, including infection (hepatitis B) or bleeding, with good skills and quality management, risk of serious adverse effects (e.g. bleeding requiring admission, anaphylactic shock, or death) is extremely low [7].

In spite of these achievements, inequalities persist in the uptake of cancer screening, particularly with regard to socioeconomic position [8, 9]. Another overlooked dimension of disparity is disability status [10]. People with disabilities are diverse, and their ability to request and receive preventive care depends on the specific type and severity of their disability [11]. For example, some disabilities impede gastric cancer screening participation and would, thus, influence the gastric cancer screening modality (i.e., gastroscopy vs. UGIS). The diverse obstacles associated with different disabilities have important implications for the creation of tailored interventions to improve gastric cancer screening participation.

However, few data are currently available on disparities in gastric cancer screening with regard to disabilities. One previous study analyzed data from the 2005 Korean National Health and Nutrition Examination and found that disability-related factors were not significantly associated with gastric cancer screening participation [9]. However, that study used self-reported disability status, defined as limitations in general activity, walking problems, visual problems, or hearing problems, and did not consider severity information. That is, the study was not based on an objective clinical assessment of disability status and, therefore, could not examine the heterogeneity of gastric cancer screening rates among people with diverse types and severities of disability. Furthermore, it was a cross-sectional study performed in 2005. Data describing trends in gastric cancer screening among people with disabilities are still lacking. To our knowledge, no study has yet used a large-scale administrative data to address that research gap.

In Korea, universal health coverage is offered to all people, and the gastric cancer screening program is offered by the NCSP at minimal or no cost [12]. In addition, through the national disability registration system, the types and severities of disabilities are classified and registered based on medical examinations and specific criteria, providing a unique opportunity to test how specific types and severities of disabilities affect gastric cancer screening participation.

In this study, we used linked administrative data to investigate 1) how gastric cancer screening participation and its modalities differ according to the presence, type, and severity of disability; and 2) temporal trends in the gastric cancer screening rate among people with disabilities.

Methods

Study setting and data sources

Korean National Health Insurance Service (NHIS)

The NHIS is the only government insurer, offering universal health insurance that covers approximately 97% of the Korean population. The government covers the medical fees of people with the lowest incomes, and their qualification status and reimbursement are also settled by the NHIS. Therefore, the NHIS has comprehensive information about the age, sex, residential area, and income level of Koreans. In Korea, health insurance coverage is determined only by income level, not according to pre-existing health risk or disability status.

National Gastric Cancer Screening Program in Korea

The NCSP was initiated in 1999 as part of the 10-year National Cancer Control Plan [13]. Currently, the NCSP covers stomach, liver, colorectal, breast, and cervical cancer screenings for all people as indicated by age (Supplementary Table 1). Since 1999, free gastric cancer screening by gastroscopy or UGIS has been provided every other year to all Korean people aged 40 and older [4]. If gastric cancer is suspected by UGIS, gastroscopy can be offered as the next step, and biopsy is also provided at no charge [5].

Table 1 Number of eligible and screened subjects based on the presence, severity, and type of disability over a 10-year period
Full size table

All people eligible for the gastric cancer screening program receive an invitation containing information about gastric screening methods and the locations of nearby NCSP providers [14]. The Korean NHIS maintains complete information about both the eligibility for national gastric cancer screenings in a given year and the actual participation.

Disability registration system in Korea

In 1988, Korean government established a national registration system for disabilities that is categorized by type and severity for the purpose of determining welfare benefits. If an individual wishes to be registered as disabled, they must submit appropriate and validated documentation to a local National Pension Service office. The paperwork includes valid results of a disability diagnosis from a specialist physician in the relevant field in accordance with government guidelines for the specific disability. The national disability registration system recognizes 15 types of disability and 6 levels of severity (Supplementary Table 2). The level of severity for each disability is determined by the specialist physician according to pre-defined criteria by the Ministry of Health and Welfare guidelines, based on the degree of functional losses and clinical impairment. Severity is graded into six levels: from grade 1 (most severe) to 6 (least severe) [15,16,17,18]. As an example, for visual impairment, patients who have visual acuity < 0.02 in the better eye are classified as grade 1 (>85% of functional loss), and subjects who have visual acuity < 0.2 (better side) or loss of visual field > 50% in both eyes are classified as grade 5 (35–44% of functional loss) (Supplementary Table 3). For brain injuries, people who cannot perform ambulation and activities of daily living (ADL) due to quadriplegia or hemiplegia and totally need help from others are classified as grade 1. Those who need partial help in ambulation and ADL are classified as grade 3, and those who can perform ambulation and ADL perfectly by themselves but take a long time are classified as grade 6. In renal failure, people who had a kidney transplantation are classified as grade 5, and those who received hemodialysis or peritoneal dialysis for more than 3 months are classified as grade 2.

Data source and study subjects

The data used in this study were from the National Health Information Database (NHID) for 2006 to 2015. The NHID is public database containing healthcare utilization, health screening, sociodemographic, and mortality data for the whole population of South Korea. It, thus, provides an excellent platform for epidemiological and health policy studies. We described the details of the database profile elsewhere [19, 20]. Because the NCSP made some changes in its coverage and copayments during its early implementation phase (2001–2005), we have limited our analyses of gastric screening variables to 2006–2015 for consistency.

Statistical analyses

We derived age- and sex-standardized participation rates with 95% confidence intervals for each year during the study period according to the presence, type, and severity of a disability.

The 2010 Census of the Korean population was used for the age and sex standardization. We also assessed the percentage of screening participants in each year who underwent endoscopy and UGIS.

To examine factors associated with participation in gastric cancer screening, we carried out a series of multivariate logistic regressions using variables for disability (presence, severity, and type), and other sociodemographic variables (age, sex, income level, and place of residence). In Model 1, we compared the screening rate of people with disabilities with that of people without disabilities. In Model 2, the severity of disability was categorized into mild vs. severe, and the screening rates for each category were compared with rates among people with no disability. In Model 3, the odds ratio of screening according to disability grade was compared with those with no disability. In Model 4, the odds ratio of screening for people with different types of disabilities was compared with that of people with no disability.

We performed all analyses using SAS 9.3 software (Cary, NC, USA), and p < 0.05 was considered statistically significant. This study was reviewed by the Institutional Review Board of Chungbuk National University (CBNU-201708-BM-501-01).

Results

Study participants

The number of people invited to undergo gastric cancer screening increased from 10 million in 2006 to 12 million in 2015. Among these, the proportion of people who had a registered disability increased from 5.75% in 2006 to 8.06% in 2015 (Table 1).

Trends in gastric cancer screening rates according to disability status

The number of eligible and screened people in the national gastric cancer screening program and the crude and age- and sex-adjusted participation rates according to time are given in Table 1 and Supplementary Table 4. Trends in the participation rate in the national gastric cancer screening program from 2006 to 2015 are shown in Fig. 1. The age- and sex-adjusted screening rates for gastric cancer among people with disabilities increased from 25.9% in 2006 to 51.9% in 2015 (absolute change: + 26.0%). Over the same period, the screening rate among people without disabilities increased from 24.7 to 56.5% (absolute change: + 31.8%).

Fig. 1
figure 1

Gastric cancer screening rate according to the presence, severity, and type of disability from 2006 to 2015

Full size image

In terms of disability severity, people with mild disabilities showed a higher increase in screening rate (from 28.9 to 58.3%, absolute change: + 29.4%), whereas people with severe disabilities exhibited a more modest increase (from 20.3 to 40.8, absolute change: + 20.5%). Overall, that trend was linear: people with grade 1 (most severe) disabilities showed the lowest increase in screening rate (from 10.8 to 29.0%, absolute change: + 18.2%), while people with grade 6 (least severe) disabilities showed the highest increase (from 29.2 to 60.8%, absolute change: + 31.6%).

Among the disability types, both the highest screening rates and largest increases were observed among people with physical disabilities (from 28.2 to 58.1%, change: + 29.9%), facial disfigurements (from 25.4 to 55.5%, change: + 30.1%), visual disabilities (from 27.2 to 55.6%, change: + 28.4%), and hearing disabilities (from 27.8 to 54.5%, change: + 26.7%). Disabilities related to internal organ problems also showed relatively large increases, as shown in liver diseases (from 9.9 to 40.1%, change: + 30.2%) and heart problems (from 21.4 to 45.4%, change: + 24.1%). The lowest screening rate and the smallest increase in screening were observed in people with disabilities caused by autism. However, the total number of people in that group was too small, and hence, we excluded those data. Otherwise, the lowest screening rates and smallest increases were seen in people with renal failure (from 8.6 to 31.9%, change: + 23.3%), disabilities caused by brain injuries (from 17.8 to 37.0%, change: + 19.2%), intellectual disabilities (from 15.8 to 35.7%, change: + 19.9%), and disabilities caused by mental disorders (18.1 to 37.1%, change: + 19.0%) (Supplementary Table 4).

Factors associated with gastric screening

Adjusted gastric cancer screening rates for 2014–2015 are displayed by disability type and grade in Fig. 2. The patterns varied with the type of disability: overall, people with physical, facial disfigurement, epilepsy, visual, or hearing disabilities showed higher screening rates than those with disabilities related to the brain/mental disorders (autism, brain injury, intellectual disability, or mental disorder), renal failure, or ostomy.

Fig. 2
figure 2

Gastric cancer screening rate and modality by the type of disability in 2014–2015

Full size image

After adjustment for age, income level, place of residence, and calendar year, the presence of a disability was associated with a slightly lower gastric cancer screening rate [adjusted OR (aOR) 0.89, 95% confidence interval (CI), 0.88–0.89]. People with severe disabilities showed a markedly lower screening rate than people without disabilities (aOR 0.58, 95% CI 0.57–0.58); on the contrary, people with mild disabilities had higher screening rates than those without disabilities (aOR 1.11, 95% CI 1.10–1.11). As the severity of disability increased, the odds of having gastric cancer screening decreased gradually, and as the severity of disability decreased, the odds of having gastric cancer screening increased gradually (Table 2). Compared with people without a disability, people with grade 4, 3, 2, 1 disabilities had 0.93 (0.92, 0.94), 0.75 (0.74, 0.75), 0.50 (0.49, 0.50), and 0.34 (0.34, 0.35) times lower participation rates in gastric cancer screening, respectively. Compared with people without a disability, people with grade 5 or 6 disabilities had 1.09 (1.09, 1.10) and 1.24 (1.24, 1.25) times higher participation rates in gastric cancer screening, respectively.

Table 2 Factors associated with gastric cancer screening in year 2014–2015
Full size table

By disability type, people with autism (aOR 0.36, 95% CI 0.25–0.52), renal failure (aOR 0.39, 95% CI 0.38–0.39), brain injuries (aOR 0.41, 95% CI 0.40–0.41), ostomy problems (aOR 0.53, 95% CI 0.51–0.55), intellectual disabilities (aOR 0.54, 95% CI 0.53–0.54), or mental disorders (aOR 0.55, 95% CI 0.54–0.56) showed substantially lower probabilities of having received gastric cancer screening than people without a disability. On the other hand, people with physical disabilities (aOR 1.13, 95% CI 1.12–1.13), facial disfigurement (aOR 1.05, 95% CI 0.96–1.15), or epilepsy disabilities (aOR 1.03, 95% CI 0.97–1.09) had higher screening rates than people without disabilities (Table 2).

Trends in the use of gastroscopy as the initial screening modality according to disability

The trends in the number and proportion people who received gastroscopy as the initial screening modality from 2006 to 2015 are shown in Fig. 3 and Supplementary Table 5. The use of gastroscopy as the initial screening modality for gastric cancer among people with disabilities increased from 42.7% in 2006 to 76.1% in 2015 (change: + 33.4%); the same rate among those without disabilities increased from 48.4 to 83.0% (change: + 34.6%).

Fig. 3
figure 3

Trend in gastric cancer screening modality by the type of disability

Full size image

Based on disability severity, the use of gastroscopy as the initial screening modality for gastric cancer among people with mild disabilities increased from 43.6% in 2006 to 76.9% in 2015 (change: + 33.3%); the same rate among those with severe disabilities increased from 40.4 to 74.0% (change: + 33.6%). The magnitude of increase in the use of gastroscopy as the initial screening modality was also similar among various disability types.

Discussion

To the best of our knowledge, this study is the first to show disparities in gastric screening rates among people with various severities and types of disabilities. The strengths of our study include the use of a large, representative sample and accurate measurements of disability status and screening practices.

Our results indicate the presence of significant disparities in gastric cancer screening participation among people with disabilities in a setting with minimal financial barriers. We have also shown that the pattern of disparities differs significantly by the severity and type of disability. Although gastric cancer screening rates in people with disabilities increased steadily during the study period, the screening rate in people without disabilities increased more rapidly during the same period, enlarging the disparity between the two groups over time. Furthermore, the choice of initial screening modality also differs significantly by the severity and type of disability, and the proportion of people with disabilities who received gastroscopy as the initial modality is increasing, although it was consistently lower than in people without disabilities.

Previous studies suggested several barriers that could account for the disparity. People with disabilities might not receive preventive screenings (e.g., Pap tests, mammography, dental checks, cancer screenings) because of the unavailability of transportation, provider limitations (poor knowledge and negative attitudes among physicians), or patient limitations (limited access to health information or poor communication with their physicians) [21,22,23,24]. We found that having a severe disability correlated with lower screening rates for gastric cancer. Thus, people with disabilities, especially those with severe disabilities, are unlikely to take full advantage of the national free cancer screening service without an improvement in the physical, social, and attitudinal barriers to their participation [25].

On the other hand, we found that people with mild disabilities are more likely to receive gastric cancer screening than those without disabilities [23]. Some previous studies have also reported this phenomenon [26, 27], which probably occurs, because people with mild disabilities have more promoting factors (high health awareness and increased contact with health providers) than restricting factors (mobility or communication barriers) [23]. However, people with mild disabilities show a lower rate of choosing gastroscopy as the initial modality than those without disabilities, perhaps because they have more intention to do the test itself, but have some uncertain fear of the gastroscopy procedure.

People with different types of disabilities experience different cancer screening barriers: physical barriers impair access to facilities or diagnostic equipment [28, 29]; visual and hearing disabilities can limit access to screening information and communication with physicians [30, 31]; and brain-related/mental disabilities can result in limited knowledge about cancer screening [22, 24].

In our study, people with renal failure were the least likely to participate in the gastric screening program and also showed a smaller increase in the screening rate across the 10 year period. Given that people with such disabilities do not usually have mobility or communication barriers, their reasons for avoiding gastric cancer screening could be related to a lack of time (e.g., hemodialysis 3 times a week), they thought that they are already terminally ill, depression, or a fear of further medical interventions.

In another example, people with brain-related/mental disabilities (autism, brain injury, intellectual disabilities, and mental disorders) showed lower rates of gastric cancer screening and smaller increases over the 10 year period than people without disabilities. These groups are characterized by cognitive and communication impairments. They might have poor communication with their healthcare providers; limited family, social, and community resources; or difficulty in understanding the importance and procedures of cancer screening [24]. Discrimination by healthcare providers against this population could also be a barrier [32]. A Japanese study revealed extremely low gastric cancer screening rates in schizophrenic patients, and suggested psychiatric outpatient clinics could be ideal places for individual interventions, as these people might not understand cancer screening and recall their prior participation.[33]

Participation in gastric cancer screening increased steadily during the study period irrespective of the type or severity of disability. However, the overall uptake remains low (56.7%) even among people with disability, and the disparity gap between people with and without disabilities also increased. This result contradicts previous studies, which suggested that the NCSP was succeeding in encouraging cancer screening equity among groups with different age and income status in Korea [34]. We have shown that different types and severities of disability affect participation in the gastric cancer screening program, and that the most disabled people are not properly benefiting from the current NCSP. It is important to develop healthcare policies to decrease this disparity in gastric cancer screening rates. For example, the NCSP could target specific information to people with disabilities (braille for visual disabilities and audiotapes for hearing disabilities), offer transportation support, allocate additional time for visits with disabled people, address negative and defensive attitudes among healthcare providers, and encourage parent/guardian recognition and participation in gastric cancer screening [23, 35]. People who have disabilities that do not negatively affect their life expectancy need to receive gastric cancer screening at rates comparable to those without disabilities.

People with disabilities generally had lower gastroscopy rates than people without disabilities, especially when the disability was severe, and the proportion of gastroscopy differed by disability type. UGIS is generally not recommended, because numerous studies have shown that gastroscopy offers a better accuracy than UGIS in detecting cancer [8, 36]. However, we could not determine the proportion of people with disabilities who could safely be screened by gastroscopy instead of UGIS. For example, people with mental disorders or intellectual disabilities show the lowest rates of gastroscopy as an initial modality. They might have avoided gastroscopy from fear or concerns about discomfort, because they might have difficulty in understanding the gastroscopy procedure. In other cases, the healthcare provider might have preferred UGIS over gastroscopy because of its simplicity and the difficulty in getting certain patients to cooperate with the gastroscopic exam. Further studies are required to assess the appropriateness of the modality selected for gastric cancer screening of people with disabilities.

Our study has some limitations. First, we could not account for several variables that can affect gastric cancer screening participation, such as the educational level, knowledge about preventive healthcare services, guardian factors, employment, and whether a disability is congenital. Further studies are needed to discover other factors that influence participation in gastric cancer screening. Second, our study did not have information about why people with disabilities did not get gastric cancer screening. Further studies that gather qualitative data through interviews or surveys are necessary to determine those reasons and establish healthcare policies. Third, people can have multiple disabilities simultaneously, but we could not take that into consideration because of the complexity of the analysis. Fourth, because of the specifics of the current Korean healthcare system, which might not reflect worldwide trends, our findings could have limited generalizability. Population-based gastric cancer screening is rarely performed except for Korea and Japan [37]. Nonetheless, our findings can suggest and broaden the understanding needed to develop preventive healthcare services that will function equally.

In summary, in spite of the accessibility of the NCSP, significant disparities exist in gastric cancer screening participation, especially among people with severe disabilities and people with renal failure and brain-related or mental disabilities (autism, brain injury, intellectual disabilities, and mental disorders). Although participation in gastric cancer screening increased steadily in people with disabilities during the study period, regardless of the type and severity of disability, the disparity between people with and without disabilities also widened. Our findings demonstrate the need to identify the specific barriers to gastric cancer screening in this vulnerable population and develop healthcare policies and interventions to remove them.