Abstract
Cancer screening is the most cost-effective prevention strategy that enables improved cancer morbidity and mortality rates. However, there was a significant decline in cancer screening during the COVID-19 pandemic worldwide. This study aimed to measure regional differences in cancer screening by comparing screening rates before and during COVID-19 and to examine factors explaining the regional difference in cancer screening. All variables including the differences in cancer screening rate before and during the COVID-19 pandemic were collected from 229 administrative subregions in South Korea and joined with each community in ArcGIS environment. Factors explaining regional difference in cancer screening were analyzed through descriptive statistics, correlation analysis, and interactive decision tree modeling. Cancer screening rates decreased in all regions during the COVID-19 pandemic in South Korea. The findings revealed that national health insurance premiums, the number of cancer screening facilities, livelihood benefit recipients, and the population aged 65+ could be classified as variables highly affecting the decrease in cancer screening rate. These findings highlight the efforts to appraise different barriers to cancer screening in each community. In addition, it is important to promote sustainable cancer screening strategies from regional perspectives based on regional vulnerabilities and their complex impact on cancer screening.
What We Already Know
Cancer screening declined significantly around the world during the COVID-19 pandemic compared with 2019.
Individual factors associated with adherence to cancer screening included age, race, household income, health-related behavior, and health literacy.
What This Article Adds
The decline in cancer screening rates varied significantly by region, and the magnitude of the decline varied depending on the regional characteristics.
Regional factors affecting cancer screening were identified as national health insurance premiums, the number of cancer screening facilities, livelihood benefit recipients, and population aged 65 or older.
Introduction
Cancer is a major cause of morbidity and mortality in both developed and developing countries, with an estimated 19.3 million new cases and 10.0 million deaths worldwide in 2020. 1 Cancer is also the leading cause of death in South Korea. 2 More than 250 000 patients were newly diagnosed with cancer and 81 203 deaths were due to cancer in 2019. 3 Given the heavy health and social burden of cancer, broad strategies are needed to alleviate the burden both nationally and globally. 4
Cancer screening is considered the most cost-effective prevention strategy that provides protection against possible harms of cancer. 5 It enables early detection of precancerous lesions and early-stage cancer diagnosis, consequently improving cancer mortality and morbidity rates. 6 The Korean government initiated the National Cancer Screening Program in 1999. It was designed based on the life cycle and health risks, including screenings for stomach, liver, colorectal, breast, uterine cervical, and liver cancers. 7 It has contributed to the improvement of cancer survival rate from 42.9% in 1993-1995 to 77.5% in 2013-2017 in South Korea. 8
Cancer screening, as a preventive care measure, significantly declined during the COVID-19 pandemic compared with 2019, with screening rates remaining below prepandemic levels, despite some rebound in 2021. 9 Previous studies have identified the impact of the COVID-19 pandemic on cancer screening across cancer. Routine screening for breast and colorectal cancers declined by over 90% of eligible individuals during 2020 in the United States. 10 Impacts of COVID-19 on Australian screening program have been observed, showing reduced screening rates of breast, bowel, and cervical cancers. 11 Likewise, there was a significant decline in cancer screening for stomach, colorectal, breast, and cervical cancers during the pandemic in South Korea. 12 Women’s cancer screening for breast and cervical cancers was also lower in 2020 than that in 2019. 13
Studies conducted before the COVID-19 pandemic mainly focused on individual factors associated with adherence to cancer screening, such as age, race, household income, health-related behavior, and health literacy.14 -17 After the outbreak of COVID-19, recent studies have shown direct and indirect effects of COVID-19 in reducing cancer screening. As a result of national health system reconfiguration in response to COVID-19, cancer screening was suspended or disrupted in many countries. 18 For cancer screening that requires an in-person visit or procedure, people tended to avoid face-to-face services more. 12 However, these studies did not examine whether decreased cancer screening rates were driven by the environment surrounding the individual.
Preexisting regional conditions such as population density and health care system should be considered when explaining health-related behaviors in the COVID-19 situation because these could be shared within or have a common influence on the community. COVID-19 has significantly affected poor urban communities and the vulnerable population living in low health care capacity, exacerbating COVID-19 mortality and other health status.19,20 Cancer screening might also be affected by regional vulnerabilities. However, little research has directly addressed these issues during COVID-19. Therefore, this study aimed to measure regional differences in cancer screening by comparing screening rates between 2019 (before COVID-19) and 2020 (during COVID-19) and examine factors explaining regional difference in cancer screening using decision tree analysis.
Methods
Data Source and Preparation
Data used in this study were collected from 229 administrative regions across 16 provinces in South Korea using various data sources. Cancer screening rate was obtained from National Screening Program Statistics for 2019 and 2020 provided by National Health Insurance Corporation (NHIC). Administrative data were primarily collected from Korean central government statistics portal, including 2020 Population Census of Ministry of the Interior and Safety, 2019 Infrastructure, and Transport, land ownership of Ministry of Land, 2018 Traffic accessibility Indicator of Korea Transport Institute, 2019 Regional Healthcare Use Statistics of National Health Insurance Corporation, and 2020 Community Health Survey of Korea Centers for Disease Control and Prevention and COVID-19 Statistics of Korean Government. All variables were extracted at the community level and joined with the communities using ArcGIS environment.
Variables
The dependent variable was the difference in cancer screening rate (percentage point; %p) before and during the COVID-19 pandemic. It was calculated by subtracting the rate in 2020 from the rate in 2019 for each administrative region. Cancer screening rate was calculated as a proportion of the population that participated in cancer screening actually among eligible population for stomach, colorectal, breast, or cervical cancer. Explanatory variables describing the characteristics of the region were defined into five categories based on COVID-19 Community Vulnerability Index (CCVI) and COVID-19 Pandemic Vulnerability Index (PVI).21,22 Five categories included COVID-19-related, demographics, socioeconomics, health status, and health care system. Indicators for each category were selected by considering data availability in South Korea (Supplementary Table 1).
Data Analysis
Descriptive statistics were used to summarize differences in the cancer screening rate before and during COVID-19 pandemic for each region. Correlations between variables were analyzed, and an interactive decision tree model was used to examine the factors explaining regional difference in cancer screening. A decision tree can be used to identify the most significant relationship with the dependent variable among explanatory variables. It splits the sample into progressively smaller subgroups by employing a specific criterion from the root of the decision tree to each branch of the tree. Descriptive and correlation analyses were performed using SAS 9.4. Interactive decision tree modeling was done with SAS Enterprise Miner 15.1.
Results
Cancer Screening Rate Difference
Descriptive statistics of the cancer screening rate difference between 2020 and 2019 are presented in Table 1. Compared with 2019, screening rates decreased in all regions during the COVID-19 pandemic. Regional distribution of the decreased screening rate is shown in Supplementary Figure 1. Supplementary Table 2 shows top 10 regions with the highest rate of decrease, all of which are nonmetropolitan areas.
Descriptive Statistics of Cancer Screening Rate Differences Between 2020 and 2019.
Correlation Analysis
Table 2 summarizes correlations between cancer screening rate changes and regional characteristics. The degree of decrease in the cancer screening rate increased with higher proportions of the population aged 65+, disabled, and less educated individuals, as well as with higher prevalence of chronic disease and lower health care access. However, conversely, the decrease in the cancer screening rate lessened with higher proportions of the population aged 20 to 39 years, national health insurance premiums, adherence to healthy living practices, availability of cancer screening facilities, hospital beds, and health care personnel. There was no significant correlation between decreased cancer screening rate and COVID-19-related characteristics.
Correlations Between Variables.
Factors Affecting Cancer Screening Differences Using Decision Tree
Figure 1 presents a decision tree that describes pathways and factors associated with differences in cancer screening rate before and during the COVID-19 pandemic. The decision tree analysis revealed that national health insurance premiums, the number of cancer screening facilities, livelihood benefit recipients, and population aged 65+ could be classified as variables highly affecting the decrease of cancer screening rate. Among these variables, national health insurance premiums (less than 80 720 Won) was the most powerful determining factor, which was located at the root of the decision tree and presented as the first-level split of two initial branches. The number of cancer screening facilities per 100 000 population (less than 4.13) was located on the second-level split. Estimated percentage of basic livelihood benefit recipients and the population of those aged 65+ were located on the third-level split.
Factors affecting cancer screening differences using a decision tree.
Discussion
The aim of this study was to identify differences in cancer screening between pre-COVID-19 and post-COVID-19 and regional factors of these differences. Twenty-three explanatory indicators under five categories were explored at the level of 229 administrative regions in South Korea and a decision tree model was employed in this study.
We found reductions in cancer screening across all regions due to the COVID-19 pandemic. The cancer screening rate decreased by up to 15.0% compared with the rate before COVID-19. Previous studies also reported declines of multiple types of cancer screenings during the pandemic as a result of cancelations of elective procedures and preventive services.23 -25 The probable reason is patients’ fear of visiting in-person services because of COVID-19 infection through medical facilities. In addition, COVID-19 affected the national public health priority. Public interest in cancer screenings decreased noticeably after the outbreak of COVID-19. 26 This result raises concerns about the impact of delayed or canceled cancer screening on overall cancer care such as diagnosis at advanced stages, delayed treatment, and increased mortality.
In a decision tree analysis, we found evidence suggesting that regional socioeconomic status could decrease cancer screening rate. Magnitude of the decline also varied depending on the socioeconomic level. Cancer screening rates declined significantly for regions paying less national health insurance premiums, that is, regions with low monthly household incomes. In addition to this, decreases in cancer screening were also confirmed in regions with a high percentage of basic livelihood benefit recipients. There were geographic disparities between rural and urban individuals in cancer screening even before the COVID-19 pandemic. 27 However, COVID-19 has exacerbated health inequality not only at the individual level, but also at the community level. 20 Similar to a previous study, this study showed that COVID-19 had a negative impact on communities with low socioeconomic status than on communities with high socioeconomic status. Considering that regular cancer screenings are provided at no additional charge in South Korea, it is difficult to interpret whether these reductions are directly caused by individual economic factors. Therefore, efforts are needed to appraise and improve different barriers to cancer screening in each community with a low socioeconomic status.
Interestingly, results of this study showed a significant decrease in regions with fewer cancer screening facilities during COVID-19 in both regions with high and low national health insurance premiums. These findings align with those of previous studies showing that lack of nearby screening facilities is a barrier to cancer screening.28,29 The distribution of medical facilities is highly dependent on the geographical location and socioeconomic status of regions. It is inevitably uneven. Despite South Korea’s relatively high accessibility to medical facilities, access to health care is having an impact on cancer screening in public health disasters like COVID-19. More seriously, the degree of negative impact was remarkable in regions with low socioeconomic status. This structure could lead to the lack of medical and social capital resources for residents in vulnerable regions, further alienating them. Thus, they would not be able to access appropriate health care services, especially during the COVID-19 situation. Thus, it is highly recommended to minimize poor access to health care in these regions and improve the distribution of medical resources in national public health strategies.
In regions with low socioeconomic status but high access to cancer screening facilities, cancer screening decreased further in regions with a low proportion of 65-year-old population. This result indicated that young people were more passive in cancer screening than the elderly because they were less susceptible to the risk of cancer. Cancer screening is cost-effective to begin younger. Thus, it is important to encourage young people to participate in routine cancer screening. 30 For regions with high socioeconomic status and low access to cancer screening facilities, cancer screening decreased further in regions with a high incidence of COVID-19. For possible reasons, we suspect that these regions are more likely to be metropolitan cities with a high population density. The high concentration of individuals increased fear of COVID-19 infection, which hindered cancer screening. In addition, travel restrictions and social distancing would have been strongly applied to these regions.
This study had several limitations. First, the study design was cross-sectional. Thus, the causal effect of regional characteristics could not be inferred. The lack of available data that could accurately reflect the regional context was another limitation. Only limited variables were used in this study due to a lack of data sources collected at 229 administrative district levels. In addition, political and regional changes according to the progress of COVID-19 were not measurable, for example, social distancing policy or temporary suspension of cancer screening. Inclusion of these changes could lead to more accurate results. Thus, future studies should consider contextual changes over time and temporal factors. Despite these limitations, to the best of our knowledge, this is the first national-based attempt to demonstrate regional factors based on decision tree modeling of cancer screening differences before and during the COVID-19 pandemic. Further studies will be necessary to explore a variety of potential regional factors and define characteristics that can vary by region in relation to cancer screening.
Conclusion
Fully understanding the regional characteristics of cancer screening is the key to implementing targeted large-scale screening programs for a community, especially in the context of a public health crisis. Given that each region has unique vulnerabilities affecting cancer screening, identifying these factors will facilitate the development of a new, community-driven public health approach to cancer management. It should be noted that the effectiveness of population-based cancer screening depends on a high participation rate. Our findings highlighted declined cancer screening due to COVID-19 and differences in regional conditions that affected the reduction of cancer screening using a decision tree model. Therefore, it remains crucial to identify regional vulnerabilities and understand their complex impact on cancer screening. In addition, promoting sustainable cancer screening strategies from regional perspectives is essential, particularly when responding to public health disasters.
Supplemental Material
sj-docx-1-aph-10.1177_10105395241282981 – Supplemental material for Regional Impact of the COVID-19 Pandemic on Cancer Screening
Supplemental material, sj-docx-1-aph-10.1177_10105395241282981 for Regional Impact of the COVID-19 Pandemic on Cancer Screening by Jongho Park, Yeaeun Kim and Jae-Hyun Park in Asia Pacific Journal of Public Health
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Supplemental Material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
