Are there disparities across identity groups for cancer mortality rates, and if there are, do these inequalities stack across multiple identities?
For this digital humanities project, we have chosen to use the Cancer CSV File from the CORGIS Dataset Project to discover potential disparities across identity groups and potential places of compounding inequalities. Cancer is a major public health issue in the United States and has a significant impact on families, individuals, and healthcare systems. Per the CDC, cancer is the second leading cause of death in the country with approximately 600,000 dying in the United States each year. Because of its impact in society, it’s important for us to discover any disparities across various identity groups and whether these disparities stack as we discuss the ideas of belonging to multiple identity groups in relation to intersectionality.
This dataset consists of “information about the rates of cancer deaths in each state is reported. The data shows the total rate as well as rates based on sex, age, and race. Rates are also shown for three specific kinds of cancer: breast cancer, colorectal cancer, and lung cancer.” (From the CORGIS site “Overview” section). Utilizing this dataset, we seek to reveal any potential areas of inequities that are based on age, gender, race, or geographical location. We seek to achieve this goal using visualizations that examine statistics on a multidimensional basis to analyze any discrepancies between the death rates of different types of cancer included in this project (total, breast cancer, colorectal cancer, and lung cancer) to not only race, gender, location, and age in these categories alone but to take all of these variables into consideration simultaneously when doing analysis and visualizations. This is to ultimately reveal potential disparities between different identity groups and reveal potential existence of compounding inequalities in relations to our discussions regarding critical race theory and critical disability studies as we are “studying and transforming the relationship among race, racism, and power” (Delgado, Stefancic “Critical Race Theory: An Introduction”) in regards to public health and our healthcare system to “understand how race and (dis)ability continue to be mutually constitutive in our contemporary moment” (Schalk, “Critical Disability Studies as Methodology”), and intersectionality and how membership in multiple groups that are not part of a given “historically dominant group” within different identities can produce further discrimination or barriers for such individuals.
In applying the ideas of data feminism and data Marxism in conjunction, we sought to question our data, specifically the format of presentation by the CORGIS Dataset Project as well as accounting for unusual values in the data in a responsible manner as these “unusual” values such as missing values could be an issue of inequality that visualizations cannot show. We then use this data to create visualizations to potentially challenge the status quo of hierarchies in certain identities as we discover the existence of disparities and seek solutions to break the status quo to resolve these inequalities. We appreciated the fact that the dataset presents the data in a manner that is consistent with the ideas of intersectionality, critical race theory, and critical disability theory and supports our cause in discovering potential compounding inequalities that exist in public health and healthcare. This data was collected from 2007-2013 and presents death rates at a rate per 100,000 population in multiple groupings that is helpful for visualization.
Each row of the data represents a state, covering all fifty states plus the District of Columbia. Using the Total.Rate column which represents the total cancer deaths per 100,000 population, we create the following bar chart.
We note that Utah has a significantly lower cancer death rate than the other states with 98.5 deaths per 100,000 compared to the nationwide average of 190.7 deaths per 100,000. We questioned whether this geographical disparity was related to potential socioeconomic inequalities. However, upon further research, Utah has actually been known to have not only cancer deaths, but cases of cancer diagnoses in general. According to the scientific article “Low cancer incidence and mortality in Utah” published by J. L. Lyon et. al., “Utah had 18% fewer cases of cancer than expected based on the Third National Cancer Survey, and 24% fewer cancer deaths than expected based on national mortality data” from cancer mortality data for the years 1950-1969 and morbidity data for the years 1966-1970, which come from decades prior to the data from our data set. The paper accounts for Utah’s unusually low cancer death rate relative to the rest of the U.S. with the fact that “cancer sites associated with cigarette smoking and alcohol use accounted for nearly half of these differences” and another paper “Cancer in Utah: risk by religion and place of residence” by J. L. Lyon et. al. explains these differences with “personal habits such as smoking and drinking and reproductive factors” associated with Mormonism in Utah as “non-Mormon urban men had a 34% higher risk of cancer compared with their rural counterparts.” With this knowledge, we come to the conclusion for this visualization that demographic differences account for this disparity where Utah’s Mormon population are likely to have personal habits that accounts for lower risks of certain cancers, rather than direct socioeconomic inequalities.
The data also provides total cancer, breast cancer, colorectal cancer, and lung by race in each state. We decided to use the mean across all states to get an average across the United States. We came to realize that some states have 0 values for certain populations, but we decided not to use median because we cannot make assumptions for why those values are zeros and continue to use them as there are multiple possible explanations for them such as ones that suggest a deeper inequality in undercounting racial minority groups or the fact that some states have a very small racial minority population to be accounted for (see data critique section for more details). We seek to use visualizations to discover whether there are disparities across different races and to research explanations for them. Below are the visualizations representing the average cancer deaths per 100,000 population across the U.S. by race.
“Dr. Lawrence noted that the disparity in deaths likely reflects systemic and preventable barriers to getting quality care. Whether it’s screening for cancer, timely diagnosis, or the receipt of proven treatments, he explained, “Black individuals continue to have a delay in care or receive poorer care than their White counterparts.” (https://www.cancer.gov/news-events/press-releases/2022/cancer-death-rates-black-people)
Black people experience higher mortality rates among race groups, which can be due to many factors. An intersectional framework can seek to identify, and account for, the issues and disparities that contribute to this inequity.
“Black people are also more likely to live in neighborhoods with poor access to specialists, to see physicians with limited clinical resources, and to live in communities with greater exposure to environmental carcinogens such as air pollution, the researchers said. According to the Kaiser Family Foundation, “the higher mortality rate among Black people partly reflects a later stage of disease at diagnosis among Black patients” which is attributed to limited clinical resources. Another factor is that there is “underrepresentation of Black and Hispanic adults and other people of color in oncology clinical trials” which “may contribute to cancer treatment and mortality disparities.” From the article, “research has identified multiple barriers to participation in clinical trials for people of color, including lack of understanding and information about trials, fear and stigma of participating, and time and resource constraints associated with trial participation (including financial burden, time commitment, transportation, and compensation).” This disparity clearly reflects the state of our society in being unequal as we clearly see that these disparities can be identified as being related to socioeconomic disparities in relations to data marxism of oppressive socioeconomic groups in society and its relation to race in explaining discrepancies between the environments and access to quality care. In addition, scientists and clinicians must seek to play a role in closing this disparity in applying ethics and inclusion not only in their studies to play a role in providing quality care that is supported from diverse clinical data sources, but to also include diverse voices in clinical teams to create a healthcare system with extensive perspectives so we would not have disparities in the ability to treat diseases in different demographics.
One limitation of the study is the broad groupings used for race and ethnicity, which could make it harder to tease out differences among people who are racially categorized as Black. Another limitation is the potential misclassification of race and ethnicity and of underlying cause of death recorded on death certificates.” (https://www.cancer.gov/news-events/press-releases/2022/cancer-death-rates-black-people)
The dataset provides total cancer and breast cancer death rates by state for different age groups. We seek to use visualizations to discover whether there are disparities across different ages and to research explanations for them to discover which ages are the most vulnerable and ultimately apply these groups in intersectionality later on in discussing compounding inequalities. We apply averages to each state to get a national average which are summarized into the bar charts below.
“In the United States of America more than 60% of the cancer cases are seen in old individuals aged 65, and over [8].
Prolonged exposure to carcinogenic agents, DNA damage accumulation, tumor suppressor gene defects, impairment of cellular repair mechanisms, oncogenic activation, and attenuation of immunity have been held responsible for higher incidence of cancer in older individuals [3]. Since carcinogenesis is a very long process, emergence of cancer in advanced ages is a natural event.” (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5175057/) (Cancer in the elderly)
Intuitively, prolonged contact with carcinogens will result in increased cancer diagnoses, and as a result, mortalities. Additionally, as individuals age, their immune systems become significantly weaker, dramatically affecting their ability to fight diseases. All of these factors compound to create the result we see in elders accounting for the vast majority of cancer mortalities in the U.S.
Finally, the data columns that deal with sex provide an excellent opportunity for intersectional analysis as there is data for total cancer death rates across age and sex, colorectal and lung cancer death rates across age and sex groups, and cancer death rates across race and sex groups. We apply averages to each state to get national averages which are summarized into the bar charts below which we can more clearly see whether there are disparities across different identities and groups and to research explanations for them.
The above findings are consistent with earlier conclusions regarding how elders are more at risk for cancer deaths than other age groups. However, we see from these visualizations that males are more likely to develop colorectal cancer and lung cancer, as well as having a higher death rate for all cancers. According to the scientific article “Sex Differences in Cancer: Epidemiology, Genetics and Therapy” by Hae-In Kim et. al., “men are more prone to die from cancer, particularly hematological malignancies” and “sex difference in cancer incidence is attributed to regulation at the genetic/molecular level and sex hormones such as estrogen.” This suggests that these disparities are due to biological reasons rather than other factors, at least not directly, such as socioeconomic factors. However, we note that Black and White Americans are much more likely to die from cancer than other groups and although Black females (Hispanic and Non-Hispanic) have comparable statistics to White females (Hispanic and Nonhispanic), the death rate for Black (Non-Hispanic and Hispanic) men are much noticeably higher than the death rate for White (Non-Hispanic and Hispanic) men, which is consistent with our earlier discussions regarding the disparities of race in cancer deaths. This visualization demonstrates the importance of intersectionality as we discover that although men are more likely to develop cancer than women, this would not explain why for example, Asian males are less likely to die from cancer than Black females. We must look at such data from a multidimensional perspective in analyzing multiple aspects of identity, including race, age, and sex all at once to be able to tell the full story and reveal such inequalities.
Using our visualizations, we have revealed many disparities across different groups of identities. Some of them are not explained directly by factors of inequalities discussed in class such as socioeconomic disparities at least not directly which includes our results for disparities in states or sex, while other disparities such as those in race to reveal injustices of our healthcare system and how socioeconomic factors directly affect one’s livelihood. The ideas of intersectionality come into play when we analyze data regarding gender, which allows us to look at gender and sex simultaneously as well as race and sex simultaneously in seeing the full picture of compounding inequalities that stacks with age, race, and sex in tying ideas of critical race theory and critical disability studies together. We acknowledge various limitations in our data. This includes systemic limitations such as the fact that there are broad groupings used for race and ethnicity which limits our scope of analysis but is how data is usually collected regarding race in the U.S., as well as potential misclassification of race and ethnicity and of underlying cause of death recorded on death certificates. There are also other limits that are not systemic but rather comes from what data was being collected. An important limitation we would like to note is that although we try to analyze the relationship between socioeconomic status with identities and cancer death rates, the data contains no data regarding income level which could be a critical measure for socioeconomic status. Instead, we sought out outside research to make such connections. We hope that the findings in this data would encourage people to challenge this status quo as although there are some disparities that are due to biological factors that likely cannot be changed, other disparities, especially those regarding race, require change to tackle. This includes increasing access to quality healthcare and screening resources for diverse populations, as well as having a responsible and diverse scientific community that makes decisions that benefit all.
The increased efforts over time to combat cancer as a national public health concern have led to notable reductions in cancer mortality across the country. The timeline includes initiatives focused on raising awareness about cancer prevention, reducing mortality rates, and improving cancer care. The addition of screening programs emphasized the importance of regular screenings for various cancers in order to detect them at an early stage when they are more treatable. Additionally, the launch of initiatives like the Cancer Genome Anatomy Project advanced cancer research to contribute to the understanding of cancer and facilitated the development of targeting therapies and treatment approaches by developing better models for studying the disease. Lastly, healthcare policy developments, such as the ACA, expanded access to health insurance coverage and preventative services including cancer screenings. By making cancer treatment and screening resources more widely available despite socio-economic conditions, the ACA, and similar policies, work to reduce the disparities that exist in cancer mortality.