Hi, I'm Dr. Tanjala Purnell. I'm one of the co-directors of the course. I'm also an assistant professor in the School of Medicine, with a joint faculty appointment in the School of Public Health. I'm Dr. Deidra Crews. I'm an associate professor in the School of Medicine, with an appointment in the School of Public Health. Today we're going to talk about using social epidemiology research methods to inform the development of health equity interventions. So let's get started. First, we're going to cover some background about social epidemiology. So as you know, health and health care disparities persist worldwide. As shown here on the table, historically disadvantaged populations include: racial and ethnic minority groups, rural and urban residents, immigrants and adults with low income or low literacy and numeracy. We often include race and ethnicity in studies to assess health and health care disparities. Therefore, it is important to ensure that we are all on the same page about the definitions of these terms. So first, I'm sure you're all aware that race and ethnicity are social and political constructs. As shown here on the slide, the US Census Bureau collects racial and ethnic data in accordance with guidelines provided by the US Office of Management and Budget. These data are based on self-identification. So the racial categories included in the census generally reflect a social definition of race recognized in this country, and do not attempt to define race biologically or genetically. In addition to this, it is recognized that these categories of race include racial and national origin or sociocultural group. Many people may choose to report more than one race. In addition, the primary categories of race often include five minimum categories. These include White, Black or African-American, American Indian or Alaska Native, Asian, Native Hawaiian, or other Pacific Islander. These categories are described more in detail on the table. So another question we often ask in these studies to assess disparities, is whether or not race or ethnicity actually cause poor health and health care outcomes. However, as I'm sure we've all learned in our epidemiology courses, causation is often very difficult to test and prove. As shown on the slide, the Bradford Hill criteria, otherwise known as Hill's criteria for causation, are a group of nine principles including temporal relationship, replication of the findings, and biologic plausibility. These principles were established in 1965 and are often used in establishing evidence of a causal relationship between a presumed cause and the effect, and have been widely used in public health research studies. So because causation is often difficult to prove, there are other methods that we use in health equity research. The first is causal inference, which focuses on testing the negation of the causal hypothesis. Or in other words, the null hypothesis that the exposure does not have a causal relationship. The second method is social epidemiology, which focuses on the effects of social and structural factors on states of health. Social epidemiology also assumes that the distribution of advantages and disadvantages in a society reflects the distribution of health and disease in that society. This figure highlights key proximate factors, intermediate factors, and distal factors that all influence disparate health outcomes. So as shown on the figure, examples of proximate factors include biologic genetic pathways and individual risk behaviors or social factors. Examples of intermediate factors include physical, social, and health care context. While distal factors may include social conditions and policies such as poverty, culture, or discrimination. These factors all work together in helping to influence the disparities that we often observe. There are various ways to measure proximate factors, intermediate factors, and distal factors. One key factor that we often include in health equity studies is socioeconomic status. So socioeconomic status or SES can be assessed through individual measures as shown on the slide. These include education, income, or wealth. Or we can measure SES through neighborhood measures such as zip code or a census tract poverty levels. Other examples are shown on the slide. There are also several datasets with social measures that are available to use in health equity research. These include national surveys such as NHANES, or the National Health and Nutrition Examination Surveys, community-based cohorts such as the Jackson Heart Study, and data from clinical trials conducted here at Johns Hopkins, such as Project ReD CHiP or the project reducing disparities and controlling hypertension in primary care study. There are various analytical approaches that we use in social epidemiology research. Often, in regression models, we will test or adjust for potential confounders or mediators, They may influence primary results. Confounders may be associated with the exposure and the outcome, but is not caused by the exposure. For example, you may hear terms such as "controlled for" in regression analysis to minimize bias. Examples of potential confounders include coffee drinking, smoking, and pancreatic cancer, and looking at the associations of coffee drinking with pancreatic cancer while controlling for differences in smoking. On the other hand, mediators are mechanisms on the causal pathway by which the exposure might cause or influence the outcome. One example of this may be looking at the association of race with outcomes in diabetes mediated through differences in access to health care. In this example, the mediator would be health care access. So one example may be a study in which you want to examine racial differences in diabetes outcomes. So we may find that these differences are actually influenced by differences in access to health care. It may not be that race is causing differences in diabetes outcomes, but the differences in access to primary care or a high-quality health care may be what's influencing the racial differences that we find in diabetes outcomes. In this example, the mediator would be the differences in health care access.
