EPI 222 – Homework Week #8
- It is often mentioned that racial/ethnic differences in health should not be investigated without consideration of socioeconomic position/status. Describe two ways to account for SES in an analytic model and the interpretations/implications of each approach.
As discussed in Holmes et al., a multilevel approach to studying health disparities is needed, such that we cannot just conclude from census-collected data that racial, ethnic, and socioeconomic variations are solely responsible for the found health associations. Furthermore, it is not enough to suggest that these disparities are potentially driven by social factors in the discussion without actually investigating their context. Multilevel research can examine how individual behaviors that influence disease risk are influenced by larger societal factors such as access to health care, social networks, poverty, and neighborhood resources.
However, it is important to not succumb to ecological fallacy, which arises when inferences about low-level factors (i.e., SES), are made from high-level factors (i.e., zipcode).
In order to avoid this, both individual and contextual measurements should be made. For instance, socioeconomic status is conventionally collected as income and education level, which may be reported on the individual level. These can be analyzed as continuous or categorical variables that effectively create a gradient to which you can juxtapose groups based on the defined categories of socioeconomic status. However, the environment that one is in may also impact one’s socioeconomic status, including the quality of the school district to which the individual attends, the societal class and status of the job that the individual holds, and the characteristics of the neighborhood in which that individual lives. These variables may not be as easily collected and quantified, and may require collateral data gathering, individual-level impressions of the quality of these items, and moderately-subjective categorizing to fit an analytical model, which will need to be justified in the discussion.
2. Select a research question investigating associations between multi-level social factors (operating at least two levels) and a health outcome. State the exposures and outcomes, the additional study covariates that would be included in the analytic model, and a discussion of the analytic considerations in a multi-level investigation.
One of my research questions of interest addresses how racial/ethnic variations mediate the effect that a provider discussing birth control plans has on a young woman developing an infection prevention plan. Exposures would include having had a prior discussion with their provider about birth control methods, whether or not they developed a plan, and how satisfied they were with that plan and if they felt they could follow that plan. Outcomes include developing an infection prevention plan, their satisfaction and feeling that they could follow through with that plan. Mediators of interest include how the outcome variables differ by race and ethnicity, perhaps alluding to a racial bias in provider counseling or implication that providers are including infection prevention plan development in their quality family planning care. Covariates that will need to be controlled for include age, education, household income, birth control methods used, obstetrical history, and sexual risk factors like number of sexual partners.