Evaluating disparities in child occupant protection using a proportion-eliminated approach to mediation

While there are clear racial/ethnic disparities in child restraint system (CRS) use, to date no studies have identified mediators that quantitatively explain the relationship between race and CRS use. Therefore, the objective of this study was to provide an example of how a proportion-eliminated approach to mediation may be particularly useful in understanding the complex relationship between race and CRS use.

Sixty-two mothers with a child between 4-8 years old completed a survey and had their CRS use assessed by a Child Passenger Safety Technician using a structured assessment based on the 2018 American Academy of Pediatrics' Best Practice guidelines. Recruitment and data collection occurred in Birmingham, Alabama between June 2018 and January 2019. We used chi-squared tests, logistic regressions, and a proportion-eliminated approach to mediation to compare our variables of interest and to estimate the amount of the association between racial group membership and errors in restraint use that may be explained by sociodemographic, psychosocial, and parenting variables.

Before mediation, Nonwhite mothers in this sample had a 7.38 greater odds of having an error in CRS use than White mothers. Mediation analyses indicated that being married and self-reported seatbelt use explained 47% and 35% of the effect of race on CRS use errors, respectively.

A proportion-eliminated approach to mediation may be particularly useful in child passenger safety research aiming to inform the development of interventions tailored for racial minority populations.