Mediation and Suppression of Racial Disparities in Reporting Need for Family Support Services among Families Having Children with Special Health Care Needs

Background and Purpose An extensive body of research has established pervasive and far-reaching racial disparities in health care. However, little is know about the extent to which Medicaid reduces or exacerbates such disparities. Further, little is known about the extent to which Black and Latino families raising children with special health care needs (CSHCN) have their support service needs met. Past research has established that family support services like family and child mental health services and professional care coordination are critically important to enable families to care for their children with disabilities at home. We hypothesized that racial disparities in receiving needed services was mediated or worsened by a lack of access to medical services, or suppressed by access to services and consultation available to Medicaid participants.

Methods We analyzed the 2001 National Survey of CSHCN (n= 26,119 CSHCN), which collected information on the health care and demographic characteristics of nearly 40,000 CSHCN and their families. We used a mediation analysis framework to decompose the total racial disparity in reported need for family support services into a component directly related to race/ethnicity, and an indirect component. The indirect component was hypothesized to be either a mediation (worsening) of the direct racial disparity due to not having medical insurance, or a suppression (improvement and possible reversal) of the racial disparity due to having Medicaid. Within the mediation analysis framework and given an observed racial disparity in the direct effect, a positive indirect effect would indicate that suppression was present, and a negative indirect effect that mediation was present. We used the Sobel test to determine the significance of the indirect effect. All analyses were conducted using logistic regression modeling, with multiple imputation used to address missing data.

Results A significant direct racial disparity was present for Black and Latino families with CSHCN in reporting the need for family support services – these families were much less likely to report needing care than white families (up to 35% for Blacks and 21% for Latinos). Medicaid acted as a significant suppressor of this disparity for Black families, reducing the disparity by one-third, but not for Latino families. Being uninsured did not significantly mediate (worsen) the disparity for Blacks or Latinos. In contravention of the hypothesis, uninsured Latino families were more likely to report a need for professional care coordination, which operated as a significant suppressor instead of a mediator, reducing the disparity by nearly one-fifth.

Conclusions and Implications These findings suggest that for Black families raising CSHCN, Medicaid reduces racial disparities in receipt of needed family support services. For Latino families, however, Medicaid did not serve the same function, indicating that Latino families likely face greater burdens in having their support needs met, burdens which may include language barriers and a lack of culturally competent care providers. Policy makers and advocates interested in eliminating racial disparities might look toward further leveraging Medicaid to do so, perhaps using quality of care targets tied to reimbursement to improve outcomes for racial minorities raising CSHCN.