Evaluating unjust contexts and children's health: An introduction to multilevel modeling in public health research

Introduction: Multilevel models (MLMs) allow researchers the ability to examine individual and contextual predictors of health and simultaneously investigate what predicts variance within and across contexts. For example, a family's income and factors associated with a family's neighborhood may influence whether a child has a medical home. However, the potential of MLMs has gone under used in health research, partly because few applied examples demonstrate them. To address this, I take a non-mathematical approach and introduce the key interpretative concepts in MLMs using a social justice issue related to children's health.

Methods: I used MLMs and data from the 2005-2006 National Survey of Children with Special Health Care Needs (NS-CSHCN: n = 40,723) to investigate whether individual-level sociodemographic variables (race/ethnicity, family income) and contextual level variables (proportion of families in poverty and income inequality within a state) predict whether CSHCN have a medical home.

Results: Poor minority children living in states with more income inequality and more families in poverty were least likely to have a medical home. However, the strength of this relationship differed across contexts and substantial cross-context variance remained unexplained.

Conclusions: Results show how simultaneously investigating individual and contextual level predictors of health and partitioning the variance into individual and contextual level components leads to more fully informed pictures of health. They demonstrate that the strength of the relationship between a social injustice (eg., economic depravity) and outcomes varies across contexts. In my discussion, I provide practical advice describing how researchers should fit and interpret MLMs.