NEIGHBORHOOD VARIATION IN THE RATE OF CHILD WELFARE CONTACT
Lloyd, Stacey Williams
MetadataShow full item record
In the United States, the child welfare system serves a vulnerable population of children with extensive health needs. With momentum building for place-based interventions to promote community health, population-level evidence is needed to identify critical elements of interventions and inform potential collaboration across service sectors. Through a systematic review of small-area ecological research on neighborhood effects (Aim 1), we framed the literature on neighborhood context and child welfare contact through a population health lens. Four constructs describing the neighborhood structure (economic disadvantage, percent of the population from racial/ethnic minority group, social disadvantage, and residential instability) and two constructs describing neighborhood processes (alcohol access, drug arrests) were positively associated with the rate of child welfare contact in multiple studies. Evidence on neighborhood processes was identified as a priority for future research and guidance for improving study design was provided. Using existing observational data from the Neighborhood Inventory for Environmental Typology (NIfETy), we developed area-level indicators for six specific constructs within the context of neighborhood processes (Aim 2). Three neighborhood process indicators were accurate for identifying areas with high levels of risk (criterion validity) and associated with all area-level measures of the neighborhood structure and youth population health outcomes included in the assessment (construct validity): physical disorder index, drug and alcohol index and violence index. We examined the relationship between neighborhood disadvantage, violence, drug and alcohol activity and the rate of child welfare investigation (Aim 3). While both the violence index and drug and alcohol index were strongly associated with the outcome in bivariate analysis, only violence was associated with a significant increase in the rate of child welfare investigation in the multivariable regression analysis. Applying concepts from spatial epidemiology, several important methodological improvements were illustrated, including person years of observation, age-adjusted rates, and the use of negative binomial regression models. Focusing child maltreatment prevention interventions in areas with the greatest density of child welfare contact is an avenue by which interventions can reduce both the incidence of child maltreatment and the rate of child welfare involvement. Considering the high rate of child welfare contract in Baltimore City, the need to reduce the burden on the child welfare system, and growing attention for the need to prevent child maltreatment in high risk neighborhoods, child welfare services may benefit from further coordinating their prevention efforts with other public sectors serving children and youth at risk of maltreatment. Collaborative efforts between hospitals, public service sectors, and community-based resources are likely to be both effective and efficient methods for targeting resources to the most vulnerable children and families in the city. While the current research sheds light on the relationship between violence, substance use activity, and the rate of child welfare contact, further evidence on neighborhood processes is needed. Small-area ecological research on other neighborhood processes, such as social cohesion and collective efficacy, is imperative to informing place-based efforts in child welfare.