Method: This study analyzed unlinked child welfare and vital records during a 66-month timeframe starting in March 2017 to generate our two dependent variables: (1) county-level monthly rates of screened-in CPS reports involving IPSE per 1,000 births and (2) county-level monthly proportion of screened-in IPSE reports that resulted in foster placement. We fitted a Bayesian Poisson mixed effects model with a log-link using the number of births as an offset term to estimate the rates of IPSE reports per birth at each timepoint for each of Connecticut’s seven counties (n = 517), controlling for other risk factors from the birth records (e.g., WIC participation) and COVID-19 hospitalizations. An autoregressive random effect of order one was included to account for temporal dependence. The interaction between a linear trend and policy implementation was included to test the effect of the policy on report rate while controlling for covariates. A model with the same structure was used to analyze the rate of foster care placement per IPSE report.
Results: The statewide average rate of reports involving IPSE fell from 21.3 to 18.7 per 1000 births after the policy change. Similarly, the statewide average proportion of IPSE reports that resulted in foster care placement fell from 54.5% to 43.1% after the policy change. The mixed effects model revealed that county-level IPSE report rates per birth decreased 12.7% [95% CI: 10.5-15.0%] after the policy change. In addition, county-level foster care placement rate per IPSE report decreased 21.9% [95% CI: 19.4-24.1%]. Depending on the outcome, significant covariates included rates of neonatal abstinence syndrome, maternal marital status, and familial risk level. Unit-level increases in neonatal abstinence syndrome diagnoses at birth increased both rate of reports [95% CI: 33-562%] and foster placements [95% CI: 17-976%]. Conversely, the proportion of married mothers decreased report rates [95% CI: 9-31%] and the proportion of low- and moderate-risk CPS referrals decreased foster care placement rates [95% CI: 23-72% and 8-65%, respectively].
Implications: Our findings reveal that Connecticut’s new reporting policy was associated with a reduction in both child maltreatment reports and foster care placements of IPSE. In a policy landscape that is historically associated with increasingly punitive approaches, Connecticut’s approach appears to achieve its public health goals.