Factors Associated with Child Welfare Diversion for Substance-Exposed Infants

Background/Purpose: The Child Abuse Prevention and Treatment Act’s (CAPTA) provisions concerning hospital and child protective services (CPS) response to substance-exposed infants (SEI) were revised in 2016 to address the impact of the opioid epidemic. Responsive to this legislation, in 2019, Connecticut unveiled a statewide policy designed to divert SEI without safety concerns from CPS. Connecticut is the first state to implement an online system that captures public health data on the occurrence of the birth (a “CAPTA notification”) and guides the person making the notification through a series of questions that determine whether the case warrants a separate maltreatment report. Connecticut’s policy also mandates that plans of safe care (POSC), intended to connect dyads to needed health and family services, be developed for all babies at or by the time of notification. Infants without POSC are ineligible for diversion. Thus, this system results in three distinct outcomes: diversion with a POSC, maltreatment report with POSC, or maltreatment report without POSC.

Method: We analyzed all CAPTA notifications made between March 15, 2019, and July 21, 2021. We used notification and birth data to determine rates per 1000 births. We examined descriptive characteristics of all notifications, plans of safe care contents, and the three outcomes. We utilized multinomial logistic regression analyses with relative risk ratios to understand factors associated with the three outcomes among the full sample (n=4763) and multivariate log- binomial regressions to further explore differences between cases with POSC who were diverted or not (n=3186) and reported cases with and without POSC (n=2310).

Results: During the first 28 months of policy implementation, over 5.5% of total births received CAPTA notifications. Of these, more than half were diverted (2.9% of total births). Less than 1% of total births were reported with a POSC and 1.8% were reported without a POSC. At the bivariate level, significantly more non-Hispanic white dyads reported with POSC, more non-Hispanic Black dyads reported without POSC, and more Hispanic dyads diverted. Diverted cases’ POSC included significantly fewer substance use services and significantly more mental health and parenting services compared to reported cases’ POSC. Our multinomial logistic regression revealed that substance exposure type was significantly associated with outcomes controlling for maternal age and race/ethnicity. Relative risk for diversion was highest for prescription drug exposures, including medications for opioid use disorder (MOUD) and lowest for illegal and polysubstance exposures. In the first binomial regression among cases with POSC, illegal and polysubstance exposures were nearly 2.5 times more likely to be reported—an effect larger than the impact of specific service needs identified in the POSC. In the second model among reported cases, MOUD exposures were 1.3 times more likely to be reported with POSC compared to marijuana exposures.

Implications: Although CAPTA’s intention is to foster a universal public health approach, our study reveals that both the POSC and CPS diversion depend on type of substance exposure. Achieving CAPTA’s goals will require additional policy implementation support as well as future research into the effectiveness of POSC to prevent maltreatment.