Rates of Infants Reported to CPS at Birth for Prenatal Substance Exposure: A Panel Data Analysis

Background and Purpose: The opioid epidemic has revitalized attention to infants with prenatal substance exposure (PSE). Unfortunately, documented rates of infants reported to child protective services (CPS) at birth due to PSE are limited. This study sought to 1) estimate rates per 1,000 births of infants reported to CPS due to PSE for every U.S. state, 2) calculate rates by race/ethnicity by state, and 3) explore the relationship between a CPS policy requiring PSE to be reported with reporting rates across time and states.

Methods: Data came from the National Child Abuse and Neglect Data System (NCANDS) for all 50 states and Washington DC. A PSE CPS report was defined as a report made within 14 days of birth with a parent or child substance concern indicated. Rates were calculated by state, year, and race/ethnicity using the number of births from Kids Count. A binary, time-varying variable captured whether the state required PSE to reported to CPS in that year. The panel data was augmented with state-level covariates on the state’s political, welfare generosity, and poverty characteristics from the Kentucky Poverty Center. We ran ordinary least square (OLS), fixed effects, and random effects models with and without the covariates. We used an F-test to identify if a fixed effects model was a better fit than the OLS model and a Haussman test to identify if the random effects was a better fit than the fixed effects model.

Results: Wide variation of PSE reports by states was identified with overall increases over time. In 2011, Maine (14.3), Oklahoma (12.8), and New Hampshire (11.4) had the highest reporting rates per 1,000 births. In 2019, nine states had rates above 20 per 1,000 births (Oklahoma, New Hampshire, Louisiana, New Mexico, Arizona, Maine, Georgia, Minnesota, and Alaska) and 18 states had above 10 per 1,000 births. For most states, Non-Hispanic White and Hispanic infants had lower rates of PSE reports than Non-Hispanic Black infants. The full model with the time fixed effects was found to be the best fit for the data. The presence of a CPS policy requiring the reporting of PSE was not found to be a statistically significant predictor of the number of children reported for PSE. However, increases in a state’s unemployment rate and rate of Medicaid recipients in the population were positively associated with its PSE reporting rate.

Conclusions and Implications: This study presents new estimates of the rates of which children with concerns of PSE are reported to CPS, including differences in the PSE reporting rates by race/ethnicity. We identified increased rates of PSE reporting across all years. It is unclear if the increases over time are reflections in actual PSE reporting rates or increased specificity in documentation in the data. The finding of these policies not having an effect within a state may be an indication that these policy changes may not impact the reporting practices immediately within the state and that their impact may take more time to influence reporting practices.