Association of state child welfare policies pertaining to substance use in pregnancy and infant mortality Authors: Hilal J, Bogen DL, Bodnar LM, Krans EE, Stein BD, Jarlenski M ABSTRACT Importance: U.S. states have adopted varying policies that shape how child welfare agencies address substance use in pregnancy. The association between such policies and infant mortality rates is unknown. Objective: To determine the association between state child welfare policies pertaining to maternal substance use on rates of infant mortality due to external causes, SIDS, and all causes. Design, Setting, and Participants: Time-series analysis of state child welfare policies pertaining to maternal substance use and state infant mortality rates in all U.S. states from 1999-2017. Conditional state fixed-effects negative binomial regression models were fit to determine the adjusted association between such policies and infant mortality rates. Exposures: State adoption of any of 3 types of child welfare laws: requiring healthcare provides to report maternal substance use to child welfare agencies; requiring a plan of safe care for children affected by in utero substance exposure; or defining maternal substance use as child neglect/abuse. Main outcomes and measures: State infant mortality rates due to external causes, SIDS, and all cause mortality. Results: Among all 50 states observed for 19 years, 48 had a mandatory reporting policy in place at any time point, 8 had a plan of safe care policy in place at any time point, and 17 had a neglect/abuse policy in place at any time point. There was an increased in external cause related mortality, and a decrease in SIDS related mortality, from 1999-2000. State mandatory reporting policies were not associated with changes in external causes mortality rates (IRR: 1.09, 99% CI: 0.98,1.21) or SIDS mortality rates (IRR: 1.07; 99% CI: 0.94,1.22). Plans of safe care policies were not associated with a change in external causes mortality rates (IRR: 1.10; 99% CI: 0.99,1.35) but were associated with an increased rate of SIDS mortality (IRR: 1.25; 99% CI: 1.04,1.50). Neglect/abuse policies were not associated with changes in external causes (IRR: 1.03; 99% CI: 0.93,1.16) or SIDS (IRR: 0.97; 99% CI: 0.83,1.13) mortality rates. Findings were similar for all-cause mortality. Conclusions and Relevance: States’ adoption of any of 3 policies pertaining to how child welfare agencies address substance use in pregnancy was not associated with significant changes in infant mortality rates due to external causes or all cause mortality. State adoption of plans of safe care policy was associated with an increase in the rate of SIDS mortality.

Figure 1

Mean infant mortality rates before and after adoption of any of 3 child welfare policies pertaining to substance use in pregnancy

Lines depict mortality rates (per 1000 births for all causes, and 100,000 births for external causes and SIDS), and are associated with the axis on the left. Bars depict N number of states and are associated with the axis on the right. The x-axis represents years from policy enactment.

Figure 1

Mean infant mortality rates before and after adoption of any of 3 child welfare policies pertaining to substance use in pregnancy

Lines depict mortality rates (per 1000 births for all causes, and 100,000 births for external causes and SIDS), and are associated with the axis on the left. Bars depict N number of states and are associated with the axis on the right. The x-axis represents years from policy enactment.

Close modal
Table 1

Association between child welfare policies pertaining to maternal substance use and infant mortality rates, 1999-2017

From a multivariable negative binomial regression model with state fixed effects; adjusted for the following annual state-level covariates: drug overdose rate; percent female; percent <18 years; percent non-white race; state government spending on incarceration, education, transportation, Medicaid, and public assistance; and state government elected official political ideology. Models for SIDS and External Cause mortality rates included all states with available data (states with <10 deaths from these causes in a given year are excluded for that year). Confidence intervals are adjusted according to the Bonferroni correction to account for 9 statistical tests (alpha=0.005).

Table 1

Association between child welfare policies pertaining to maternal substance use and infant mortality rates, 1999-2017

From a multivariable negative binomial regression model with state fixed effects; adjusted for the following annual state-level covariates: drug overdose rate; percent female; percent <18 years; percent non-white race; state government spending on incarceration, education, transportation, Medicaid, and public assistance; and state government elected official political ideology. Models for SIDS and External Cause mortality rates included all states with available data (states with <10 deaths from these causes in a given year are excluded for that year). Confidence intervals are adjusted according to the Bonferroni correction to account for 9 statistical tests (alpha=0.005).

Close modal