Opioid Exposed Infants and Child Welfare System Involvement: A Multilevel Analysis

Background and Purpose:

Opioid use and misuse has increased considerably in the past decade (Saha et al., 2016). Further, research has indicated that county-level factors play a role in this prescription drug use epidemic (Wright, et al., 2014). For infants who are exposed to opioids in utero, there is potential for serious harms; the child protective system (CPS) is a common intervention for these families. What is not clear is if the county-level factors affecting substance abuse rates also contribute to the birth rates of opiate-exposed infants and the CPS responses to these families. Thus, the research questions for this study are 1) what are the individual and county-level predictors of opioid-exposure births and 2) what are the individual and county-level predictors of CPS removals among opioid-exposure births?

Methods:

All birth records for children born in Washington State from 2006-2013 were included in the study (n=706,882). Birth records were linked with CPS records. Birth records were examined for opioid exposure in utero indicated through ICD-9 codes. The study employed a multi-level analysis to simultaneously examine the relationships of individual level factors (e.g. race, prenatal care start, birth payment) and county-level factors (e.g. child poverty rates, race composition, unemployment rate, substance abuse prevention program presence) on opioid-exposure births. Then using the same covariates, analysis was conducted to examine out-of-home placement among just opioid-exposed infants.

Results:

Results indicate that infants whose births were paid through public insurance were 5.4 times more likely to born exposed to opioids (OR: 5.40, 95% CI:4.86, 5.99). Native American (compared to white) infants (OR: 2.70, CI: 2.37, 3.07) and infants whose mothers did not receive prenatal care (OR: 2.69, CI: 2.40, 3.02) were also more likely to be exposed to opioids. The county-level predictor of having a substance abuse prevention program also increased the odds of an infant being exposed to opioids (OR: 4.45, CI: 1.11, 1.90) while the other county-level predictors were not statistically significant.

For opioid-exposed infants, similar results are found where public birth pay (OR: 1.49, CI: 1.09, 2.04), Native American maternal race (OR: 1.61, CI: 1.18, 2.18), and no prenatal care (OR: 1.44, CI: 1.06, 1.95) all increased the odds that an infant would be placed in out-of-home care within 90 days. Counties with higher rates of child poverty were also more likely to remove an opioid-exposed infant (OR: 1.05, CI: 1.03, 1.07).

Conclusions and Implications:

This study expands the current knowledge base by identifying that both individual and county-level predictors matter for infant opioid-exposure and subsequent removal from their parents. We discuss the policy and practice implications as this high-risk population interacts with CPS.