Maternal mortality, defined as the death of a woman while pregnant or within 42 days of termination of pregnancy, has more than tripled in the past three decades in the United States despite declining rates worldwide.Reference Hoyert1 Maternal mortality involves a complex interplay of many contributing factors. Substance use during pregnancy is a major risk factor for maternal mortality.Reference Schiff2 In fact, fatal maternal overdose rates more than doubled in the US from 2007 to 2017.Reference Cleveland3
Though researchers have suggested that substance use is a surrogate for other factors that influence mortality (e.g. reduced healthcare utilization, poor nutrition, unstable housing, and exposure to violence), states continue to develop policies to combat substance use during pregnancy.4 The number of states that have enacted policies considering prenatal substance use as a form of child abuse rose from 12 states in 2000 to 24 states in 2020.Reference Faherty5 It has recently been posited that punitive laws regarding substance use during pregnancy might contribute to increased rates of maternal mortality.6 Fear of seeking care due to social stigma and legal ramifications are common barriers to prenatal care for pregnant women with substance use disorders.Reference Gopman7 Laws punishing prenatal substance use are often disproportionately applied, such that low income Black women who live in Southern states are more likely to be reported by hospital staff, subjected to drug testing, charged with a felony, and arrested.Reference Paltrow and Flavin8 Research has begun to explore the impacts of punitive prenatal substance use laws on infant outcomes.9 For instance, one study found that enactment of punitive policies across eight states was associated with greater rates of neonatal abstinence syndrome (NAS), caused when an infant withdraws from certain drugs after prenatal exposure.10
To our knowledge, no studies have examined the impact of state-level policies regarding substance use during pregnancy on maternal mortality across all 50 states. This study utilizes a legal epidemiological approach, which is an emerging area of literature that provides a method to study the effects of laws on health-related outcomes.Reference Ramanathan11 This study examined the impact of state-level policies on rates of maternal mortality, including: (a) whether substance use during pregnancy is considered child abuse; (b) whether substance use during pregnancy requires mandated testing; (c) whether suspected substance use during pregnancy requires mandated reporting; and (d) whether substance use during pregnancy is grounds for civil commitment (see Table 1 for a complete list). We hypothesized that states that enacted each of these punitive laws would demonstrate higher rates of maternal mortality while controlling for correlates that have been shown in prior research to increase risk of maternal mortality, including: race, poverty, and rates of neonatal abstinence syndrome, prenatal care utilization, substance use in pregnant women, and health insurance.
Methods
Procedure
Publicly available state-level data from 2018 were combined from several sources including: (1) the Guttmacher Institute state-level coding prenatal substance use laws;6 (2) United Health Foundation’s data on rates of maternal mortality;12 (3) Healthcare Cost and Utilization Project data on rates of NAS among newborn hospitalizations;13 (4) Centers for Disease Control and Prevention data on rates of prenatal care in the first trimester;Reference Osterman and Martin14 (5) U.S. Census estimates of race, poverty levels, and health insurance (2010 to 2019);15 and (6) the National Survey on Drug Use and Health data on estimates of substance use during pregnancy.16 Data from all 50 states were included in analyses. See Table 2 for a complete list of population characteristics.
Data Analysis
Bivariate correlations were examined for all variables (Table 3). Four separate hierarchical multiple regression models were built to examine the impact of each type of state-level law on maternal mortality. A backward stepwise approach was utilized to select salient correlate variables while reducing the likelihood of overfitting with variables that did not contribute to the variance in maternal mortality. The backwards stepwise regression began with a full, saturated model that included all potential population correlates informed by extant literature. Correlates that contributed the least amount of variance were removed one-by-one until a parsimonious model that predicted maternal mortality from population correlates was detected. In the second step of each model, the state-level policy was included.
*** p < .001, **p < .01, * p < .05, N = 50.
Results
The percent of population that identified as Black, percent of population below the federal poverty line, and percent of women receiving prenatal care in the first trimester were significant predictors of maternal mortality, while the percent of infants with NAS and percent of pregnant women who used illicit drugs, tobacco, or alcohol in the past month were not significant predictors of maternal mortality (Table 3).
The percent of population that identified as Black, percent of population below the federal poverty line, and percent of women receiving prenatal care in the first trimester were significant predictors of maternal mortality, while the percent of infants with NAS and percent of pregnant women who used illicit drugs, tobacco, or alcohol in the past month were not significant predictors of maternal mortality.
The backwards stepwise approach resulted in a final regression model which controlled for state-level percent of population that identifies as Black, percent of the population that falls below the poverty line, percent of infants with NAS, percent of women receiving prenatal care in the first trimester, and percent of pregnant women who used illicit drugs, tobacco, or alcohol in the past month. In the first step of the model (excluding state-level policy), no correlates significantly predicted rates of maternal mortality. In the second step of the model, the presence of mandated testing laws significantly predicted higher rates of maternal mortality; additionally, the percent of women receiving prenatal care in the first trimester was a significant predictor of maternal mortality (Table 4). The remaining three types of laws were not significant in separate models controlling for population correlates.
** p <.01, * p <.05, N = 50.
Discussion
Results revealed that the presence of mandated testing laws significantly predicted increased rates of maternal mortality after controlling for state-level population correlates, including race, poverty, NAS, prenatal care, and prenatal substance use. The model accounted for 35.6% of the variance in mortality. The results of this study add to prior work demonstrating that enactment of punitive policies were associated with worsened infant outcome (NAS).17
Substance use during pregnancy may increase the odds of receiving inadequate prenatal care, which increases maternal mortality.18 It is possible that women may avoid prenatal care based on fear of legal consequences, including fear of child protective service involvement or criminal charges. In our mandated testing model, lack of prenatal care was related to higher maternal mortality. This provides preliminary support that reduced rates of prenatal care may be related to greater maternal mortality in light of punitive practices related to substance use. Future research utilizing medical records or Medicaid data could further examine this potential mechanism.
Findings provide preliminary support for the relationships between mandated testing laws, reduced rates of prenatal care, and increased rates of maternal mortality. Healthcare providers should be cognizant of these relationships and enact practices that bolster prenatal care utilization among pregnant women who use substances. Specifically, a multidisciplinary approach, in which clients are referred to providers with expertise in maternal substance use and treatment, has the potential to reduce risk for maternal mortality. Clinicians should engage in collaborative therapeutic discussions with clients to enhance health equity. Still, state-level policy reform, focused on best practices for treatment rather than legal consequences, could potentially reduce maternal mortality.
It is also important to note that race, specifically the percentage of the population that identifies as Black, was positively related to maternal mortality (as has been found in prior workReference Noursi19); however, in our final regression model, race was not a significant predictor, suggesting that there is shared variance between race and other correlate and legal variables in the model (i.e., mandated testing laws, poverty, NAS, prenatal care, and prenatal substance use). Bivariate correlations revealed that states with a higher proportion of Black residents tended to have lower levels of prenatal care utilization and higher rates of unemployment and poverty. This finding likely represents structural racism, such that Black people may experience disparities with regards to healthcare access in these states.Reference Howell20 This is consistent with research demonstrating that Black women are more likely to experience negative pregnancy-related health outcomes and healthcare discrimination, including increased pregnancy-related deaths, discrimination during prenatal care, and decreased access to healthcare.Reference Vintzileos21 Future research should expand on work examining inequitable implementation of these laws among racial minorities.22
The study is not without limitations. First, analyses were conducted with state-level data from several publicly available federal and state datasets. Thus, the results examined trends at the population level rather than the impact of policies at the individual level. Future work in this area should utilize diverse methodology to examine individualized experiences with policies, healthcare, and legal systems to better understand the impact of stigma or legal consequences. Second, the cross-sectional nature of this study does not allow for causal inference or trends over time. Further research can utilize a longitudinal approach to examine how maternal mortality is impacted by changes in policy. Third, there are inconsistences in how local jurisdictions document substance use during pregnancy and file petitions for court interventions and child removal.Reference Ondersma23 Thus, research accounting for nuances in policy enforcement across states could complement the present findings.
This study is the first known to utilize a legal epidemiological approach to examine the impact of state-level policies punitive of substance use during pregnancy with maternal mortality across all states. Results provide preliminary support for the relationship between mandated testing laws, reduced rates of prenatal care, and increased rates of maternal mortality.
Notes
The work described in this article was funded in part by NIH grants T32DA019426 and T32DA037202. The views and opinions expressed are those of the authors. The funding organizations had no further role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.