Substance Use Disorder-Related Deaths and Maternal Mortality in New Mexico, 2015–2019

In recent decades, maternal mortality has increased significantly across the U.S. The Centers for Disease Control and Prevention (CDC) has launched efforts to combat these increases through jurisdiction-level Maternal Mortality Review Committees (MMRCs) to comprehensively review pregnancy-associated deaths and make recommendations to prevent future deaths (CDC, 2022a). MMRCs have strengthened surveillance efforts to track leading causes of pregnancy-associated deaths. Substance use has been identified as a leading cause of pregnancy-associated deaths (Smoots et al., 2021).

A review of 22 states from 2007 to 2016 found that pregnancy-associated mortality involving opioids more than doubled, from 1.3 deaths to 4.2 deaths per 100,00 live births (Gemmill et al, 2019). The leading cause of death among females 15–49 years old in the U.S. from 2015 to 2019 was unintentional injury; overdoses made up 61% of these injuries (CDC, 2022b). In Utah, the drug-induced pregnancy-associated mortality ratio in 2014 was 11.7 deaths per 100,000 live births (Smid et al, 2019). Few states have published SUD-related maternal mortality findings exploring risk factors and substance use patterns among SUD-related deaths.

In New Mexico (NM), the fatal overdose death rate increased from 24.0 deaths per 100,000 people in 2015 to 33.3 deaths per 100,000 people in 2019 (Bureau of Vital Records and Health Statistics, 2019). Additionally, NM had the highest age-adjusted alcohol-attributable death rate in the country from 2011 to 2015, with 53.1 deaths per 100,000 people compared to the national average of 28.0 deaths per 100,000 people (Esser et al, 2020). While NM rates of substance use and death due to substance use in the general population are high, experiences of pregnant and postpartum people who have Substance Use Disorder (SUD)-related deaths have not been previously evaluated in NM.

The aim of this study was to assess ratios of SUD maternal mortality and investigate risk factors associated with SUD-related and non-SUD-related deaths among all pregnancy-associated deaths in New Mexico from 2015 to 2019. This involved comparing demographics, pregnancy experiences, circumstances surrounding deaths, treatment of mental health conditions, and experiences of social stressors of SUD-related deaths to non-SUD-related deaths. We also explored patterns of substance use among both groups.

We assessed 87 pregnancy-associated deaths in New Mexico from 2015 to 2019. Of these deaths, 49% were SUD-related deaths (n = 43) and 51% were non-SUD-related deaths (n = 44). The pregnancy-associated mortality ratio for 2015–2019 was 72.5 deaths per 100,000 live births. The SUD-related mortality ratio was 35.8 deaths per 100,000 live births. The SUD-related mortality ratio varied from 2015 to 2018, with 31.1 (95% CI 13.4–61.3) deaths per 100,000 live births in 2015 to 47.7 (95% CI 23.8–85.4) deaths per 100,000 births in 2018 but decreased in 2019 to 30.5 (CI 12.3–62.8) deaths per 100,000 people. Our time trend findings were not statistically significant (p = 0.62) (Fig. 1).

People with SUD-related deaths were more likely to have died 43–365 days postpartum (81% vs. 45%, p-value = 0.002), have had mental health conditions as a contributing factor to their death (81% vs. 11%, p-value < 0.001), have had mental health conditions as their primary cause of death (47% vs 10%, p < 0.001), and have died of an overdose (41% vs. 8%, p-value = 0.002) compared to people with a non-SUD-related death (Table 1). We found no statistical differences in the other factors evaluated in Table 1. There were no statistically significant differences between SUD-related and non-SUD-related deaths in preventability (79% of all pregnancy-associated deaths in NM were preventable) or death from suicide. Among the SUD-related deaths, 54% of people were Hispanic, 95% were on Medicaid, 86% had delivered, and 17% were suicides.

People with SUD-related deaths were more likely to have received treatment before pregnancy for anxiety disorder (37% vs. 11%, p-value = 0.005), depression (37% vs. 9%, p-value = 0.002), SUD (40% vs. 2%, p-value = 0.01), and any MHCs (61% vs. 16%, p-value < 0.001) compared to people with non-SUD-related deaths. People with SUD-related deaths were more likely to have received treatment for depression (23% vs. 7%, p-value = 0.04) after pregnancy compared to people with non-SUD-related deaths. They were also more likely to have received treatment for anxiety disorder (44% vs. 11%, p-value = 0.001), depression (47% vs. 11%, p-value < 0.001), SUD (49% vs. 2%, p-value < 0.001), or any MHCs (72% vs. 18%, p-value < 0.001) at any point before, during, or after pregnancy compared to non-SUD-related deaths. Of all people with SUD-related deaths, fewer than half received treatment for SUD before pregnancy (40%) or during pregnancy (37%). Fewer than one-third of people with a SUD-related death received SUD treatment after pregnancy (28%), while the majority of these SUD-related deaths occurred 43–365 days postpartum (Table 2).

People with SUD-related deaths were more likely to have had CPS involvement (54% vs 14%, p-value = 0.01), have been unemployed (49% vs. 14%, p-value = 0.03) and were more likely to have experienced any social stressors (86% vs 30%, p < 0.001) compared to people with non-SUD-related deaths (Table 3). The most common stressors among people with SUD-related deaths were CPS involvement (54%), unemployment (49%), domestic violence (41%), and previous suicide attempt (32%). Fewer than one-third of people with SUD-related deaths had substance use treatment as a documented stressor (27%).

The substances most used at the time of death among SUD-related deaths were amphetamines (70%), opiates/opioids (35%), alcohol (30%), and benzodiazepines (26%). Seventy-four percent of all SUD-related deaths had at least one substance detected in their toxicology, compared to 30% of non-SUD-related deaths. Just under half (49%) of all SUD-related deaths had evidence of substance use documented during prenatal care. Most SUD-related deaths engaged in polysubstance use (63%), with 42% of SUD-related-deaths using three or more substances at the time of death (Table 4).

We found more SUD-related deaths among people with Medicaid insurance, and people with SUD-related deaths were more likely to die 43–365 days postpartum than people with non-SUD-related deaths. We additionally found a high frequency of treatment for MHCs, experiences of CPS involvement and unemployment as social stressors, polysubstance use, and a lack of SUD detected during prenatal care among people with SUD-related deaths.

Our finding that SUD-related deaths were more likely to occur 43–365 days postpartum is consistent with previous studies. An analysis of drug-induced deaths in Utah found that 80% of deaths occurred in this same “late postpartum” period of 43–365 days after pregnancy (Smid et al, 2019). Additionally, while high rates of prenatal abstinence are found across all substances, rates of return to use in the postpartum period are high—a study focused on opioid use found that rates of opioid relapse, overdose, and death increased in the year following a birth among people with Opioid Use Disorder (Schiff et al, 2018). An additional study of people with Opioid Use Disorder receiving buprenorphine treatment at 6 months postpartum found that 51.7% of people had substance use recurrence after delivery (Ellis et al, 2022). Qualitative studies have found that reasons for returning to use include stress of caring for a newborn, lack of sleep, lack of social and financial support, shame, low self-esteem, and experiences with domestic violence (Ripley-Moffitt et al, 2008; Notley et al, 2015; Pallatino et al., 2021). Return to substance use, particularly opioids including fentanyl, is dangerous as there is a higher risk of an overdose due to a reduced tolerance following abstinence (Warner-Smith et al, 2001). Previous studies have found that people with strong social supports, including having spouses or significant others that did not use drugs and having positive family relationships postpartum, were less likely to return to use after delivery (Ellis et al., 2004). Qualitative studies have also found that social support systems, such as online forums or peer support workers in perinatal substance use programs, can be helpful during treatment and in assisting patients to find continued social supports after leaving the programs (Moore et al, 2017; Olding et al, 2022). Expanding peer support networks may improve support and access to resources for people in their postpartum periods, which may reduce stress and likelihood of returning to use.

People represented in the SUD-related deaths were more likely to have been covered by Medicaid for healthcare in the perinatal period. In many states—including NM, until 2022—Medicaid for pregnant people only covered 60 days postpartum, and up to one-third of postpartum people have faced disruptions in insurance statuses (Shah & Friedman, 2022). Uninsured people in the postpartum period face poorer outcomes, particularly in treating chronic diseases—a known contributor to maternal mortality (Villavicencio et al, 2020).

New Mexico pregnancy-related Medicaid was extended in April 2022 to cover people up to one year postpartum. Implementation of Medicaid extension beyond 60 days in Texas resulted in an increase in uptake of mental and behavioral health services (Wang et al, 2022). Previous studies have found that, as the Medicaid eligibility threshold increased to include more people, SUD-related deaths reduced by 6.5% from 2002 to 2015, and states that expanded Medicaid mental health plans found increases in enrollment in treatment programs (Snider et al, 2019). Based on results of expansion of Medicaid, we expect to see similar improvements in maternal outcomes due to Medicaid extension. Until we complete case reviews for 2022 deaths, we are not able to determine the impact of Medicaid extension on preventing deaths; however, we hope to see an increase in receipt of SUD treatment and a reduction in preventable, SUD-related deaths for our at-risk population of people who are on Medicaid and between 43 and 365 days postpartum.

We found a greater likelihood of treatment for MHCs and experiences of social stressors among people with SUD-related deaths compared to people with non-SUD-related deaths. Previous studies have found high rates of depression, anxiety, and other MHCs among people with SUD-related deaths and pregnant people with SUD (Arnaudo et al, 2017; Smid et al, 2019). CPS involvement and unemployment were leading social stressors, followed by domestic violence and a previous suicide attempt. Experiences of chronic stress can increase rates of substance use, particularly among people with comorbidities of MHCs and SUD, and among people with experiences of abuse or violence (Brady et al., 2005; Sinha, 2008). Substance use treatments must be trauma-informed and take into consideration the common co-occurring morbidities to address all aspects of the person’s health comprehensively and effectively, rather than exclusively focusing on clinical or acute addiction issues.

We found that fewer than half of people (49%) with SUD-related deaths received SUD treatment at any point before, during, or after pregnancy. Universal screening of all people at their first prenatal visit with a validated tool, following The American College of Obstetricians and Gynecologists (ACOG) guidelines, could reduce stigma often associated with substance use during pregnancy. It would also facilitate receipt of opioid agonist pharmacotherapy—the therapy recommended by ACOG, a component of Medication Assisted Treatment (MAT) and Medications for Opioid Use Disorder (MOUD)—and support to pregnant people using substances prenatally and in the postpartum period (ACOG Committee Opinion, 2017). According to a New Mexico Department of Health Gap Analysis in 2020 on Substance Use Treatment, only 38% of SUD treatment locations in New Mexico offered MAT, and eight counties out of 33 did not have any facility that offered this treatment. Additionally, many facilities reported that they rely on one provider to prescribe MAT and stated that patients will often cancel appointments because they cannot afford co-pays. Of 308 facilities in New Mexico, 240 (78%) accepted pregnant patients, and 51 (17%) treatment locations allowed parents to bring a child with them to sessions or transitional living houses (New Mexico Department of Health, 2020). This indicates that there are treatment resources for pregnant people, but not as many that provide MAT, the more effective treatment, and fewer that accept parenting individuals. New Mexico needs to increase the number of providers that have the certification required to prescribe MAT so that it is more accessible, particularly for pregnant and postpartum people and people in rural areas with fewer treatment options. New Mexico also needs to expand affordable treatment options with reduced co-pays to improve patient retention.

A majority of people with SUD-related deaths engaged in polysubstance use (63%). These findings are consistent with previous studies that have found between 64 and 89% of pregnant people with Opioid Use Disorder engaging in polysubstance use (Jarlenski et al, 2017, 2020; Smid et al, 2019). Treatment programs must account for polysubstance use to treat SUD more effectively.

While amphetamines and opiates/opioids were the most used substances, alcohol made up 30% of substances used among SUD-related deaths. This is consistent with findings that NM has had the highest age-adjusted alcohol attributable death rate in the country. As treatment for other SUDs are highlighted amidst the opioid crisis, treatment for alcohol use disorder still must remain a priority for SUD treatment in NM.

Just under half of SUD-related deaths (49%) had any evidence of substance use in their prenatal care record, indicating that patients were either not using substances during pregnancy but still had a SUD-related death, providers were unaware of their patient’s substance use during pregnancy, or that these patients did not receive any prenatal care. Providers may be unaware of substance use if a patient does not disclose information on their use, possibly due to a fear of stigma or punishments such as CPS involvement, child separation, or legal ramifications (Falleta et al., 2018; Stone, 2015).

Our study had several limitations. We had a small sample size of 87 deaths total, with 43 SUD-related deaths, and this may have limited our ability to detect statistically significant differences and draw definitive conclusions from our findings. Our small sample size restricted us to descriptive, univariate analyses. These findings are not generalizable to all birthing people in New Mexico, or to all people who die SUD-related deaths. These data represent deaths among pregnant and postpartum people with SUD in New Mexico and may not be generalizable to other state or national pregnant and postpartum populations.

Inconsistency of data entry and overall missingness of data may have contributed to misclassification in our analyses. If there was no documented treatment for a MHC, abstractors left the variable blank, meaning that the missing category includes both people with unknown documentation of MHCs and people who were known to be not treated for MHCs.

Nearly one-third (27%) of non-SUD-related deaths did not have toxicology completed, either because an autopsy was not completed, or the Medical Investigator did not deem one necessary. This amount of missing data limits what conclusions we can draw from these findings.

Additionally, information on treatments received for SUD—such as MAT, counseling, and inpatient or outpatient treatment—the duration of treatment, dosage of medications prescribed, or other information indicating resources pregnant and postpartum people accessed was ultimately unavailable and thus limited our conclusions on these topics.

Substance use is a leading contributing factor to death in the United States. SUD-related deaths are more likely to occur postpartum, and nearly half of those who died a SUD-related death in NM were not properly screened or detected for SUD when receiving prenatal care. Providers, health departments, and community organizations must prioritize support and treatment for people using substances both during and after pregnancy to prevent death and improve quality of life for pregnant and postpartum people and their families.