Polysubstance Use Among Pregnant Women With Opioid Use Disorder in the United States, 2007–2016

Marian P Jarlenski; Nicole C Paul; Elizabeth E Krans

doi:10.1097/AOG.0000000000003907

. Author manuscript; available in PMC: 2021 Sep 1.

Published in final edited form as: Obstet Gynecol. 2020 Sep;136(3):556–564. doi: 10.1097/AOG.0000000000003907

Polysubstance Use Among Pregnant Women With Opioid Use Disorder in the United States, 2007–2016

Marian P Jarlenski ¹, Nicole C Paul ², Elizabeth E Krans ³

PMCID: PMC7483798 NIHMSID: NIHMS1581649 PMID: 32769641

Abstract

Objective:

To assess trends in polysubstance use among pregnant women with opioid use disorder in the United States.

Methods:

We conducted a time trend analysis of pooled, cross-sectional data from the National Inpatient Sample, an annual nationally representative sample of US hospital discharge data. Among 38.0 million females aged 15–44 years with a hospitalization for delivery from 2007 to 2016, we identified 172,335 pregnant women with an ICD-9 or ICD-10 diagnosis of opioid use disorder. Polysubstance use among pregnant women with opioid use disorder was defined as ≥1 co-occurring diagnosis of other substance use, including alcohol, amphetamine, cannabis, cocaine, sedative, or tobacco. We fit weighted multivariable logistic regression models to produce nationally representative estimates, including an interaction between year and rural vs. urban county of residence; controlled for age, race, and insurance type. Average predicted probabilities and 95% CIs were derived from regression results.

Results:

Polysubstance use among women with opioid use disorder increased from 60.5% (95% CI: 58.3%, 62.8%) to 64.1% (95% CI: 62.8%,65.3%). Differential time trends in polysubstance use among women with opioid use disorder were found in rural versus urban counties. Large increases in amphetamine use occurred among those in both rural and urban counties (255.4%; 95% CI: 90.5%,562,9% and 150.7%; 95% CI: 78.2%,252.7%, respectively), similarly to tobacco use (30.4%; 95% CI: 16.9%,45.4% and 23.2%; 95% CI: 15.3%,31.6%, respectively). Cocaine use diagnoses declined among women with opioid use disorder at delivery among those in rural (−70.5%; 95% CI: −80.4%,−55.5%) and urban (−61.9%; 95% CI: −67.6%,−55.1%) counties. Alcohol use diagnoses among those with opioid use disorder declined −57% (95% CI: −70.8%,−37.7%) among those in urban counties but did not change among those in rural counties.

Conclusion:

Over the past decade, polysubstance use among pregnant women with opioid use disorder has increased more rapidly in rural versus urban counties in the U.S., with amphetamines and tobacco use increasing most rapidly.

PRECIS

Polysubstance use among pregnant women with opioid use disorder has increased more rapidly in rural compared with urban counties in the United States, with amphetamines and tobacco use increasing most rapidly.

INTRODUCTION

Opioid use disorder is a major contributor to maternal and neonatal morbidity and mortality in the United States.¹ While the prevalence of opioid use disorder and overdose during pregnancy has escalated over the past 20 years, the nature of the opioid epidemic in the population as a whole has continued to evolve over time.² Characterized as three distinct waves, the first 10 years of the epidemic was characterized by the rise in overprescribing and overdose deaths among adults caused primarily from prescription opioid use.² During the second wave, starting in 2010, heroin replaced prescription opioids as the leading cause of overdose death but was soon surpassed by fentanyl, in the third wave starting in 2013.³ While synthetic opioids remain the primary contributor to overdose mortality among adults, the co-occurring rise in polysubstance use over the past decade has received considerably less attention, and may represent a fourth wave of the rapidly evolving crisis. In a survey of over 15,000 people entering treatment for opioid use disorder, 96.4% reported the use of at least 1 non-opioid drug in the past month.⁴ When co-occurring substance use was evaluated over time, methamphetamine use had the greatest observed increase (85%) among people with opioid use disorder between 2011 and 2018.⁴

Over 89% of women of reproductive age (18–44 years) who use illicit opioids also use at least one non-opioid drug with tobacco, binge drinking of alcohol and cannabis the most common substances used.⁵ Despite the rise of polysubstance use among people with opioid use disorder, the effect of this trend on substance use during pregnancy is poorly understood, which hinders the development of effective public health strategies to respond to the changing dynamic of the U.S. opioid epidemic. Further, understanding trends in polysubstance use among women with opioid use disorder has important clinical implications as tobacco, alcohol and amphetamine use during pregnancy are well-established independent risk factors for adverse maternal and neonatal health outcomes.⁶

To address the gap in our knowledge of how patterns of substance use during pregnancy have changed over time among women with opioid use disorder, we conducted a time trend analysis of nationally representative hospital discharge data from 2007 to 2016 to evaluate for the prevalence of co-occurring substance use. To identify differential patterns in substance use by geographic region, differences in polysubstance use during pregnancy were also evaluated among women residing in rural compared to urban counties.

METHODS

We used data from the National Inpatient Sample (NIS) in 2007–2016, which is an annual, nationally representative sample of hospital discharges administered by the U.S. Agency for Healthcare Research and Quality.⁷ The NIS includes a 20% stratified sample of discharges from U.S. hospitals, excluding rehabilitation and long-term care hospitals, and includes uniform data elements on diagnosis and procedure codes, diagnosis related groups, severity and comorbidities, and patient characteristics. The NIS provides sample weights so that data can be analyzed to produce nationally representative estimates and appropriate standard errors. To identify delivery hospitalizations, we used Diagnosis Related Group (DRG) codes and ICD-9-CM diagnosis and procedure codes for females ages 15–44 years, following an established algorithm.⁸ We then crosswalked the ICD-9 codes to ICD-10 codes for all hospitalizations occurring after Oct 1, 2015, the date of the U.S. healthcare system’s transition from ICD-9 to ICD-10 coding (see Appendix 1, for specific codes used). Weighted and unweighted counts of delivery hospitalizations were consistent before and after the ICD-9 to ICD-10 transition (see Appendix 2). We identified a weighted count of 38.0 million delivery hospitalizations across 10 years of data. To identify a sample of women with opioid use disorder at delivery, we used ICD-9 and ICD-10 diagnosis codes to identify diagnoses for OUD, including a broad range of codes for abuse or dependence of any opioids.⁹ The final analytic sample included a weighted count of 172,335 delivery hospitalizations among women with opioid use disorder.

Our outcome of interest, polysubstance use, was defined as the diagnosis of any additional non-opioid substance use during the delivery hospitalization. We used ICD-9 and ICD-10 diagnoses for abuse or dependence of the following substances: tobacco, alcohol, cannabis, cocaine, sedatives, and amphetamines. Sedatives included any sedatives, hypnotic, or anxiolytic substances. Appendix 3, shows diagnostic codes used to identify substance use.

Covariates included patient age category (15–18 years, 19–34 years, or >=35 years), patient race (White, Black, Asian, Other, Missing), patient ethnicity (Hispanic vs not Hispanic), and insurance type (Medicaid vs other insurance). Nearly 12% of hospital discharge data were missing patient race, as some state data sources do not provide race data for privacy or other administrative reasons. Our primary analyses included an indicator of missing race; we also considered analyses excluding data with missing patient race and the results were unchanged. The data included a variable classifying the urbanicity of the county of patient residence (although data do not identify specific counties). Patient residence in a rural or urban county was defined by the rural-urban classification scheme developed by the National Center for Health Statistics (NCHS) to define rural and urban residence.¹⁰ The NCHS classifies counties into 1 of 6 categories: large metropolitan counties with populations>1 million, fringe counties to large metropolitan areas, medium metropolitan counties with populations of 250,000–999,999, small metropolitan counties with populations of 50,000–249,999, micropolitan counties with urban populations of 10,000–49,999, and noncore counties that are neither metropolitan or micropolitan. Following prior research,¹¹ we dichotomized the NCHS classification into urban counties (large metropolitan, fringe metropolitan, medium metropolitan, and small metropolitan) and rural counties (defined as micropolitan and noncore counties).

All analyses incorporated NIS weights to provide nationally representative estimates and appropriate standard errors. Because the NIS redesign that began in 2012 resulted in some disruptions in trend analyses,¹² NIS developed new discharge-level weights that can be applied to data prior to 2012 to provide consistent results over time. As recommended, we applied the trend weights to data from 2007–2011 to be consistent with weighted data from 2012–2016.¹³

Time trend analyses were conducted considering year as an ordered categorial variable. Because the NIS data are fully de-identified, and it is not possible to follow the same person over time across multiple deliveries, the unit of analysis was each hospitalization. First, we calculated weighted descriptive characteristics of the study sample, overall and stratified by opioid use disorder only vs polysubstance diagnosis, and by rural vs. urban county of residence. Next, we analyzed time trends in the prevalence of any co-occurring substance use diagnosis among women with opioid use disorder. To do this, we fit a multivariable logistic regression model where an indicator of any additional non-opioid substance use diagnosis was the outcome, and included an ordinal variable indicating year, an indicator of rural vs. urban patient county of residence, and an interaction between year and rural residence. Crude and adjusted odds ratios and related 95% confidence intervals from the regression models are shown in Appendix 4. We used marginal standardization methods to estimate time trends in the prevalence of co-occurring substance use diagnoses among women with opioid use disorder at delivery.¹⁴ Under this approach, we derived average predicted probabilities and related 95% confidence intervals of having any co-occurring substance diagnosis for each year, overall and for both rural and urban counties, from the regression models. The percent changes in the outcome from 2007 to 2016 and related 95% confidence intervals, in both rural and urban counties, were also derived from the regression results. We replicated this approach to evaluate for specific changes across each substance among women with opioid use disorder at delivery. The University of Pittsburgh Institutional Review Board determined that this research was exempt because it consisted of secondary data analysis using fully de-identified hospital discharge data.

RESULTS

The final analytic sample included data on a weighted 172,235 delivery hospitalizations in the U.S. between 2007–2016 in which opioid use disorder was diagnosed, 20.4% in which patients resided in rural counties and 79.6% in which patients resided in urban counties (TABLE 1). The majority of the study sample was aged 19–34 years at delivery (88.9%), was Non-Hispanic white (74.4%), and had Medicaid insurance coverage (80.5%). Among women diagnosed with opioid use disorder at delivery, 62.2% were diagnosed with any polysubstance use. Overall, the most prevalent non-opioid substances used were tobacco (53.5%), cannabis (9.8%), and cocaine (7.0%). Among those with polysubstance use, women residing in rural counties had higher rates of tobacco and amphetamine use, while women in urban areas had higher rates of alcohol and cocaine use.

Table 1.

Weighted descriptive characteristics of females age 15–44 years with opioid use disorder (OUD) at delivery, 2007–2016

	All OUD	OUD Only		Polysubstance
	All OUD	Rural^a	Urban^a	Rural^a	Urban^a
Weighted n	172,235	13,490	51,685	25,545	81,515
Age Group, %
15–18 y	1.0	1.1	0.9	1.2	0.9
19–34 y	88.9	90.9	87.8	91.3	88.5
>34 y	10.1	8.1	11.3	7.6	10.5
Race and Ethnicity, %^b
Non-Hispanic White	74.4	77.4	71.2	80.2	74.1
Non-Hispanic Black	6.0	2.2	6.9	1.7	7.3
Non-Hispanic Asian	1.8	3.3	1.5	2.7	1.4
Other race	1.7	1.1	2.1	1.4	1.7
Missing race	9.5	11.3	8.3	10.9	9.5
Hispanic ethnicity	6.7	4.7	10.0	3.2	6.1
Socioeconomic Status
Medicaid insurance coverage	80.5	81.3	75.0	85.2	82.3
Reside in low-income area^c	36.4	49.6	29.3	52.0	33.4
Substance use diagnoses, %
OUD and other substance use^d	62.2	0.0	0.0	100.0	100.0
Tobacco	53.5	--	--	89.0	85.1
Alcohol	1.8	--	--	1.7	3.2
Cannabis	9.8	--	--	15.9	15.7
Cocaine	7.0	--	--	6.4	12.9
Sedatives^e	2.5	--	--	4.5	3.9
Amphetamines^f	4.1	--	--	6.2	6.7

Open in a new tab

Rural and urban areas defined according to the National Center for Health Statistics classification scheme for U.S. counties

Race not available from some state data sources

Defined as residing in the lowest income quartile of U.S. zip codes

Defined as diagnosis of OUD in addition to at least one of the following: Tobacco, alcohol, cannabis, cocaine, sedatives, amphetamines

Includes diagnoses related to use, dependence, or abuse of sedatives, hypnotic, or anxiolytic substances

includes diagnoses related to use, dependence, or abuse of amphetamines

Among women with opioid use disorder, the average predicted probability of having any co-occurring substance use diagnosis at delivery increased significantly in 2016 relative to 2007 (FIGURE 1). From 2007–2016, polysubstance use among those with opioid use disorder increased from 60.5% (95% CI: 58.3%, 62.8%) to 64.1% (95% CI: 62.8%, 65.3%). Following a significant decline from 2007 to 2009, polysubstance use diagnoses remained stable in 2010–2013, and then were statistically significantly greater in 2014, 2015, and 2016, relative to 2007.

Figure 1. — Adjusted prevalence of polysubstance use diagnosis among those with opioid use disorder at delivery in the United States, 2007–2016. Average predicted probabilities and 95% CIs are derived from a weighted logistic regression controlling for rural or urban residence, age, race or ethnicity, and Medicaid insurance coverage. Rural and urban areas defined according to the National Center for Health Statistics classification scheme for U.S. counties. Polysubstance use diagnosis was statistically significantly lower in 2008 and 2009, relative to 2007; polysubstance use was statistically significantly greater in 2014, 2015, and 2016, relative to 2007.

The average predicted probability of polysubstance use among those with opioid use disorder diagnosis at delivery increased at a faster rate in rural counties, relative to urban counties (FIGURE 2). Increases in polysubstance use relative to 2007 were statistically significantly greater among rural vs urban residents in 2008, 2013, 2014, and 2016. From 2007–2016, the percentage of women with opioid use disorder and any additional substance use increased from 60.4% (95% CI: 55.8%, 65.1%) to 68.7% (95% CI: 66.1%, 71.3%) among those residing in rural counties and increased from 60.6% (95%CI: 58.0%, 63.2%) to 62.7% (95% CI: 61.3%, 64.1%) among those residing in urban counties. For those in rural counties, polysubstance use diagnosis was statistically significantly greater in 2014, 2015, and 2016, relative to 2007. For those in urban counties, polysubstance use diagnosis was statistically significantly lower in 2008 and 2009, relative to 2007; and was statistically significantly greater in 2015, relative to 2007.

Figure 2. — Adjusted prevalence of polysubstance use diagnosis among those with opioid use disorder at delivery residing in rural and urban counties in the United States, 2007–2016. Average predicted probabilities and 95% CIs are derived from a weighted logistic regression controlling for rural or urban residence, age, race or ethnicity, and Medicaid insurance coverage. Rural and urban areas defined according to the National Center for Health Statistics classification scheme for U.S. counties. For those in rural counties, polysubstance use diagnosis was statistically significantly greater in 2014, 2015, and 2016, relative to 2007. For those in urban counties, polysubstance use diagnosis was statistically significantly lower in 2008 and 2009, relative to 2007; and was statistically significantly greater 2015, relative to 2007. Increases in polysubstance use relative to 2007 were statistically significantly greater among rural vs urban residents in 2008, 2013, 2014, and 2016.

FIGURE 3 shows the time trends in the diagnoses of specific substances used among pregnant women with opioid use disorder at delivery. The largest relative increase from 2007–2016 was in use of amphetamines, which rose from an average predicted probability of 2.4% (95% CI: 1.7%, 3.1%) to 6.6% (5.9%, 7.2%). Tobacco use diagnoses in addition to opioid use disorder also increased markedly over time, from 45.3% (95% CI: 43.0%, 47.6%) to 56.3% (95% CI: 55.1%, 57.6%). Diagnosed use of cannabis and sedatives in addition to opioid use disorder were relatively stable over time. Diagnosed use of cocaine in addition to opioid use disorder declined from 15.3% (95% CI: 13.6%, 16.9%) to 5.7% (95% CI: 5.1%, 6.3%). Diagnosed use of alcohol in addition to opioid use disorder also declined from 2.9% (95% CI: 2.1%, 3.6%) to 1.3% (95% CI: 1.0%, 1.5%)

FIGURE 4 shows the estimated percent changes from 2007–2016 of any polysubstance use diagnosis and the estimated percent change in the diagnoses of specific substances used among pregnant women with opioid use disorder. The rate of any polysubstance use diagnosis among women with opioid use disorder at delivery increased more among those women residing in rural (13.8% increase; 95% CI: 4.4%, 24.1%) vs urban counties (3.5% increase; 95% CI: −1.4%, 8.7%). Diagnosed use of amphetamines and OUD more than doubled among those residing in both rural (255.4% increase; 95% CI: 90.5%, 562.9%) and urban (150.7% increase; 95% CI: 78.2%, 252.7%) counties. Similarly, diagnosed tobacco use and opioid use disorder increased among those in both rural (30.4% increase; 95% CI: 16.9%, 45.4%) and urban (23.2% increase, 95% CI: 15.3%, 31.6%) counties. Diagnosed use of alcohol and opioid use disorder did not change among those residing in rural counties, although a decline was observed among those residing in urban counties (57.4% decline; 95% CI: −70.8%, −37.7%). Diagnosed use of cocaine and opioid use disorder declined in both rural (70.5% decline, 95% CI: −80.4%, −55.5%) and urban (61.9% decline; 95% CI: −67.6%, −55.1%) counties.

DISCUSSION

In this nationally representative study, we found that polysubstance use diagnoses are common among pregnant women with opioid use disorder and have increased at a faster rate in rural (13.8% increase) relative to urban (3.5% increase) counties. The changing trends in polysubstance use among pregnant women with opioid use disorder inform efforts to improve screening for substance use disorders during pregnancy by emphasizing the need to screen pregnant women for all licit and illicit substances inclusive of tobacco, alcohol, marijuana and non-opioid illicit substances. Among pregnant women with opioid use disorder, evolving trends in substance use behaviors during pregnancy reinforce the urgency to expand the scope and intensity of treatment services for pregnant and parenting women with beyond opioid-specific interventions, such as medications for opioid use disorder, to comprehensively address substance use in the US.. Medications for opioid use disorder, such as methadone and buprenorphine, are already underused during pregnancy and our findings highlight not only the need to expand access to opioid-specific medications, but the need to take a comprehensive approach to substance use treatment that incorporates medical, psychosocial and behavioral interventions to address co-occurring substance use and associated factors (i.e. life stressors, poverty, co-occurring psychiatric disorders).^15,16 Further, because this study and others have shown that the primary source of insurance for women with substance use disorders is Medicaid,^5,17 state Medicaid programs have an important role in improving access to comprehensive treatment services by developing payment structures that provide reimbursement for services beyond medications for opioid use disorder to address co-occurring substance use (i.e. tobacco session, intensive outpatient therapy, cognitive behavioral therapy).^18–20

The prevalence of polysubstance use during pregnancy in our sample increased disproportionately in rural versus urban counties with the largest increases found in amphetamine use (250%) among pregnant women living in rural counties. While rural-urban disparities in substance use across the US have been well documented in prior research, current findings highlight the need to expand the content of physician training initiatives to include the management of co-occurring substance use and the need to increase the number of physicians trained to care for pregnant women with co-occurring substance use in underserved, rural areas.²

The largest increases in polysubstance use among pregnant women with opioid use disorder occurred among those in rural counties where amphetamine use increased more than 250%. Over the past decade, methamphetamine use has been increasing across the US and in Western parts of the country, has developed into a parallel epidemic.²¹ From 2011 to 2017, methamphetamine use among people with opioid use disorder increased from 18.8% to 34.2% with more significant increases observed among women (+97.8%) compared to men (+81.8%).²¹ Similarly, the prevalence of amphetamine use among pregnant women without opioid use disorder has also been on the rise. From 2014–2015, 1% of deliveries in the rural West were complicated by amphetamine use which exceeds the incidence of OUD during pregnancy in many parts of the country.²²

Methamphetamine exposure in utero are both independently associated with an increased risk of growth restriction, placental abruption, and preterm birth^23–26 and may exacerbate high rates of adverse birth outcomes among women with opioid use disorder.^27,28 Polysubstance use among pregnant women with opioid use disorder has also been associated with increased neonatal abstinence syndrome (NAS) severity.²⁹ In an evaluation of factors that contribute to NAS severity among 41 women with OUD using buprenorphine during pregnancy, co-occurring illicit substance use (i.e. methamphetamines, cocaine, marijuana) was the most significant predictor of NAS severity.³⁰ Despite these findings, the effects of polysubstance use on short and long-term maternal and neonatal health outcomes is largely unknown and warrant increased attention to the both the additive and cumulative effects of co-occurring substance use on health outcomes for pregnant women and children. Additional research is also needed to identify more effective pharmacologic and non-pharmacologic interventions for amphetamine and other stimulant use disorders.

Tobacco use diagnoses also increased significantly over time, although less dramatically than amphetamine use, and tobacco use remains a leading modifiable cause of adverse pregnancy and birth outcomes.²⁵ Notably, the rate of tobacco use diagnoses (53.5%) in our analysis is significantly greater than the rate of tobacco use in the general population of pregnant women which has been relatively stable at about 12%, with no significant changes over time.³¹ While the true prevalence of tobacco use may be under-estimated in our sample due to under-coding at the delivery hospitalization, these findings are consistent with high rates of tobacco use (>50%) found among pregnant women with opioid use disorder in prior studies.^32,33 Due to the high prevalence of co-occurring tobacco use during pregnancy, evidence-based pharmacologic and non-pharmacologic interventions to decrease tobacco use during pregnancy and postpartum should be universally employed in both prenatal and substance use treatment settings.

Findings from this study should be interpreted in light of certain limitations. First, the NIS data rely on diagnoses recorded in hospital discharge data, which may under-represent the true prevalence of substance use at delivery. Such codes are dependent on screening for and documentation of substance use during prenatal care or at delivery, which are not universal. Additionally, patients may not report substance use. In the NIS, it is not possible to differentiate prescribed medication use from illicit substance use. Nevertheless, the NIS data provide a unique opportunity to obtain nationally representative estimates of time trends in opioid use disorder and other substance use at the time of delivery. Diagnostic codes are often dependent on screening for and the documentation of substance use during pregnancy, which is not consistent across institutions. Even when substance use screening occurs, limitations in the sensitivity and specificity of screening using both validated self-reported screening instruments and biologic testing may limit an accurate assessment of the true prevalence of substance use.^34–36 Second, it is possible that the time trends in substance use we observed may be related to secular trends in hospital-based screening and diagnosis of such conditions. However, this limitation may be mitigated in the population of women diagnosed with opioid use disorder at delivery, as healthcare professionals who diagnose opioid use disorder are perhaps more likely to consistently diagnose other substance use behaviors in women with a disorder. Third, it is not possible to differentiate prescribed medication use from illicit substance use in NIS datasets and it remains unknown how much, if any, the increase in amphetamine use in pregnancy is attributable to prescription amphetamine use. However, in spite of this limitation, this misclassification error does not likely differ by rural and urban county of residence and our results mirror rural-urban differences in substance use during pregnancy in other published research.²² Fourth, data limitations precluded us from following an individual patient over time, so we were unable to study different substance use patterns in the same person over time. Fifth, we were unable to assess differences in trends between specific U.S. states because the NIS is not designed to produce state-specific estimates. Sixth, because the sample for this study is drawn from hospital discharge data, results should not be generalized to those women who delivery outside of a hospital setting. An estimated 1.4% of U.S. births occur outside of hospitals, with out-of-hospital births occurring more commonly among non-Hispanic white women.³⁷

Findings of the present study highlight the need to take a more holistic view of substance use in the US to effectively confront the evolving nature of the US opioid epidemic. Current public health efforts to expand treatment services for pregnant women with opioid use disorder should include system (i.e. comprehensive screening), healthcare and social services professional (i.e. education and training) and payer-level (i.e. reimbursement) initiatives that allow for the effective identification and management of polysubstance use during pregnancy. Despite the increases we observed, particularly in amphetamine use among pregnant women with opioid use disorder, remarkably little data are available on the effects of co-occurring substance use on maternal and neonatal health outcomes prompting the need for more clinical research. Finally, effective treatments for amphetamine and other stimulant use disorders are limited and research designed to develop addition pharmacologic and non-pharmacologic interventions is critically needed. Most people with a substance use disorder, including those who are pregnant, do not use substances in isolation.^5,17 As such, we must take a more comprehensive view of substance use behaviors, beyond a particular substance used, to fully understand the social, economic, medical and behavioral health interventions that are necessary to mitigate the current public health crisis.

Supplementary Material

Supplemental Digital Content_1

NIHMS1581649-supplement-Supplemental_Digital_Content_1.pdf^{(405.8KB, pdf)}

Supplemental Digital Content_2

NIHMS1581649-supplement-Supplemental_Digital_Content_2.pdf^{(181.5KB, pdf)}

Acknowledgements:

The authors thank Leah Klocke for project management support.

Funding Disclosure: Research reported in this publication was supported by the National Institute on Drug Abuse under Award Number R01DA045675 (Krans and Jarlenski).

Footnotes

Financial Disclosure

Elizabeth E. Krans reports that money was paid to her institution from Gilead and Merck. The other authors did not report any potential conflicts of interest.

Each author has confirmed compliance with the journal’s requirements for authorship.

Contributor Information

Marian P. Jarlenski, Department of Health Policy and Management, University of Pittsburgh, Pittsburgh, Pennsylvania.

Nicole C. Paul, University of Pittsburgh, Pittsburgh, Pennsylvania

Elizabeth E. Krans, Department of Obstetrics, Gynecology & Reproductive Sciences, Magee-Womens Research Institute, University of Pittsburgh, Pittsburgh, Pennsylvania.

REFERENCES

1.Reddy UM, Davis JM, Ren Z, Greene MF, Opioid Use in Pregnancy NAS, Childhood Outcomes Workshop Invited S. Opioid Use in Pregnancy, Neonatal Abstinence Syndrome, and Childhood Outcomes: Executive Summary of a Joint Workshop by the Eunice Kennedy Shriver National Institute of Child Health and Human Development, American College of Obstetricians and Gynecologists, American Academy of Pediatrics, Society for Maternal-Fetal Medicine, Centers for Disease Control and Prevention, and the March of Dimes Foundation. Obstetrics and gynecology. 2017;130(1):10–28. [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Cicero TJ, Ellis MS, Surratt HL, Kurtz SP. The changing face of heroin use in the United States: a retrospective analysis of the past 50 years. JAMA Psychiatry. 2014;71(7):821–826. [DOI] [PubMed] [Google Scholar]
3.Centers for Disease Control and Prevention. Understanding the Epidemic. https://www.cdc.gov/drugoverdose/epidemic/index.html. Published 2020. Accessed Jan 13, 2020.
4.Cicero TJ, Ellis MS, Kasper ZA. Polysubstance Use: A Broader Understanding of Substance Use During the Opioid Crisis. American journal of public health. 2020;110(2):244–250. [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Jarlenski M, Barry CL, Gollust S, Graves AJ, Kennedy-Hendricks A, Kozhimannil K. Polysubstance Use Among US Women of Reproductive Age Who Use Opioids for Nonmedical Reasons. American journal of public health. 2017:e1–e3. [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Behnke M, Smith VC, Committee on Substance Abuse, Committee on Fetus and Newborn. Prenatal substance abuse: short- and long-term effects on the exposed fetus. Pediatrics. 2013;131(3):e1009–1024. [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Healthcare Cost and Utilization Project. Overview of the National (Nationwide) Inpatient Sample (NIS). Agency for Healthcare Research and Quality; https://www.hcup-us.ahrq.gov/nisoverview.jsp. Published 2015. Accessed July 20, 2015. [Google Scholar]
8.Kuklina EV, Whiteman MK, Hillis SD, et al. An enhanced method for identifying obstetric deliveries: implications for estimating maternal morbidity. Matern Child Health J. 2008;12(4):469–477. [DOI] [PubMed] [Google Scholar]
9.Haight SC, Ko JY, Tong VT, Bohm MK, Callaghan WM. Opioid Use Disorder Documented at Delivery Hospitalization - United States, 1999–2014. MMWR Morb Mortal Wkly Rep. 2018;67(31):845–849. [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Ingram DD, Franco SJ. 2013 NCHS Urban-Rural Classification Scheme for Counties. Vital Health Stat 2. 2014(166):1–73. [PubMed] [Google Scholar]
11.Garcia MC, Rossen LM, Bastian B, et al. Potentially Excess Deaths from the Five Leading Causes of Death in Metropolitan and Nonmetropolitan Counties - United States, 2010–2017. Morbidity and mortality weekly report Surveillance summaries. 2019;68(10):1–11. [DOI] [PubMed] [Google Scholar]
12.Healthcare Cost and Utilization Project. Nationwide Inpatient Sample Redesign Report. https://www.hcup-us.ahrq.gov/db/nation/nis/reports/NISRedesignFinalReport040914.pdf. Published 2014. Accessed December 28, 2019.
13.Healthcare Cost and Utilization Project. Trend weights for HCUP NIS data. https://www.hcup-us.ahrq.gov/db/nation/nis/trendwghts.jsp. Published 2019. Accessed December 28, 2019.
14.Muller CJ, MacLehose RF. Estimating predicted probabilities from logistic regression: different methods correspond to different target populations. Int J Epidemiol. 2014;43(3):962–970. [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Krans EE, Kim JY, James AE 3rd, Kelley D, Jarlenski MP. Medication-Assisted Treatment Use Among Pregnant Women With Opioid Use Disorder. Obstetrics and gynecology. 2019;133(5):943–951. [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Krans EE, Campopiano M, Cleveland LM, et al. National Partnership for Maternal Safety: Consensus Bundle on Obstetric Care for Women With Opioid Use Disorder. Obstetrics and gynecology. 2019;134(2):365–375. [DOI] [PubMed] [Google Scholar]
17.Winkelman TNA, Villapiano N, Kozhimannil KB, Davis MM, Patrick SW. Incidence and Costs of Neonatal Abstinence Syndrome Among Infants With Medicaid: 2004–2014. Pediatrics. 2018;141(4). [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Coviello DM, Alterman AI, Rutherford MJ, Cacciola JS, McKay JR, Zanis DA. The effectiveness of two intensities of psychosocial treatment for cocaine dependence. Drug Alcohol Depend. 2001;61(2):145–154. [DOI] [PubMed] [Google Scholar]
19.Maude-Griffin PM, Hohenstein JM, Humfleet GL, Reilly PM, Tusel DJ, Hall SM. Superior efficacy of cognitive-behavioral therapy for urban crack cocaine abusers: main and matching effects. J Consult Clin Psychol. 1998;66(5):832–837. [DOI] [PubMed] [Google Scholar]
20.McLellan AT, Hagan TA, Meyers K, Randall M, Durell J. “Intensive” outpatient substance abuse treatment: comparisons with “traditional” outpatient treatment. J Addict Dis. 1997;16(2):57–84. [DOI] [PubMed] [Google Scholar]
21.Ellis MS, Kasper ZA, Cicero TJ. Twin epidemics: The surging rise of methamphetamine use in chronic opioid users. Drug Alcohol Depend. 2018;193:14–20. [DOI] [PubMed] [Google Scholar]
22.Admon LK, Bart G, Kozhimannil KB, Richardson CR, Dalton VK, Winkelman TNA. Amphetamine- and Opioid-Affected Births: Incidence, Outcomes, and Costs, United States, 2004–2015. American journal of public health. 2018:e1–e7. [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Cleary BJ, Donnelly JM, Strawbridge JD, et al. Methadone and perinatal outcomes: a retrospective cohort study. American journal of obstetrics and gynecology. 2011;204(2):139 e131–139. [DOI] [PubMed] [Google Scholar]
24.Cohen JM, Hernandez-Diaz S, Bateman BT, et al. Placental Complications Associated With Psychostimulant Use in Pregnancy. Obstetrics and gynecology. 2017;130(6):1192–1201. [DOI] [PMC free article] [PubMed] [Google Scholar]
25.In: The Health Consequences of Smoking-50 Years of Progress: A Report of the Surgeon General. Atlanta (GA) 2014. [Google Scholar]
26.Cnattingius S The epidemiology of smoking during pregnancy: smoking prevalence, maternal characteristics, and pregnancy outcomes. Nicotine Tob Res. 2004;6 Suppl 2:S125–140. [DOI] [PubMed] [Google Scholar]
27.Wright TE, Schuetter R, Tellei J, Sauvage L. Methamphetamines and pregnancy outcomes. Journal of addiction medicine. 2015;9(2):111–117. [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Kalaitzopoulos DR, Chatzistergiou K, Amylidi AL, Kokkinidis DG, Goulis DG. Effect of Methamphetamine Hydrochloride on Pregnancy Outcome: A Systematic Review and Meta-analysis. Journal of addiction medicine. 2018;12(3):220–226. [DOI] [PubMed] [Google Scholar]
29.Huybrechts KF, Bateman BT, Desai RJ, et al. Risk of neonatal drug withdrawal after intrauterine co-exposure to opioids and psychotropic medications: cohort study. BMJ. 2017;358:j3326. [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Jansson LM, Velez ML, McConnell K, et al. Maternal buprenorphine treatment and infant outcome. Drug Alcohol Depend. 2017;180:56–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Tong VT, Dietz PM, Morrow B, et al. Trends in smoking before, during, and after pregnancy--Pregnancy Risk Assessment Monitoring System, United States, 40 sites, 2000–2010. Morbidity and mortality weekly report Surveillance summaries. 2013;62(6):1–19. [PubMed] [Google Scholar]
32.Jones HE, Heil SH, Tuten M, et al. Cigarette smoking in opioid-dependent pregnant women: neonatal and maternal outcomes. Drug Alcohol Depend. 2013;131(3):271–277. [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Jones HE, Heil SH, O’Grady KE, et al. Smoking in pregnant women screened for an opioid agonist medication study compared to related pregnant and non-pregnant patient samples. Am J Drug Alcohol Abuse. 2009;35(5):375–380. [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Ondersma SJ, Chang G, Blake-Lamb T, et al. Accuracy of five self-report screening instruments for substance use in pregnancy. Addiction. 2019;114(9):1683–1693. [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Coleman-Cowger VH, Oga EA, Peters EN, Trocin KE, Koszowski B, Mark K. Accuracy of Three Screening Tools for Prenatal Substance Use. Obstetrics and gynecology. 2019;133(5):952–961. [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Johnson-Davis KL, Sadler AJ, Genzen JR. A Retrospective Analysis of Urine Drugs of Abuse Immunoassay True Positive Rates at a National Reference Laboratory. J Anal Toxicol. 2016;40(2):97–107. [DOI] [PubMed] [Google Scholar]
37.MacDorman MF, Matthews TJ, Declercq E. Trends in out-of-hospital births in the United States, 1990–2012. NCHS Data Brief. 2014(144):1–8. [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Digital Content_1

NIHMS1581649-supplement-Supplemental_Digital_Content_1.pdf^{(405.8KB, pdf)}

Supplemental Digital Content_2

NIHMS1581649-supplement-Supplemental_Digital_Content_2.pdf^{(181.5KB, pdf)}

[R1] 1.Reddy UM, Davis JM, Ren Z, Greene MF, Opioid Use in Pregnancy NAS, Childhood Outcomes Workshop Invited S. Opioid Use in Pregnancy, Neonatal Abstinence Syndrome, and Childhood Outcomes: Executive Summary of a Joint Workshop by the Eunice Kennedy Shriver National Institute of Child Health and Human Development, American College of Obstetricians and Gynecologists, American Academy of Pediatrics, Society for Maternal-Fetal Medicine, Centers for Disease Control and Prevention, and the March of Dimes Foundation. Obstetrics and gynecology. 2017;130(1):10–28. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R2] 2.Cicero TJ, Ellis MS, Surratt HL, Kurtz SP. The changing face of heroin use in the United States: a retrospective analysis of the past 50 years. JAMA Psychiatry. 2014;71(7):821–826. [DOI] [PubMed] [Google Scholar]

[R3] 3.Centers for Disease Control and Prevention. Understanding the Epidemic. https://www.cdc.gov/drugoverdose/epidemic/index.html. Published 2020. Accessed Jan 13, 2020.

[R4] 4.Cicero TJ, Ellis MS, Kasper ZA. Polysubstance Use: A Broader Understanding of Substance Use During the Opioid Crisis. American journal of public health. 2020;110(2):244–250. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R5] 5.Jarlenski M, Barry CL, Gollust S, Graves AJ, Kennedy-Hendricks A, Kozhimannil K. Polysubstance Use Among US Women of Reproductive Age Who Use Opioids for Nonmedical Reasons. American journal of public health. 2017:e1–e3. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R6] 6.Behnke M, Smith VC, Committee on Substance Abuse, Committee on Fetus and Newborn. Prenatal substance abuse: short- and long-term effects on the exposed fetus. Pediatrics. 2013;131(3):e1009–1024. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R7] 7.Healthcare Cost and Utilization Project. Overview of the National (Nationwide) Inpatient Sample (NIS). Agency for Healthcare Research and Quality; https://www.hcup-us.ahrq.gov/nisoverview.jsp. Published 2015. Accessed July 20, 2015. [Google Scholar]

[R8] 8.Kuklina EV, Whiteman MK, Hillis SD, et al. An enhanced method for identifying obstetric deliveries: implications for estimating maternal morbidity. Matern Child Health J. 2008;12(4):469–477. [DOI] [PubMed] [Google Scholar]

[R9] 9.Haight SC, Ko JY, Tong VT, Bohm MK, Callaghan WM. Opioid Use Disorder Documented at Delivery Hospitalization - United States, 1999–2014. MMWR Morb Mortal Wkly Rep. 2018;67(31):845–849. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R10] 10.Ingram DD, Franco SJ. 2013 NCHS Urban-Rural Classification Scheme for Counties. Vital Health Stat 2. 2014(166):1–73. [PubMed] [Google Scholar]

[R11] 11.Garcia MC, Rossen LM, Bastian B, et al. Potentially Excess Deaths from the Five Leading Causes of Death in Metropolitan and Nonmetropolitan Counties - United States, 2010–2017. Morbidity and mortality weekly report Surveillance summaries. 2019;68(10):1–11. [DOI] [PubMed] [Google Scholar]

[R12] 12.Healthcare Cost and Utilization Project. Nationwide Inpatient Sample Redesign Report. https://www.hcup-us.ahrq.gov/db/nation/nis/reports/NISRedesignFinalReport040914.pdf. Published 2014. Accessed December 28, 2019.

[R13] 13.Healthcare Cost and Utilization Project. Trend weights for HCUP NIS data. https://www.hcup-us.ahrq.gov/db/nation/nis/trendwghts.jsp. Published 2019. Accessed December 28, 2019.

[R14] 14.Muller CJ, MacLehose RF. Estimating predicted probabilities from logistic regression: different methods correspond to different target populations. Int J Epidemiol. 2014;43(3):962–970. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R15] 15.Krans EE, Kim JY, James AE 3rd, Kelley D, Jarlenski MP. Medication-Assisted Treatment Use Among Pregnant Women With Opioid Use Disorder. Obstetrics and gynecology. 2019;133(5):943–951. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R16] 16.Krans EE, Campopiano M, Cleveland LM, et al. National Partnership for Maternal Safety: Consensus Bundle on Obstetric Care for Women With Opioid Use Disorder. Obstetrics and gynecology. 2019;134(2):365–375. [DOI] [PubMed] [Google Scholar]

[R17] 17.Winkelman TNA, Villapiano N, Kozhimannil KB, Davis MM, Patrick SW. Incidence and Costs of Neonatal Abstinence Syndrome Among Infants With Medicaid: 2004–2014. Pediatrics. 2018;141(4). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R18] 18.Coviello DM, Alterman AI, Rutherford MJ, Cacciola JS, McKay JR, Zanis DA. The effectiveness of two intensities of psychosocial treatment for cocaine dependence. Drug Alcohol Depend. 2001;61(2):145–154. [DOI] [PubMed] [Google Scholar]

[R19] 19.Maude-Griffin PM, Hohenstein JM, Humfleet GL, Reilly PM, Tusel DJ, Hall SM. Superior efficacy of cognitive-behavioral therapy for urban crack cocaine abusers: main and matching effects. J Consult Clin Psychol. 1998;66(5):832–837. [DOI] [PubMed] [Google Scholar]

[R20] 20.McLellan AT, Hagan TA, Meyers K, Randall M, Durell J. “Intensive” outpatient substance abuse treatment: comparisons with “traditional” outpatient treatment. J Addict Dis. 1997;16(2):57–84. [DOI] [PubMed] [Google Scholar]

[R21] 21.Ellis MS, Kasper ZA, Cicero TJ. Twin epidemics: The surging rise of methamphetamine use in chronic opioid users. Drug Alcohol Depend. 2018;193:14–20. [DOI] [PubMed] [Google Scholar]

[R22] 22.Admon LK, Bart G, Kozhimannil KB, Richardson CR, Dalton VK, Winkelman TNA. Amphetamine- and Opioid-Affected Births: Incidence, Outcomes, and Costs, United States, 2004–2015. American journal of public health. 2018:e1–e7. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R23] 23.Cleary BJ, Donnelly JM, Strawbridge JD, et al. Methadone and perinatal outcomes: a retrospective cohort study. American journal of obstetrics and gynecology. 2011;204(2):139 e131–139. [DOI] [PubMed] [Google Scholar]

[R24] 24.Cohen JM, Hernandez-Diaz S, Bateman BT, et al. Placental Complications Associated With Psychostimulant Use in Pregnancy. Obstetrics and gynecology. 2017;130(6):1192–1201. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R25] 25.In: The Health Consequences of Smoking-50 Years of Progress: A Report of the Surgeon General. Atlanta (GA) 2014. [Google Scholar]

[R26] 26.Cnattingius S The epidemiology of smoking during pregnancy: smoking prevalence, maternal characteristics, and pregnancy outcomes. Nicotine Tob Res. 2004;6 Suppl 2:S125–140. [DOI] [PubMed] [Google Scholar]

[R27] 27.Wright TE, Schuetter R, Tellei J, Sauvage L. Methamphetamines and pregnancy outcomes. Journal of addiction medicine. 2015;9(2):111–117. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R28] 28.Kalaitzopoulos DR, Chatzistergiou K, Amylidi AL, Kokkinidis DG, Goulis DG. Effect of Methamphetamine Hydrochloride on Pregnancy Outcome: A Systematic Review and Meta-analysis. Journal of addiction medicine. 2018;12(3):220–226. [DOI] [PubMed] [Google Scholar]

[R29] 29.Huybrechts KF, Bateman BT, Desai RJ, et al. Risk of neonatal drug withdrawal after intrauterine co-exposure to opioids and psychotropic medications: cohort study. BMJ. 2017;358:j3326. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R30] 30.Jansson LM, Velez ML, McConnell K, et al. Maternal buprenorphine treatment and infant outcome. Drug Alcohol Depend. 2017;180:56–61. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R31] 31.Tong VT, Dietz PM, Morrow B, et al. Trends in smoking before, during, and after pregnancy--Pregnancy Risk Assessment Monitoring System, United States, 40 sites, 2000–2010. Morbidity and mortality weekly report Surveillance summaries. 2013;62(6):1–19. [PubMed] [Google Scholar]

[R32] 32.Jones HE, Heil SH, Tuten M, et al. Cigarette smoking in opioid-dependent pregnant women: neonatal and maternal outcomes. Drug Alcohol Depend. 2013;131(3):271–277. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R33] 33.Jones HE, Heil SH, O’Grady KE, et al. Smoking in pregnant women screened for an opioid agonist medication study compared to related pregnant and non-pregnant patient samples. Am J Drug Alcohol Abuse. 2009;35(5):375–380. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R34] 34.Ondersma SJ, Chang G, Blake-Lamb T, et al. Accuracy of five self-report screening instruments for substance use in pregnancy. Addiction. 2019;114(9):1683–1693. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R35] 35.Coleman-Cowger VH, Oga EA, Peters EN, Trocin KE, Koszowski B, Mark K. Accuracy of Three Screening Tools for Prenatal Substance Use. Obstetrics and gynecology. 2019;133(5):952–961. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R36] 36.Johnson-Davis KL, Sadler AJ, Genzen JR. A Retrospective Analysis of Urine Drugs of Abuse Immunoassay True Positive Rates at a National Reference Laboratory. J Anal Toxicol. 2016;40(2):97–107. [DOI] [PubMed] [Google Scholar]

[R37] 37.MacDorman MF, Matthews TJ, Declercq E. Trends in out-of-hospital births in the United States, 1990–2012. NCHS Data Brief. 2014(144):1–8. [PubMed] [Google Scholar]

PERMALINK

Polysubstance Use Among Pregnant Women With Opioid Use Disorder in the United States, 2007–2016

Marian P Jarlenski, PhD, MPH

Nicole C Paul

Elizabeth E Krans, MD, MSc