Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Apr 1.
Published in final edited form as: Prev Med. 2024 Feb 24;181:107914. doi: 10.1016/j.ypmed.2024.107914

Maternal Opioid Use Disorder and Infant Mortality in Wisconsin, United States, 2010–2018

David C Mallinson 1,*, Hsiang-Hui Daphne Kuo 2, Russell S Kirby 3, Yi Wang 4, Lawrence M Berger 5, Deborah B Ehrenthal 6,7
PMCID: PMC10947857  NIHMSID: NIHMS1970301  PMID: 38408650

Abstract

Objective

Differences in infant health outcomes by maternal opioid use disorder (OUD) status is understudied. We measured the association between maternal OUD during pregnancy and infant mortality and investigated whether these associations differ by infant neonatal opioid withdrawal syndrome (NOWS) or maternal receipt of medication for OUD (MOUD) during pregnancy.

Methods

We sampled 204,543 Medicaid-paid births from Wisconsin, United States (2010–2018). The primary exposure was any maternal OUD during pregnancy. We also stratified this exposure on NOWS diagnosis (no OUD; OUD without NOWS; OUD with NOWS) and on maternal MOUD receipt (no OUD; OUD without MOUD; OUD with <90 consecutive days of MOUD; OUD with 90+ consecutive days of MOUD). Our outcome was infant mortality (death at age <365 days). Demographic-adjusted logistic regressions measured associations with odds ratios (OR) and 95% confidence intervals (CI).

Results

Maternal OUD was associated with increased odds of infant mortality (OR 1.43; 95% CI 1.02–2.02). After excluding infants who died <5 days post-birth (i.e., before the clinical presentation of NOWS), regression estimates of infant mortality did not significantly differ by NOWS diagnosis. Likewise, regression estimates did not significantly differ by maternal MOUD receipt in the full sample.

Conclusions

Maternal OUD is associated with an elevated risk of infant mortality without evidence of modification by NOWS nor by maternal MOUD treatment. Future research should investigate potential mechanisms linking maternal OUD, NOWS, MOUD treatment, and infant mortality to better inform clinical intervention.

Keywords: Analgesics, opioid, infant mortality, Medicaid, neonatal opioid withdrawal syndrome, opioid-related disorders, pregnancy

INTRODUCTION

Over the past twenty-five years, opioid-related morbidity and mortality surged in the United States (U.S.) (Lyden and Binswanger, 2019), including diagnosed maternal opioid use disorder (OUD) during pregnancy (Desai et al., 2014; Krans and Patrick, 2016). Indeed, the national prevalence of maternal OUD escalated from 1.5 cases to 8.2 cases per 1000 delivery hospitalizations from 1999 to 2017 (Haight et al., 2018; Hirai et al., 2021). Maternal OUD is associated with adverse infant health outcomes, including preterm birth, low birth weight, respiratory complications, and neonatal opioid withdrawal syndrome (NOWS) (Tobon, Habecker and Forray, 2019). Because these conditions are risk factors for infant mortality (death within 365 days post-birth) (Almli et al., 2020; Behrman and Butler, 2007; Witt et al., 2017), maternal OUD may exacerbate such risk.

The link between maternal OUD and infant mortality has only recently been studied. In Ontario, Canada during 2002–2014, the infant mortality incidence among prenatally opioid-exposed infants was approximately three-times greater than that of Ontario’s full population of births (Brogly et al., 2017). A study of Texas Medicaid births spanning 2010–2014 found that maternal OUD was associated with a doubled odds of infant mortality, and this association was stronger for infants without a NOWS diagnosis (Leyenaar et al., 2021). The authors speculated that services provided at discharge to NOWS-diagnosed infants may prevent mortality. However, a subsequent Tennessee-based study found that the association between maternal OUD and postneonatal mortality (death between 28–365 days post-birth) did not notably vary by NOWS diagnosis (Grossarth et al., 2023). With clinical guidelines recommending medication for opioid use disorder (MOUD) to manage OUD during pregnancy (American College of Obstetricians and Gynecologists, 2017; Rodriguez and Klie, 2019; Substance Abuse and Mental Health Services Administration, 2018), a study of NOWS-diagnosed infants born in North Carolina during 2016–2018 found that infant mortality was slightly lower with maternal MOUD receipt (1.0%) relative to no treatment (1.3%) (Austin et al., 2022).

Prior work is limited in several dimensions. The Ontario and North Carolina studies described differences in infant mortality in relatively small cohorts (<5,000 infants) without adjustment for demographic confounders of opioid use and infant mortality (Austin et al., 2022; Brogly et al., 2017; Mohamoud, Kirby and Ehrenthal, 2019; Nguyen et al., 2023). Additionally, differential survival time may explain heterogeneity in the association between maternal OUD and infant mortality by NOWS diagnosis, as infants may die before presenting NOWS symptoms (Jansson, Velez and Harrow, 2009). To address these gaps, our study investigates associations between maternal OUD and infant mortality using data from Wisconsin, U.S. during 2010–2018. Further, we examine whether associations vary by NOWS diagnosis and by MOUD receipt during pregnancy.

METHODS

Data and Sample

We analyzed data from the Big Data for Little Kids (BD4LK) project. BD4LK generates a population-based longitudinal birth cohort from Wisconsin, U.S. by linking several data sources through the Wisconsin Administrative Data Core (Institute for Research on Poverty, 2022): birth records for live, in-state, resident deliveries; infant death records; Medicaid claims for mothers and infants; and Medicaid enrollment files. Records include unique maternal and infant identifiers to generate family clusters. Data linkage for earlier BD4LK waves has been described elsewhere (Larson et al., 2019; Mallinson and Ehrenthal, 2019).

We sampled infants born in 2010–2018 who met inclusion criteria: Medicaid delivery coverage; gestational age 22+ weeks (i.e., within the period of viability (Patel et al., 2017)); mothers had 5+ months of continuous Medicaid coverage immediately before delivery; and complete information on analyzed variables. Without other insurance data, these restrictions ensured that we could observe claims for most of a pregnancy’s duration. We also required sampled infants to have complete information on maternal characteristics (age; race/ethnicity; marital status; education; resident county; cigarette smoking), birth order, plurality, and the Kotelchuck Index of prenatal care utilization from birth records, which were regression covariates. From 237,622 Medicaid-paid deliveries (2010–2018), we sequentially excluded records on gestational age (N=269), Medicaid coverage during pregnancy (N=25,687), and data missingness (N=7,123), yielding a sample of 204,543 infants (full sampling procedure in Figure A.1, Appendix). Table A.1 (Appendix) lists the billing codes to identify Medicaid-paid deliveries. In 74.1% of sampled infants, mothers had full Medicaid coverage throughout pregnancy (Figures A.2 and A.3, Appendix). The University of Wisconsin-Madison and The Pennsylvania State University institutional review boards approved this project.

Measures

Medicaid claims captured opioid medications and diagnoses. Table A.1 (Appendix) lists the International Classification of Diseases codes for OUD and NOWS diagnoses. We assessed maternal OUD during the prenatal period, defined as the duration from the estimated date of conception (measured by subtracting the gestational age in completed weeks from the birthdate—both from birth records) and the birthdate. Additionally, we identified NOWS diagnosis within 30 days of the birthdate. The United States Food and Drug Administration approves buprenorphine, methadone, and naltrexone for MOUD (Oesterle et al., 2019; Centers for Medicare and Medicaid Services, 2021). We identified MOUDs with generic drug names in the Medicaid pharmaceutical data.

Our primary dichotomous exposure variable was maternal OUD. We coded the infant as maternal OUD-exposed if there was an OUD diagnosis claim or an MOUD prescription fill for the mother during the prenatal period. Consistent with prior work (Leyenaar et al., 2021), we assumed that individuals only received MOUD if they had a clinical diagnosis of OUD (American College of Obstetricians and Gynecologists, 2017; Rodriguez and Klie, 2019; Substance Abuse and Mental Health Services Administration, 2018).

We also constructed OUD-NOWS, an exposure variable that stratified maternal OUD on NOWS diagnosis within 30 days of birth (no OUD; OUD without NOWS; OUD with NOWS). We identified NOWS with ICD codes (Table A.1, Appendix). Pharmacologic and non-pharmacologic treatment for NOWS may improve infant health and prevent infant mortality (Mangat, Schmölzer and Kraft, 2019; Weller et al., 2021). Stratifying maternal OUD on NOWS allowed us to examine whether survival differed between these groups. However, NOWS symptoms typically appear after 24–48 hours of life, so only infants who survive at least 2–5 days typically receive NOWS diagnoses (Jansson, Velez and Harrow, 2009). Therefore, our primary OUD-NOWS analyses included only infants who survived 5+ days (N=204,335 infants). Supplemental analyses with the OUD-NOWS exposure included infants who survived 2+ days (N=204,383 infants), representing a lower-bound survival time necessary for NOWS screening.

Finally, we constructed OUD-MOUD, an exposure variable that stratified maternal OUD on MOUD receipt during the prenatal period (no OUD; OUD without MOUD; OUD with MOUD (<90 consecutive days of treatment); OUD with MOUD (90+ consecutive days of treatment)). We stratified MOUD exposure length on this duration because 90 days of continuous treatment is associated with improved health outcomes and abstinence from illicit opioid use (McCarthy et al., 2015; Substance Abuse and Mental Health Services Administration, 2023a). Current clinical guidelines support MOUD treatment for OUD in pregnancy to prevent relapse and nonmedical opioid use (American College of Obstetricians and Gynecologists, 2017; Substance Abuse and Mental Health Services Administration, 2018). Yet, there are important differences across MOUD regimens. Both buprenorphine (a partial opioid receptor agonist) and methadone (a full opioid receptor agonist) can be administered daily through a registered opioid treatment program (American College of Obstetricians and Gynecologists, 2017; Kinsella et al., 2022; Substance Abuse and Mental Health Services Administration, 2018). Stable treatment provision may prevent misuse and adverse outcomes from self-administration—particularly if dosage requires adjustment—and improve prenatal care engagement (American College of Obstetricians and Gynecologists, 2017; Harris et al., 2023; Kinsella et al., 2022; Mozurkewich and Rayburn, 2014; Rodriguez and Klie, 2019; Substance Abuse and Mental Health Services Administration, 2018). Buprenorphine prescriptions were available in outpatient offices with a Drug Addiction Treatment Act of 2000 waiver (the waiver requirement existed during the study period but was since removed) (Substance Abuse and Mental Health Services Administration, 2023b). Compared to methadone, buprenorphine is associated with fewer drug interactions, a lower risk of overdose, and a lower risk of adverse neonatal outcomes (Nørgaard, Schou Nielsson and Heide-Jørgensen, 2015; Saia et al., 2016; Suarez et al., 2022). Naltrexone (an opioid receptor antagonist) is not currently recommended for treating maternal OUD due to limited knowledge of its effects on pregnancy health outcomes, although it is sometimes used (American College of Obstetricians and Gynecologists, 2017; Harris et al., 2023; Krans et al., 2019; Saia et al., 2016; Substance Abuse and Mental Health Services Administration, 2018). Stratifying OUD on MOUD treatment allowed us to determine if MOUD receipt is associated with improved survival of opioid-exposed infants.

Our outcome variable was infant mortality (death at age <365 days [1 year]). Among infant mortalities, we calculated survival time in days by subtracting the birthdate from the date of death. Additional variables abstracted from birth records included maternal characteristics (age; race/ethnicity; education; marital status; resident county urbanicity using the National Center for Health Statistics urban-rural classification scheme (Centers for Disease Control and Prevention, 2017); cigarette smoking during pregnancy) and infant characteristics (first-born status; plurality; gestational age in completed weeks; birth weight in grams). We also constructed a maternal indicator for using any mental health medications (anticonvulsants; psychotherapeutic agents; anxiolytics, sedatives, or hypnotics) during pregnancy, as maternal mental health is associated with OUD and infant mortality (American College of Obstetricians and Gynecologists, 2017; Ban et al., 2012; Rodriguez and Klie, 2019; Substance Abuse and Mental Health Services Administration, 2018; Zhong et al., 2018). We identified medications in Medicaid claims with American Hospital Formulary Service codes (Table A.2, Appendix).

Statistical Analysis

We tabulated sample characteristics and mortality incidences overall and by all maternal OUD exposures. To explore temporal patterns in OUD and infant mortality, we graphed Kaplan-Meier survival plots of mortality incidence within 365 days of birth by maternal OUD exposures and tabulated time to mortality by maternal OUD exposures among infants who died. In our main analyses, mortality incidences and Kaplan-Meier plots by OUD-NOWS exposure were estimated using only infants who survived 5+ days because these infants would have survived long enough to exhibit the signs and symptoms of NOWS (Jansson, Velez and Harrow, 2009). In supplemental descriptive analyses, we report mortality incidences and Kaplan-Meier plots by OUD-NOWS among infants who survived 2+ days. Among mortalities, we tabulated causes of death with 5 categories: congenital malformations; sudden unexpected death in infancy (SUDI); perinatal conditions; all other causes; or missing (Bairoliya and Fink, 2018).

Our primary logistic regressions estimated the association between each exposure (OUD; OUD-NOWS; OUD-MOUD) and infant mortality, in separate models, and we report odds ratios (OR) and corresponding 95% confidence intervals (CI). No OUD was the reference group for each model. We estimated unadjusted regressions and regressions that were adjusted for maternal characteristics (age; race/ethnicity; marital status; education; resident county urbanicity; cigarette smoking; mental health medication receipt), the infant’s first-born status and plurality, and the Kotelchuck Index of prenatal care utilization. We hypothesized that adjusted regressions would limit bias by controlling for potential confounders of maternal OUD’s impact on infant mortality (American College of Obstetricians and Gynecologists, 2017; Ban et al., 2012; Leyenaar et al., 2021; Nguyen et al., 2023; Mohamoud, Kirby and Ehrenthal, 2019; Rodriguez and Klie, 2019; Substance Abuse and Mental Health Services Administration, 2018; Zhong et al., 2018). Wald tests evaluated whether estimates significantly differed by NOWS or MOUD status among opioid-exposed births (StataCorp, 2023a). We clustered the standard errors in all models at the mother-level because some mothers had multiple deliveries during the observation period. For sensitivity analyses, we computed E-values following adjusted logistic regressions to determine the necessary minimum strength of an unmeasured confounder to explain estimated OUD-mortality associations (Haneuse, VanderWeele and Arterburn, 2019; Linden, Mathur and VanderWeele, 2020).

We conducted four supplementary analyses. First, we repeated all adjusted regressions while controlling for gestational age and birth weight. Attenuated estimates would signal that adverse birth outcomes—specifically, preterm birth or low birth weight—may partially explain the relationship between maternal OUD and infant mortality (American College of Obstetricians and Gynecologists, 2017; Rodriguez and Klie, 2019; Substance Abuse and Mental Health Services Administration, 2018). Second, we repeated all adjusted regressions after including all infants with missing information on the Kotelchuck Index of prenatal care utilization (N=209,637 infants). Of the 7,123 infants who were excluded from the primary analytic sample for missing information on covariates, 5,094 infants (71.5%) exclusively had missing information on the Kotelchuck Index. We repeated regressions after including these infants to determine if their exclusion notably affected estimates. Further, we did not adjust for the Kotelchuck Index in these regressions. Third, we repeated the adjusted OUD-NOWS regressions after expanding the sample to include infants who survived 2+ days (a lower-bound survival time during which NOWS may present). This test demonstrates whether OUD-NOWS estimates depend on excluding infants who may have died before NOWS screening. Lastly, we repeated adjusted OUD-NOWS regressions and only counted NOWS diagnoses that occurred within 5 days of birth. We captured NOWS claims within 30 days of delivery for maternal OUD-exposed infants, although NOWS diagnoses generally occur within the first 5 days of life (Jansson, Velez and Harrow, 2009). Indeed, of the 2,327 infants in the OUD with NOWS group, 156 infants (6.7%) received diagnoses beyond the first 5 days of life. It is possible that some infants who died between 5–30 days following birth would have received NOWS diagnoses but died beforehand. We recoded these infants to the OUD without NOWS group to determine if our 30-day span for capturing NOWS diagnoses influenced regression estimates.

We conducted all analyses in Stata Statistical Software: Release 18 (StataCorp, 2023b). Statistical tests were two-sided. We considered P <0.05 to be statistically significant.

RESULTS

Across all 204,543 infants, 4,672 (2.3%) were exposed to maternal OUD (Table 1). Of infants in the any maternal OUD group, approximately half had NOWS and about two-thirds had a mother who received MOUD during the pregnancy, with a majority having 90+ days of MOUD fills. In most MOUD-exposed cases, mothers used buprenorphine (47.0%) or methadone (44.2%) exclusively, while few used naltrexone exclusively (2.4%) or had mixed MOUD use (6.4%). MOUD-type estimates mirror prior studies of MOUD use among Medicaid pregnancies (Krans et al., 2019). Maternal OUD diagnosis was associated with the mother being non-Hispanic white, unmarried, using cigarettes, or receiving mental health medications.

Table 1.

Descriptive statistics for the full sample and by maternal opioid use disorder status, Medicaid-covered deliveries in Wisconsin, United States, 2010–2018 (N=204,543 infants)

ANY OUD OUD-NOWS OUD-MOUDa
VARIABLE OVERALL No Yes OUD w/o NOWS OUD w/ NOWS OUD w/o MOUD OUD w/MOUD (<90d) OUD w/MOUD (90+d)
BIRTHS, N (% OF SAMPLE) 204,543 (100) 199,871 (97.72) 4,672 (2.28) 2,345 (1.15) 2,327 (1.14) 1,408 (0.69) 937 (0.46) 2,327 (1.14)
MATERNAL CHARACTERISTICS
Age (years), mean (SD) 26.35 (5.54) 26.31 (5.55) 28.32 (4.71) 28.38 (4.79) 28.27 (4.64) 28.44 (5.17) 28.20 (4.61) 28.31 (4.46)
Race/ethnicity, N (%)
 American Indian/Alaska Native (NH) 4,099 (2.00) 3,791 (1.90) 308 (6.59) 155 (6.61) 153 (6.57) 74 (5.26) 65 (6.94) 169 (7.26)
 Asian/ Pacific Islander (NH) 9,877 (4.83) 9,870 (4.94) [Suppressed] [Suppressed] [Suppressed] [Suppressed] [Suppressed] [Suppressed]
 Black (NH) 39,532 (19.33) 39,182 (19.60) 350 (7.49) 231 (9.85) 119 (5.11) 192 (13.64) 75 (8.00) 83 (3.57)
 Hispanic 33,311 (16.29) 32,973 (16.50) 338 (7.23) 150 (6.40) 188 (8.08) 93 (6.61) 80 (8.54) 165 (7.09)
 White (NH) 111,055 (54.29) 107,553 (53.81) 3,502 (74.96) 1,715 (73.13) 1,787 (76.79) 976 (69.32) 682 (72.79) 1,844 (79.24)
 Multiple race (NH) 5,560 (2.72) 5,396 (2.70) [Suppressed] [Suppressed] [Suppressed] [Suppressed] [Suppressed] [Suppressed]
 Other (NH) 1,109 (0.54) 1,106 (0.55) [Suppressed] [Suppressed] [Suppressed] [Suppressed] [Suppressed] [Suppressed]
Education, N (%)
 No HS degree 40,857 (19.97) 39,960 (19.99) 897 (19.20) 413 (17.61) 484 (20.80) 268 (19.03) 177 (18.89) 452 (19.42)
 HS degree/equivalent 84,791 (41.45) 82,784 (41.42) 2,007 (42.96) 1,031 (43.97) 976 (41.94) 624 (44.32) 406 (43.33) 977 (41.99)
 Some college or greater 78,895 (38.57) 77,127 (38.59) 1,768 (27.84) 901 (38.42) 867 (37.26) 516 (36.65) 354 (37.78) 898 (38.59)
Currently married, N (%)
 No 138,388 (67.66) 134,500 (67.29) 3,888 (83.22) 1,920 (81.88) 1,968 (84.57) 1,181 (83.88) 778 (83.03) 1,929 (82.90)
 Yes 66,155 (32.34) 65,371 (32.71) 784 (16.78) 425 (18.12) 359 (15.43) 227 (16.12) 159 (16.97) 398 (17.10)
NCHS classification of residence county, N (%)
 Large central metro 59,707 (29.19) 58,515 (29.28) 1,192 (25.51) 563 (24.01) 629 (27.03) 390 (27.70) 260 (27.75) 542 (23.29)
 Large fringe metro 17,292 (8.45) 16,814 (8.41) 478 (10.23) 245 (10.45) 233 (10.01) 161 (11.43) 81 (8.64) 236 (10.14)
 Medium metro 26,680 (13.04) 26,062 (13.04) 618 (13.23) 249 (10.62) 369 (15.86) 166 (11.79) 122 (13.02) 330 (14.18)
 Small metro 49,086 (24.00) 48,047 (24.04) 1,039 (22.24) 538 (22.94) 501 (21.53) 294 (20.88) 187 (19.96) 558 (14.18)
 Micropolitan 25,160 (12.30) 24,521 (12.27) 639 (13.68) 332 (14.16) 307 (13.19) 186 (13.21) 140 (14.94) 313 (13.45)
 Non-core 26,618 (13.01) 25,912 (12.96) 706 (15.11) 418 (1.83) 288 (12.38) 211 (14.99) 147 (15.69) 348 (14.95)
Cigarette smoking during pregnancy, N (%)b
 Not reported 147,006 (71.87) 145,950 (73.02) 1,056 (22.60) 624 (26.61) 432 (18.56) 382 (27.13) 238 (25.40) 436 (18.74)
 Reported 57,537 (28.13) 53,921 (26.98) 3,616 (77.40) 1,721 (73.39) 1,895 (81.44) 1,026 (72.87) 699 (74.60) 1,891 (81.26)
Mental health medications, N (%)c
 None 172,922 (84,57) 170,713 (85.41) 2,279 (48.78) 1,122 (47.85) 1,157 (49.72) 668 (47.44) 441 (47.07) 1,170 (50.28)
 Any 31,551 (15.43) 29,158 (14.59) 2,393 (51.22) 1,223 (52.15) 1,170 (50.28) 740 (52.56) 496 (52.93) 1,157 (49.72)
Kotelchuck Index, N (%)
 Adequate plus 17,301 (8.46) 12\6,389 (8.20) 912 (19.52) 443 (18.89) 469 (20.15) 317 (22.51) 215 (22.95) 380 (16.33)
 Adequate 21,888 (10.70) 21,159 (10.59) 729 (15.60) 340 (14.50) 389 (16.72) 206 (14.63) 170 (18.14) 353 (15.17)
 Intermediate 34,666 (16.95) 33,932 (16.98) 734 (15.71) 390 (16.63) 344 (14.78) 210 (14.91) 141 (15.05) 383 (16.46)
 Inadequate 129,472 (63.30) 127,257 (63.67) 2,215 (47.41) 1,144 (48.78) 1,071 (46.02) 654 (46.45) 383 (40.88) 1,178 (50.62)
 No prenatal care 1,216 (0.59) 1,134 (0.57) 82 (1.76) 28 (1.19) 54 (2.32) 21 (1.49) 28 (2.99) 33 (1.42)
INFANT CHARACTERISTICS
First-born, N (%)
 No 136,072 (66.52) 132,449 (66.27) 3,623 (77.55) 1,858 (79.23) 1,765 (75.85) 1,065 (75.64) 720 (76.84) 1,838 (78.99)
 Yes 68,471 (33.48) 67,422 (33.73) 1,049 (22.45) 487 (20.77) 562 (24.15) 343 (24.36) 217 (23.16) 489 (21.01)
Plurality, N (%)
 Singleton 198,923 (97.25) 194,393 (97.26) 4,530 (96.96) 2,249 (95.91) 2,281 (98.02) 1,344 (95.45) 915 (97.65) 2,271 (97.59)
 Plural 5,620 (2.75) 5,478 (2.74) 142 (3.04) 96 (4.09) 46 (1.98) 64 (4.55) 22 (2.35) 56 (2.41)
Open in a new tab
a

“OUD w/ MOUD (<90d)” indicates that the mother had <90 consecutive days of MOUD receipt during pregnancy. “OUD w/ MOUD (90+d)” indicates that the mother had 90+ consecutive days of MOUD receipt during pregnancy.

b

For maternal cigarette smoking, 2010 birth records include an indicator for self-reported smoking at any point, whereas 2011–2018 birth records include indicators for self-reported pre-pregnancy smoking and self-reported smoking at each trimester of the pregnancy. Measures were combined to create an indicator for any reported cigarette smoking.

c

Mental health medications during pregnancy include anticonvulsants, psychotherapeutic agents, anxiolytics, sedatives, and hypnotics.

Notes: A mother was coded as exposed to opioid use disorder if they had an opioid use disorder diagnosis claim during pregnancy or if they received a claim for medication for opioid use during pregnancy. Additionally, an infant was coded as having neonatal opioid withdrawal syndrome if they had a claim for the diagnosis following birth. We suppressed cells with small sample sizes.

Abbreviations: “HS” high school, “metro” metropolitan, “MOUD” medication for opioid use disorder, “NH” non-Hispanic, “NOWS” neonatal opioid withdrawal syndrome, “OUD” opioid use disorder, “w/” with, “w/o” without.

The infant mortality rate among maternal OUD-exposed infants was 8.8 per 1000 births compared to 4.2 per 1000 births among mothers without OUD (Table 2). Among maternal OUD-exposed infants, the infant mortality incidence rose incrementally with greater levels of MOUD exposure, but—after restricting the sample to infants who survived 5+ days—infant mortality was balanced regardless of NOWS diagnosis (~8 per 1000 births). NOWS-stratified metrics remained stable with a less restrictive sample of infants who survived 2+ days (Table A.3, Appendix). SUDI was the most common cause of death among non-OUD infants (38.4%) and any OUD infants (63.4%), regardless of NOWS or MOUD status (Table A.4, Appendix).

Table 2.

Incidence of mortality outcomes, overall and by maternal opioid use disorder status, Medicaid-covered deliveries in Wisconsin, United States, 2010–2018 (N=204,543 infants)

SAMPLE POPULATION BIRTHS, N INFANT MORTALITY, N (DEATHS/1000 BIRTHS) SURVIVAL TIME AMONG DEATHS (DAYS), MEAN (SD)
Overall 204,543 884 (4.32) 69.21 (77.88)
Any OUD
 No 199,871 842 (4.21) 68.33 (78.06)
 Yes 4,672 41 (8.78) 87.29 (72.67)
OUD-NOWS a
 OUD w/o NOWS 2,342 19 (8.11) 79.11 (75.17)
 OUD w/ NOWS 2,327 19 (8.17) 109.26 (65.06)
OUD-MOUD b
 OUD w/o MOUD 1,408 11 (7.81) 108.18 (72.49)
 OUD w/ MOUD (<90d) 937 8 (8.54) 85.63 (75.94)
 OUD w/ MOUD (90+d) 2,327 22 (9.45) 77.45 (72.82)
Open in a new tab
a

The OUD-NOWS crude mortality measures include only infants who survived 5+ days (204,335 infants overall; 4,669 infants with any OUD exposure), thus representing infants who survived the period when infants may exhibit symptoms of neonatal opioid withdrawal syndrome.

b

“OUD w/ MOUD (<90d)” indicates that the mother had <90 consecutive days of MOUD receipt during pregnancy. “OUD w/ MOUD (90+d)” indicates that the mother had 90+ consecutive days of MOUD receipt during pregnancy.

Notes: A mother was coded as exposed to opioid use disorder if they had an opioid use disorder diagnosis claim during pregnancy or if they received a claim for medication for opioid use during pregnancy. Additionally, an infant was coded as having neonatal opioid withdrawal syndrome if they had a claim for the diagnosis following live birth.

Abbreviations: “MOUD” medication for opioid use disorder, “NOWS” neonatal opioid withdrawal syndrome, “OUD” opioid use disorder, “SD” standard deviation, “w/” with, “w/o” without.

Kaplan-Meier plots showed that survival was consistently lower for maternal OUD-exposed infants than for unexposed infants (Figure 1). Among infants who lived 5+ days, survival within 120 days post-birth was lower for infants with maternal OUD and NOWS compared to those with maternal OUD but no NOWS, although survival was more equal for the remaining duration of the first year of life. Including infants who survived 2+ days did not notably alter survival by NOWS diagnoses among OUD-exposed infants (Figure A.4, Appendix). Among maternal OUD-exposed infants, survival was incrementally lower with increasing levels of MOUD exposure. Log-rank tests consistently demonstrated that the survival curves significantly differed between maternal OUD-exposed and unexposed infants, regardless of NOWS or MOUD status (Tables A.5 and A.6, Appendix).

Figure 1.

Figure 1.

Open in a new tab

Kaplan-Meier survival plots of infant mortality by maternal opioid use disorder status, Medicaid-covered deliveries in Wisconsin, United States, 2010–2018 (N=204,543 infants)

Notes: A mother was coded as exposed to opioid use disorder if they had an opioid use disorder diagnosis claim during pregnancy or if they received a claim for medication for opioid use during pregnancy. Additionally, an infant was coded as having neonatal opioid withdrawal syndrome if they had a claim for the diagnosis following live birth. OUD-NOWS plot only includes infants who survived 5 days, thus representing infants who survived the period when infants may exhibit symptoms of neonatal opioid withdrawal syndrome. “OUD w/ MOUD (<90d)” indicates that the mother had <90 consecutive days of MOUD receipt during pregnancy. “OUD w/ MOUD (90+d)” indicates that the mother had 90+ consecutive days of MOUD receipt during pregnancy. Abbreviations: “MOUD” medication for opioid use disorder, “NOWS” neonatal opioid withdrawal syndrome, “OUD” opioid use disorder, “w/” with, “w/o” without.

Table 3 displays results from our primary logistic regressions. In adjusted regressions, maternal OUD was associated with a nearly 50-percent greater odds of infant mortality (OR 1.4; 95% CI 1.0–2.0). Among maternal OUD-exposed infants who survived 5+ days, Wald tests indicated that the odds of infant mortality did not significantly differ between those with (OR 1.7; 95% CI 1–0 2.8) and without (OR 1.7; 95% CI 1.1–2.8) a NOWS diagnosis. Compared to no OUD exposure, OUD with 90+ days of MOUD receipt during pregnancy was associated with a higher odds of infant mortality (OR 1.6; 95% CI 1.0–2.6), whereas OUD without MOUD or OUD with 1–90 days of MOUD were not associated with infant mortality. However, Wald tests indicated that associations did not significantly differ across levels of MOUD among OUD-exposed infants. Unadjusted regressions yielded larger point estimates, although Wald tests still indicated no significant difference in the association between OUD and infant mortality by NOWS diagnosis or MOUD intensity among OUD-exposed infants. E-values indicated that an unmeasured confounder would have to be associated with the exposure and outcome by an odds ratio of 2.2- to 2.9-fold each to explain away significant OUD-mortality associations (Table A.7, Appendix) (Linden, Mathur and VanderWeele, 2020). For context, prior research suggests that maternal smoking—a covariate in our regressions—has similar or weaker associations with infant mortality outcomes (Anderson et al., 2019; Sun et al., 2023). Thus, our estimates are likely robust to unmeasured confounding.

Table 3.

Logistic regression results of associations of opioid use disorder, NOWS, and MOUD statuses with infant mortality (death within 365 days post-birth), Medicaid-covered deliveries in Wisconsin, United States, 2010–2018 (N=204,543 infants)

UNADJUSTED ADJUSTEDa
MODEL OR 95% CI P Value OR 95% CI P Value
Any OUD
 No 1.00 -- -- 1.00 -- --
 Yes 2.09 1.52–2.88 <0.001 1.43 1.02–2.02 0.04
OUD-NOWS b
 No OUD 1.00 -- -- 1.00 -- --
 OUD w/o NOWS 2.55 1.61–4.03 <0.001 1.73 1.07–2.79 0.03
 OUD w/NOWS 2.57 1.59–4.15 <0.001 1.71 1.04–2.81 0.04
 Wald testc -- -- 0.98 -- -- 0.97
OUD-MOUD d
 No OUD 1.00 -- -- 1.00 -- --
 OUD w/o MOUD 1.86 1.02–3.38 0.04 1.20 0.65–2.21 0.57
 OUD w/MOUD (<90d) 2.03 1.01–4.09 0.05 1.36 0.66–2.77 0.41
 OUD w/MOUD (90+d) 2.25 1.45–3.51 <0.001 1.63 1.03–2.57 0.04
 Wald testc -- -- 0.87 -- -- 0.70
Open in a new tab

Notes: Standard errors were clustered at the mother level. A mother was coded as exposed to opioid use disorder if they had an opioid use disorder diagnosis claim during pregnancy or if they received a claim for medication for opioid use during pregnancy. Additionally, an infant was coded as having neonatal opioid withdrawal syndrome if they had a claim for the diagnosis following live birth.

a

Included covariates for maternal characteristics at delivery (age; race/ethnicity; education; marital status; urbanicity of residence county; any cigarette smoking; mental health medication receipt), for the Kotelchuck Index of prenatal care utilization, and for the infant’s first-born status and plurality.

b

The OUD-NOWS analysis includes only infants who survived 5 days (N=204,335 infants), thus representing infants who survived the period when infants may exhibit symptoms of neonatal opioid withdrawal syndrome.

c

Wald tests evaluated the equality of coefficients across strata of NOWS or MOUD exposure among maternal OUD-exposed infants. P <0.05 was used to determine that coefficients significantly differed.

d

“OUD w/ MOUD (<90d)” indicates that the mother had <90 consecutive days of MOUD receipt during pregnancy. “OUD w/ MOUD (90+d)” indicates that the mother had 90+ consecutive days of MOUD receipt during pregnancy.

Abbreviations: “CI” confidence interval, “MOUD” medication for opioid use disorder, “NOWS” neonatal opioid withdrawal syndrome, “OR” odds ratio, “OUD” opioid use disorder, “w/” with, “w/o” without.

Controlling for gestational age and low birth weight attenuated point estimates, suggesting that birth outcomes explain part of the association between maternal OUD and infant mortality (Table A.8, Appendix). Additionally, regression estimates modestly increased after including infants with missing information on the Kotelchuck Index (Table A.9, Appendix), although the pattern of the associations between OUD and infant mortality were consistent. Focusing on adjusted regressions with the OUD-NOWS exposure group, results were generally consistent after including infants who survived 2+ days (Table A.10, Appendix). Only capturing NOWS diagnoses within the first 5 days widened the absolute differences in ORs between the OUD without NOWS group and the OUD with NOWS group (Table A.11, Appendix). Nonetheless, Wald tests still indicated that the association between maternal OUD and infant mortality did not significantly vary by NOWS diagnosis.

DISCUSSION

Maternal OUD during pregnancy may elevate the risk of adverse obstetric and perinatal outcomes (Behrman and Butler, 2007; Tobon, Habecker and Forray, 2019; Witt et al., 2017), including infant mortality (Brogly et al., 2017; Leyenaar et al., 2021). Whether the relation between maternal OUD and infant mortality differs by NOWS diagnosis or by MOUD receipt during pregnancy is unclear. This study investigated the association between maternal OUD and infant mortality and its variation by NOWS and MOUD status.

Consistent with prior research, maternal OUD was associated with a higher odds of infant mortality. However, we found no evidence of significant differential association by NOWS diagnosis. A study of Texas Medicaid births found that, after adjusting for maternal sociodemographic and health characteristics and excluding infants who survived <5 days, the association between maternal OUD and infant mortality was qualitatively greater for infants without a NOWS diagnosis (OR 2.31) than for infants with a NOWS diagnosis (OR 1.73) (Leyenaar et al., 2021). The authors suggested that additional services and support provided to infants identified with NOWS may prevent mortality among opioid-exposed infants. However, the study did not test whether those estimates differed statistically. Among OUD-exposed infants in our sample, the unadjusted average time to mortality was longer among NOWS-diagnosed infants relative to undiagnosed infants. Independently, this could favor the Texas study’s hypothesis. Yet, we found no statistically significant evidence that survival in the first year of life differs by NOWS diagnosis among maternal OUD-exposed infants. Consequently, our results do not suggest that additional support for NOWS-diagnosed infants improved survival.

We also found no evidence that the association between maternal OUD and infant mortality varied by MOUD receipt during pregnancy, including between low (<90 consecutive days of treatment) and high levels of MOUD receipt (90+ consecutive days of treatment). It is recommended that OUD-diagnosed pregnant individuals receive MOUD to prevent withdrawal or relapse of nonmedical opioid use (American College of Obstetricians and Gynecologists, 2017; Substance Abuse and Mental Health Services Administration, 2018). Thus, we hypothesized MOUD treatment would improve survival. A recent descriptive study of NOWS-diagnosed infants born in North Carolina during 2016–2018 found that the incidence of infant mortality was slightly lower if the mother received MOUD treatment during pregnancy (1.0%) than if she did not (1.3%) (Austin et al., 2022). The authors did not indicate whether these incidences significantly differed, and estimates were unadjusted for confounders. While our results do not align with the North Carolina study, they are not contradictory. Among maternal OUD-exposed infants in our cohort, the unadjusted incidence of infant mortality incrementally and modestly increased with greater levels of maternal MOUD receipt. However, the association between maternal OUD and infant mortality did not significantly vary by maternal MOUD receipt—even between low and high levels of MOUD receipt—which does not support our hypothesis that MOUD may improve survival among maternal OUD-exposed infants.

It is important to note that these findings do not detract from the benefits of MOUD for pregnant individuals with OUD. Following current clinical guidelines, adherent MOUD use prevents relapse of nonmedical opioid use, limits obstetric complications associated with OUD, and lowers the risks of preterm birth and low birth weight (American College of Obstetricians and Gynecologists, 2017; Behrman and Butler, 2007; Piske et al., 2021; Rodriguez and Klie, 2019; Substance Abuse and Mental Health Services Administration, 2018; Tobon, Habecker and Forray, 2019; Witt et al., 2017). Moreover, our study focuses only on infant mortality. Future research to better understand whether and how MOUD treatment, and patterns therein, is associated with better outcomes for prenatally opioid exposed infants who survive is warranted.

Presently, the mechanisms linking maternal OUD and infant mortality are poorly understood (Mascarenhas et al., 2024). Maternal OUD is associated with adverse pregnancy complications and birth outcomes—specifically, preterm birth and low birth weight—that elevate the risk of infant mortality (Stover and Davis, 2015; Yazdy, Desai and Brogly, 2015). While not confirmatory, attenuated OUD-infant mortality associations after controlling for gestational age and birth weight align with this hypothesized pathway. The socio-environmental factors in this relationship are less clear. The U.S.’s opioid epidemic—which coincided with rising maternal OUD (Desai et al., 2014; Krans and Patrick, 2016)—disproportionately affected socioeconomically vulnerable populations and, in turn, amplified poverty, neighborhood deprivation, and housing insecurity (Ghertner and Groves, 2018; Maclean et al., 2022; van Draanen et al., 2020), all of which are known risk factors of infant mortality (Reece, 2021; Sullivan et al., 2023; Mohamoud, Kirby and Ehrenthal, 2019). It is uncertain how these societal influences shape the interplay between maternal OUD and infant mortality, making it difficult to pinpoint a nexus of causality and, consequently, intervention. While beyond our study’s purview, the biologic and socio-environmental mechanisms underlying maternal OUD and infant mortality merits investigation.

Our analysis boasts several strengths. large statewide cohort of linked birth records, infant death records, and Medicaid claims allowed us to measure a robust set of opioid-related variables and sociodemographic or health-related confounders. Likewise, we could assess the timing of NOWS diagnoses and capture the continuity of MOUD receipt throughout pregnancy. This breadth of administrative data is not commonly available to study maternal opioid use in the U.S. However, we note some limitations. We could not account for pregnancies that did not result in live delivery. Maternal OUD during pregnancy may increase the risk of stillbirth (Maeda et al., 2014; Varner et al., 2014), although it is uncertain if this varies by MOUD receipt. The generalizability of our findings using only Medicaid-covered births in Wisconsin may be limited vis-a-vis populations both within and outside of Wisconsin. If the relation between maternal OUD and infant mortality varies by income, then results may not apply to non-Medicaid (i.e., higher income) populations. Further, Wisconsin is notably less racially diverse compared to the entire U.S. yet has relatively large racial disparities in pregnancy outcomes (Gordon, 2019; United States Census Bureau, [no date]a; United States Census Bureau, [no date]b; Wisconsin Department of Health Services, 2022). With considerable racial variation in maternal OUD and infant mortality across the nation (Ko et al., 2020; Mohamoud, Kirby and Ehrenthal, 2019), findings may not generalize to other states. Lastly, our measures of OUD, NOWS, and MOUD may be prone to misclassification that bias estimates toward the null. We measured OUD diagnose during pregnancy, although physicians may have diagnosed OUD based on pre-pregnancy symptoms, and few individuals may receive MOUD for non-OUD indications (American College of Obstetricians and Gynecologists, 2017; Rodriguez and Klie, 2019; Substance Abuse and Mental Health Services Administration, 2018). While NOWS diagnoses claims are accurate (Maalouf et al., 2019), they may indicate screening rather than diagnosis.

In conclusion, this study contributes to growing evidence that maternal OUD during pregnancy is associated with an increased risk of infant mortality. Future research with larger cohorts should investigate mechanisms linking maternal OUD, NOWS, MOUD, infant mortality, and survival time to enhance interventions of infant mortality. As the U.S.’s opioid epidemic evolves, researchers should track trends in opioid use during pregnancy and its association with maternal and infant health outcomes.

Supplementary Material

Appendices

ACKNOWLEDGEMENTS

We thank the Wisconsin Department of Children and Families and the Wisconsin Department of Health Services for the use of the data. The design and conduct of the study are solely the responsibility of the authors and does not necessarily represent the official view of supporting agencies. Additionally, supporting agencies do not certify the accuracy of the analyses presented.

FUNDING SOURCE

This work was supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development [R01HD102125] and by the Health Resources and Services Administration through the University of Wisconsin Primary Care Research Fellowship [T32HP10010].

Footnotes

DECLARATION OF COMPETING INTERESTS

The authors have no conflicts of interest to disclose.

Financial Disclosures: All authors have no financial disclosures.

DATA AVAILABILITY

The data used for this study are not publicly available.

REFERENCES

  1. Almli LM, Ely DM, Ailes EC, et al. , 2020. Infant mortality attributable to birth defects – United States, 2003–2017. Morb. Mortal. Wkly. Rep 69 (2), 25–29. 10.15585/mmwr.mm6902a1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. American College of Obstetricians and Gynecologists, 2017. Opioid use and opioid use disorder in pregnancy. Committee opinion No. 711. Obstet. Gynecol 130 (2), e81–e94. 10.1097/aog.0000000000002235 [DOI] [PubMed] [Google Scholar]
  3. Anderson TM, Lavista Ferres JM, Ren SY, et al. , 2019. Maternal smoking before and during pregnancy and the risk of sudden unexpected infant death. Pediatrics 143(4), e20183325. 10.1542/peds.2018-3325 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Austin AE, Di Bona V, Cox ME, Proescholdbell SK, Naumann RB, 2022. Differences in mortality among infants with neonatal opioid withdrawal syndrome. Am. J. Prev. Med 63 (4), 619–623. 10.1016/j.amepre.2022.03.018 [DOI] [PubMed] [Google Scholar]
  5. Bairoliya N, Fink G, 2018. Causes of death and infant mortality rates among full-term births in the United States between 2010 and 2012: an observational study. PLoS Med 15 (3), e1002531. 10.1371/journal.pmed.1002531 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Ban L, Tata LJ, West J, et al. , 2012. Live and non-live pregnancy outcomes among women with depression and anxiety: a population-based study. PLoS One 7, e43462. 10.1371/journal.pone.0043462 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Behrman RE, Butler AD (eds.), 2007. Preterm Birth: Causes, Consequences, and Prevention National Academies Press, Washington, DC. [PubMed] [Google Scholar]
  8. Brogly SB, Turner S, Lajkosz K, et al. , 2017. Infants born to opioid-dependent women in Ontario, 2002–2014. J. Obstet. Gynaecol. Can 39 (3), 157–165. 10.1016/j.jogc.2016.11.009 [DOI] [PubMed] [Google Scholar]
  9. Centers for Disease Control and Prevention, 2017. NCHS urban-rural classification scheme for counties Available at: https://www.cdc.gov/nchs/data_access/urban_rural.htm (Accessed: 11 November 2023).
  10. Centers for Medicare and Medicaid Services, 2021. Opioid Treatment Programs (OTPs) Medicare Billing & Payment Available at: https://www.cms.gov/sites/default/files/2021-12/2021_12_MLN8296732_OTP_Billing_Payment_FINAL_0.pdf (Accessed: 11 November 2023).
  11. Desai RJ, Hernandez-Diaz S, Bateman B, Huybrechts KF, 2014. Increase in prescription opioid use during pregnancy among Medicaid-enrolled women. Obstet. Gynecol 123 (5), 997–1002. 10.1097/aog.0000000000000208 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Ghertner R, Groves L, 2018. The Opioid Crisis and Economic Opportunity: Geographic and Economic Trends United States Department of Health and Human Services, Washington, DC. [Google Scholar]
  13. Gordon S, 2019. Race in the Heartland: Equity, Opportunity, and Public Policy in the Midwest. Economic Policy Institute, Washington, DC. [Google Scholar]
  14. Grossarth S, Osmundson SS, Wiese AD, et al. , 2023. Maternal opioid use disorder and the risk of postneonatal infant mortality. JAMA Pediatr 177 (7), 675–683. 10.1001/jamapediatrics.2023.1047 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Haight SC, Ko JY, Tong VT, Bohm MK, Callaghan WM, 2018. Opioid use disorder documented at delivery hospitalization – United States, 1999–2014. Morb. Mortal. Wkly. Rep 67 (31), 845–849. 10.15585/mmwr.mm6731a1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Haneuse S, VanderWeele TJ, Arterburn D, 2019. Using the E-value to assess the potential effect of unmeasured confounding in observational studies. JAMA 321 (6), 602–603. 10.1001/jama.2018.21554 [DOI] [PubMed] [Google Scholar]
  17. Harris M, Schiff DM, Saia K, Muftu S, Standish KR, Wachman EM, 2023. Academy of Breast Feeding Medicine Clinical Protocol #21: Breastfeeding in the setting of substance use and substance use disorder (revised 2023). Breastfeed. Med 18 (10), 715–733. 10.1089/bfm.2023.29256.abm [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Hirai AH, Ko JY, Owens PL, Stocks C, Patrick SW, 2021. Neonatal abstinence syndrome and maternal opioid-related diagnoses in the US, 2010–2017. JAMA 325 (2), 146–155. 10.1001/jama.2020.24991 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Institute for Research on Poverty, 2022. WADC Data Documentation. WADC Overview and Data Structure University of Wisconsin-Madison, Madison, WI. [Google Scholar]
  20. Jansson LM, Velez M, Harrow C, 2009. The opioid exposed newborn: assessment and pharmacologic management. J. Opioid Manag 5 (1), 47–55. 10.5055/jom.2009.0006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Kinsella M, Halliday LOE, Shaw M, Capel Y, Nelson SM, Kearns RJ, 2022. Buprenorphine compared with methadone in pregnancy: a systematic review and meta-analysis. Subst. Use Misuse 57 (9), 1400–1416. 10.1080/10826084.2022.2083174 [DOI] [PubMed] [Google Scholar]
  22. Ko JY, D’Angelo DV, Haight SC, et al. , 2020. Vital signs: prescription opioid pain reliever use during pregnancy – 32 U.S. jurisdictions, 2019. Morb. Mortal. Wkly. Rep 69 (28), 897–903. 10.15585/mmwr.mm6928a1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Krans EE, Kim JY, James III AE, Kelley D, Jarlenski MP, 2019. Medication-assisted treatment utilization among pregnant women with opioid use disorder. Obstet. Gynecol 133 (5), 943–951. 10.1097/AOG.0000000000003231 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Krans EE, Patrick SW, 2016. Opioid use disorder in pregnancy: health policy and practice in the midst of an epidemic. Obstet. Gynecol 128 (1), 4–10. 10.1097/aog.0000000000001446 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Larson A, Berger LM, Mallinson DC, Grodsky E, Ehrenthal DB, 2019. Variable uptake of Medicaid-covered Prenatal Care Coordination: the relevance of treatment level and service context. J. Community Health 44 (1), 32–43. 10.1007/s10900-018-0550-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Leyenaar JK, Schaefer AP, Wasserman JR, Moen EL, O’Malley J, Goodman DC, 2021. Infant mortality associated with prenatal opioid exposure. JAMA Pediatr 175 (7), 706–714. 10.1001/jamapediatrics.2020.6364 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Linden A, Mathur MB, VanderWeele TJ, 2020. Conducting sensitivity analysis for unmeasured confounding in observational studies using E-values: the evalue package. Stata J 20 (1), 162–175. 10.1177/1536867X20909696 [DOI] [Google Scholar]
  28. Lyden J, Binswanger IA, 2019. The United States opioid epidemic. Semin. Perinatol 43 (3), 123–131. 10.1053/j.semperi.2019.01.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Maalouf FI, Cooper WO, Stratton SM, et al. , 2019. Positive predictive value of administrative data for neonatal abstinence syndrome. Pediatrics 143 (1), e20174183. 10.1542/peds.2017-4183 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Maclean JC, Mallatt J, Ruhm CJ, Simon K, 2022. The opioid crisis, health, healthcare, and crime: a review of quasi-experimental economic studies. Ann. Am. Acad. Pol. Soc. Sci, 703 (1), 15–49. 10.1177/00027162221149285 [DOI] [Google Scholar]
  31. Maeda A, Bateman BT, Clancy CR, Creanga AA, Leffert LR, 2014. Opioid use and dependence during pregnancy: temporal trends and obstetrical outcomes. Anesthesiology 121 (6), 1158–1165. 10.1097/aln.0000000000000472 [DOI] [PubMed] [Google Scholar]
  32. Mallinson DC, Ehrenthal DB, 2019. Classification of Medicaid coverage on birth records in Wisconsin, 2011–2012. Public Health Rep 134 (5), 542–551. 10.1177/0033354919860503 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Mangat AK, Schmölzer GM, Kraft WK, 2019. Pharmacological and non-pharmacological treatments for neonatal abstinence syndrome. Semin. Fetal Neonatal Med 24 (2), 133–141. 10.1016/j.siny.2019.01.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Mascarenhas M, Wachman EM, Chandra I, Xue R, Sarathy L, Schiff DM, 2024. Advances in the care of infants with prenatal opioid exposure and neonatal opioid withdrawal syndrome. Pediatrics 153 (3), e2023062871. 10.1542/peds.2023-062871 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. McCarthy JJ, Leamon MH, Willits NH, Salo R, 2015. The effect of methadone dose regimen on neonatal abstinence syndrome. J. Addict. Med 9 (2), 105–110. 10.1097/adm.0000000000000099 [DOI] [PubMed] [Google Scholar]
  36. Mohamoud YA, Kirby RS, Ehrenthal DB, 2019. Poverty, urban-rural classification and term infant mortality: a population-based multilevel analysis. BMC Pregnancy Childbirth 19, 40. 10.1186/s12884-019-2190-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Mozurkewich EL, Rayburn WF, 2014. Buprenorphine and methadone for opioid addiction during pregnancy. Obstet. Gynecol. Clin. North Am 41 (2), 241–253. 10.1016/j.ogc.2014.02.005 [DOI] [PubMed] [Google Scholar]
  38. Nguyen RHN, Knapp EA, Li X, et al. , 2023. Characteristics of individuals in the United States who used opioids during pregnancy. J. Women’s Health 32 (2), 161–170. 10.1089/jwh.2022.0118 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Nørgaard M, Schou Nielsson M, Heide-Jørgensen U, 2015. Birth and neonatal outcomes following opioid use in pregnancy: a Danish population-based study. Subst. Abuse Res. Treat 9 (Suppl 2), 5–11. 10.4137/sart.s23547 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Oesterle TS, Thusius NJ, Rummans TA, Gold MS, 2019. Medication-assisted treatment for opioid-use disorder. Mayo Clin. Proc 94 (10), 2072–2086. 10.1016/j.mayocp.2019.03.029 [DOI] [PubMed] [Google Scholar]
  41. Patel RM, Rysavy MA, Bell EF, Tyson JE, 2017. Survival of infants born at periviable gestational ages. Clin. Perinatol 44 (2), 287–303. https://doi.org/10.1016%2Fj.clp.2017.01.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Piske M, Homayra F, Min JE, et al. , 2021. Opioid use disorder and perinatal outcomes. Pediatrics 148 (4), e2021050279. 10.1542/peds.2021-050279 [DOI] [PubMed] [Google Scholar]
  43. Reece J, 2021. More than shelter: housing for urban maternal and infant health. Int. J. Environ. Res. Public Health 18 (7), 3331. https://doi.org/10.3390%2Fijerph18073331 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Rodriguez CE, Klie KA, 2019. Pharmacological treatment of opioid use disorder in pregnancy. Semin. Perinatol 43 (3), 141–148. 10.1053/j.semperi.2019.01.003 [DOI] [PubMed] [Google Scholar]
  45. Saia KA, Schiff D, Wachman EM, et al. , 2016. Caring for pregnancy women with opioid use disorder in the USA: expanding and improving treatment. Curr. Obstet. Gynecol. Rep 5 (3), 257–263. 10.1007/s13669-016-0168-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. StataCorp, 2023a. test – test linear hypotheses after estimation Available at: https://www.stata.com/manuals/rtest.pdf (Accessed: 11 November 2023).
  47. StataCorp, 2023b. Stata Statistical Software: Release 18 StataCorp LLC, College Station, TX. [Google Scholar]
  48. Stover MW, Davis JM, 2015. Opioids in pregnancy and neonatal abstinence syndrome. Semin. Perinatol 39 (7), 561–565. https://doi.org/10.1053%2Fj.semperi.2015.08.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Suarez EA, Huybrechts KF, Straub L, et al. , 2022. Buprenorphine versus methadone for opioid use disorder in pregnancy. N. Engl. J. Med 387 (22), 2033–2044. 10.1056/nejmoa2203318 [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Substance Abuse and Mental Health Services Administration, 2018. Clinical Guidance for Treating Pregnancy and Parenting Women with Opioid Use Disorder and their Infants United States Department of Health and Human Services, Washington, DC. [Google Scholar]
  51. Substance Abuse and Mental Health Services Administration, 2023a. Methadone Available at: https://www.samhsa.gov/medications-substance-use-disorders/medications-counseling-related-conditions/methadone (Accessed: 11 November 2023).
  52. Substance Abuse and Mental Health Services Administration, 2023b. Waiver elimination (MAT Act) Available at: https://www.samhsa.gov/medications-substance-use-disorders/waiver-elimination-mat-act (Accessed: 6 November 2023).
  53. Sullivan BA, Doshi A, Chernyavskiy P, et al. , 2023. Neighborhood deprivation and association with neonatal intensive care unit mortality and morbidity for extremely premature infants. JAMA Netw. Open 6 (5), e2311761. 10.1001/jamanetworkopen.2023.11761 [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Sun J, Liu X, Zhao M, Magnussen CG, Xi B, 2023. Dose-response association between maternal smoking during pregnancy and the risk of infant death: a nationwide, population-based, retrospective cohort study. eClinicalMedicine 57, 101858. 10.1016/j.eclinm.2023.101858 [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Tobon AL, Habecker E, Forray A, 2019. Opioid use in pregnancy. Curr. Psychiatry Rep 21 (12), 118. 10.1007/s11920-019-1110-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. United States Census Bureau, [no date]a. Quick facts United States Available at: https://www.census.gov/quickfacts/fact/table/US/PST045221 (Accessed: 6 November 2023).
  57. United States Census Bureau, [no date]b. Quick facts Wisconsin Available at: https://www.census.gov/quickfacts/fact/table/WI/PST04522 (Accessed: 6 November 2023).
  58. van Draanen J, Tsang C, Mitra S, Karamouzian M, Richardson L, 2020. Socioeconomic marginalization and opioid-related overdose: a systematic review. Drug Alcohol Depend 214, 108127. 10.1016/j.drugalcdep.2020.108127 [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Varner MW, Silver RM, Rowland Hogue CJ, et al. , 2014. Association between stillbirth and illicit drug use and smoking during pregnancy. Obstet. Gynecol 123 (1), 113–125. 10.1097/aog.0000000000000052 [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Weller AE, Crist RC, Reiner BC, Doyle GA, Berrettini WH, 2021. Neonatal opioid withdrawal syndrome (NOWS): a transgenerational echo of the opioid crisis. Cold Spring Harb. Perspect. Med 11 (3), a039669. 10.1101/cshperspect.a039669 [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Wisconsin Department of Health Services, 2022. African Americans in Wisconsin: an overview Available at: https://www.dhs.wisconsin.gov/minority-health/population/afriamer-pop.htm (Accessed: 14 November 2023).
  62. Witt CE, Rudd KE, Bhatraju P, Rivara FP, Hawes SE, Weiss NS, 2017. Neonatal abstinence syndrome and early childhood morbidity and mortality in Washington state: a retrospective cohort study. J. Perinatol 37 (10), 1124–1129. 10.1038/jp.2017.106 [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Yazdy MM, Desai RJ, Brogly SB, 2015. Prescription opioids in pregnancy and birth outcomes: a review of the literature. J. Pediatr. Genet 4 (2), 56–70. https://doi.org/10.1055%2Fs-0035-1556740 [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Zhong Q-Y, Gelaye B, Fricchione GL, et al. , 2018. Adverse obstetric and neonatal outcomes complicated by psychosis among pregnant women in the United States. BMC Pregnancy Childbirth 18, 120. 10.1186/s12884-018-1750-0 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

Appendices
NIHMS1970301-supplement-Appendices.pdf (514.3KB, pdf)

Data Availability Statement

The data used for this study are not publicly available.

RESOURCES