Mental health and substance use disorders (MSUD) during pregnancy and postpartum are a public health concern. The proportion of delivery hospitalizations with mental health condition diagnoses such as prenatal depression, anxiety, bipolar disorders, and psychotic disorders has increased significantly in recent years (Haight et al., 2019; Logue et al., 2022; McKee et al., 2020; Weiss et al., 2022). Prevalence of diagnosed perinatal depression ranged from 6.5%–12.9% (Gavin et al., 2005). Prevalence of self-reported depression during pregnancy was 14.8% and prevalence of self-reported postpartum depressive symptoms was 13.4% in 2019 (Centers for Disease Control and Prevention, 2022). Prevalence of self-reported prenatal anxiety symptoms was 18.2% and prevalence of a clinical diagnosis of any anxiety disorder was 15.2%, with postpartum anxiety prevalence estimated at 9.9% and postpartum anxiety symptoms prevalence estimated at 15.0% (Dennis et al., 2017). Pooled prevalence of bipolar disorder during the perinatal period among women with no previously known psychiatric illness was 2.6% (Masters et al., 2022). About 18.4% of pregnant females aged 15–4 years had past-month use of illicit drugs, tobacco products, or alcohol in 2019 in the United States (US) (Substance Abuse and Mental Health Services Administration, 2019). The prevalence of maternal opioid-related diagnoses at delivery hospitalization significantly increased from 2010 to 2017 (Hirai et al., 2021). MSUDs are associated with adverse perinatal, neonatal, and child outcomes (Chang, 2020; Forray and Foster, 2015; Sithisarn et al., 2012). Select obstetric risk factors were more common among delivery stays involving at least one mental health disorder than among delivery stays with no mental health disorders (Weiss et al., 2022). Mental health conditions, which include deaths of suicide, overdose/poisoning related to substance use disorder, and other deaths determined to be related to a mental health condition, including substance use disorder are a leading underlying cause of pregnancy-related death in the US, accounting for 22.7% of pregnancy-related deaths from year 2017 to 2019 (Trost et al., 2022). Nearly three-quarters of persons with a pregnancy-related mental health cause of death had a history of depression, and more than two-thirds had past or current substance use (Trost et al., 2021).

Professional and clinical organizations have issued recommendations on screening and management of MSUDs during pregnancy and postpartum, which are known to improve maternal and infant health (Substance Abuse and Mental Health Services Administration, 2018; O’Connor et al., 2016). However, there remain unmet needs and barriers to MSUD care during pregnancy and postpartum (Ko and Haight, 2020; Raffi et al., 2021; Webb et al., 2021). Readmissions data can inform interventions to improve discharge planning and follow-up care and reporting such data was included as a bundle element to improve health and well-being of postpartum women (Stuebe et al., 2021). Anticipated care needs can take into account the reasons for readmissions and factors that are associated with readmissions (Girsen et al., 2022). Studies of MSUD readmissions after delivery have focused on all-cause readmissions (Clapp et al., 2016; Kumar et al., 2019; Salemi et al., 2020), specific conditions such as opioid use disorder (Wen et al., 2019), or grouped psychiatric conditions (Wen et al., 2021). However, these studies used International Classification of Diseases, Ninth Edition, Clinical Modification (ICD-9-CM) diagnosis codes for a subset of MSUDs.

Our study assessed MSUDs based on grouped ICD-10-CM diagnosis codes to (1) characterize delivery hospitalizations with and without MSUD diagnoses; (2) assess 90-day MSUD and non-MSUD readmission rates and reasons for readmissions; and (3) estimate how MSUD type, patient, hospital, and clinical factors at delivery were associated with MSUD readmissions.

Of 2,697,605 weighted delivery hospitalizations from January to September 2019 included in our analysis, an estimated 11.8% of delivery hospitalizations recorded one or more MSUDs. Among these delivery hospitalizations recorded with MSUDs, an estimated 8.8% had one MSUD recorded; 61.4% had two; and 29.8% had three or more. The most frequently recorded MSUDs at delivery were miscellaneous mental and behavioral disorders/conditions (9.4%, Table 1), anxiety and fear-related disorders (5.7%), depressive disorders (4.4%), other specified substance-related disorders (2.9%), cannabis-related disorders (1.6%), bipolar and related disorders (0.9%), opioid-related disorders (0.7%), trauma- and stressor-related disorders (0.6%), stimulant-related disorders (0.6%), schizophrenia spectrum and other psychotic disorders (0.1%) and alcohol-related disorders (0.1%). Other MSUD categories each accounted for less than 0.1% of deliveries.

Among the most frequently recorded MSUD at delivery, “miscellaneous mental and behavioral disorders/conditions.”, the specific ICD-10-CM codes that collectively accounted for over 90% of “miscellaneous mental and behavioral disorders/conditions” included O99344, O99343, O99345, O99340, O99342, O99341 (other mental disorders complicating pregnancy, childbirth, and the puerperium), F53 (Mental and behavioral disorders associated with the puerperium, not elsewhere classified), and F99 (Mental disorder, not otherwise specified). The specific ICD-10-CM codes that collectively accounted for over 90% of “other specified substance related disorders” included O99324 (drug use complicating childbirth), O99323 (drug use complicating pregnancy, third trimester), F1910 (other psychoactive substance abuse, uncomplicated), and F1990 (other psychoactive substance use, unspecified, uncomplicated).

As presented in Table 1, distribution of characteristics among deliveries with or without MSUD diagnoses recorded was statistically different for all examined characteristics except hospital bed size. Compared to delivery hospitalizations with no MSUD recorded, a higher proportion of delivery hospitalizations with MSUDs recorded listed Medicaid as their primary expected payer (49.5% vs 39.5%, P<0.001), residency at locations other than large metropolitan areas (49.8% vs 46.1%, P<0.001), and median household income at zip codes of residence belonging to the lowest quartile (28.4% vs 27.0%, P<0.001). A higher percentage of deliveries with MSUD diagnoses were by cesarean compared to those deliveries without MSUD recorded (34.9% vs 31.3%, P<0.001). Delivery hospitalizations with MSUD recorded on average had longer length of stay (3.1 vs 2.6 days, P<0.001) and younger age at delivery than delivery hospitalizations without MSUD recorded (28.7 vs 29.1, P<0.001).

The top 10 primary CCSR body system chapters based on principal diagnosis recorded on all-cause 90-day readmissions after delivery are in Fig. 2. Pregnancy-related readmissions accounted for 77.4% of readmissions among deliveries without MSUD diagnoses, in comparison to 66.7% among deliveries with MSUD diagnoses. Mental, behavioral, and neurodevelopmental disorders chapter accounted for 11.5% of readmissions among deliveries with MSUD diagnoses, in comparison to 2.3% among deliveries without MSUD diagnoses. The 90-day all-cause, MSUD, and non-MSUD readmission rates differed by MSUD status at delivery hospitalization (Table 2). Regardless of MSUD status at delivery, the number of non-MSUD readmissions decreased after the first 30 days after delivery discharge. In contrast, the number of MSUD readmissions increased over time after delivery (Table 2). For example, among deliveries with MSUD diagnoses, the number of MSUD readmissions increased from 338 in the first 30 days to 508 in the 61–90 days after delivery discharge. Among deliveries without MSUD diagnoses, the number of MSUD readmissions increased from 218 in the first 30 days to 481 in the 61–90 days after delivery discharge.

In multivariable regression among deliveries with MSUD, MSUD type at delivery were differentially associated with 90-day MSUD readmissions (Table 3). Among the 11 types of MSUDs evaluated, eight MSUDs were associated with significantly higher odds for postpartum MSUD readmissions, with the highest odds for schizophrenia (OR = 7.30, 95%CI: 5.44–9.78), bipolar (OR = 2.27, 95%CI: 1.70–3.02), and stimulant-related disorders (OR =2.15, 95%CI: 1.52–3.05). Other types of MSUDs associated with increased odds of MSUD readmissions included depressive disorders (OR =1.35, 95%CI:1.07 −1.72), trauma- and stressor-related disorders (OR =1.65, 95%CI:1.28 −2.12), miscellaneous mental and behavioral disorders/conditions (OR =1.77, 95% CI:1.24 −2.54), alcohol-related disorders (OR =1.75, 95%CI: 1.04–2.94), and other specified substance-related disorders (OR =1.54, 95%CI:1.08 −2.20). Medicare or Medicaid insurance (vs private insurance) as primary expected payer at delivery and residence in a zip code with the lowest household income quartile (vs the highest income quartile) were significantly associated with higher odds for MSUD readmissions. Odds of MSUD readmission also increased with longer length of delivery hospitalization stay and younger age at delivery. In a separate multivariable regression analysis, where specific MSUD categories were replaced by the number of MSUDs at delivery, three or more MSUDs (vs one MSUD) at delivery was associated with significantly higher odds for MSUD readmissions (OR=3.27, 95% CI:2.10–5.09) (Data not shown).

Eight MSUDs were associated with significantly higher odds for postpartum MSUD readmissions. This is consistent with previous findings that women with psychiatric diagnoses at delivery hospitalization were more likely to be readmitted for postpartum psychiatric conditions compared with those without psychiatric comorbidity, although that study grouped all psychiatric diagnoses together (Wen et al., 2021). Although screening for depression, bipolar disorder, anxiety, trauma and intimate partner violence, and substance use disorders (for example, opioid use disorders and stimulant use disorder) is recommended by professional organizations, implementation of screening and referral to treatment can still benefit from improvement, such as evidence-based, individualized, non-judgmental care and the capacity for integrating screening and intervention for multiple, co-occurring MSUDs (Byatt et al., 2020; Ecker et al., 2019; Gulbransen et al., 2022). Furthermore, we found that severe mental illness such as schizophrenia, bipolar and related disorders, noted at delivery hospitalization was associated with the highest odds for MSUD readmissions, even when controlling for the presence of other MSUDs. Continuity of care post-discharge is especially important for individuals with these conditions (Conejo-Galindo et al., 2022; Jones et al., 2014). We noted a large proportion of MSUDs have indicators for nonspecific miscellaneous mental and behavioral disorders/conditions or other specified substance-related disorders. This may reflect difficulties in confirming a diagnosis or transferring diagnosis to coding for MSUD conditions in the billing infrastructure (Kim et al., 2012; Howell et al.,2021).

As expected, we found a higher rate of MSUD readmissions for delivery hospitalizations with MSUD diagnoses than for those without MSUD diagnoses. The number of MSUD readmissions increased over time after delivery for both groups regardless of whether MSUD was present at delivery, although the MSUD readmissions numbers were small for both groups. This finding may reflect the continued need for screening and referral to treatment for MSUD beyond the traditional postpartum visit at 6 weeks after delivery. The American College of Obstetricians and Gynecology has increased awareness of the postpartum period or “fourth trimester” of pregnancy and recommended that women with chronic medical conditions, including mood disorders and substance use disorders, should be counseled regarding the importance of timely follow-up with their obstetrician-gynecologists or primary care providers for ongoing coordination of care (McKinney et al., 2018).

Our results showed that 51.4% of delivery hospitalizations with MSUD diagnoses had Medicaid or Medicare as their primary expected payer, in comparison to 40.2% of delivery hospitalizations without MSUD diagnoses. MSUDs are among the most prevalent chronic conditions and incur high health service use during pregnancy through one year postpartum among persons who were covered by Medicaid at delivery (Pollack et al., 2022; Salahuddin et al., 2022). Structural and individual barriers to care among Medicaid and Medicare beneficiaries are wide ranging, such as lack of provider continuity, lack of transportation, and stigma related to insurance status or substance use (Bellerose et al., 2022; Pollack et al., 2022). In multivariable analysis, our result showed that Medicare or Medicaid insurance (vs private insurance) as primary expected payer was associated with higher odds of MSUD readmissions. Other studies have shown that insurance coverage can affect postpartum care access and Medicaid insurance is associated with obstetric hospital readmissions (Matthews et al., 2022; Reddy et al., 2021). To address insurance instability, which is a barrier to health care access in the postpartum period, particularly for those insured by Medicaid at delivery, states can implement strategies to extend Medicaid coverage up to 1 year postpartum (Centers for Medicare & Medicaid Services, 2021). The extended postpartum coverage option through the American Rescue Plan offers states an opportunity to improve the continuity of care for chronic conditions, including MSUDs (Centers for Medicare & Medicaid Services, 2021).

In addition to insurance coverage, our results highlight how other social determinants of health, such as living in the lowest quartile of median household income at zip code level, were associated with higher odds of MSUD readmissions. A recent study on assessment of multiple social needs (e.g., food, utilities, transportation) among Medicaid beneficiaries found that each of the 10 examined social needs was consistently associated with MSUDs (McQueen et al., 2021). MSUD clinical care can be improved by integrating interventions for meeting unmet social needs (Barfield, 2021; Byatt et al., 2020; Horbar et al., 2020).

Our study documented MSUDs at delivery hospitalization and readmissions and assessed factors associated with MSUD readmissions. Factors associated with higher odds of readmission, such as type of MSUD, primary expected payer and community income levels, may present considerations for anticipatory care needs at the delivery hospitalization to reduce readmissions. Strategies to improve care for persons with MSUD at delivery can include improved follow-up screening and management of MSUDs in the postpartum period, equitable access to quality postpartum care, expanded Medicaid postpartum coverage, and addressing social determinants of health.