Prevalence of Psychosocial Issues Among Pregnant Women Who Do and Do Not Use Illicit Substances

not, reinforcing recommendations for multidisciplinary approaches to care


Introduction
The American College of Obstetricians and Gynecologists (ACOG) recommends prenatal care be initiated in the rst trimester and continue at regular intervals throughout pregnancy to reduce the risk of adverse outcomes (American College of Pediatrics and American College of Obstetricians and Gynecologists, 2017). Pregnant women who report use of illicit substances initiate prenatal care later than those who do not (second vs. rst trimester, respectively) and are more likely to receive inadequate prenatal care (Simmons & Austin, 2022;Wu et al., 2013). One common explanation for these differences is that pregnant women who report use of illicit substances have a higher prevalence of psychosocial issues than pregnant women who do not, which make it more di cult for them to access and engage in care.
The extant data on the prevalence of psychosocial issues among pregnant women who report use of illicit substances frequently come from studies on other topics where select issues are reported on when characterizing the population and from studies using convenience samples. Using weighted estimates from a nationally representative population health survey in the United States, this report aims to quantify and compare the prevalence of a range of psychosocial issues faced by pregnant women who did vs. did not report past-month illicit substance use.

Design
Data covered the period 2015 to 2019 and came from the public use les of the National Survey of Drug Use and Health (NSDUH). NSDUH is an annual epidemiological survey of the prevalence of substance use and mental health issues in the US civilian noninstitutionalized population aged 12 or older (SAMHDA, 2019). NSDUH interviews are carried out using primarily computer-assisted personal interviewing and for more sensitive questions related to illicit substance use, using audio computerassisted self-interviewing. Using a strati ed, multi-stage cluster sampling design, approximately 70,000 respondents were interviewed annually (68,073 in 2015, 67,942 in 2016, 68,032 in 2017, 67,791 in 2018, and 67,625 in 2019). Weighted response rates for each year 2015 to 2019 were 69.7%, 68.4%, 67.1%, 66.6% and 64.9%, respectively. Sampling weights were provided by NSDUH to account for unit-and individual-level non-response and adjusted to ensure estimates were consistent with estimates provided by the US Census (SAMHDA, 2019). For the current analysis, respondents were restricted to those who were currently pregnant at the time of the interview (N = 3,657).

Measures Pregnancy
Women were classi ed as pregnant if they responded a rmatively to the question "Are you pregnant?"

Illicit substance use
We used the recoded variable in the NSDUH survey referring to any reported illicit substance use in the past month (binary: used in past month vs. did not report use in past month) (see Supplementary   Table 1). This variable covers past-month use (or misuse in the case of licit/medically prescribed substances) of cocaine, hallucinogens, heroin, inhalants, methamphetamine, marijuana, pain relievers, sedatives, stimulants, and tranquilizers.
Psychosocial issues ACOG recommends that antepartum screening include assessment of multiple psychosocial issues, namely a patient's desire for pregnancy, tobacco use, substance use, depression, safety, intimate partner violence, stress, barriers to care, unstable housing, communication barriers, and nutrition (American Academy of Pediatrics and American College of Obstetricians and Gynecologists, 2017). We searched the NSDUH codebook for variables that related to these issues. The seven binary (yes/no) variables we identi ed (Supplementary Table S1) and, in parentheses, the issues they represent are: past-month cigarette smoking (tobacco use), past-month alcohol use (substance use), serious past-month distress (both depression and stress), no health insurance (barriers to care), moved 3 + times in the past year (unstable housing), does not speak English well (communication barriers), and household receives food stamps (nutrition). Past-month distress was measured using the clinically-validated Kessler Psychological Distress Scale (K6) (Kessler et al., 2002(Kessler et al., , 2010Prochaska et al., 2012). This six-item measure screens for individuals with a high likelihood of being diagnosed with a mental illness. A sum score with a possible range from 0 to 24 is calculated, with scores of 13 and higher indicating serious distress that warrants intervention. As far as we are aware, there are no variables in NSDUH that relate to a patient's desire for pregnancy, or safety/intimate partner violence.

Analysis
All analyses were carried out in R version 4.2.2, using the survey and tidyverse packages (Lumley, 2004;Wickham et al., 2019). Weighted sample characteristics are presented for all pregnant women, and for those who did vs. did not report past-month illicit substance use. We then report weighted prevalence statistics with 95% CI for each psychosocial variable among pregnant women who did vs. did not report past-month illicit substance use. To assess the association between each psychosocial issue (binary dependent variable; issue not present (referent) vs. issue present) and past-month illicit substance use (no (referent) vs. yes), separate weighted and age-adjusted logistic regression models for each issue were also run.
Given variation in the legal status of recreational marijuana use between states (Pacula & Smart, 2017), and its relatively higher prevalence compared with other substances in the US (SAMHDA, 2019), we also conducted a sensitivity analysis looking at the prevalence of each psychosocial issue according to use of any illicit substance other than marijuana in the past month.

Transparency And Openness
All analyses were carried out in R version 4.2.2, using the survey and tidyverse packages (  Ns are weighted to refer to the population of the US that respondents in the survey represent. The rates of all but one of the psychosocial issues examined were higher among those reporting recent illicit substance use ( Table 2). Brie y, among pregnant women who reported using illicit substances in the past month around half reported smoking, one-third reported alcohol use, and one-quarter reported serious psychological distress, respectively. These women had ve-to seven-times higher odds of reporting these issues compared with women who did not report using illicit substances (p < .001).  Adjusted odds ratios (AORs) based on weighted sample, adjusted for age.
In addition, almost half of pregnant women who reported using illicit substances in the past month reported they or someone in their household was receiving food stamps and around one in ten reported that they did not have any health insurance and had moved at least three times in the past year. The odds of these issues were twice as high in the group who did vs. did not report illicit substance use in the past month, although the measure of unstable housing was compatible with their being no association. In contrast, the rate of not speaking English well was higher among those who did not report using illicit substances in the past month. This is likely a re ection of Hispanic women having the highest rate of poor English pro ciency (~ 19%) but relatively low rates of past-month illicit substance use (~ 3%) (SAMHDA, 2019).
Results from sensitivity analyses assessing the prevalence of psychosocial issues among those reporting use of any illicit substance other than marijuana in the past month were concordant with the primary analyses, albeit with higher point prevalence estimates for some factors such as cigarette smoking, pastmonth distress, and housing instability (Supplementary Table S2).

Discussion
In a representative sample of pregnant women in the US between 2015-2019, about 1 of every 15 reported recent use of illicit substances. The prevalence of all but one of the psychosocial issues examined was higher in this group as compared to pregnant women who did not report recent use of illicit substances and typically was at least twice as prevalent, if not more. Although we are not able to quantify how many psychosocial issues each pregnant woman reporting recent use of illicit substances endorsed, the high prevalence of several issues, and the known correlations between them, means it is likely that many women were experiencing more than one and perhaps multiple issues concurrently. The World Health Organization's most recent guidelines for the identi cation and management of substance use and substance use disorders (SUDs) in pregnancy say that services for these women "should have a level of comprehensiveness that matches the complexity and multifaceted nature of substance use disorders and their antecedents" (WHO, 2014, p. 7). The present results reinforce long-standing recommendations for multidisciplinary care for this population.
As noted in the Introduction, it is commonly thought that the higher prevalence of many psychosocial issues makes it more di cult for women who use illicit substances to initiate and engage in prenatal care, leading to later and inadequate prenatal care in this population (Simmons & Austin, 2022;Wu et al., 2013). In addition, psychosocial issues may also indirectly contribute to later and inadequate prenatal care. Pregnant women with SUDs often report they feel judged by health care providers (Renbarger et al., 2020) and perinatal providers have reported feeling angry and frustrated with pregnant patients with illicit SUDs and described them as uncaring (Raeside, 2003;Romisher et al., 2018;Seybold et al., 2014). Providers would seem to often be making the fundamental attribution error, attributing women's actions to their character rather than to external situational factors. As Radcliffe (2011) observed, "few concessions would be seen to be made for the fact that these women...are often contending with the di cult social circumstances that accompany a drug-using or former drug-using lifestyle" (p. 505). Broader appreciation of the di cult circumstances so many of these women are in and compassion for their situation could lead to more women seeking timely and consistent care and in turn, better outcomes.
There are several limitations to the current analysis. First, to maximize the sample for analysis, we computed estimates in those who reported using illicit substances in the past month, rather than only those classi ed as having an SUD. In addition, most reported substance use refers to marijuana use, which is no longer illicit in many jurisdictions across the US. Nevertheless, results of the sensitivity analysis excluding all reported marijuana use lead to similar, if somewhat less precise, conclusions. Second, accurate measurement of illicit substance use in priority populations in population surveys is challenging due to issues of selective non-response, small case numbers, and under sampling of certain populations such as the unsheltered homeless (Reuter et al., 2021). However, this means the present estimates are likely conservative. Relatedly, two psychosocial issues listed in the ACOG screening guidance that are believed to be more prevalent among women using illicit substances, namely desire for pregnancy and intimate partner violence, do not currently have a related variable in NSDUH. Still, we were able to provide much more precise estimates of many other psychosocial issues by using data from this large population-based survey.
In summary, pregnant women who report using illicit substances experience higher rates of most psychosocial issues compared to those who do not, reinforcing recommendations for multidisciplinary care that matches the complexity of their needs.

Declarations
Author Note This work was supported in part by grants R01DA036670, R01DA047867, and T32DA07242 from the National Institute on Drug Abuse. The authors have no con icts of interest to declare.
All analyses were carried out in R version 4.2.2, using the survey and tidyverse packages (Lumley, 2004;Wickham et al., 2019) and all analysis code are publicly available at the open science framework https://osf.io/394qy/. The data reported in this manuscript were obtained from publicly available data (SAMHSA , https://www.data les.samhsa.gov/dataset/national-survey-drug-use-and-health-2020-nsduh-2020-ds0001). A bibliography of journal articles, working papers, conference presentations, and dissertations using the National Survey on Drug Use and Health is available at https://www.data les.samhsa.gov/bibliography-search. The relationships examined in the present article have not been examined in any previous or current articles, or to the best of our knowledge in any papers that will be under review soon.