Abstract
Introduction:
Screening and referral for substance use are essential components of prenatal care. However, little is known about barriers to participation in substance use interventions that are integrated within prenatal care.
Methods:
Our study examines demographic and clinical correlates of participation in an initial assessment and counseling intervention integrated into prenatal care in a large healthcare system. The sample comprised Kaiser Permanente Northern California pregnant women with a live birth in 2014 or 2015 who screened positive for prenatal substance use via a self-reported questionnaire and/or urine toxicology test given as part of standard prenatal care (at ~8 weeks gestation).
Results:
Of the 11,843 women who screened positive for prenatal substance use (median age =30 years; 42% white; 38% screened positive for alcohol only, 20% for cannabis only, 5% nicotine only, 17% other drugs only, and 19% ≥2 substance categories), 9,836 (83%) completed the initial substance use assessment and counseling intervention. Results from multivariable logistic regression analyses indicated that younger age, lower income, single marital status, and a positive urine toxicology test predicted higher odds of participation, while other/unknown race/ethnicity, greater parity, receiving the screening later in pregnancy, and screening positive for alcohol only or other drugs only predicted lower odds of participation (all Ps<.05).
Discussion:
Findings suggest that integrated substance use interventions can successfully reach vulnerable populations of pregnant women (e.g., younger, lower income, racial/ethnic minorities). Future research should address whether differences in participation are due to patient (e.g., type of substance used, perceived stigma) or provider factors (e.g., working harder to engage traditionally underserved patients).
Substance use during pregnancy is an important public health problem (Committee on Obstetric Practice, 2017a, 2017b; Substance Abuse and Mental Health Services Administration, 2013), and the prenatal period provides a unique window of opportunity to reduce adverse fetal and neonatal outcomes through early identification and treatment of prenatal substance use (Haug, Duffy, & McCaul, 2014). National guidelines consider substance use screening, brief intervention, and referral essential components of prenatal care (American College of Obstetricians Gynecologists Committee on Health Care for Undeserved Women, 2006; Committee on Obstetric Practice, 2017a; Paine & Garceau, 1999) and prenatal interventions of various intensities are associated with improved perinatal outcomes (Corrarino et al., 2000; Goler, Armstrong, Taillac, & Osejo, 2008; Haug et al., 2014; Krans et al., 2018; Ordean & Kahan, 2011; Sweeney, Schwartz, Mattis, & Vohr, 2000). However, pregnant women face barriers to utilizing substance use interventions, including social stigma, mental health issues, uncertainty about what treatment entails, cost, lack of availability of programs, and competing demands on time (Clark, Dee, Bale, & Martin, 2001; Howell & Chasnoff, 1999; Jackson & Shannon, 2012; Sutter, Gopman, & Leeman, 2017). Further, women may have valid concerns about legal consequences as some states consider prenatal substance use child abuse under civil child-welfare statutes or grounds for civil commitment. Thus, there continues to be an unmet need for help with substance use among pregnant women (Terplan, McNamara, & Chisolm, 2012).
Existing research on predictors of prenatal substance use counseling utilization has mostly focused on utilization of programs external to prenatal care. These studies suggest that participation in outpatient substance use programs is more likely among pregnant women with prior treatment experiences (Clark et al., 2001; Messer, Clark, & Martin, 1996), greater mental health problems, greater legal severity, African American race (Messer et al., 1996), prenatal use of illicit drugs (Messer et al., 1996), and more severe drug addiction (Haller, Miles, & Dawson, 2003). However, it is unknown whether these factors are associated with participation in interventions available without a secondary cost-share as part of standard prenatal care.
Obstetrics/gynecology departments are well-suited to provide counseling for prenatal substance use linked with prenatal visits, potentially reducing barriers to participation. Kaiser Permanente Northern California’s (KPNC) prenatal substance use program, Early Start, screens women for prenatal substance use by self-report and urine toxicology testing and provides counseling with no secondary cost-share as part of comprehensive prenatal care (Armstrong et al., 2001). Licensed counselors (social workers, marriage and family therapists, and psychologists) are located within obstetrics and gynecology departments, and a comprehensive hour-long psychosocial assessment and substance use counseling intervention, which includes a detailed interview about past and current substance use, education and early intervention, diagnosis, referrals to addiction treatment or follow-up appointments (as needed). In California, healthcare providers are not mandated to report prenatal substance use to law enforcement or social services agencies, allowing for increased confidentiality.
While this integrated prenatal substance use screening and intervention program has had great success in improving prenatal outcomes (Armstrong et al., 2003; Goler et al., 2008) and lowering overall healthcare costs (Goler et al., 2012), not all women participate (Armstrong et al., 2003; Goler et al., 2008). Our prior research suggests that at-risk women who do not participate in the initial psychosocial assessment and counseling intervention (~15%) have worse neonatal outcomes and higher healthcare costs, warranting special attention (Armstrong et al., 2003; Goler et al., 2008). Little is known, however, about the correlates of participation in this initial hour-long substance use assessment and counseling intervention conveniently integrated into prenatal care and tailored to pregnant women. Further, it is unknown whether participation rates vary with the method by which women screen positive for use in pregnancy (e.g., self-report only, urine toxicology test only, or both).
The current study uses data from a large healthcare organization with universal prenatal substance use screening by self-report and urine toxicology testing to examine whether patient characteristics are associated with participation in an hour-long substance use assessment and counseling intervention linked with prenatal care. Findings will improve understanding of whether socio-demographic and clinical characteristics predict utilization of prenatal substance use interventions after removing key barriers to participation and can be used to develop more tailored outreach.
Methods
Study setting
KPNC is a nonprofit healthcare delivery system serving ~4 million members, and covering >40% of the region’s commercially insured population. Patients are racially and socioeconomically diverse and representative of the northern California population. Standard prenatal care includes universal prenatal drug screening by a self-administered questionnaire and urine toxicology testing at the first prenatal visit (~8 weeks gestation). KPNC’s Institutional Review Board approved this study with waiver of informed consent.
Study participants
Our sample comprised KPNC pregnant women with a live or still birth between 2014 and 2015 who completed the prenatal screening questionnaire and screened positive for prenatal substance use by self-report and/or urine toxicology testing. To avoid non-independent observations, we included only a woman’s first pregnancy during the study period.
Measures
Data on participation in the comprehensive, hour-long substance use assessment and counseling intervention, which includes a detailed interview about past and current substance use, education and early intervention, diagnosis, and referrals to addiction treatment or follow-up appointments (as needed), were extracted from the electronic health record (EHR).
The prenatal screening included questions about any substance use during pregnancy (i.e., not just substance use disorders). Prenatal substance use was defined based on: 1) any self-reported use of alcohol, nicotine, cannabis, or other drugs (cocaine, methamphetamine, opioids/pain medication, heroin, methadone/buprenorphine, sleep medication, anxiety medication, other street drugs) on the prenatal screening questionnaire, and/or 2) a confirmed positive universal urine toxicology test for ethanol, cannabis, or other drugs (cocaine, amphetamine/methamphetamines, MDA/MDMA, opioids, benzodiazepines, or barbiturates); heroin and nicotine were assessed only by self-report and not by urine toxicology testing. Toxicology tests from 2 weeks prior to 8 weeks after the prenatal questionnaire screening date were included. If patients had multiple toxicology tests, the test closest to the screening questionnaire was included.
While not validated, the prenatal screening questionnaire has been used clinically in KPNC for over 20 years across all medical centers, and we have published extensively using data from this questionnaire (Armstrong et al., 2003; Armstrong et al., 2001; Goler et al., 2012; Goler et al., 2008; Young-Wolff, Sarovar, Tucker, Avalos, et al., 2019; Young-Wolff et al., 2018; Young-Wolff, Sarovar, Tucker, Conway, et al., 2019; Young-Wolff et al., 2017). We categorized women based on the screening method that was positive for substance use (positive self-report, negative toxicology; positive self-report, unknown toxicology; negative self-report, positive toxicology; positive self-report, positive toxicology). We combined information from self-report and toxicology tests and grouped substance use into: alcohol only, cannabis only, nicotine only, other drugs only, and ≥2 of the categories above (e.g., alcohol and cannabis).
Depressive disorders, anxiety disorders, and substance use disorders were extracted from the EHR using ICD codes during the year prior to conception through the date of the screening questionnaire. The Patient Health Questionnaire (PHQ-9) is given at the first prenatal visit across KPNC medical centers via universal screening and results are recorded in the EHR. Severity of depression was categorized into: none (0–4), mild (5–9), moderate (10–14), and moderately-severe to severe (15–27). Data on emergency room utilization during the year before questionnaire came from the EHR.
Data on age, race/ethnicity, parity, and median household neighborhood family income came from the EHR. Patients’ marital status and education were assessed as part of the screening questionnaire.
Statistical analysis
We compared differences in the demographic and clinical characteristics among women who did and did not participate in the initial hour-long substance use assessment and counseling intervention. Chi-squared tests were used to examine categorical variables and non-parametric tests for non-normally distributed continuous variables. Multivariable logistic regression models that included demographic and clinical characteristics in one model were conducted to identify factors related to participation.
Results
Of 70,031 eligible women, 11,926 (17%) screened positive for prenatal substance use via self-report and/or toxicology, including 11,843 who had a live birth and 83 who had a stillbirth. Due to small sample size, pregnancies that ended with a stillbirth were not included. Over 90% of study participants (10,767/11,843) completed both the self-reported questionnaire and the toxicology test. Of these, 84% completed both screenings during their first trimester, 52% of toxicology tests were completed at the same visit as the questionnaire, and 95% were completed within 14 days.
Table 1 shows the distribution of demographic and clinical characteristics overall and by completion of the initial hour-long assessment and counseling intervention. Women had a median age of 30, 42% were White, 65% had graduated from high school, and median neighborhood household income was $67,033. In the past year, more than a third (36%) had an emergency room visit, 12% had an anxiety disorder, 12% a depressive disorder, and 12% a substance use disorder. Most women completed the questionnaire during the first trimester (86%). Alcohol use only was most common (38%), followed by cannabis only (17%), other drugs only (17%), and nicotine only (5%); 19% used ≥2 of the substance categories above; 71% screened positive for substance use only by self-report, 16% only by urine toxicology testing, and 13% by both.
Table 1.
Characteristics of Women with Positive Substance Use Screening in Pregnancy, by Participation in the Assessment and Counseling Intervention
| Characteristics, N (%) | Total N = 11,843 | Did Not Participate N = 1,980 (17%) | Participated N = 9,863 (83%) | p |
|---|---|---|---|---|
| Median Age [IQR], Years | 30 [25–34] | 31 [27–34] | 29 [25–33] | <.0001 |
| Age at Screening, Years | <.0001 | |||
| <18 | 151 (1.3) | 17 (0.9) | 134 (1.4) | |
| 18–24 | 2,591 (21.9) | 291 (14.7) | 2,300 (23.3) | |
| 25–34 | 6,825 (57.6) | 1,189 (60.1) | 5,636 (57.1) | |
| ≥35 | 2,276 (19.2) | 483 (24.4) | 1,793 (18.2) | |
| Race/Ethnicity | <.0001 | |||
| White | 5,016 (42.4) | 854 (43.1) | 4,162 (42.2) | |
| Asian | 1,167 (9.9) | 232 (11.7) | 935 (9.5) | |
| African-American | 1,123 (9.5) | 136 (6.9) | 987 (10.0) | |
| Hispanic | 3,170 (26.8) | 494 (25.0) | 2,676 (27.1) | |
| Other/unknown | 1,367 (11.5) | 264 (13.3) | 1,103 (11.2) | |
| Neighborhood Income, $ | <.0001 | |||
| <$48,534 | 2,940 (24.8) | 406 (20.5) | 2,534 (25.7) | |
| $48,534–$67,033 | 2,960 (25.0) | 465 (23.5) | 2,495 (25.3) | |
| $67,034–$89,583 | 2,979 (25.2) | 503 (25.4) | 2,476 (25.1) | |
| >$89,583 | 2,942 (24.8) | 601 (30.4) | 2,341 (23.7) | |
| Unknown | 22 (0.2) | 5 (0.3) | 17 (0.2) | |
| Marital Status | <.0001 | |||
| Not married | 4,142 (35.0) | 550 (27.8) | 3,592 (36.4) | |
| Married/Common law/Domestic partner | 5,521 (46.6) | 1,110 (56.1) | 4,411 (44.7) | |
| Unknown | 2,180 (18.4) | 320 (16.2) | 1,860 (18.9) | |
| Education | <.0001 | |||
| ≤High school graduate | 2,074 (17.5) | 280 (14.1) | 1,794 (18.2) | |
| >High school graduate | 7,658 (64.7) | 1,388 (70.1) | 6,270 (63.6) | |
| Unknown | 2,111 (17.8) | 312 (15.8) | 1,799 (18.2) | |
| Parity | <.0001 | |||
| 0 | 4,744 (40.0) | 681 (34.4) | 4,063 (41.2) | |
| 1 | 3,216 (27.1) | 596 (30.1) | 2,620 (26.6) | |
| >1 | 2,109 (17.8) | 419 (21.2) | 1,690 (17.1) | |
| Unknown | 1,774 (15.0) | 284 (14.3) | 1,490 (15.1) | |
| Gestational Age at Screening | <.0001 | |||
| 1st trimester | 10,165 (85.8) | 1,551 (78.3) | 8,614 (87.3) | |
| 2nd trimester | 1,273 (10.8) | 264 (13.3) | 1,009 (10.2) | |
| 3rd trimester | 405 (3.4) | 165 (8.3) | 240 (2.4) | |
| Anxiety Disorders | .14 | |||
| No | 10,409 (87.9) | 1,760 (88.9) | 8,649 (87.7) | |
| Yes | 1,434 (12.1) | 220 (11.1) | 1,214 (12.3) | |
| Depressive Disorders | .003 | |||
| No | 10,440 (88.2) | 1,784 (90.1) | 8,656 (87.8) | |
| Yes | 1,403 (11.9) | 196 (9.9) | 1,207 (12.2) | |
| Substance Use Disorders | <.0001 | |||
| No | 10,462 (88.3) | 1,834 (92.6) | 8,628 (87.5) | |
| Yes | 1,381 (11.7) | 146 (7.4) | 1,235 (12.5) | |
| Emergency Room Visit | <.0001 | |||
| No | 7,601 (64.2) | 1,365 (68.9) | 6,236 (63.2) | |
| Yes | 4,242 (35.8) | 615 (31.1) | 3,627 (36.8) | |
| Depression Symptoms | <.0001 | |||
| None | 6,277 (53.0) | 1,037 (52.4) | 5,240 (53.1) | |
| Mild | 2,977 (25.1) | 442 (22.3) | 2,535 (25.7) | |
| Moderate | 975 (8.2) | 146 (7.4) | 829 (8.4) | |
| Moderately severe/severe | 438 (3.7) | 56 (2.8) | 382 (3.9) | |
| Unknown | 1,176 (9.9) | 299 (15.1) | 877 (8.9) | |
| Substance Use | <.0001 | |||
| Alcohol only | 4,532 (38.3) | 895 (45.2) | 3.637 (36.9) | |
| Nicotine only | 602 (5.1) | 85 (4.3) | 517 (5.2) | |
| Cannabis only | 2,403 (20.3) | 244 (12.3) | 2,159 (21.9) | |
| Other drug(s) a only | 2,018 (17.0) | 496 (25.1) | 1,522 (15.4) | |
| ≥2 substances | 2,288 (19.3) | 260 (13.1) | 2,028 (20.6) | |
| Substance Use Screening Method | <.0001 | |||
| Positive self-report, negative toxicology | 7,360 (62.2) | 1,376 (69.5) | 5,984 (60.7) | |
| Positive self-report, unknown toxicology | 1,076 (9.1) | 282 (14.2) | 794 (8.1) | |
| Negative self-report, positive toxicology | 1,844 (15.6) | 204 (10.3) | 1,640 (16.6) | |
| Positive self-report, positive toxicology | 1,563 (13.2) | 118 (6.0) | 1,445 (14.7) | |
Notes.
Including anxiety medication, cocaine, heroin, methadone/buprenorphine, methamphetamine, opioids/pain medication, other street drugs, and sleep medication.
Of women who screened positive for other drugs only via self-report and/or the urine toxicology testing, 1.8% screened positive for cocaine (by self-report or urine toxicology positive for cocaine), 1.5% for heroin (by self-report only), 2.1% for methamphetamine (by self-report or urine toxicology positive for amphetamine/methamphetamine), 28.8% for anxiety medications (by self-report or urine toxicology positive for benzodiazepines), 55.6% for pain medication (by self-report or urine toxicology positive for prescription opioids), 23.4% for sleep medication (by self-report or urine toxicology positive for barbiturates), and 1.7% for other street drugs (self-report or urine toxicology positive for MDA/MDMA).
Overall, 83% participated in the initial substance use assessment and counseling intervention. Women who participated were younger, more likely to be African-American or Hispanic race/ethnicity, had lower neighborhood income, lower parity, and a greater likelihood of single marital status, not graduating from high school, depressive and substance use disorders, emergency room visits, and substance use screening in the first trimester (Table 1). They were also more likely to screen positive for substance use via a urine toxicology test, and to use cannabis only, nicotine only, and ≥2 substance categories, while those who did not participate were more likely to screen positive only for alcohol or only other drugs.
In multivariable logistic regression models, women ages 18–24 and 25–34 (versus >34) who had lower income, single marital status, and a positive urine toxicology test (with or without self-reported substance use) had higher odds of participation, while greater parity, completion of the screening during the second or third trimester, and other/unknown race/ethnicity, unknown depression symptoms, unknown toxicology test, and screening positive for alcohol only or other drugs only were associated with lower odds of participation (Table 2).
Table 2.
Adjusted Odds Ratios for Participation in the Assessment and Counseling Intervention among 11,843 Women Who Screened Positive for Substance Use in Pregnancy
| Characteristics | Odds Ratio | 95% Confidence Interval | P |
|---|---|---|---|
| Age at Screening, Years | |||
| <18 | 1.25 | 0.72–1.10 | .43 |
| 18–24 | 1.39 | 1.15–1.67 | .0007 |
| 25–34 | 1.13 | 1.00–1.28 | .05 |
| >34 | 1 | -- | -- |
| Race/Ethnicity | |||
| White | 1 | -- | -- |
| Asian | 0.93 | 0.78–1.10 | .37 |
| African-American | 1.07 | 0.86–1.32 | .55 |
| Hispanic | 1.04 | 0.91–1.19 | .56 |
| Other/unknown | 0.83 | 0.71–0.98 | .02 |
| Neighborhood Income, $ | |||
| <$48,534 | 1.28 | 1.11–1.49 | .001 |
| $48,534-$67,033 | 1.20 | 1.04–1.38 | .01 |
| $67,034-$89,583 | 1.15 | 1.01–1.32 | .04 |
| >$89,583 | 1 | -- | -- |
| Unknown | 1.21 | 0.40–3.69 | .73 |
| Marital Status | |||
| Married/Common law/Domestic partner | 1 | -- | -- |
| Not married | 1.16 | 1.02–1.33 | .03 |
| Unknown | 1.20 | 0.92–1.27 | .21 |
| Educational Level | |||
| ≤High school graduate | 1.08 | 0.92–1.27 | .34 |
| >High school graduate | 1 | -- | -- |
| Unknown | 1.12 | 0.84–1.49 | .44 |
| Parity | |||
| 0 | 1 | -- | -- |
| 1 | 0.80 | 0.71–0.91 | .0007 |
| ≥2 | 0.76 | 0.66–0.88 | .0003 |
| Unknown | 0.93 | 0.79–1.09 | .37 |
| Gestational Age at Screening | |||
| 1st Trimester | 1 | -- | -- |
| 2nd Trimester | 0.62 | 0.53–0.72 | <.0001 |
| 3rd Trimester | 0.25 | 0.20–0.31 | <.0001 |
| Anxiety Disorder | 0.98 | 0.83–1.17 | .85 |
| Depressive Disorder | 1.14 | 0.95–1.36 | .17 |
| Substance Use Disorder | 1.17 | 0.95–1.43 | .14 |
| Emergency Room Visit | 1.00 | 0.89–1.12 | .97 |
| Depression Symptoms | |||
| None | 1 | -- | -- |
| Mild | 1.01 | 0.89–1.14 | .89 |
| Moderate | 0.99 | 0.81–1.20 | .88 |
| Moderately severe/severe | 1.11 | 0.82–1.51 | .49 |
| Unknown | 0.61 | 0.52–0.71 | <.0001 |
| Substance Use | |||
| ≥2 substances | 1 | -- | -- |
| Alcohol only | 0.72 | 0.61–0.85 | .0001 |
| Nicotine only | 1.02 | 0.77–1.36 | .88 |
| Cannabis only | 0.91 | 0.70–1.17 | .45 |
| Other drug(s)a only | 0.51 | 0.43–0.61 | <.0001 |
| Substance Use Screening Method | |||
| Positive self-report, negative toxicology | 1 | -- | -- |
| Positive self-report, unknown toxicology | 0.71 | 0.61–0.83 | <.0001 |
| Negative self-report, positive toxicology | 1.31 | 1.02–1.70 | .04 |
| Positive self-report, positive toxicology | 2.01 | 1.60–2.54 | <.0001 |
Note.
Including anxiety medication, cocaine, heroin, methadone/buprenorphine, methamphetamine, opioids/pain medication, other street drugs, and sleep medication. Results from multivariable logistic regression models that include all variables listed above in one model.
Discussion
Our study of factors associated with pregnant women’s participation in an initial substance use assessment and counseling intervention integrated into prenatal care has two key findings. First, most pregnant women who screened positive for substance use via universal prenatal screening participated (83%). Second, women who entered prenatal care in the first trimester, those with lower parity, and those from groups who have systematically experienced greater obstacles to healthcare (e.g., those with lower income) were more likely to participate (Department of Health and Human Services, 2008). These findings suggest that removing barriers to engagement by embedding substance use interventions within prenatal care at no cost-share may reduce disparities in access, contributing to health equity.
Early intervention can make a critical difference in the course of prenatal substance use and its associated outcomes. Although many women discontinue substance use when they learn that they are pregnant, some are unable to quit and continue to use substances well into the second and third trimesters (Salas-Wright, Vaughn, Ugalde, & Todic, 2015). Research has shown that earlier discontinuation of prenatal substance use may decrease risks to the woman and the developing fetus, particularly during the first trimester (Polakowski, Akinbami, & Mendola, 2009) and interventions to prevent continued prenatal substance use significantly improve the health of mothers and their children. Our prior research indicates that among women at risk for prenatal substance use, those who do not participate in the initial substance use assessment and substance use counseling intervention have poorer perinatal outcomes and greater healthcare costs (Armstrong et al., 2003; Goler et al., 2008). Results from this study provide novel data indicating that integrated psychosocial assessment and counseling interventions for substance use offered at no cost-share within prenatal care can successfully reach traditionally underserved and vulnerable populations of pregnant women.
Results also highlight the potential need to improve linkage of pregnant women with greater parity and higher socioeconomic status with substance use interventions. Women with children were significantly less likely to participate, possibly due to competing demands on time or lack of childcare. Lower participation among higher socioeconomic women may be due to patient factors such as greater availability of other resources and better access to addiction treatment external to KPNC, greater perceived stigma, or less interest in receiving resources. Alternatively, women with lower socioeconomic status may be more concerned about potential legal consequences and child protective services involvement and might believe that participation might mitigate future consequences. Further, lower participation among women with higher socioeconomic status may be due in part to clinician factors and their delivery of services. For example, clinicians may perceive that women with lower socioeconomic status have more acute substance use problems and might work harder to link them to the intervention. If women of higher socioeconomic status are ultimately less likely to get care they need, organizational strategies may be needed to better engage these patients and their providers to enhance linkage with prenatal interventions. Additional research is needed to understand the mechanisms underlying lower engagement in substance use assessment and counseling embedded within prenatal care in this population.
Interestingly, compared to women who screened positive for more than one substance use category, those who screened positive for alcohol use only or other drug use only had lower odds of participation. This may be because women who use more substances are in greater need of assistance and thus more likely to participate. Conversely, women who only screened positive for alcohol use may have used infrequently or in low quantities or could be reporting on use before they realized they were pregnant, and may not perceive themselves to be at risk or in need of intervention (e.g., if they stopped drinking upon learning they were pregnant).
Our category of other drugs included both illicit drugs (e.g., cocaine, methamphetamine, heroin) and prescription drugs (e.g., opioid/pain medications, sleep medications, anxiety medications). In our sample, the majority of women who only screened positive for other drug use used opioid/pain medications (55.6%), barbiturates/sleep medications (23.4%), or benzodiazepines/anxiety medications (28.8%), and few screened positive for illicit substance use. While we do not have data on whether these medications were misused or diverted from others, women may have had a legal prescription to use these drugs for a clear medical indication, and might not have considered themselves at risk for misuse or in need of additional substance use services during pregnancy. Further research is needed to understand differences in participation in substance use interventions integrated into prenatal care among women who use different types of substances.
We also found that relative to women who disclosed substance use but had a negative toxicology test, those who had a positive toxicology test were more likely to participate regardless of whether they self-reported use. This suggests that the intervention reaches some women who would not have been identified based only on self-reported substance use. Greater participation among women who screen positive by urine toxicology testing may be due to greater severity of substance use problems, greater motivation to get help, or concerns about possible consequences of a positive toxicology test and participation in an attempt to mitigate later consequences (e.g., child protective services involvement). It is also notable that among women who screened positive for prenatal substance use, those with an “unknown” status for depression symptoms, “other” or “unknown” race/ethnicity, or “unknown” toxicology test results were significantly less likely to participate. Women who are more cautious about disclosing their personal information with the healthcare system may be more private in general and thus less inclined to consent to a urine toxicology test or discuss their substance use with their clinician. Missing data might also reflect lower engagement with the healthcare system in general or factors related to non-participation in prenatal substance use counseling, such as late entry into prenatal care. Additional research is needed to understand this gap in care.
This study has several important limitations. Our study took place within a large, integrated healthcare system in California and findings may not be generalizable to pregnant women without access to healthcare or to other states with more punitive laws around prenatal substance use. We are limited to data from the EHR and additional unmeasured factors (e.g., motivation to quit) likely also contribute to participation. Although we include prenatal substance use via both self-report and urine toxicology testing at approximately 8 weeks gestation (except nicotine and heroin which were self-reported only), women may not disclose prenatal substance use, and the duration substances are detectable in urine after last use varies by drug type and other factors such as heaviness of use. Further, our use of a dichotomous measure of any substance use does not reflect frequency or quantity of use, or whether prescription drugs were prescribed for clear medical indications and used as directed, and additional studies are needed to better understand how these factors relate to participation in prenatal substance use interventions integrated into prenatal care.
Conclusions
Our results indicate that a no cost-share substance use assessment and counseling intervention embedded within prenatal care reaches most women who screen positive for substance use in pregnancy, representing an important pathway for overcoming traditional healthcare obstacles. Findings suggest that substance use assessment and counseling integrated into prenatal care may be an effective way to potentially reduce health disparities by engaging pregnant women from traditionally underserved populations, particularly when women are able to enter prenatal care and receive counseling early in pregnancy.
Acknowledgments
This study was supported by a NIH NIDA K01 Award (DA043604) and a grant from the Kaiser Permanente Northern California Community Benefit Program.
Footnotes
Disclosure: The authors report no conflicts of interest.
References
- American College of Obstetricians Gynecologists Committee on Health Care for Undeserved Women (2006). ACOG Committee Opinion No. 343: Psychosocial risk factors: Perinatal screening and intervention. Obstetrics and Gynecology, 108, 469–477. [DOI] [PubMed] [Google Scholar]
- Armstrong MA, Gonzales Osejo V, Lieberman L, Carpenter DM, Pantoja PM, & Escobar GJ (2003). Perinatal substance abuse intervention in obstetric clinics decreases adverse neonatal outcomes. Journal of Perinatology, 23, 3–9. [DOI] [PubMed] [Google Scholar]
- Armstrong MA, Lieberman L, Carpenter DM, Gonzales VM, Usatin MS, Newman L, et al. (2001). Early Start: an obstetric clinic-based, perinatal substance abuse intervention program. Quality Management in Health Care, 9, 6–15. [DOI] [PubMed] [Google Scholar]
- Clark KA, Dee DL, Bale PL, & Martin SL (2001). Treatment compliance among prenatal care patients with substance abuse problems. American Journal of Drug and Alcohol Abuse, 27, 121–136. [DOI] [PubMed] [Google Scholar]
- Committee on Obstetric Practice (2017a). Committee opinion no. 711: Opioid use and opioid use disorder in pregnancy. Obstetrics and Gynecology, 130, e81–e94. [DOI] [PubMed] [Google Scholar]
- Committee on Obstetric Practice (2017b). Committee opinion no. 722: Marijuana use during pregnancy and lactation. Obstetrics and Gynecology, 130, e205–e209. [DOI] [PubMed] [Google Scholar]
- Corrarino JE, Williams C, Campbell WS 3rd, Amrhein E, LoPiano L, & Kalachik D (2000). Linking substance-abusing pregnant women to drug treatment services: a pilot program. Journal of Obstetric, Gynecologic, and Neonatal Nursing, 29, 369–376. [DOI] [PubMed] [Google Scholar]
- Department of Health and Human Services. (2008). The Secretary’s Advisory Committee on National Health Promotion and Disease Prevention Objectives for 2020, Phase I report: Recommendations for the framework and format of Healthy People 2020. Retrieved January 31, 2020, from https://www.healthypeople.gov/sites/default/files/PhaseI_0.pdf. [Google Scholar]
- Goler NC, Armstrong MA, Osejo VM, Hung YY, Haimowitz M, & Caughey AB (2012). Early start: a cost-beneficial perinatal substance abuse program. Obstetrics and Gynecology, 119, 102–110. [DOI] [PubMed] [Google Scholar]
- Goler NC, Armstrong MA, Taillac CJ, & Osejo VM (2008). Substance abuse treatment linked with prenatal visits improves perinatal outcomes: a new standard. Journal of Perinatology, 28, 597–603. [DOI] [PubMed] [Google Scholar]
- Haller DL, Miles DR, & Dawson KS (2003). Factors influencing treatment enrollment by pregnant substance abusers. American Journal of Drug and Alcohol Abuse, 29, 117–131. [DOI] [PubMed] [Google Scholar]
- Haug NA, Duffy M, & McCaul ME (2014). Substance abuse treatment services for pregnant women: psychosocial and behavioral approaches. Obstetrics and Gynecology Clinics of North America, 41, 267–296. [DOI] [PubMed] [Google Scholar]
- Howell EM, & Chasnoff IJ (1999). Perinatal substance abuse treatment. Findings from focus groups with clients and providers. Journal of Substance Abuse Treatment, 17, 139–148. [DOI] [PubMed] [Google Scholar]
- Jackson A, & Shannon L. (2012). Barriers to receiving substance abuse treatment among rural pregnant women in Kentucky. Maternal and Child Health Journal, 16, 1762–1770. [DOI] [PubMed] [Google Scholar]
- Krans EE, Bobby S, England M, Gedekoh RH, Chang JC, Maguire B, et al. (2018). The Pregnancy Recovery Center: A women-centered treatment program for pregnant and postpartum women with opioid use disorder. Addictive Behaviors, 86, 124–129. [DOI] [PubMed] [Google Scholar]
- Messer K, Clark KA, & Martin SL (1996). Characteristics associated with pregnant women’s utilization of substance abuse treatment services. American Journal of Drug and Alcohol Abuse, 22, 403–422. [DOI] [PubMed] [Google Scholar]
- Ordean A, & Kahan M. (2011). Comprehensive treatment program for pregnant substance users in a family medicine clinic. Canadian Family Physician, 57, e430–435. [PMC free article] [PubMed] [Google Scholar]
- Paine LL, & Garceau LM (1999). Health behaviors during pregnancy: risks and interventions In McCormick MC & Siegel JE (Eds.), Prenatal Care Effectiveness and Implementation (pp. 33–62). New York, NY: Cambridge University Press. [Google Scholar]
- Polakowski LL, Akinbami LJ, & Mendola P. (2009). Prenatal smoking cessation and the risk of delivering preterm and small-for-gestational-age newborns. Obstetrics and Gynecology, 114, 318–325. [DOI] [PubMed] [Google Scholar]
- Salas-Wright CP, Vaughn MG, Ugalde J, & Todic J. (2015). Substance use and teen pregnancy in the United States: evidence from the NSDUH 2002–2012. Addictive Behaviors, 45, 218–225. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Substance Abuse and Mental Health Services Administration. (2013). Results from the 2012 National Survey on Drug Use and Health: Summary of National Findings. NSDUH Series H-46, HHS Publication No. (SMA) 13–4795. Retrieved January 31, 2020, from http://www.samhsa.gov/data/sites/default/files/NSDUHresults2012/NSDUHresults2012.pdf. [Google Scholar]
- Sutter MB, Gopman S, & Leeman L. (2017). Patient-centered care to address barriers for pregnant women with opioid dependence. Obstetrics and Gynecology Clinics of North America, 44, 95–107. [DOI] [PubMed] [Google Scholar]
- Sweeney PJ, Schwartz RM, Mattis NG, & Vohr B. (2000). The effect of integrating substance abuse treatment with prenatal care on birth outcome. Journal of Perinatology, 20, 219–224. [DOI] [PubMed] [Google Scholar]
- Terplan M, McNamara EJ, & Chisolm MS (2012). Pregnant and non-pregnant women with substance use disorders: the gap between treatment need and receipt. Journal of Addictive Diseases, 31, 342–349. [DOI] [PubMed] [Google Scholar]
- Work Group on Women, Drinking, and Pregnancy of the Interagency Coordinating Committee on Fetal Alcohol Syndrome,, National Institute on Alcohol Abuse and Alcoholism, & American Legacy Foundation. (2008). Preventing alcohol, tobacco, and other substance-exposed pregnancies: A community affair. Retrieved January 31, 2020, from http://www.niaaa.nih.gov/about-niaaa/our-work/ICCFASD/proceedings/2008. [Google Scholar]
- Young-Wolff KC, Sarovar V, Tucker LY, Avalos LA, Alexeeff S, Conway A, et al. (2019). Trends in marijuana use among pregnant women with and without nausea and vomiting in pregnancy, 2009–2016. Drug and Alcohol Dependence, 196, 66–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Young-Wolff KC, Sarovar V, Tucker LY, Avalos LA, Conway A, Armstrong MA, et al. (2018). Association of nausea and vomiting in pregnancy with prenatal marijuana use. JAMA Internal Medicine, 178, 1423–1424. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Young-Wolff KC, Sarovar V, Tucker LY, Conway A, Alexeeff S, Weisner C, et al. (2019). Self-reported daily, weekly, and monthly cannabis use among women before and during pregnancy. JAMA Network Open, 2, e196471. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Young-Wolff KC, Tucker LY, Alexeeff S, Armstrong MA, Conway A, Weisner C, et al. (2017). Trends in self-reported and biochemically tested marijuana use among pregnant females in California from 2009–2016. JAMA, 318, 2490–2491. [DOI] [PMC free article] [PubMed] [Google Scholar]