Independent risk factors for chronic illicit substance use during pregnancy

Abstract

Objective:

We aimed to determine the incidence of chronic illicit substance use during pregnancy and to identify associated risk factors.

Design:

A 2-year time-matched retrospective maternal quality control database (n = 4,470) analysis of parturients with chronic illicit substance use compared to controls.

Setting:

A tertiary academic medical center located in a rural setting.

Results:

The rate of chronic illicit substance use was 1.95 percent. Demographic factors associated with chronic illicit substance use in pregnancy-included lower body mass index (BMI; OR: 0.93; 95 percent CI: 0.89–0.96, p < 0.0001), higher gravidity (OR: 1.24; 95 percent CI: 1.13–1.36, p < 0.0001), higher parity (OR: 1.38; 95 percent CI: 1.22–1.57, p < 0.0001), and more live births (OR: 1.30; 95 percent CI: 1.16–1.46, p < 0.0001). A history of smoking (OR: 10.51; 95 percent CI: 5.69–19.42, p < 0.0001), alcohol use (OR: 48.98; 95 percent CI: 17.33–138.40, p < 0.0001), anxiety (OR: 1.88; 95 percent CI: 1.16–3.05, p = 0.01), depression (OR: 2.44; 95 percent CI: 1.55–3.85, p = 0.0001), transfer on admission (OR: 2.12; 95 percent CI: 1.16–3.87, p = 0.01), payor insurance (OR: 2.12, 95 percent CI: 2.10–5.04, p < 0.0001), and Apgar scores <7 at 1 minute (OR: 0.50; 95 percent CI: 0.25–1.00, p = 0.049) were significant. Multiple variable logistic regression-revealed BMI, smoking, alcohol use, and Apgar score <7 at 1 minute as significant factors.

Conclusions:

Awareness of these factors can assist in identifying and treating parturients with chronic illicit substance use.

INTRODUCTION

Opioid consumption has been increasing over recent years in all demographics, including pregnant women.¹ From 2000 through 2014, almost half a million people died from drug overdoses and most of the deaths involved opioids.² The Council of Economic Advisors estimates the true cost to society of drug abuse is over $500 billion annually, which amounted to 2.8 percent of the US GDP in 2015.³

The World Health Organization defines substance abuse as the harmful or hazardous use of psycho-active substances, including alcohol and illicit drugs.⁴ Substance abuse during pregnancy is a significant public health issue and places a substantial burden on the healthcare system. In recent years, the rates of substance use disorders among pregnant women has been increasing, posing a health risk to the mother, fetus, neonate, and developing child.⁵ Research in this area has been limited due to widespread underreporting of maternal substance abuse.⁵ Limited data and lack of specifically targeted research studies have not only made it difficult to estimate the prevalence of substance abuse in pregnancy but also limits the ability of obstetric providers to effectively screen for and manage this condition during pregnancy. The aim of this study was to examine associations between sociodemographic factors, mental health characteristics, and chronic illicit drug use during pregnancy.

METHODS

The West Virginia University Institutional Review Board approved this investigation of pregnant women aged 14–47 years from the West Virginia University Maternal Quality Control Assurance database involving years 2017 through 2018. Demographic-measured variables in quality control database included maternal age, gestation, gravida, parity, prior live births, height, weight, body mass index (BMI), race, ethnicity, type of medical insurance payer, and rural county. Maternal medical-measured variables included smoking status, alcohol use, history of opioid use, intravenous drug abuse, heroin addiction, polysubstance abuse, alcohol abuse, depression, and anxiety. The two most common abused intravenous drugs include synthetic opioids (mainly fentanyl) and heroin.⁶

Obstetrical maternal-measured variables included source of admission, admission and discharge time, anesthesia type (epidural, spinal general, combined spinal-epidural, local, and none), delivery method (vaginal, instrumental vaginal, and cesarean), obstetrical delivery procedures (tubal/salpingectomy, total abdominal hysterectomy, B-Lynch sutures, and none), and total length of hospital stay. Neonatal-measured variables included Apgar scores at 1 and 5 minutes after delivery and live birth status after delivery. Our quality control coordinator (Jennifer Ball) maintains and codes the information in the Maternal Quality Control Assurance database based on the deliveries documented within our electronic medical record. Measured variable information in the West Virginia University Maternal Quality Control Assurance database was based on electronic medical record diagnosis and not International Classification of Diseases or the diagnostic and statistical manual of mental disorders codes. For the purpose of this study, chronic illicit substance use is defined as having one or more of four maternal measured study variables (opioid abuse, intravenous abuse, heroin addiction, and polysubstance use disorder).

Statistical analysis and power

Statistical analysis was performed using SAS v. 9.4 software (SAS Institute, Inc., 2013. Cary, North Carolina). Categorical variables were described using valid percentages and frequencies; continuous variables with means and standard deviations and ordinal variables with medians and interquartile range. Differences between demographics and other variables of interest by chronic illicit substance use was tested using χ²-square or Fisher’s exact test. The primary research question of interest was further tested using standard multivariable logistic regression; parameter estimates were exponentiated and presented as odds ratios with 95 percent confidence intervals. Alpha was set to 0.05 unless otherwise stated.

Appropriate sample size was determined based on prior research suggesting illicit substance use in all pregnant women to be approximately 6 percent.⁷ Given the data are taken at delivery with four defined areas of substance use (opioids, IV use, heroin, and polysubstance), the percentage is expected to be slightly lower. At 2.5 percent, a sample of 4,000 would give approximately 100 women with illicit substance use; at 90 percent power would be ample to detect odd ratios (ORs) as small as 1.4. A final sample of 4,470 was obtained.

RESULTS

Maternal demographic data are presented in Table 1. Out of the sample of 4,470 women, 87 were classified as having chronic illicit substance use (1.95 percent). Within the chronic illicit substance use group, 48 were classified as having opioid abuse (1.07 percent), 17 with intravenous abuse (0.38 percent), nine with heroin addiction (0.20 percent), and 20 with polysubstance use disorder (0.45 percent). Of the chronic illicit substance use group, six patients were classified in more than one category (0.13 percent). No differences were noted with respect to age, gestation, ethnicity, race, or rural county.

Table 1.

Maternal data by chronic illicit substance use group, N = 4470

	Chronic illicit substance use (n = 87)	Nonillicit substance use (n = 4,383)	p-value
Chronic substance use group	87	0	-
Opioid use	48	0
IV drug abuse	17	0
Heroin addiction	9	0
Polysubstance disorder	20	0
Age (years)1	28.0 ± 4.9	28.5 ± 5.7	0.39
BMI (kg/m2)1	29.3 ± 6.2	32.8 ± 7.7	<0.001
Gestation (weeks)1	35.5 ± 4.5	35.8 ±4.2	0.54
Gravida2	3 [2–4]	2 [1–3]	<0.001
Parity2	1 [1–2]	1 [0–1]	<0.001
Live births2	2 [1–3]	2 [0–2]	<0.001
Race (white)	77 (89.5 percent)	3926 (89.8 percent)	0.36
Nonhispanic ethnicity	85 (98.8 percent)	4323 (98.8 percent)	0.86
Smoker	73 (85.9 percent)	1603 (36.7 percent)	<0.001
Alcohol use	8 (0.2 percent)	7 (8.1 percent)	<0.001
History of anxiety	24 (27.9 percent)	721 (16.5 percent)	0.005
History of depression	30 (34.9 percent)	765 (17.5 percent)	<0.001
Medical assistance	36 (41.9 percent)	806 (18.4 percent)	<0.001
Admission transfer	13 (15.1 percent)	343 (7.9 percent)	0.01
Rural county area	28 (32.6 percent)	1484 (33.9 percent)	0.79
Anesthesia type			0.65
Epidural	48 (59.3 percent)	2259 (53.9 percent)
Spinal	15 (18.5 percent)	980 (23.4 percent)
General	5 (6.2 percent)	149 (3.6 percent)
Combined spinal-epidural	2 (2.5 percent)	110 (2.6 percent)
Local	0 (0 percent)	83 (2.0 percent)
None	10 (12.4 percent)	540 (12.9 percent)
Delivery method			0.39
CD	25 (29.1 percent)	1542 (35.3 percent)
Vaginal	57 (66.3 percent)	2697 (61.7 percent)
Instrumental vaginal	4 (4.7 percent)	135 (3.1 percent)
OB delivery procedures			0.68
Tubal/salpingectomy	8 (10.4 percent)	309 (7.7 percent)
TAH	0 (0 percent)	7 (0.2 percent)
B-lynch sutures	0 (0 percent)	5 (0.1 percent)
None	69 (89.6 percent)	3674 (91.7 percent)
Apgar 1 < 7	9 (10.7 percent)	855 (19.7 percent)	0.04
Apgar 5 < 9	19 (22.9 percent)	1205 (27.7 percent)	0.33

Legend:

¹Means and SD, p value based on T-test.

²Median [IQR]; p value based on Mann-Whitney test. All other p values based on chi-square or Fisher’s exact test with Monte Carlo estimation.

Table 1 also presents maternal, neonatal, and anesthetic procedures. No differences were noted with respect to types of anesthesia (epidural, spinal general, combined spinal-epidural, local, and none), delivery method (cesarean delivery, vaginal, and instrumental vaginal), and obstetrical delivery procedures (tubal/salpingectomy, total abdominal hysterectomy, B-Lynch sutures, and none). Regarding neonatal outcome, no differences were noted with Apgar scores <9 at 5 minutes.

The chronic illicit substance use group had a lower BMI (p = 0.001), a higher gravidity (p < 0.001), higher parity (p < 0.001), and more live births (p < 0.001) compared to controls (Table 1). A history of smoking (p < 0.001), alcohol use (p < 0.001), anxiety (p = 0.002), depression (p < 0.001), transfer of admission (p < 0.001), and having medical assistance payor insurance (p < 0.001) were also higher in the chronic illicit substance use group (Table 1). The chronic illicit substance use group had better Apgar scores <7 at 1 minute (p = 0.04) compared to the control group (Table 1).

Controlling for all significant bivariate associations (gravidity, parity, BMI, payor status, live births, history of smoking, history of alcohol use, history of depression and anxiety, and Apgar score at 1 minute), multiple variable logistic regression revealed BMI, smoking, alcohol use, and Apgar score <7 at 1 minute to be significant after controlling for all other associations (Table 2).

Table 2.

Bivariate and multivariable logistic regression results, associations with chronic substance use group, n = 4,459

Parameter	Unadjusted OR	p value	Adjusted OR	Lower Adj. OR 95 percent CI	Upper Adj. OR 95 percent CI	p value
BMI (kg/m2)	0.93	<0.0001	0.93	0.89	0.96	0.0001
Live births	1.30	<0.0001	0.95	0.71	1.29	0.75
Gravida	1.24	<0.0001	1.10	0.91	1.31	0.33
Parity	1.38	<0.0001	1.19	0.82	1.75	0.36
Smoker	10.51	<0.000l	7.34	3.75	14.39	<0.0001
Alcohol use	48.98	<0.0001	26.33	8.39	82.55	<0.0001
Depression	2.44	0.0001	1.59	0.92	2.74	0.10
Anxiety	1.88	0.01	1.18	0.67	2.10	0.56
Medical assistance	3.25	<0.0001	1.32	0.80	2.19	0.25
Admission transfer	2.12	0.01	1.57	0.79	3.12	0.20
Apgar <7 at 1 minute	0.50	0.049	0.43	0.20	0.91	0.03

DISCUSSION

Previous research has shown that 5 percent of pregnant women use one or more addictive substances.⁵ We found that the rate of chronic illicit substance use to be 2.0 percent in our pregnant patient population. Labor and delivery are the most common reason for hospitalization in the United States.⁸ As most patients receive an opioid during inpatient hospitalization, childbirth is likely to be the initial exposure to opioids for many women.⁸ This initial exposure may serve as the gateway for chronic substance use. Although providers have a responsibility to treat pain, it is now becoming clear that aggressive opioid prescription practices have contributed to an epidemic of abuse.²

Some women will require opioid therapy in labor and delivery, especially those already opioid dependent. Proper risk assessment before initiating or continuing opioid therapy is necessary.² A multimodal approach consisting of two or more analgesic agents or local anesthetic techniques that act by different analgesic mechanisms will provide improved pain relief while limiting adverse outcomes.⁹ In general, pregnant women and women of childbearing age should be counseled about the harmful effects of all medications.¹ This includes a discussion about risks and benefits of the prescribed medications, dose changes, dose tapering if necessary to avoid withdrawal symptoms, and additional support and resources to aid them in discontinuing if needed.^5,9

Nearly, two-thirds of women who smoke during pregnancy also use some other form of medication(s) or drug(s).¹⁰ Oga et al.¹⁰ found that smoking in pregnancy was associated with a positive urine drug screen with an adjusted odds ratio (aOR) of 4.7 (95 percent CI 2.6–8.3) for current smokers. In our study, we found the rate of smoking to be 85 percent in the chronic illicit drug use group compared to just 38 percent (p < 0.001) in the control group with an aOR of 7.34 (Table 2). Understanding that greater than 66 percent of women who smoke during pregnancy are also using some other drugs is important because this information should lead providers to provide screening and education to women using tobacco about other substance use and its adverse maternal and fetal effects.¹⁰

In our study, medical assistance payor insurance (OR: 3.25), anxiety (OR: 2.44), and depression (OR: 1.88) and were all found to be associated with chronic illicit drug use in pregnancy (Table 2). This is of clinical significance because depression and anxiety are linked to inadequate nutrition and weight gain, increased alcohol, and other substance abuse, poor adherence to perinatal care and adverse neonatal physical and behavioral outcomes.¹¹

The use of alcohol in pregnancy is also a contributor to chronic illicit drug use.^12,13 We found the rate to be 9 percent in the chronic substance group compared to 0.4 percent (p < 0.001) overall with an aOR of 26.33 (Table 2). This association between mental health and substance use disorders highlights the need to better treat psychiatric illness to reduce drug use during pregnancy. Treatment of maternal depression during pregnancy and afterward at the lowest effective dose is uniformly recommended despite the potential side effects to the fetus and newborn.¹⁴ Parturients with high-risk potential need to be drug screened and referred to mental health treatment programs if needed.

An interesting finding of our study was that no differences were noted with respect to intrapartum maternal, neonatal, and anesthetic outcomes between the two groups. This finding contradicts certain findings within the current literature. Opioidpolydrug use during pregnancy has been associated with a myriad of adverse outcomes for mothers and infants, such as placental abruption, preterm labor, cesarean delivery, transfusion, and extended hospital stays >7 days.¹⁵ Apgar scores <7 at 1 minute was less in the chronic illicit substance use group compared to controls both independently and on multivariable logistic regression. Recent data suggest that hospital delivery volume, healthcare provider patient volume, and specialized care, such as that received at an academic medical center, can all affect maternal outcomes.¹⁶ As this study was performed at a high-volume specialized tertiary care center, the mothers and neonates likely benefited from a higher acuity of care. The better 1-minute Apgar scores in the chronic illicit substance use group may be a result of increased vigilance and supportive care because these neonates are at higher risk for complications such as neonatal abstinence syndrome.

This study has several limitations. Retrospective studies are susceptible to incomplete data and code misclassification. Potential confounders may have influenced selection of provider types. Furthermore, the study was limited to a single institution with standard postpartum analgesia management.

CONCLUSION

Chronic illicit drug use during pregnancy is an alarming and often overlooked problem in our society that can lead to adverse perinatal and neonatal outcomes. A history of smoking, alcohol use, anxiety, depression, transfer on admission, and payor insurance are factors associated with chronic illicit drug use during pregnancy. Awareness of these factors can assist in focusing on more targeted prevention efforts aimed at educating women of the risk of prenatal substance use, and the need for nonstigmatizing treatment approaches for women who misuse opioids and other illegal drugs during all stages of pregnancy.

Data availability:

Data supporting the results reported is available upon request.