Adverse effects on birth weight of parental illegal drug use during pregnancy and within two years before pregnancy.

This study investigates possible links between maternal illegal drug use during pregnancy and up to two years before pregnancy with birth weight (BW), and explores the potential role of paternal illegal drug use on low birth weight. A population-based retrospective cohort study was conducted that linked four national databases in Taiwan. A total of 1,698 subjects with a criminal record of schedule I or II illegal drug use within two years before pregnancy were enrolled as the drug-exposed group, and 16,980 matched subjects were enrolled as the unexposed group. Multivariate analysis of BW found a decrease of 108.63 g (95% CI: -172.29, -44.96), 79.67 g (95% CI: -116.91, -42.43), and 69.78 g (95% CI: -106.71, -32.84) in newborns whose mothers used illegal drugs only during pregnancy (period I), only within one year before pregnancy (period II), and only within the second year before pregnancy (period III), respectively. Paternal use of illegal drugs before maternal pregnancy was significantly associated with low birth weight. The paternal effect on low birth weight was opposite the maternal effect. The adverse effect of illegal drug use on birth weight existed even if the mother did not use drugs during pregnancy but had ever used drugs during the two years before pregnancy. Paternal factors' contribution to low birth weight persisted, and the decrement of BW was even greater than the maternal effect within one or two years before pregnancy. Maternal and paternal illegal drug use may have a lasting effect on their offspring's birth weight.

1. Introduction

Drug abuse has become a critical public health issue in many parts of the world, and a growing number of women use illegal drugs even during pregnancy. According to a national survey in the United States in 2012, 5.9% of pregnant women used illicit drugs [1]. Prenatal illegal drug use can cause deleterious consequences in terms of pregnancy outcomes, including preterm labor, small-for-gestational age neonates, and low birth weight (LBW) [2-6]. As LBW accounts for most neonatal mortality and morbidity [7], researchers are becoming increasingly interested in the association between drug use and LBW, especially the timing of drug use [8,9], the paternal effect on LBW [10], and an adequate indicator to measure the exposure to illegal drug use [11].

Most studies to date have established the effects of time periods of illegal drug use through recorded data for short time periods, such as early pregnancy [12], six months before pregnancy [13], and three months before pregnancy [14]. Until now, to the best of our knowledge, no study has investigated the association of illegal drug use with LBW so far in advance of pregnancy as two years.

To determine the timing of illegal drug use, self-reporting, maternal or infant hair samples, and infant meconium have been used to clarify the association of illegal drug use with LBW, but there are some disadvantages to these measures [11,13,15-18]. To avoid our study being affected by these disadvantages, it used drug-related criminal records (DCRs) held by police departments as an indicator to determine whether women used illegal drugs or not.

Most studies aimed at the association between illegal drug use and birth outcomes recruited health-seeking samples and therefore are susceptible to selection bias, information bias, and disparities in health care, especially in vulnerable populations [19,20]. Pregnant women with illegal drug history often experience more severe social and structural barriers in seeking medical services due to social stigma, low social support, and disadvantaged socioeconomic condition [21,22]. Therefore, it’s critical to tackle the reproductive outcome for this hidden population.

The majority of prior research has emphasized the impact of maternal characteristics on birth weight (BW). Many experts believe that maternal factors are more significant than paternal factors [10,23,24]; however, the effects of paternal drug use on BW remain controversial [25]. Therefore, it would be useful to understand the adverse effects of paternal drug use on BW.

The current study took these methodological limitations into account, and is one of the first to focus on the relationship between illegal drug use within one or two years before pregnancy and LBW in a large population-based cohort. We analyzed the nationwide Integrated Illegal Drug Database (IIDD), which included DCRs of schedule I and schedule II drugs from 2004 to 2017 in Taiwan. The aims of this study were to determine (1) the association between maternal illegal drug use and LBW, (2) whether the impact of illegal drug use on BW exists even long before pregnancy (up to two years), and (3) the contribution of paternal factors to LBW.

2. Methods

2.1. Data sources

In this study, four national databases, all of which are under the supervision of the Ministry of Health and Welfare, Taiwan, were linked. Those databases were as follows: (1) the 2004–2014 Taiwan Maternal and Child Health database, in which parent and child IDs are linked together; (2) the National Health Insurance Research Database (NHIRD), which is derived from the National Health Insurance Administration and contains 99% of the total population’s medical care claims data since 1995 [26]; (3) the Birth Notification System (BNS) from 2004 to 2014, under which it is mandatory to report to the government newborns weighing more than 500 g or with a gestational age greater than 20 weeks [27]; and (4) the IIDD from the Health and Welfare Data Science Center, which consists of 20 illegal-drug-related databases from different governmental agencies. The IIDD provided information on national drug scheduling and types of DCR (e.g., drug consumption, possession, manufacturing, trafficking, and distribution) from 2001 to 2017. All four databases are derived from administrative databases of Taiwan’s government; the first two have been especially useful resources for population health sciences studies [28,29].

This study was approved by the Institutional Review Board of the National Health Research Institutes, Taiwan (IRB number: EC 1070601-E). All personal data obtained were anonymized before analysis, and informed consent was thus waived.

2.2. Study population

In this retrospective cohort study, 1,272,723 pregnant women undergoing their first childbirth were selected in the first step to link the Taiwan Maternal and Child Health dataset with the NHIRD, and work was then carried out to link this dataset with the BNS from 2004 to 2014. A total of 81,114 (6.4%) of the 1,272,723 subjects of pregnant women were excluded: those under 12 years old (n = 2), those who had multiple births (n = 22,653), and those missing a paternal identification number (n = 58,459). The remaining 1,191,609 subjects were linked with the IIDD to determine whether the pregnant women had any DCR. In total, exactly 7,500 subjects were confirmed as having used drugs; the remaining 1,184,109 subjects had no DCR. Among the 7,500 subjects, a total of 1,698 had used a schedule I or schedule II illegal drug at least once during pregnancy or within two years before pregnancy; and 16,980 subjects were randomly selected from the non-illegal drug users to match illegal drug users (1:10) according to maternal age and year of childbirth (Fig. 1).

Fig. 1

Flow diagram of the study subjects.

2.3. Measures

2.3.1. Exposure information

Drug exposure was defined as having at least one DCR of schedule I or schedule II drug use during pregnancy or within two years before pregnancy. Those with DCRs of only drug possession, manufacturing, trafficking, or distribution were excluded from the analysis. The drug users must have tested positive during urine drug screening and then recorded as having a DCR by a police department. In addition, to explore the relationship between maternal ever use of illegal drugs and BW, the time of drug exposure was categorized into three periods. Period I was defined as having at least one DCR during pregnancy but without any record within two years before pregnancy. Period II was defined as at least one DCR within one year before pregnancy but without any record during pregnancy or within the second year before pregnancy. Period III was defined as at least one DCR use within the second year before pregnancy but without any record during pregnancy or within one year before pregnancy. Other combinations of DCR and period before or during pregnancy are possible; however, this study examines only periods I, II, and III to maintain clear-cut durations for comparison.

2.3.2. Outcome information

BW was retrieved from birth certificates in the BNS, for which weight was measured using a digital scale at birth and recorded in grams [27].

2.3.3. Confounding variable information

Information regarding income, urbanization level of residence (urban, suburban, rural), maternal comorbidities (diabetes mellitus, hypertension, hyperlipidemia, chronic obstructive pulmonary disease (COPD) [30-32], and pregnancy-related complications (gestational hypertension, pre-eclampsia, gestational diabetes mellitus) [33-35] were retrieved from the NHIRD. Maternal comorbidities were defined as applying to those who had inpatient or outpatient care with the diagnosis of diabetes mellitus, hypertension, hyperlipidemia or COPD during the two years before pregnancy. Pregnancy-related complications were defined as those of subjects who had inpatient or outpatient care with a diagnosis of gestational hypertension, pre-eclampsia, or gestational diabetes mellitus during pregnancy. Income was categorized into four levels using the payroll bracket table published by the Ministry of Health and Welfare’s National Health Insurance Administration, and recorded in Taiwan dollars (TWD) [36]. Information regarding gestational age (≥37 weeks, <37 weeks) [3] and mode of delivery (vaginal, cesarean) were retrieved from the BNS. Paternal drug use was defined as at least one DCR of schedule I or schedule II drug use during his wife’s pregnancy or within two years before that time.

2.4. Statistical analysis

Data are presented as means ± standard deviations (SD) for continuous variables and as numbers (percentages) for categorical variables. The independent t-test for continuous variables and Pearson’s χ2 test for categorical variables were used for analysis. Multivariate linear regression models were used to determine the adverse effects of maternal illegal drug use on the risk of LBW, as shown by the beta coefficient (β) with a 95% confidence interval (CI). Finally, in order to assess the robustness of the results, we repeated analyses of factors associated with BW according to the time period of drug exposure. All data were analyzed using SAS statistical software version 9.4 (SAS Institute, Cary, NC, USA). A p value <0.05 was considered to indicate statistical significance.

3. Results

Table 1 shows the basic characteristics of the study population. A higher percentage of drug-exposed subjects had a significantly lower socioeconomic status (66.5%) (p < 0.001) and lived in regions of lower urbanization (51.5%) (p < 0.001). In the drug-exposed group, 33.9% (p < 0.001) of all parents were drug users, and 60.1% (p < 0.001) of the subjects had a vaginal delivery. The prevalences of maternal comorbidities and pregnancy-related complications among the drug-exposed group were comparable to the matched unexposed group, except the prevalences of hyperlipidemia (0.6% vs. 1.2%, p = 0.033) and gestational diabetes mellitus (0.6% vs. 2.3%, p < 0.001) were lower in the drug-exposed group than in the matched unexposed group. In addition, the drug-exposed subjects were more likely than the matched unexposed group to have a baby that was small for the gestational age (12.7% vs. 6.5%, p < 0.001). Of the 1,698 drug-exposed subjects, the rates of drug use were 16.4%, 33.0%, and 30.9% for periods I, II, and III, respectively.

Table 1

Characteristics of study subjects by drug exposure.

Characteristic	Drug-exposed	Matched unexposed	p-value
	(n = 1698)	(n = 16980)
	n (%)	n (%)
Maternal age (years)			1.000
<25	599 (35.3)	5990 (35.3)
25–29	579 (34.1)	5790 (34.1)
30–34	368 (21.7)	3680 (21.7)
≥35	152 (9)	1520 (9)
Year of childbirth			1.000
2004	66 (3.9)	660 (3.9)
2005	257 (15.1)	2570 (15.1)
2006	254 (15)	2540 (15)
2007	171 (10.1)	1710 (10.1)
2008	210 (12.4)	2100 (12.4)
2009	153 (9)	1530 (9)
2010	116 (6.8)	1160 (6.8)
2011	126 (7.4)	1260 (7.4)
2012	118 (6.9)	1180 (6.9)
2013	112 (6.6)	1120 (6.6)
2014	115 (6.8)	1150 (6.8)
Paternal drug use			<0.001
No	1123 (66.1)	16906 (99.6)
Yes	575 (33.9)	74 (0.4)
Income (TWD)			<0.001
≤15840	1129 (66.5)	3053 (18)
15841–28800	497 (29.3)	9218 (54.3)
28801–45800	59 (3.5)	3511 (20.7)
>45800	13 (0.8)	1198 (7.1)
Urbanization of residence a			<0.001
Urban	823 (48.5)	9916 (58.4)
Suburban	298 (17.6)	2346 (13.8)
Rural	576 (33.9)	4711 (27.8)
Mode of delivery			<0.001
Vaginal delivery	1020 (60.1)	11666 (68.7)
Cesarean section	678 (39.9)	5314 (31.3)
Maternal comorbidities
Diabetes mellitus	8 (0.5)	81 (0.5)	0.973
Hypertension	13 (0.8)	109 (0.6)	0.546
Hyperlipidemia	10 (0.6)	196 (1.2)	0.033
COPD	44 (2.6)	475 (2.8)	0.622
Pregnancy-related complications
Gestational hypertension	10 (0.6)	100 (0.6)	1.000
Pre-eclampsia	22 (1.3)	197 (1.2)	0.621
Gestational diabetes mellitus	11 (0.6)	399 (2.3)	<0.001
Gestational age			<0.001
≥37 weeks	1482 (87.3)	15875 (93.5)
<37 weeks	216 (12.7)	1105 (6.5)
Exposure period
Period I	278 (16.4)
Period II	561 (33.0)
Period III	524 (30.9)

Missing:8.

Multivariable analysis showed that maternal drug-exposed subjects (−100.25 g, 95% CI: −123.77, −76.74), those whose husbands had ever used illegal drugs (−64.33 g, 95% CI: −99.78, −28.87), those with hypertension (−87.37 g, 95% CI: −158.38, −16.36), those with pre-eclampsia (−269.40 g, 95% CI: −321.93, −216.86), and those whose neonate had a gestational age less than 37 weeks (−658.89 g, 95% CI: −680.68, −637.10) had a lower mean neonate BW as compared with their counterparts (Table 2). In addition, the neonatal BWs of mothers aged 25–29, 30–34, and ≥35 years were, respectively, 20.41 (95% CI: 6.80, 34.02), 42.28 (95% CI: 26.40, 58.17), and 54.68 g (95% CI: 33.07, 76.30), greater than those for infants of mothers aged <25, and BW declined by 4.87 g (95% CI: −6.77, −2.97) per year. However, pregnant women with a cesarean delivery, diabetes, or gestational diabetes were associated with a higher neonatal BW (Table 2).

Table 2

Multivariable linear regression of birth weight (n = 18,678).

Variables	Beta (β) coefficient	95% Confidence limits		p value
Mother used drugs	−100.25	−123.77	−76.74	<0.001
Father used drugs	−64.33	−99.78	−28.87	<0.001
Maternal age (years)
<25	reference
25–29	20.41	6.80	34.02	0.003
30–34	42.28	26.40	58.17	<0.001
≥35	54.68	33.07	76.30	<0.001
Year of childbirth	−4.87	−6.77	−2.97	<0.001
Income (TWD)
≤15840	reference
15841–28800	1.07	−13.72	15.86	0.887
28801–45800	−0.97	−19.58	17.64	0.919
>45800	15.73	−10.40	41.86	0.238
Urbanization of residence
Urban	reference
Suburban	1.74	−14.79	18.28	0.836
Rural	−6.11	−18.97	6.76	0.352
Gestational age
≥37 weeks	reference
<37 weeks	−658.89	−680.68	−637.10	<0.001
Mode of delivery
Vaginal delivery	reference
Cesarean section	50.48	38.42	62.54	<0.001
Maternal comorbidities
Diabetes mellitus	148.10	65.18	231.03	0.001
Hypertension	−87.37	−158.38	−16.36	0.016
Hyperlipidemia	30.52	−24.40	85.45	0.276
COPD	−13.95	−47.59	19.70	0.417
Pregnancy-related complications
Gestational hypertension	−46.40	−118.94	26.15	0.210
Pre-eclampsia	−269.40	−321.93	−216.86	<0.001
Gestational diabetes mellitus	94.23	55.92	132.54	<0.001

To explore whether illegal drug use long before pregnancy had an impact on neonatal BW, multi-variable regression models of BW by period of illegal drug use were performed. The results (Table 3) showed a gradient of decreasing BW with duration of maternal illegal drug use. Subjects exposed to illegal drugs had a 108.63 g decrease (95% CI: −172.29, −44.96) in BW in period I, a 79.67 g decrease (95% CI; −116.91, −42.43) in period II, and a 69.78 g decrease (95% CI: −106.71, −32.84) in period III. Moreover, paternal use of illegal drugs was significantly associated with LBW in both period II (−115.05 g, 95% CI: −223.89, −6.21) and period III (−128.33 g, 95% CI: −254.05, −2.60). The trend of BW decrease affected by paternal factor was opposite that of the maternal effect (Table 3).

Table 3

Multivariate analysis of factors associated with low birth weight.

Variables	Period I (n = 3058)				Period II (n = 6171)				Period III (n = 5764)
	Beta (β) coefficient	95% Confidence limits		p value	Beta (β) coefficient	95% Confidence limits		p value	Beta (β) coefficient	95% Confidence limits		p value
Mother used	−108.63	−172.29	−44.96	0.001	−79.67	−116.91	−42.43	<0.001	−69.78	−106.71	−32.84	<0.001
Father used a	−34.17	−128.99	60.65	0.480	−115.05	−223.89	−6.21	0.038	−128.33	−254.05	−2.60	0.045
Maternal age (years)
<25	reference				reference				reference
25–29	59.97	24.68	95.26	0.001	−17.63	−41.41	6.16	0.146	35.64	11.69	59.59	0.004
30–34	69.51	28.06	110.95	0.001	6.06	−22.04	34.16	0.673	64.48	35.05	93.92	<0.001
≥35	145.12	91.38	198.87	<0.001	34.21	−6.44	74.86	0.099	69.05	27.79	110.31	0.001
Year of childbirth	−7.47	−12.72	−2.22	0.005	−5.11	−8.34	−1.87	0.002	−4.08	−7.45	−0.71	0.018
Income (TWD)
≤15840	reference				reference				reference
15841–28800	9.64	−27.46	46.73	0.611	4.70	−21.57	30.97	0.726	6.56	−19.43	32.55	0.621
28801–45800	22.76	−24.02	69.54	0.340	5.46	−27.66	38.57	0.747	3.86	−30.26	37.98	0.824
>45800	59.48	−6.38	125.34	0.077	58.98	12.03	105.94	0.014	20.97	−27.54	69.48	0.397
Urbanization of residence
Urban	reference				reference				reference
Suburban	9.51	−31.98	51.00	0.653	−2.63	−32.26	27.00	0.862	−6.56	−36.27	23.15	0.665
Rural	4.85	−28.17	37.87	0.773	−11.57	−34.12	10.98	0.315	1.03	−22.21	24.26	0.931
Gestational age
≥37 weeks	reference				reference				reference
<37 weeks	−666.19	−721.27	−611.11	<0.001	−687.34	−727.11	−647.57	<0.001	−742.14	−782.37	−701.91	<0.001
Mode of delivery
Vaginal delivery	reference				reference				reference
Cesarean section	58.51	27.87	89.15	<0.001	46.79	25.23	68.36	<0.001	37.73	15.88	59.58	0.001
Maternal comorbidities
Diabetes mellitus	273.56	75.56	471.56	0.007	147.18	−11.06	305.43	0.068	22.97	−130.88	176.81	0.770
Hypertension	−15.51	−195.09	164.06	0.866	−34.60	−177.63	108.43	0.635	−137.98	−287.56	11.61	0.071
Hyperlipidemia	108.73	−28.15	245.60	0.119	91.61	−26.02	209.23	0.127	60.62	−43.70	164.93	0.255
COPD	31.89	−61.23	125.01	0.502	−36.73	−95.07	21.62	0.217	−1.39	−63.33	60.56	0.965
Pregnancy-related complications
Gestational hypertension	−63.13	−221.74	95.48	0.435	−115.59	−243.06	11.88	0.076	−45.15	−166.72	76.42	0.467
Pre-eclampsia	−441.31	−586.51	−296.10	<0.001	−223.57	−319.59	−127.55	<0.001	−192.43	−284.48	−100.38	<0.001
Gestational diabetes mellitus	64.67	−30.18	159.52	0.181	71.70	4.72	138.68	0.036	187.24	111.07	263.40	<0.001

Defined as having at least one schedule I or schedule II drug record during his wife’s pregnancy.

4. Discussion

Using a population-based retrospective dataset formed by linkage of 4 useful government registries and administrative databases, we found that BW decreased by an average of 108 g if the mother was exposed to illegal drugs during pregnancy, and there was a decreasing trend of BW with the period of maternal illegal drug use. These results are consistent with those of most other studies, showing that the risk of LBW was significantly higher in women who used illegal drugs during pregnancy [2,3,37,38].

Despite interest in the association between BW and duration of drug use, no one, to the best of our knowledge, has traced back participants’ drug-use history so long before pregnancy (up to two years). The most striking point of our study was that we adopted DCR to establish that BW can still be affected even when the maternal illegal drug exposure occurred only within one year before pregnancy, and without any DCR during pregnancy (period II). The effects persisted even when drug users were exposed to drugs within two years before pregnancy and without any DCR thereafter (period III). In our study, we should sound a note of caution with regard to those subjects without DCR, because some probably still used drugs but did not happen to be arrested for this. Because the drug-exposure effect is likely an underestimation, the statistically significant association between drug exposure and LBW means the true decrease in BW would be greater than the value presented in Table 3.

Using DCR as the indicator to determine drug-exposure history has some advantages. One is that the validity of DCR is very high, especially in Taiwan, because the drug users not only were caught red-handed but also tested positive in a urine screening. Another advantage is that DCR could compensate for the limitations that arise from self-reporting, hair sampling, and infant meconium testing. Self-reporting has a very low sensitivity [39-41] and is not an adequate measure of drug use in a pregnant population [13]. Maternal and infant hair samples have some weaknesses, including incomplete removal by washing and drug concentration changed by cosmetics [17,18,42]. In comparison with the abovementioned methods, DCR appears to be a relatively reliable indicator to measure drug-exposure history. According to the 2014 National Survey of Substance Use in Taiwan, lifetime prevalences were 0.23% for heroin users and 0.6% for methamphetamine users [43]; these results are very low compared with Western countries [44,45]. Our findings would seem to imply that DCR provides an important opportunity to examine the association between maternal and paternal illegal drug use on BW in a non-Western country. However, using DCR to determine drug-exposure history still has a disadvantage: the probability of being arrested by police is not random, especially for illegal drug users. Drug users are caught more often at drug-involved scenes/settings (e.g., karaoke bars, internet cafes, or nightclubs) than at home [46].

The causal relationship between maternal illegal drug use and LBW is very complex and affected by multiple risk factors, including socio-demographic, medical, and behavioral ones [47-49]. Almost all researchers have tried to control as many as risk factors as possible, but no study could control for a comprehensive set of associated factors, especially for information related to illegal drugs [49]. Since our study used information from DCRs, variables such as lifestyle or risk behaviors (e.g., smoking) cannot be asserted and adjusted. Therefore, the association might be unfavorable lifestyle factors [50,51] or inadequate access to healthcare services among drug users [50,52-55]. Our study also found that the rate of pregnant women in the drug-exposed group who did not receive any prenatal visits was 16.5%, compared to just 9% in the matched unexposed group (p < 0.05, data not shown).

The impact of maternal drug use on BW is better established by previous studies than paternal factors. BW correlations were found to be higher for mother–child than for father–child [23]. Some studies did not find that paternal cannabis use [12] or paternal involvement [25] were associated with fetal growth. Although maternal drug use during pregnancy has a larger effect on BW than paternal factors, our study showed that after adjustment for confounding factors, paternal drug exposure was still significantly associated with BW and a greater BW loss than maternal effect in periods II and III. Our research could not identify the reasons for the higher contribution of paternal factors than maternal factors to LBW in periods II and III; we believe important factors related to LBW might be due to paternal low socioeconomic status [56,57], paternal low educational attainment [10,57-61], paternal lifestyle factors [62], intimate partner violence [63-69], and the relationship between fathers and mothers [24].

Our research may have three limitations. First, our subjects were limited to married couples, excluding unmarried domestic partners. Because one of our study’s aims was to investigate the parental factors related to LBW, we had to exclude 4.6% of subjects missing a paternal identification number. Second, lifestyle behaviors were not available in the four national databases that we used. Lifestyle behaviors (e.g., cigarette smoking) are believed to be an important risk factor for LBW. Some studies found that the risk of LBW was significantly higher for pregnant women who smoked cigarettes than for those who used drugs [12,58,70-73]. Other studies found the effect of smoking on BW was less than that of illegal drug use [74-77]. The smoking rate of women in Taiwan is 2.4% [78], which is much lower than that in the United States (15.9%) and Britain (19.5%) [79]. Our study might slightly overestimate the impact of illegal drug use on BW by not taking into account tobacco smoking; nevertheless, the estimated association did not have appreciable differences once we used maternal COPD as a proxy for tobacco use in multivariate analyses.

The third limitation is, because of pervasive multidrug use and potential interactions between different illegal drugs, it is difficult to differentiate the consequences of each individual illegal drug on BW [9]. This is the reason our study did not analyze individual drug effects on BW but used schedule I and schedule II drug offenders as a group to explain the association between illegal drug use and LBW.

5. Conclusion

Our work has led us to conclude that maternal illegal drug exposure was associated with increased risk of lower birth weight, and that there is a gradient of decreasing BW corresponding to the period of maternal illegal drug exposure. The adverse effect on BW continues to exist even if the mother did not use illegal drugs during pregnancy but had ever used drugs during the one or two years before pregnancy. The contribution of paternal factors to LBW persists, and the decrement of BW is even greater than the maternal effect within one or two years before pregnancy. We provide further evidence that maternal and paternal drug use may have a lasting effect on their offspring’s birth weight.

Our research highly suggests a window of intervention opportunity to provide readily accessible integrated addiction treatment and reproductive health services for drug-involved women when they encounter the law enforcement and correctional systems.