CONTRACEPTIVE USE IN APPALACHIAN WOMEN WHO USE DRUGS AND WERE RECRUITED FROM RURAL JAILS

Abstract

This study describes a secondary data analysis of contraceptive use across the lifetime and within the six months prior to incarceration in a sample of 400 currently incarcerated women recruited from rural, Appalachian jails, who were using drugs prior to incarceration. Phase 1 (baseline) data from an NIH funded study were used to examine rates of contraceptive use, reasons for non-use of condoms, and correlates of condom use. Results indicate that the majority (96.5%) of respondents reported lifetime use of contraceptives, and most (70.5%) had a history of using multiple methods, with male condoms, oral contraceptive pills, and contraceptive injections being the most commonly used methods. Almost 69% of respondents reported non-use of contraceptives within the last six months, despite high rates of involvement in risky, intimate male partnerships prior to incarceration. Contraceptive use was found to be historically acceptable in this sample, in stark contrast to rates of use within the last six months prior to incarceration, suggesting that reproductive justice-informed, social work interventions to help improve current contraceptive use are warranted as a harm-reduction approach.

Introduction

Appalachia is a unique, often remote, and mountainous region of the Eastern United States (US) that encompasses all of West Virginia, most of Pennsylvania, and parts of Alabama, Georgia, Kentucky, Maryland, Mississippi, New York, North Carolina, Ohio, South Carolina, Tennessee and Virginia (Appalachian Regional Commission, 2019). The region is comprised of 420 counties and almost 205,000 square miles, and is home to a population of 25 million people (Appalachian Regional Commission, 2019). Appalachia has many regional strengths, including a resilience around problem solving in the context of difficult circumstances (Smith, Moore, Anderson & Siderelis, 2012), yet it is primarily rural and often under-resourced. Residents experience problems with health care access, poor preconception health care indicators, unintended pregnancy, and higher rates of poverty, in ways similar to what is experienced in lower-resourced nations, in contrast to other geographic areas of the US (Broecker, Jurich & Fuchs, 2016; Huttlinger, Schaller-Ayers & Lawson, 2004; Pollard & Jacobsen, 2017; Short, Oza-Frank & Conrey, 2012). Further, rates of injection drug use and opiate addiction have been increasing in the region, and the associated negative impacts include involvement with the criminal justice system, risk of unintended pregnancy, and sexually transmitted infections, such as hepatitis C and HIV, among other risks (Staton-Tindall et al, 2015; Staton-Tindall, Webster, Oser, Havens & Leukefeld, 2015; Young & Havens, 2015; Zibbel et al, 2015). Contraceptive use, especially use of male condoms, in the context of active drug use, is critical for sexual health, including being useful for reducing sexually transmitted infection (STI) transmission and preventing unintended pregnancy that can lead to public health problems such as neonatal drug exposure (Hathazi, Lankenau, Sanders & Bloom, 2009; Miller et al, 2003; Warren, Miller, Traylor, Bauer & Patrick, 2015).

While a body of literature related to the overall health risk factors within the Appalachian population exists, there remains a scarcity of research around reproductive health overall in Appalachia, and there is a dearth of information around contraceptive use in this population, particularly in the context of criminal justice involvement and recent drug use (Shannon, Nash & Jackson, 2016). Given this gap in the literature, this paper explores, for the first time, rates of contraceptive use in a sample of Appalachian women recruited from rural jails who were using drugs prior to becoming incarcerated.

Background

Women in rural, Appalachian settings often experience significant disparities in access to reproductive health services and high rates of unintended pregnancy (Broecker, Jurich & Fuchs, 2016; Finer & Henshaw, 2006; Finer & Zolna, 2011; Hall, Moreau & Trussel, 2012; Moody, Satterwhite & Bickel, 2017; Reel, 2001). Research suggests an overall lack of knowledge about and access to family planning services exists in Appalachia due to interconnected barriers to seeking such services in the region (Hansen & Moloney, 2019; Swan et al, 2020).

Substance use is a known risk factor for increased maternal morbidity and mortality in rural Appalachia (Hansen & Moloney, 2019). Research suggests that up to 90% of pregnancies occurring in the context of opiate abuse are unintended (Eyo & Chenoy, 2014; Heil et al, 2011). Women who use substances are also less likely to use contraceptives, and they are less likely to use the most effective forms of contraception when compared to women who do not have substance use disorders, which suggests a potential need for access to longer acting reversible methods (LARCs) (Black & Day, 2016; Terplan, Hand, Hutchinson, Salisbury-Afshar, & Heil, 2015). Women’s injection drug use is associated with poor health and mental health outcomes, physical and sexual partner violence, involvement with risky and drug-injecting partners, exchanging sex for drugs, involvement with greater numbers of male partners, and increased risk of sexually transmitted infections (Iversen, Page, Madden & Maher, 2015; Staton, Strickland, Tilson, Leukefeld, Webster & Oser, 2017; Staton et al, 2018; Topp, Iversen, Conroy, Salmon & Maher, 2008), among other risks. Despite often elevated rates of substance use and the co-occurring risk factors, rural areas have only a fraction of the health and treatment resources that are available in urban settings, causing significant disparities in access (Dew, Elifson & Dozier, 2007; Keyes, Cerdá, Brady, Havens & Galea, 2014). There is also considerable stigma associated with drug-using populations, which may prevent vulnerable Appalachians from seeking health care (Benoit et al, 2014).

Pregnancy prevention efforts in Appalachia may be impeded by economic barriers to contraceptive access, including lack of insurance coverage, high out-of-pocket expenses for obtaining intrauterine devices (IUDs)(Broecker, Jurich & Fuchs, 2016). Other identified barriers to reproductive health care in the region include a lack of knowledge about contraception, problems accessing care, problems trusting providers, and problems being able to get appointments or physically access a provider, among other problems (Fagan, Boussios, Moore & Galvin, 2006; Swan et al, 2020). Rural areas are also known to be underserved by physicians, (Balance & Kornegay, 2009), which likely contributes to problems accessing and using contraception for rural populations. Evidence suggests that there are differences between cultural groups in terms of attitudes about and commitment towards prevention of unintended pregnancy in the context of substance use (Gutierres & Barr, 2003), which has implications for working with Appalachian populations.

Cases of neonatal abstinence syndrome (NAS), which is a post-birth drug withdrawal condition, are elevated in Appalachian counties, suggesting a need for primary prevention of unintended pregnancy in Appalachian women who are active drug users (Driscoll & Ely, 2019; Warren, Miller, Traylor, Bauer & Patrick, 2015). Appalachian incidences of NAS are estimated at 33.4 cases per 1000 live births (National Institute on Drug Abuse, 2018). Findings from one study show that NAS rates in rural and Appalachian Kentucky counties were approximately two and a half times greater compared to rates in urban and non-Appalachian Kentucky counties (Brown, Goodin & Talbert, 2018). In another study, researchers characterize the rates of NAS in the East Tennessee region of Appalachia to be at epidemic levels, as NAS cases accounted for 28.8 out of 1000 live births in 2014 (Erwin, Meschke, Ehrlich & Lindley, 2017). Efforts to reduce the NAS rates in Appalachia may be hampered by health service access problems, including a lack of available substance abuse programs equipped to serve populations who are pregnant or parenting (Swan et al, 2020).

While the period of incarceration is not the focus of the current study, incarcerated women in general have high rates of drug use and trauma histories (Jones, Worthen, Sharp & MacLeod, 2018). While the ability to engage in intimate relations is halted during incarceration, many common prescription contraceptives (i.e. oral birth control pills and contraceptive injections) require active engagement by the user, and access to these contraceptive medications may be denied or delayed during the incarceration period. Such delays can interrupt the contraceptive cycle and elevate risk for unintended pregnancy post-incarceration. These issues are highlighted in one study where incarcerated women identified being at greater risk for unintended pregnancy, as they reported high rates of pre-incarceration inconsistent contraceptive use, a desire to avoid pregnancy in the near future, and an intention to engage in sexual intercourse post-incarceration (Hale et al, 2009). Moreover, research suggests that there is a risk of rape or assault during incarceration, as staff in prisons and jails report that sexual violence is a problem of serious concern (Owen & Moss, 2009), and this, unfortunately, further substantiates the importance of contraceptive continuation during incarceration.

All of these factors coalesce to create a perfect storm that decreases the likelihood that active or recent drug users in rural and Appalachian regions will be able to effectively obtain, use, and maintain contraceptives to prevent unintended pregnancy, particularly when their health behaviors are interrupted by involvement with the criminal justice system. However, information about contraceptive practices in Appalachia, especially for populations who experience additional vulnerabilities is severely lacking, despite the need to enhance the prevention of unintended pregnancy in the context of regional substance abuse. Given this dearth of information, the purpose of the current study was to conduct a secondary data analysis of contraceptive use in Appalachian women recruited from jails who had been actively using drugs prior to incarceration. Data for the current study was drawn from an NIH funded study originally designed to test a Hepatitis C prevention program for women in rural jails who had been recently using drugs prior to incarceration.

The Current Study

The first aim of this paper is to examine lifetime use rates of various types of contraception, and contraceptive use within the six months prior to incarceration, among this sample of rural Appalachian women who were recruited to participate in this study while incarcerated in a rural jail. In the United States, jail incarcerations differ from prison incarcerations, in that they are usually short term, and a person in jail has been charged with a crime, but not necessarily convicted or sentenced. In the second aim, we identify correlates of condom use with main and casual partners, with whom the women were partnered prior to incarceration. In the United States, people in jail are typically not permitted to engage in intimate partner relations while incarcerated. Condom use was the primary focus of the second aim because it was the only method of contraception quasi-frequently used within the past year, and because reasons for non-use were only reported in relation to condoms.

Methods

Participants and Procedure

This analysis is based on secondary data from the phase 1 (baseline) wave of an NIH funded study of rural women from Appalachia who had been actively using drugs prior to incarceration and were recruited for the original study from jails where they were currently incarcerated (N=400). Only women identified as being drug users prior to the incarceration were recruited, and the majority of participants (almost 60%) had been using intravenous drugs within the last year, and over 75% reported using intravenous drugs in their lifetimes (Staton et al, 2017). The original study was conducted from 2012-2015, and this baseline data was collected in 2012. This data was collected solely for research purposes and was not part of any jail-related assessment. IRB approval for the original study was obtained from the institution where the study originated, which included prisoner review for vulnerable populations. The research staff who collected the original data were trained in confidentiality procedures, and data was collected in private settings with no correctional staff present. Prior to conducting this secondary data analysis, a reciprocal IRB agreement was put in place between the institution of origin, and the institution where the current study was conducted, which allowed for review of this data that did not involve contact with human subjects.

Study participants and procedures are described in detail elsewhere (Staton, Strickland, Tilson, Leukefeld, Webster & Oser, 2017). To summarize, women who met NIDA-modified ASSIST Quick Screen (National Institute on Drug Abuse, NIDA, 2012) criteria were randomly selected from three rural jails in the Appalachian region of one Central Appalachian, Southern state to undergo an interview with study personnel where the data used for this current study was originally obtained. Participants were interviewed in the jail by trained, regional research assistants, and the data collection instrument focused on past substance use, health, and health service utilization. Participants in the original study were paid $25 for their interview time.

Measures

Contraceptive Use.

Participants were asked which contraceptive methods they had used across their lifetime and in the six months prior to incarceration, and how often they used condoms with their main partner, any casual partners, and when exchanging sex for drugs or other resources (bartering), within the six months prior to their incarceration. Response options were never, sometimes, quite a bit, or all the time.

Condom use with main and casual partners.

Participants were asked to distinguish between main and casual sexual partners and to describe how frequently they had used condoms or other types of barrier protection (i.e. female condoms, dental dams) with partners of each type, across their lifetimes, and within the last six months prior to their current incarceration. Condom use frequency was rated on a four-point scale (never, sometimes, quite a bit, all the time). We elected, for some analyses, to combine the sometimes, quite a bit, all the time responses to form a dichotomous scale.

Reasons for not using condoms.

Participants responded to 15 dichotomous (yes/no) items about reasons why they did not use condoms with their main partner within six months prior to their incarceration. The list of reasons included perceived effects on partner (less romantic, makes it look like I think partner has AIDS, partner will think I’m having an affair), being present in the moment (do not think about it, get high and forget, lose control in the heat of the moment), physical/practical (costs too much, partner has trouble maintaining an erection, less sensation, having sex with only one partner, partner doesn’t want to use condoms), being fearful of requesting partner’s use, and religious reasons.

Relationship Control Scale.

Women’s power in their sexual relationships with main partners within six months prior to incarceration was assessed by the Relationship Control Subscale (RCS), a part of the Sexual Relationship Power Scale (Pulerwitz, Amaro, De Jong, Gortmaker & Rudd, 2010). The RCS consists of 15 items (current sample alpha = .85) assessing the women’s perception that their partner controls the relationship. Items are rated on a four point scale (strongly agree, agree, disagree, strongly disagree).

Decision Making Dominance.

The perception of which person (respondent or partner) usually has the final say in relationships with main partners (within six months prior to incarceration) was assessed by the Decision Making Dominance Subscale (DMDS), a part of the Sexual Relationship Power Scale (Pulerwitz, Amaro, De Jong, Gortmaker & Rudd, 2010). The DMDS consists of eight items (alpha = .63) assessing who in the relationship has more input on social activities and sexual interactions. Item responses are my partner, both of us equally, and me.

Reducing Risky Relationships Thinking Myths Scale (RRR-TMS).

Items from this scale were adapted for the current study from the original RRR-TMS scale that was designed to measure thinking myths in relation to HIV risk behavior knowledge (see Leukefeld et al, 2012). The RRR-TMS is an 11 item scale and the RRR-TMS used in the current study consists of 12 items that explored attitudes on each thinking myth (e.g., Sex without protection will strengthen my relationship). Responses are made using a ten-point scale with the endpoints labeled as definitely true and definitely not true.

Analysis

In order to address the first aim of the study, counts and percentages of respondents using each type of contraceptive method were calculated for lifetime use and use in the past six months prior to incarceration. To address the second aim, frequencies, counts, and percentages of condom use and other barrier methods with main partner and with casual partners within six months prior to incarceration were calculated. Correlates of condom use with main and casual partners were also examined. Spearman’s correlations between Relationship Control Scale items and condom use with main partner were then examined. In order to explore the association between DMDS items and condom use, responses on this scale were dichotomized into responses that represented that the behavior occurred “never” or the behavior occurred “some or all of the time” with the main partner, and then percentages and chi squares were calculated and correlations were examined. Lastly, correlations were examined between RCS, DMDS, and RRR-TMS items and condom use with main partner. Correlations were deemed significant if they were ≤ .05.

Results

Demographics

The average age of the 400 women in this study was 32.81 (SD: 8.24; range: 18-61). Women were most often white (99%) and heterosexual (79.3%). Past six months income ranged from 0 (none) to 210,000 dollars, with 90% of women reporting less than 18,000 dollars.

Contraceptive use

The majority of the sample (96.5%) reported using one or more forms of birth control (Table 1) over the course of their lifetimes and most (70.5%) used multiple methods of contraceptives during their lifetimes, most frequently two or three methods, but as many as six different types of methods. The most frequently used type of birth control method was a male condom, followed by oral contraceptives (birth control pills), contraceptive injections (i.e. Depo-Provera), and intra-uterine devices (IUDs). The most frequently reported method combinations were male condoms and birth control pills (27.8%) and male condoms, birth control pills, and contraceptive injections (20.3%). While a high percentage of women reported lifetime use of some form of contraceptive, less than one-third of the sample used some form of contraceptive in the last six months prior to their incarceration, during the time they were recently using drugs. Of the 125 women using a birth control method within the six months prior to their incarceration, the overwhelming majority, 76% (n = 95), used male condoms and considerably fewer used oral contraceptives (birth control pills) or IUDs (both 14.4%) or contraceptive injections (13.6%). Of the women who used multiple methods in the past year, 20 of the 23 combined male condoms with IUDs, contraceptive injections, and/or oral contraceptive pills.

Table 1.

Counts of birth control method used across lifetime and the past six months.

	Lifetime (N = 400)		Past Six Months (N = 399)
Birth Control Method	Count	Percentage [95% CI]	Count	Percentage [95% CI]
Male Condom	368	92.0% [89.3, 94.5]]	95	23.8% [19.5, 28.3]
Female Condom	7	1.8% [0.5, 3.3]	2	0.5% [0.00, 1.3]
Dental Dam	3	0.8% [0.00-1.8]	1	0.3% [0.00, 0.8]
Birth Control Pills	256	64.0% [59.5, 68.5]	18	4.5% [2.5, 6.8]
Diaphragm/cervical Cap	7	1.8% [0.5, 3.3]	1	0.3% [0.00, 0.8]
Contraceptive Injection	134	33.5% [29.0, 38.0]	17	4.3% [2.3, 6.3]
Contraceptive Implant	17	4.3% [2.3, 6.5]	1	0.3% [0.00, 0.8]
Intra-Uterine Device (IUD)	43	10.8% [7.8, 14.0]	18	4.5% [2.8, 6.5]
No birth control methods used	14	3.5% [2.0, 5.5]	274	68.7% [63.5, 73.0]

Participants were asked about their lifetime frequency of using condoms or other barrier methods, such as female condoms or dental dams—both of which were nearly never used as reported in Table 1—with main and with casual partners. As Table 2 shows, nearly 70% of participants never used a condom with their main partners and just over one-third did so with their casual partners.

Table 2.

Lifetime frequency of using condoms or other protection during sex with main and casual partners prior to incarceration

	Main Partner (N = 399)a		Casual Partner (N = 267)b
Frequency	Count	Percentage [95% CI]	Count	Percentage [95% CI]
Never	277	69.3% [65.2, 73.9]	95	35.6% [25.6, 36.0]
Sometimes	88	22.0% [18.0, 25.8]	90	33.7% [24.4, 33.8]
Quite a bit	26	6.5% [4.3, 9.0]	35	13.1% [7.5, 14.9]
All the time	8	2.0% [0.8, 3.5]	47	17.6% [11.4, 19.5]

Note. Other protection methods included female condoms and dental dams.

^aOne participant refused to answer.

^bForty-one participants denied having had a casual sex partner and 92 were coded as not applicable.

In addition, women were asked whether they had ever in their lifetimes had sex with a partner in exchange for drugs or resources (bartering) and 174 (43.5%) women reported that they had. During such bartering exchanges, these women reported their frequency of use of condom or other barrier methods (female condoms or dental dams) as never (37.9%), sometimes (27.6%), quite a bit (17.8%), and all the time (16.7%). In addition, we compared the male condom use percentages for women who had and had not used sex to barter. Women who had ever bartered in their lifetimes were asked if they had also done so in the past six months and 71 (40.8%) of the 174 who had ever bartered had also bartered within the past six months prior to incarceration. The condom use percentage for these women during these recent bartering exchanges was 37.9%. Compared to the percentage (19.0%) for women who had not bartered in the six months prior to incarceration, even if they had done so previously, the difference was significant (χ2 (1, N = 398) = 14.98, p < .001). However, compared to the percentage (25.4%) for women who had not bartered in the previous six months but had done so previously, the difference was not significant (χ2 (1, N = 174) = 2.99, p = .084).

Correlates of Condom Use

Reasons for non-use of condoms.

The four most frequently cited reasons for not using condoms across the lifetime, all cited by ten or more percent of participants, were (a) having sex with only one partner (59.6%), (b) do not think about it (39.1%), (c) get high and forget (26.1%), and (d) partner doesn’t want to use condoms (11.8%). We then computed the association between endorsement of the reason and frequency of condom use, recoded as never vs. some to all the time, with the main partner. None of the associations attained significance.

Relationship Control Scale.

We computed the correlation between RCS items and the recoded frequency of condom use with the main partner across the lifetime. A positive correlation means that the participant strongly disagrees with the RCS items and uses condoms some of the time to all the time. Table 3 presents the results, which indicate that participants who were more likely to use a condom were also more likely to report that their partners would not get violent, angry, suspicious, or exert control in their relationship. At the same time, women who used condoms more frequently did not see themselves as quiet in relation to their partner.

Table 3.

Correlation between Relationship Control Scale items and Condom Use with Main Partner (N = 365)

Relationship Control Scale Items	Spearman’s Correlation	p value
1) If I asked my partner to use a condom, he would get violent	.181***	.000
2) If I asked my partner to use a condom, he would get angry	.141**	.007
3) Most of the time, we do what my partner wants to do	.033	.532
4) My partner won't let me wear certain things	.146**	.005
5) When my partner and I are together, I'm pretty quiet	.108*	.039
6) My partner has more say than I do about important decisions that affect us	.077	.140
7) My partner tells me who I can spend time with	.081	.123
8) If I asked my partner to use a condom, he would think I'm having sex with other people	.148**	.005
9) I feel trapped or stuck in our relationship	.032	.543
10) My partner does what he wants, even if I do not want him to	.098+	.061
11) I am more committed to our relationship than my partner is	.082	.115
12) When my partner and I disagree, he gets his way most of the time	.061	.241
13) My partner gets more out of our relationship than I do	.031	.553
14) My partner always wants to know where I am	.037	.478
15) My partner might be having sex with someone else	.074	.158

Note. A positive correlation indicates that women who strongly disagree with the item are more likely to also use a condom some or all the time.

^*p < .05

^**p < .01

^***p < .001

⁺p < .10.

Decision Making Dominance.

We examined the DMDS items, which explored relationship power, in relation to condom use across the lifetime. As shown in Table 4, the most frequent response (42% to 72%) to items was both of us equally with the remaining responses about equally divided between my partner and me (the respondent). As Table 4 also shows, the percentages of those reporting they never used condoms also reported that the both of us equally and me responses were similar and sometimes nearly equal. These were both lower than the never percentage for the my partner response. We therefore elected to omit cases with the both of us equally response and compute the correlation for the remaining cases. This result is reported in the far right column of the table. A positive correlation means that participants who say I decide (i.e. the respondent decides herself) are less likely to report that they never used condoms with a main partner. The correlation is significant and positive with respect to decisions about whose friends are socialized with, couple’s leisure activities, condom use, and overall power.

Table 4.

Association between DMDS items and Condom Use Never versus Some/All the time with Main Partner (N = 382)

Item	Item text	Percent Never used condoms (n)	Chi Square	p value	Spearman’s Correlation	p value
1	Who usually has more say about whose friends to go out with?		7.008*	.030	.131*	.012
	My partner decides	82.6% (69)
	Both of us equally	67.9% (243)
	Me (woman decides)	62.7% (51)
2	Who usually has more say about whether you have sex?		3.370	.185	.085	.105
	My partner	74.0% (77)
	Both of us equally	70.7% (239)
	Me	59.2% (49)
3	Who usually has more say about what you do together?		4.942+	.085	.108*	.040
	My partner	80.6% (72)
	Both of us equally	68.0% (241)
	Me	64.7% (51)
4	Who usually has more say about how often you see one another?		1.772	.412	.036	.492
	My partner	76.3% (59)
	Both of us equally	68.0% (275)
	Me	73.3% (30)
5	Who usually has more say about when you talk about serious things?		0.716	.669	−.006	.908
	My partner	72.3% (65)
	Both of us equally	68.5% (232)
	Me	73.1% (67)
6	In general, who do you think has more power in your relationship?		4.962+	.084	.116*	.026
	My partner	77.1% (109)
	Both of us equally	69.2% (159)
	Me	62.9% (97)
7	Who usually has more say about whether you use condoms?		9.757**	.008	.146**	.005
	My partner	76.4% (55)
	Both of us equally	72.7% (242)
	Me	54.4% (68)
8	Who usually has more say about what types of sexual acts you do?		1.974	.373	.070	.181
	My partner	77.4% (53)
	Both of us equally	69.3% (254)
	Me	65.5% (58)

Note.

^*p < .05

^**p < .01

^***p < .001

⁺p < .10.

Thinking Myths Scale.

Analysis of the RRR-TMS item frequencies revealed that with three exceptions (items 4 and 6, which are reversed items, and item 12) participants endorsed the definitely not true response in percentages ranging from 62% to 86%. Endorsement of the other endpoint, definitely true, ranged from 3% to 12%. The percentages for items 4 and 6 showed the reversed pattern and item 12 showed somewhat balanced percentages for the two endpoints. We elected to recode the items to contrast the definitely not true response to all other responses and compute correlations between these items and recoded condom use frequency (Table 5). A negative correlation means that participants endorsing the definitely not true response were more likely to report they never used a condom across the lifetime.

Table 5.

Correlation between RRR-TMS items and Condom Use with Main Partner (N = 382)

RRTMS item	Correlation	p value
1) Having sex without protection will strengthen my relationship	−.099+	.054
2) Using drugs with my partner will strengthen my relationship	−.144**	.005
3) I only think good things about myself when I am in a relationship, even if it is a risky relationship	−.058	.258
4) I think good things about myself even when I'm not in a risky relationship	.005	.915
5) I can use drugs and always make healthy choices about protection	−.096+	.060
6) I don't make healthy choices about HIV protection when I use drugs	.052	.305
7) I know my partner is safe from HIV by the way my partner looks	−.156**	.002
8) I know my partner is safe from HIV by the way my partner talks	−.167***	.001
9) I know my partner is safe from HIV by the way my partner acts	−.147**	.004
10) I know my partner is safe by the way my partner looks, talks, and/or acts	−.154**	.003
11) I will not get HIV because I'm not really at risk	−.006	.907
12) I will get HIV because I am at risk	−.008	.870

^*p < .05

^**p < .01

^***p < .001

⁺p < .10.

Discussion

While most of the sample had a lifetime history of contraceptive use, current use within the six months prior to incarceration, when most participants had been recently using intravenous drugs, was astonishingly low in this sample, as almost 69% reported not using contraceptives during this time, even in the context of having risky sexual partnerships, including high rates of bartering for drugs or other services. In contrast, the reportedly high rates of lifetime use (only 3.5% reported lifetime histories of non-use) in this sample indicates that contraceptive use was acceptable for these women at other times in their lives. This suggests the presence of situations that were potentially impeding their contraceptive use within the last six months, which is a significant public health concern in light of their vulnerabilities and risky lifestyles. Information from others suggests that lack of contraceptive use increases risk for a multitude of negative health impacts, including unintended and/or drug-exposed pregnancies, and STI exposure (Heil et al, 2012). These results are also consistent with work from others showing that those with substance use disorders used contraceptives less frequently than those who did not have substance use disorders (Terplan, Hand, Hutchinson, Salisbury-Afshar, Heil, 2015), and with a study from Russia, indicating that contraceptive use among injection drug-using women was virtually non-existent (Abdala, Kershaw, Krasnoselskikh, & Kozlov, 2011). This indicates that pregnancy prevention behaviors in drug users may be common across cultures, despite research suggesting that there are cultural differences in attitudes towards contraceptive use in the context of substance abuse (Gutierres, & Barr, 2003), although more research is needed to confirm this.

Specific to condom use, approximately 70% of respondents reported never using condoms with their main partners when engaging in sexual intercourse prior to incarceration, and only one-third report condom use with casual partners, despite the risks likely inherent in these romantic partnerships, and despite that these women were not likely to be using other contraceptives. Only about 40% of women who bartered by exchanging sex for drugs in the last six months prior to incarceration used condoms during these risky exchanges. Taken together, these results suggest a low ability for self-protection against unintended pregnancy and exposure to sexually transmitted infections in this sample, and this finding highlights the need for social workers to design and implement interventions targeted towards encouraging substance using women to use condoms as a harm reduction measure. These results were consistent with international findings from injection drug users in a Russian sample, who reportedly used condoms around 22% of the time, mostly only with casual partners (Abdala, Kershaw, Krasnoselskikh & Kozlov, 2011).

Our results also indicate that participants who were more likely to use condoms were also more likely report that they made more decisions in their relationships and that their partners were reportedly less likely to get violent, angry, suspicious, or exert control in their relationship, and that the women were less likely to perceive themselves to be “quiet” in these relationships. This suggests that women who have more positive romantic partnerships or more control in their partnerships may be more comfortable suggesting condom use within those relationships. This finding is not surprising, given that negative relationship traits (partner violence, HIV risk, risky sexual activity, and involvement with risky lifestyles) have been associated with condom refusal, and lack of condom use, while overall involvement in risky lifestyles has been previously documented in at-risk, Appalachian, drug-using, and incarcerated populations (Allen, Flaherty & Ely, 2010; Campbell, 2002; Roberts & Kennedy, 2006; Staton-Tindall et al, 2015a,b). Taken together, our results suggest that those with higher quality relationships or more input or control in their romantic partnerships may be more able to negotiate condom use, even in the context of their drug use. This is a notable strength in this sample that could potentially be built upon for intervention development moving forward.

Implications for Practice

Social workers often play a critical role in facilitating access to general health care services, as well as reproductive health care services, particularly for vulnerable, at-risk populations. As such, we are obligated as a profession to advocate for access to contraception and all forms of reproductive health care for women who are actively using drugs, and the continuation of contraceptive care for women who are at risk of contraceptive interruption after becoming incarcerated. This approach is supported by experts who argue that embracing and supporting reproductive health is a key responsibility of the social work profession (Alzate, 2009).

We recommend that this support for reproductive health care access be provided in the context of a focus on reproductive justice, which is a unifying framework compatible with social work and developed by Black women, that advocates for the right to parent, the right not to parent, and the right to parent in safe and peaceful circumstances (Liddell, 2019; Ross & Solinger, 2017). This suggestion is made in contrast to a more conventional focus on reproductive “rights”, which, when applied, typically emphasizes a right to use contraception, but does not account for the access barriers experienced by vulnerable women that can make exercising such a right virtually impossible (Liddell, 2019; Ross & Solinger, 2017). The application of the reproductive justice framework by social workers, which is supported by the National Association of Social Workers in their most recent policy statements (NASW, 2018), would allow for a more complete conceptualization of the need for social workers to help dismantle the barriers to reproductive health care access that are experienced by vulnerable populations, including the women in this sample who are active drug users and experiencing incarceration. This would help social workers answer the question of how to best assist vulnerable, at-risk women to access and use contraception, if they wish to avoid unintended pregnancy, independent of their drug-use or their incarceration status.

In light of our findings, social workers can apply harm reduction approaches, which are recommended to improve access to treatment and contraceptive services for vulnerable populations who use drugs and live in areas where health access may be difficult, as it is in much of Appalachia. This is important for the prevention of untended pregnancy, reduction of drug-exposure during pregnancy, and reduction of risk related to sexually transmitted infections, as we anticipate that many of these women returned to both active drug use and sexual activity upon release. However, it is important to note that previous research indicates that condom use in drug users is inconsistent, and interventions designed to address adverse birth outcomes may not be adequately reaching rural areas (Kahn, Berger, Hemberg, O’Neill, Dyer, & Smyrk, 2013; Kent, McClure, Zaitchik, & Gohlke, 2013). Thus, innovative, reproductive justice-informed intervention approaches are warranted. For example, community messaging to inform the public that people using drugs will not be turned over to authorities when seeking health care could be provided as an educational intervention. This is an important recommendation, given that many substance users fear that their illegal drug use will be exposed to health providers, which is a barrier to health care access (MacAfee, 2019). This recommendation is consistent with the principles of the NASW Code of Ethics, and those of reproductive justice, and in line with research indicating that culturally sensitive, community-based education programs that provide knowledge about contraceptive options and access have been shown to improve contraceptive use internationally (Daniel, Masilamani & Rahman, 2008), and may be promising for Appalachian populations.

The exploration of offering integrated interventions designed to improve contraceptive use in unconventional settings is also warranted (MacAfee et al, 2019), given the high levels of risk and low levels of condom use present in this sample, and the likely risk of returning to drug use and sexual activity after incarceration. This recommendation is consistent with research indicating that mental health prevention and HIV education should be provided in alternative venues, including rural jails, as was done in the original study where these data were drawn from, due to service limitations in Appalachia and other rural areas (Staton-Tindall, 2015a,b). Others recommend offering contraceptive programs within the jail setting, as has been done recently in East Tennessee, in a jail-based program that was estimated to successfully prevent both unintended pregnancy and cases of neonatal drug exposure from occurring (McNeely, Hutson, Sturdivant, Jabson & Isabell, 2019).

Research shows promising results from health interventions offered in conjunction with opiate maintenance medication programs that were combined with on-site free prescription contraceptives and financial incentives for attending follow-up visits (Heil, Hand, Sigmon, Badger, Meyer & Higgins, 2016), and this kind of approach is worthy of consideration for vulnerable Appalachian populations in light of recent research indicating that contraceptive access in the region is compromised due to a multitude of complex factors (Swan et al, 2020). Contraceptive programs could be more integrated into settings where family planning is not necessarily the typical focus, including in primary care practices, or in settings typically associated with social services, such as in probation and parole offices. The content of these programs could include educational initiatives combined with on-site access to prescription contraceptives, including placement of LARCs, and social work support services to assist patients with substance abuse (Eyo & Chenoy, 2014). This recommendation is made in light of other study results indicating that LARC use can be increased through promoting their availability and then providing them on site or with a quick turnaround time, at no or low cost (Broecker, Jurich & Fuchs, 2016; Eyo & Chenoy, 2014), although we wish to note that this must be done in a way that is consistent with reproductive justice by making sure that vulnerable women do not feel coerced into LARC use. Given findings from one study indicating that about 60% of intravenous drug using respondents saw a primary care provider within the last year who was aware of their drug use, a strong case can be made to expand primary care services to focus more intently on drug treatment, family planning, and on-site social work services (Dion et al, 2020). Nontraditional approaches that incorporate social workers into health settings and medical practice models are likely going to be needed if we are to impact the low rates of contraceptive use evidenced in this study, and federal, state, and local policy efforts are needed to support the implementation of such programs. These approaches could have multiple benefits, including the reduction of neonatal exposed births, which typically require extensive medical care (Patrick, Schumaker, Benneyworth, Krans, McAllister & Davis, 2012).

We also propose health policy that places community health workers (CHW), drawn from the Appalachian region, in non-traditional health or social service settings, to focus on increasing knowledge around family planning in the context of substance abuse, and that promotes the importance of condom use to reduce unintended pregnancy and to increase protection from STI exposure (Garcia-Retamero & Cokely, 2011; Kirby, Raine, Thrush, Yuen, Sokoloff & Potter, 2010). CHWs can improve the cultural sensitivity of interventions, as they tend to be less threating because they are laypeople and peers, and, because of their familiarity with community members, participants might be more receptive to their messages (Ely, Miller & Dignan, 2011; Gold, 2010). The use of CHWs to provide contraceptive education has been successful in other low resourced, rural settings, and it has been successful in Appalachia in other health contexts (Feltner, Ely, Whitler, Gross & Dignan, 2012; Scott et al, 2015; Viswanathan et al, 2012; Weidert, Gessessew, Bell, Godefay & Prata, 2017).

Limitations of Study

Data for this study were drawn from self-report surveys, which can be influenced by desirability and response bias, and which include the potential for misunderstanding of the concepts by study respondents (Rosenman, Tennekoon & Hill, 2011). Given the sensitive information subjects were asked to recall, which included drug use, injection practices, and contraceptive use history, social desirability bias may have impacted these responses. These respondents were recruited from Appalachian jails and then asked to remember past events, which may impact their recall ability, and limits the generalizability of these study results outside of this sample of women. It is also possible that confidentiality in a jail setting could have been a concern to respondents, even though everything possible was done to minimize this concern during original data collection, including conducting private interviews without the presence of law enforcement officers, and the obtainment of a Certificate of Confidentiality from staff. Moreover, these cross-sectional data do not allow us to infer causal relationships.

Conclusion

Women in rural Appalachia who use drugs and are involved in the criminal justice system are a vulnerable population at risk of several negative sexual and reproductive health outcomes, chief of which appears to be unintended pregnancies, drug-exposed births, and increased risk of STI exposure, due to extremely low rates of contraceptive use within the six months prior to incarceration when they were actively using drugs. However, evidence of a lifetime history of contraceptive use in this sample indicates that contraception was acceptable and accessible for this population at some point, which is information that could be leveraged in intervention development targeted towards elevating contraceptive use rates, and ensuring continuation of contraceptive care even during drug use and periods of incarceration.