Contraceptive choice and power amongst women receiving opioid replacement therapy: qualitative study

Abstract Background Women receiving treatment for opioid use disorder have low levels of contraception use and high rates of unintended pregnancies, abortion and children being adopted or fostered. This paper aims to understand the relationship between contraceptive choice and power amongst women receiving Opioid Replacement Therapy (ORT). Methods During 2016/17, semi-structured interviews were undertaken with 40 women (aged 22–49 years) receiving ORT in the South of England. Data relating to the latent concept of power were inductively coded and analysed via Iterative Categorisation. Findings Power manifested itself through six interconnected ‘fields’: i. ‘information about fertility and contraception’; ii. ‘access to contraception’; iii. ‘relationships with professionals and services’; iv. ‘relationships with male partners’; v. ‘relationships with sex work clients’; and vi. ‘life priorities and preferences’. Each field comprised examples of women’s powerlessness and empowerment. Even when women appeared to have limited power or control, they sometimes managed to assert themselves. Conclusions Power in relation to contraceptive choice is multi-faceted and multi-directional, operating at both individual and structural levels. Informed decision-making depends on the provision of clear, non-judgemental information and advice alongside easy access to contraceptive options. Additional strategies to empower women to make contraceptive choices and prevent unplanned pregnancies are recommended.


Introduction
Opioid replacement therapy (ORT) is an effective and evidence-based approach to treating opioid dependence that involves the prescription of pharmaceutical opioids, sometimes combined with counselling, behavioural therapies and other psychosocial supports (Bell, 2012;NICE, 2007;SAMHSA, 2018;WHO/UNODC/UNAIDS, 2004). Although ORT is a 'whole person' approach to the treatment of opioid use disorder (SAMHSA, 2018), there is evidence that women's contraceptive needs are overlooked. Statistics are difficult to obtain, but UK data have shown that only 35% of women receiving treatment for opioid use disorder employ contraception (Cornford et al., 2015;Mundt-Leach, 2014) compared with 75% of sexually active women in the general population (Family Planning Association, 2007). In addition, studies of women using opioids have indicated that seven to nine in every ten pregnancies are unintended (Black et al., 2012;Fischbein et al., 2018;Heil et al., 2011) and rates of abortion and children being adopted or placed in foster care are very high (Cornford et al., 2015;Martino et al., 2006). Women experiencing a substance use disorder also seem less likely than other women to utilise the most effective forms of contraception (Terplan et al., 2015). The primary contraceptive method used by women who depend on drugs is often condoms (Terplan et al., 2015); however, these are only 'moderately effective' and require the user to take decisive action at every sexual encounter (World Health Organization & Johns Hopkins Bloomberg School of Public Health, 2018). Illustrating this, a study of drug treatment clinics found that less than a quarter of female clients who said that they used condoms reported consistent use (Terplan et al., 2015). In contrast, implant and intrauterine devices (IUDs), which are generally considered more effective than condoms, are less likely to be used by women taking substances (Cornford et al., 2015;Fischbein et al., 2018;Griffith et al., 2017;Terplan et al., 2015). Whilst women receiving ORT may have higher rates of long-acting reversible contraception (LARC) use compared with oral contraceptive use, it is uncertain whether this reflects their preference or is simply the method most frequently offered to them by healthcare professionals (Cornford et al., 2015). Lower engagement with contraception by women taking opioids may in part relate to them underestimating their own or their partner's fertility (Harding & Ritchie, 2003;Mundt-Leach, 2014). Women who use opioids often experience amenorrhea and men who use opioids may experience erectile dysfunction. Both can lead women to wrongly assume that they are unable or unlikely to conceive and therefore do not need contraception (Elko & Jansson, 2011;Harding & Ritchie, 2003;Jessup & Brindis, 2005;Mundt-Leach, 2014). Other factors that may contribute to low contraception uptake include lack of information about the available options (Jessup & Brindis, 2005) and problems discussing contraception with sexual partners (Armstrong et al., 1991;Jessup & Brindis, 2005). In addition, women using opioids may avoid healthcare services because they are afraid of being stigmatised by staff (Armstrong et al., 1991;Black & Day, 2016) or may not want to discuss sexual activity with professionals (Edelman et al., 2014). Equally, they may be reluctant to use contraception with monogamous partners (Fischbein et al., 2018;Harvey et al., 2003); prefer to avoid contraception in case they experience side-effects (Banwell et al., 2009;Olsen et al., 2014); or not feel that contraception is a priority given other stresses in their lives (Hans, 1999).
Further to the above, contraceptive choice may be affected by elevated levels of sex work, intimate partner violence (IPV) and abuse among women using opioids (Cornford et al., 2015;Dietz et al., 1999;Edelman et al., 2014;Marchand et al., 2012). Among sex workers, condom use is generally high relative to other contraceptive methods, whilst the use of non-condom contraception is low (Duff et al., 2011;Jeal & Salisbury, 2007). This may reflect the fact that condoms tend to provide protection against sexually transmitted diseases as well as against pregnancy, and can be accessed comparatively easily and quickly (Duff et al., 2011). Among women experiencing IPV, meanwhile, take-up of contraception can often be undermined by a partner's reluctance to use condoms or by his desire for the woman to become pregnant (Gee et al., 2009;Melendez et al., 2003). There is also evidence that condom use is low amongst women who have experienced abuse, particularly violence, as they may be afraid of men becoming angry if condom use is proposed (Wingood & DiClemente, 2000;Zierler et al., 1996).
Most of the literature todate thus indicates that women receiving ORT have limited volition (or 'power') when it comes to making choices about contraception. Without underplaying this, it is important to recognise that this is not always or necessarily the case. For example, in one US study of 95 women who injected drugs or who were partners of people who injected drugs, most felt influential in their sexual relationships and reported that they actively participated in decisions about condom use, contraception and when to have sex (Harvey et al., 2003). Australian research has likewise found that women with current or past histories of injecting drug use expressed individual contraceptive preferences, had used different contraceptive methods at different points in their lives and were capable of making choices about their own reproductive health (Olsen et al., 2014). The authors of this research concluded that women who use drugs can make meaningful decisions about contraception and there is a need for treatment programs to capitalise on this by providing free, non-discriminating reproductive advice (Olsen et al., 2014).
The concept of power has been widely documented and debated within the health literature as well as within sociology and psychology more broadly (Allen, 2016;Butler, 1990;Foucault, 1963Foucault, , 1975Foucault, , 1978Foucault, , 1980Krause, 1977;Lukes, 2021;Miller, 1992;Turner, 1995). At its most general level, power refers to the ability or capacity to do something or to act in a particular way. Definitions are commonly divided into: (i) having the ability to influence others' actions (power over others) and (ii) having the ability to personally act/ change in the desired direction (power over oneself) (Ryn & Heaney, 1997), also known as self-efficacy (Bandura & Walters, 1977). Many feminists have, meanwhile, argued that power is more appropriately conceptualised as an ability or capacity to empower not only oneself but also others, and thus we should talk about 'power to' rather than 'power over' (Allen, 2016;Miller, 1992). Furthermore, 'power to' can be operationalised at the level of the individual (for example, enabling a woman to change her behaviour and/or make positive decisions about her own contraception) and the collective (for example, enabling women to act together to challenge the sexism and sexual oppression that underpin IPV, sexual abuse and inadequate contraceptive provision) (Young, 1994). Foucault (1963Foucault ( , 1975Foucault ( , 1980, who has significantly shaped discourses of power in society, has argued that power is a mobile and constantly shifting set of force relations that emerge from every social action. Moreover, power cannot be divorced from knowledge (Foucault, 1980). According to Foucault, healthcare providers routinely use medical knowledge to direct clients' healthcare choices (Foucault, 1963). Nonetheless, clients are capable of resistance (via 'micropowers') and often find unexpected ways of acting and exercising choice despite the power of the 'medical gaze' (Foucault, 1963(Foucault, , 1978(Foucault, , 1982. In this way, Foucault focused his attention on 'power over', particularly at the interpersonal level. Over the years, feminists have drawn upon, and developed Foucault's work; for example, using it as a foundation to explore how women are oppressed by men, how women police and discipline their own bodies, and even how women exercise power over each other (Allen, 2016;Bartky, 1990;Bordo, 1993;Butler, 1990). Nonetheless, feminists have also been extremely critical of Foucault; for example, highlighting how his work ignored gendered relations, overlooked structural inequalities, and ultimately failed to provide an account of empowerment that facilitates wider societal change (Allen, 2016;Deveaux, 1994;Harstock, 1990;Phelan, 1990).
Whilst personal empowerment is underpinned by constructs such as self-esteem, self-confidence, self-efficacy, selfcontrol, self-determination, and autonomy (Patrikar et al., 2014;Ryn & Heaney, 1997;Young, 1994), the empowerment of others can be operationalised through techniques such as sharing personal experiences, talking to others, and raising awareness (also known as 'consciousness raising') (Young, 1994). A further important pathway to empowerment is informed choice; that is having treatment options, knowledge about each option, and the ability to make decisions that reflect a person's culture, values, and views (Evans, 2000). Contraceptive decision-making is also affected by the nature of the relationship between sexual partners; resources such as education and income; and type of contraception used. For example, general population data from the United States indicate that women in married and cohabiting relationships have greater power to decide the type of contraception employed than women in dating relationships; higher relative education and higher relative income increase women's influence over contraceptive methods among couples living together; and women generally have more control over oral contraceptives whereas men are more dominant in relation to condom use (Grady et al., 2010).
Our aim in this paper is to better understand the relationship between contraceptive choice and power amongst women receiving ORT, including how, when and why these two phenomena interact. The salience of this issue emerged during analyses of a dataset generated as part of a qualitative study exploring community pharmacist provision of contraceptive services for women receiving ORT. The study was designed to inform a new community pharmacist-led intervention to support women in choosing and controlling when they conceive and start a family. However, preliminary readings of the data by members of the research team indicated that it was important to first understand, in a more general sense, when and how women felt both empowered and disempowered in making decisions relating to contraception. The team believed that this information could then be used by a wider range of services and professionals to better understand and support women in respect of their contraceptive choices and to prevent unplanned pregnancies that might subsequently result in terminations or care proceedings (Cornford et al., 2015;Edelman et al., 2014).

Methods
The research was conducted in the South of England between September 2016 and May 2017. It comprised semistructured interviews and overt non-participant observations with 20 community pharmacists during a typical shift (in order to explore whether and how current practice may need to change to facilitate a new community pharmacist-led contraceptive intervention), plus face-to-face semi-structured interviews with 40 women receiving ORT (in order to explore their views and experiences of contraception as well as their thoughts on receiving contraceptive information and advice within a community pharmacy setting). Only the interviews conducted with the 40 women receiving ORT are reported here. Ethical approval was granted by the Greater Manchester West Research Ethics Committee (REC Reference 16/NW/0376), with additional approvals received from the Health Research Authority (IRAS Ref: 20049) and participating local organisations and services.
Women were recruited from eight alcohol and other drug treatment services in rural, urban and inner-city areas. Eligibility criteria included being aged between 18 and 49 years old (to capture fertility years), current or recent history of ORT and English speaking. Potential participants were given written and verbal information about the study by a member of staff at the recruitment service and, if they expressed interest, were offered the opportunity to participate in an interview at a time convenient to them. All interviews were conducted either in a private room at the service from which they were recruited or in a quiet local caf e by one member of the research team (NA) who was trained in qualitative interviewing. Prior to starting any interview, NA reminded each woman of their right to decline to answer any questions or to withdraw from the study without giving a reason. She also asked them to provide written informed consent, including consent for their interview to be audiorecorded and for anonymised excerpts to be included in any study publications.
At the start of the interview, participants were asked a short series of closed questions to gather basic demographic and lifestyle data to help NA understand each woman and her circumstances, be aware of any issues that might be particularly difficult to discuss (such as sex work, terminations, or child loss) and tailor the interview accordingly. The remainder of the interview was then semi-structured in format and followed a topic guide that sought the participants' views and experiences of contraception, pregnancy and receiving contraceptive advice and pregnancy planning. Each interview lasted between 20 and 107 minutes and concluded with a short debrief during which women who had discussed experiences of violence, abuse or other matters of concern were offered further support from the service where they were recruited and given details of other organizations that might provide additional help. Participants were also offered a £15 high street shopping voucher as a token of thanks for their time.

Analysis
All interviews were labelled with a unique participant identifier, transcribed verbatim by a professional transcription service, and then imported into NVivo V11.4.3 for systematic line-by-line coding. As the main aim of the study was to design a community pharmacy-led contraceptive intervention, the initial coding frame was based on codes derived from the Theoretical Domains Framework (Cane et al., 2012) and the COM-B system (Michie et al., 2011), both of which are intended to understand behaviours that affect the successful development and implementation of new interventions and practices. Coding was undertaken by authors HF and NA, with early team discussions indicating the need for additional 'inductive' codesthat is codes reflecting topics that emerged more spontaneously from reviewing the data. One of these inductive codes was 'power', a concept to which team members were already sensitised given its relevance to the existing contraception literature and centrality to health decision-making.
During the second phase of coding, HWin consultation with JN and HFre-reviewed all 40 interview transcriptions to identify extracts that seemed to capture the concept of 'power' in relation to contraceptive choice. No formal definition of power was used since the intention was to be as inclusive as possible to minimize the chances that any manifestations of power (or powerlessness) would be overlooked. All interview extracts labelled as 'power' were then exported into a single Microsoft Word document and analysed inductively following the stages of Iterative Categorisation (IC) (Neale, 2016(Neale, , 2021. To this end, HW first read and re-read the contents of the Word document and then summarised each text extract into one or more bullet-points labelled by participant identifier. Next HW, HF and JN iteratively grouped and regrouped all the bullet points into categories, gave the categories descriptive labels, and organised the newly labelled categories into a coherent structure which was discussed and agreed upon by other team members. Quotations (pseudonymised) were then selected to represent salient points and, finally, the findings were related back to the literature and implications for practice were discussed within the team.

Participants
All participants were aged between 22 and 49 years old (mean ¼ 38 years). Most (n ¼ 25) were White British; most (n ¼ 32) defined themselves as heterosexual, and over a third (n ¼ 15) reported that they had engaged in sex work. Twenty-four said they were in a current relationship and sixteen identified as single. In total, 25 were currently prescribed methadone, 9 were prescribed buprenorphine, 2 were prescribed combined buprenorphine and naloxone (Suboxone), and 4 had no current prescribed ORT (2 were detoxing, 1 was buying ORT from the street, and 1 was about to start a new episode of ORT). Prescriptions ranged from 1 day to 27 years (missing ¼ 11) (see Table 1 for further participant characteristics).
Thirty-six of the 40 participants had ever been pregnant. Collectively, these 36 women had experienced 136 pregnancies and given birth to 84 babies. Sixteen women reported a total of 37 miscarriages, 11 reported a total of 15 terminations, and a small number reported that a child had died. Only 4 participants had never been pregnant. At the time of the interview, 20 participants were not using any form of contraception; 5 were using condoms; 2 were using the withdrawal method ('coitus interruptus'); 2 were using the progestogen-only pill ('the mini-pill'); 2 had an IUD ('coil'); 1 had an implant; 1 had a contraceptive injection (Depo-Provera); 1 was taking a combined oral contraceptive pill ('the pill'); and 1 had an expired IUD. In addition, three had been sterilised and two were pregnant. Although 29 participants were potentially at risk of pregnancy since they were either heterosexual and in a relationship or sex working, only 10 of these women were using contraception or had been sterilised (14 were not taking any precautions, 2 were pregnant, 2 were using the withdrawal method, and 1 had an expired IUD).

Findings
As our participants were never explicitly asked to state whether they felt 'empowered' or 'disempowered' when discussing their contraceptive choices, the word 'power' did not feature directly in their accounts. Power is a latent concept and so we had to rely on our collective judgement when deciding on material that seemed relevant. Accordingly, we do not claim that our findings are an absolute representation of power or lack of power; rather they offer insights that provide an increased understanding of how power operates within the context of contraceptive decision-making by women receiving ORT. With that clarification made, our analyses indicated that power manifested itself via six interconnected topics or 'fields': i. 'information about fertility and contraception'; ii. 'access to contraception'; iii. 'relationships with professionals and services'; iv. 'relationships with male partners'; v. 'relationships with sex work clients'; and vi. 'life priorities and preferences'.

Field 1: Information about fertility and contraception
Most participants reported that neither healthcare providers nor the staff at their local community drug team had ever given them any information on contraception or on ORT's impact on fertility. One woman stated that she had asked her pharmacist how ORT would affect her menses and the pharmacist had not been able to answer. Others said that general practitioners (GPs) had advised them to initiate oral contraception but without providing information on the likely side effects or other available methods. For example, one woman who had a history of depression stated that she had not been told that the combined pill and depot injection might affect her mood and, after using each, she had started to feel suicidal.
Because of this lack of information, women often reported that they felt ill-equipped to assess the different methods of contraception or to choose the approach most suitable for themselves. Furthermore, this lack of information had sometimes resulted in misunderstandings or had caused women to turn to potentially unreliable information sources, including acquaintances, family and friends. For example, some women receiving methadone or Subutex had wrongly assumed that they were unlikely to become pregnant because their libido was low, they had low body weight, or, as one woman explained, their partner's sperm was 'sleepy' because he was also receiving ORT: Other participants complained they had been given contraception without any directions on usage. Thus, one woman said that she had not initially realised that she needed to take additional precautions to prevent sexually transmitted infections (STIs) whilst taking oral contraception and another said she did not know what to do if she missed a pill. Participants also said that they had not been told when their coil or implant would have to be replaced or when they would need another injection. Illustrating this, one participant said that she had accidentally become pregnant after a misunderstanding when her injection had to be renewed and another said that she had developed a bacterial infection because she had never been told that her coil needed to be changed: And I just bled all the time and it smelt … I'd had swabs done to say there was like a bacterial infection, but still, nobody had said to me 'your coil might need changing', you know. Nobody kind of knew, nobody. (Megan, 43 years, White British, heterosexual, experience of sex work) Only two participants reported that they had been given information and advice about fertility or contraception when first starting ORT and many said that they lacked information on contraception and fertility in general. Nonetheless, others stated that they had received information (on types of contraception and their respective potential side-effects) from GPs, family planning services, staff at abortion centres, and hospital maternity departments at different times in their lives. Women generally appreciated any contraceptive information received and explained how this gave them choices and enabled them to identify a method that suited their circumstances best: I did used to go to the family planning clinic [ … ] and they did like tell me about what types of contraception I could use and stuff. That was a long time ago though. But, yeah, they was quite helpful. (Anna,43 years,White British,heterosexual,no sex work) The three participants who had been sterilised all stated that they had received detailed advice and information that had helped their decision-making and were happy with their choice. Two had been provided with information on different contraceptive methods, including sterilization, at the hospital after giving birth, and a third had opted for sterilization after consultations with a series of doctors: I went to my doctors and asked to be sterilised … I had to go through like different stages, having interviews with three different doctors, and [I] quizzed them enough and I got sterilised. (Aurora, 46 years, White British, heterosexual, no sex work) Despite this, one participant spoke about a friend who had been told by her probation officer that the only way to secure custody of her two children was to be sterilised. The friend had proceeded with this, but her children were later taken into care and she now wanted her sterilisation reversed.

Field 2: Access to contraception
Participants frequently reported that gaining access to contraception, the morning after pill, and terminations were all very difficult. Some women said that they had not been offered any contraception until after they had become pregnant or given birth when, to them, it seemed 'too late'. Meanwhile, others complained that they had not been given any practical support with contraception at critical moments, such as after giving birth or following a miscarriage or termination: Interviewer: So after your abortion which advice did you get? Participant: Nothing. Absolutely nothing … Just Paracetamol and a leaflet about bleeding. That was it. (Angela, 38 years, White British, heterosexual, experience of sex work) One participant described how hospital staff had been more concerned with contacting social services about her newborn baby than with offering her contraception. Two others said that although they were given information (after a miscarriage and birth respectively), they were not actually given any contraception. Likewise, another participant said that she had been given information about fertility after leaving an in-patient detoxification programme, but this had not been accompanied by any provision and she had become pregnant two weeks later.
Some women also remarked that condoms were expensive to buy and/or difficult to obtain, and consequently they had not always used them. One woman commented that she knew free condoms were available from needle and syringe services, but she did not use these services as she smoked heroin. Two other women reported that sex workers were often too self-conscious or caught up in their 'drug-taking lifestyle' to ask for, or seek out condoms: There is a lot of girls [sex workers] out there that don't use [condoms], because they are too embarrassed to come and ask or just, you know, too caught up in the drug life. (Sophie, 38 years, White Irish, bi-sexual, experience of sex work) Although several participants said that they had talked to doctors when particular contraceptive methods had caused them side effects or concerns (such as weight gain or blood clotting), doctors had not offered them any alternatives. As a result of this, one woman had stopped taking the pill and then became pregnant unexpectedly. Others felt that they had been given contraception (particularly the pill and condoms) that required a level of reliability and attention that was not compatible with their drug use and associated lifestyle. Again, this had resulted in some reporting unwanted pregnancies.
Despite these difficulties, several women stated that they had found it easy to access oral contraception, either directly from their doctor or via a family planning service, and others reported that free condoms were readily available from a range of services, such as drug treatment services, sex work initiatives (including a mobile van), sexual health clinics, doctors' surgeries, and services helping women in the criminal justice system. Participants said they valued services that provided condoms discreetly and without them having to ask. Reflecting this, two women who were sex workers spoke positively of the fact that condoms were usually provided, and often compulsory to use, when they worked in off-street establishments: I have worked in like massage parlour-type establishments where there is more than usually like a basket of condoms supplied like in the room … There is support there for sex workers, so I have always found it easy to access [condoms]. (Leah, 37 years, White British, bi-sexual, experience of sex work)

Field 3: Relationships with professionals and services
Many participants stated that the negative attitudes of healthcare professionals had prevented them from either seeking or receiving the contraceptive support they needed. In this regard, several participants described both doctors and doctor surgery receptionists as rude, unsympathetic, judgemental, and unhelpful. Some women explained how this had stopped them from attending services and asking about contraception or, in one case, had made them inclined to disregard the doctor's advice: I think some of them [doctors] are approachable and some of them aren't approachable … their attitude they give you back when they respond to you is not an attitude [that would make you] say 'OK, I'm going to take your advice'. (Aaliyah, 43 years, Black British, heterosexual, no sex work) Some participants also reported that they had been deterred from talking about contraception if their doctor was male (which made women feel uncomfortable) or because their doctor had a strong accent (which meant they could not always understand what was being said). A few participants additionally said that appointments with doctors were strictly time-limited and this inhibited any in-depth discussion. Indeed, one woman recounted how her GP had used a stopwatch to prevent appointments from lasting too long. In a very small number of cases, participants also stated that they had felt pressurised or even coerced into using contraception by healthcare professionals. For example, one woman who had had multiple pregnancies reported that hospital staff had insisted she has an implant before she left the hospital after her last birth.
Compounding these problems, a few participants remarked that they felt embarrassed going to family planning clinics, to the extent that one woman said that she would not attend them. Others felt that family planning services were not intended for women 'like them' or were not suitable for their needs. For example, some felt that they were for pregnancy testing or for receiving the morning after pill (rather than for accessing contraception) or were only for 'younger women', 'pregnant women' or 'women with children'. In addition, one participant pointed out that family planning clinic opening hours were unsuitable for sex workers who work at night and sleep during the day. Another participant said that she had begged for help at a family planning clinic after learning she was pregnant, but the staff had been rude and unhelpful and had only given her a leaflet on abortion: I went to the family planning and I broke down when he told me I was pregnant and all she did was give me leaflets on abortion. That was it. (Angela, 38 years, White British, heterosexual, experience of sex work) More positively, other women emphasised that professionals openly discussed and encouraged contraception use and so there was no need to feel any embarrassment or stigma. Furthermore, two women reported that they had felt more confident discussing contraception with professionals as they had become older and 'more mature'. Accordingly, several participants maintained that talking to professionals about contraception felt 'normal', 'very comfortable', and 'safe and secure':

Field 4: Relationships with male partners
Many women stated that their male partners did not engage in discussions about contraception and instead left all the decision-making and responsibility to them. Some women experienced this as burdensome and unfair. Moreover, several explained that they had sometimes forgotten contraception because they had been intoxicated. In contrast, others felt that women should be in control of their own bodies, so should be responsible: As far as I'm concerned, I'm the one that's going to fall pregnant so [contraception] was up to me. I didn't give him a say in that. (Mia,47 years,White British,heterosexual,no sex work) Several participants described how they had previously used contraception to avoid becoming pregnant by violent partners who raped them. Indeed, two women, who did not want to become pregnant, reported that they had secretly used a depot injection and the pill respectively because their partners had objected to using any form of contraception as they wanted children: In terms of condom use, however, the situation was different, with participants generally reporting that men took the lead in decision-making. Whilst some women approved of this, others complained that male partners unilaterally decided against condom use (often arguing that condoms rendered sex less pleasurable). In addition, several women explained how they had been forced into sex without condoms by partners who had abused or raped them, which for two women had resulted in unintended pregnancies and terminations. Frequently, women said that it was difficult or impossible to negotiate condom use with some men. As a result of this, some participants had unprotected sex, which for one had led to two terminations: Some [men] don't negotiate 'cause they just don't want to wear one [condom]. Or some guys would insist on wearing them. So, it just depends on the guy. (Victoria, 38 years, White British, heterosexual, experience of sex work) Despite this, several participants emphasised that they personally took control of decisions about condom use. For example, a few said that they insisted on condom use as a matter of routine when having sex with partners as this avoided repetitive negotiation, plus men then seemed less likely to resist. Furthermore, condoms protected them from both pregnancy and STIs. Indeed, one participant stated she would only have sex without condoms if her partner had been recently health checked: I usually make [partners] go to the clinic and get checked out together to make sure we are both okay and stop using [condoms]. (Anna,43 years,White British,heterosexual,no sex work) Some women also explained that they had made a conscious decision not to use condoms as they were in stable relationships, using other forms of contraception, felt that condoms reduced intimacy or passion, or did not themselves like the sensation of condoms. In addition, women who were sex workers sometimes reported that they had decided not to use condoms in their personal relationships, although they used them with clients, to highlight that their partners were 'different' and 'important'. Meanwhile, several women said that male partners shared contraceptive decision-making with them or had supported their contraceptive choices. Furthermore, two stated that their partners had taken full responsibility for birth control by having a vasectomy.
Field 5: Relationships with sex work clients Several participants who had experience of sex work said that they found it difficult to raise the topic of condoms with clients. This included one woman who described having sex with her drug dealer in exchange for substances. A few others stated that condom use was sometimes difficult to enforce because clients would argue that they were allergic to rubber or latex, 'safe' or could not hold an erection or orgasm with condoms. Additionally, some clients would offer monetary incentives for sex without condoms: Now and again [I would not use condoms with clients] … So stupid. I can't believe I have done that. You know, they give you a bit of extra money, but you think 'how many other girls have they done that with?' (Haideh, 44 years, White British, heterosexual, experience of sex work) In contrast, other women reported that they always insisted on condom use during penetrative sex with clients, and two said they additionally used condoms for oral sex. The main reason women who had experience of sex work gave for enforcing condom use was to protect themselves from STIs, with several participants clarifying that they used additional methods (such as, spermicide, a diaphragm and depot medication) to provide further protection against pregnancy. Although several women noted that some clients took condom use seriously and would provide their own condoms, a few women had developed additional strategies to ensure compliance. For example, some participants said that they put condoms on clients to prevent clients from purposefully damaging them or as part of their sex worker role and service. In addition, two women were always accompanied by an acquaintance (in one case her romantic partner with a dog) who could be called in the case of condom refusal.

Field 6: Life priorities and preferences
Several women reported that contraception had often not been a priority for them or had not 'entered their head' as they were more focused on using drugs (or trying to stop using drugs), did not care about their own health or safety, had other more pressing health issues to worry about, or simply 'stuck their head in the sand': I think I found it easier to stick my head in the sand, you know, rather than try and deal with it [contraception] properly. It is a typical drug user thing to do, stick your head in the sand and deal with it by … blocking out any pain, any emotion, by carrying on using drugs … just telling yourself one day, one day, one day maybe things will change. (Libby, 25 years, White British, heterosexual, no sex work) Importantly, however, other women explained how unintended pregnancies, miscarriages, terminations, and having children placed in care had 'jolted them' into thinking about, enquiring about, and ultimately initiating contraception: We [partner and self] weren't the best role models. That's the best way of looking at it, and I didn't think it would be fair to be a single parent to two children. So, I spoke to my family about it, and I went to the [abortion] clinic [ … ] and, when I had the termination, I had the Mirena coil fitted. (Emiliabeth, 37 years, White British, heterosexual, no sex work) Amongst women who reported a contraceptive preference, 'condoms', 'the pill', 'the depot injection' and 'the coil' were variously discussed. Those who preferred condoms generally said that they liked them because they were nonhormonal, had no side effects, did not require medical advice and were less likely to be forgotten than the pill. Those who preferred the pill valued the fact that this method was nonpermanent, straightforward, simple to understand from the information provided, and could easily be stopped when planning a pregnancy. Meanwhile, participants who preferred the depot injection or the coil described these methods as convenient, especially as there was no requirement for them to remember anything or take any action prior to having sex.

Discussion
In seeking to understand the relationship between contraceptive choice and power amongst women receiving ORT, we have identified six interconnected 'fields of power': i. 'information about fertility and contraception'; ii. 'access to contraception'; iii. 'relationships with professionals and services'; iv. 'relationships with male partners'; v. 'relationships with sex work clients'; and vi. 'life priorities and preferences'. Examples of powerlessness were prominent across all six fields, although responses tended to vary by situation and circumstance (for example, some women had different opinions about condom use with a partner versus with a sex work client or thought differently about contraception depending on their current substance use). In addition, participants often expressed different views of each contraceptive method. Importantly, however, each of the six fields of power identified was also associated with examples of women making positive choices for themselves and taking control in respect of contraceptive decision-making, even if this had to be done secretively or with the support of another (for example, a chaperone when sex-working).
Illustrating this complex interplay, many participants said that a lack of information had prevented them from assessing options and choosing appropriate contraception, whereas others reported that good information had enabled them to make informed choices and had resulted in them feeling satisfied with their decisions. Although participants often described poor access to contraception (stating that condoms were expensive and embarrassing to buy, and other options were limited, unsuitable or only offered too late), others identified easy and free access to contraception that had enabled them to avoid unwanted pregnancy. Some women spoke of disempowering interactions with professionals who had been rude, unsympathetic and judgemental, whereas others detailed constructive and helpful discussions. Although male partners and sex work clients generally took more control over condom use than other contraceptive methods, women also articulated ways in which they had successfully managed contraception even with violent or abusive men. Lastly, some women acknowledged that they had not previously prioritised contraception but had taken precautions later in their lives.
Consistent with previous research (Cornford et al., 2015;Mundt-Leach, 2014), our study found relatively low levels of contraception use (only a third of participants at risk of pregnancy were using contraception) and high rates of abortion (just over a quarter of participants reported a termination) (Cornford et al., 2015;Martino et al., 2006). Levels of sex work, intimate partner violence, and abuse were also high (Cornford et al., 2015;Dietz et al., 1999;Edelman et al., 2014;Marchand et al., 2012) and condoms were more likely to be used compared with any other contraceptive method (Terplan et al., 2015). As found by others, some women appeared to underestimate their own or partner's fertility (Mundt-Leach, 2014;Harding & Ritchie, 2003); avoided seeking contraceptive advice from healthcare services because of negative staff attitudes (Black & Day, 2016); and did not always feel that contraception was urgent given other stresses in their lives (Hans, 1999). Contraception use was additionally undermined by difficulties discussing condom use with men (Catania et al., 1992;Fullilove et al., 1990;Jessup & Brindis, 2005) and by partners' reluctance to use condoms and/or their desire for the woman to become pregnant (Dietz et al., 1999;Melendez et al., 2003).
Building on insights from earlier studies, our analyses identified additional issues of concern. Information about, and provision of, contraception was described as limited, particularly when women started receiving ORT. Participants reported a lack of awareness of the various types of contraception and also little understanding of how to use or replace what was provided to them. Many described healthcare professionals as being unable or unwilling to engage them in discussions about fertility or to proactively offer contraception, and a minority stated that professionals had pressured them into accepting contraception against their will. Sometimes women identified a lack of access to free condoms or said that family planning clinics were unsuitable for their needs. Despite this, they expressed appreciation of any advice and support they received and frequently articulated and rationalised their preferences for particular contraceptive methods. Furthermore, some reflected on how their views of contraception had changed, especially in response to unintended pregnancies, miscarriages, terminations, and children being taken into care.
In terms of the concept of power, our findings speak to definitions of both 'power over' (Foucault, 1963(Foucault, , 1975(Foucault, , 1980 and 'power to' (Allen, 2016;Miller, 1992). Beginning with 'power over', women's choices were often undermined by how they were treated by others. This included hostility and stigma from professionals, violence and rape by partners, and manipulation and exploitation by sex work clients. In respect of 'power to', factors that facilitated women's empowerment included the provision of information and advice by professionals, the sharing of contraceptive decision-making with supportive male partners, and the readiness of some sex work clients to carry and use condoms. These findings remind us that interpersonal relationships and everyday social interactions can be both disempowering and empowering. Moreover, the key actors in contraceptive decisionmaking are not only women and the healthcare professionals who provide access to contraception. Power dynamics are also played out between women and their partners in the domestic sphere and between women and clients within the context of sex work.
Within the home, a complex constellation of gendered relations will influence whether, when and how women are able to assert control over contraception use (Grady et al., 2010;Harvey et al., 2003;Jessup & Brindis, 2005;Olsen et al., 2014). Within sex work, decisions will additionally be affected by the many and diverse power relationships involved in transactional sex (Sanders & Campbell, 2007). Power in relation to contraceptive decision-making, therefore, extends beyond the interpersonal sphere. Choices are also affected by wider power dynamics that differ from one context to another, overlap, and may not always be visible in one-toone interactions. These relate to gender, income, education, employment, age, and sexual orientation, but also to race, class, and culture, etc. Reflecting this, many feminists have, in recent years, turned to intersectionality as a useful paradigm for understanding how power relationships operate (Connell, 2009;Crenshaw, 1991;Hankivsky & Cormier, 2009). Intersectionality recognises the significance of gender but does not assume that gender is the only, or necessarily the most important, axis of power. This can help explain power differences between women and why individual women may change their behaviours and choices over time and place.
Combining our empirical data and theorising enables us to envision ways that women receiving ORT might be empowered in making contraceptive decisions. At the individual level, women clearly need to be provided with information (as a basis of 'knowledge') so that they can make meaningful choices. All information needs to be given in a clear, relaxed, and non-judgemental way, and ideally include strategies for discussing and negotiating contraceptive use with reluctant or abusive partners. We note that our participants deployed their own creative strategies to reassert some control over contraceptive behaviours even if sexual partners were opposed or reluctant to engage. These included using contraception secretly, routinizing condom use, applying condoms as part of the sex act, refusing to negotiate about noncondom use, and enlisting physical protection from others. These acts are, however, best described as forms of resistance or micro-powers (Foucault, 1978(Foucault, , 1982 and, as such, constitute only relatively weak forms of empowerment. We, therefore, need wider service, system and societal level commitments and strategies to ensure that all women have easy access to contraceptive options and feel genuinely empowered in making decisions that meet their personal contraceptive needs (Young, 1994).
In the UK, contraception is freely available to those registered with a GP, although condoms generally have to be bought from shops or pharmacies. Women receiving ORT will not always have easy access to, or be willing to visit, a GP and many may therefore not access contraception via this route. Furthermore, the cost and effort of purchasing condoms can be barriers to use, particularly if women are embarrassed or afraid of being judged (Bell, 2009;McCool-Myers et al., 2019;Moore et al., 2006). Professionals, including pharmacists, working with women receiving ORT should therefore proactively raise the topic of contraception, particularly when women initiate treatment. Additionally, information and advice must be offered in hospitals and other generic healthcare settings, especially after women experience births, miscarriages and terminations. Contraceptive devices should then be given whenever and wherever interest is shown. To this end, those working with women using substances may need training in the various contraceptive options available, potential side effects, and methods of use.
Supplementing these strategies, free access to condoms should be provided in all drug treatment services, not just those targeting people who are injecting or engaged in sex work. Furthermore, staff in family planning clinics could potentially work more closely with those in specialist addiction and sex work services, even providing satellite and outof-hours clinics. Lastly, and although not explicitly discussed in our research, women may find it helpful to share their experiences with, and learn from, their peers in supportive all-female groups (Lennon-Dearing, 2008;Prendergast et al., 2011;Young, 1994). This does not necessarily mean bringing women with diverse life experiences and circumstances together just because they are using substances. A more productive and empowering approach could involve recognizing the heterogeneity of women's lives and circumstances and enabling those receiving ORT to meet in small groups based on shared characteristics and experiences such as IPV, sex work, age, and past pregnancy outcomes.

Methodological reflections
The analyses we present have both strengths and weaknesses. We interviewed a demographically diverse group of women with a range of personal experiences of contraception and pregnancy. Women were encouraged to speak openly and freely about their experiences in a supportive context with a trained qualitative researcher. Methodologically, it would have been very difficult, and probably impossible, to have asked participants to discuss such a complex and abstract concept as 'power'. Instead, we sought to identify and extract all incidents relating to power that were latent in the women's accounts. Our team comprises researchers with backgrounds in health psychology, psychology, social science, addiction science, perinatal mental health, and pharmacy. Via a line-by-line re-analysis of the data and regular team discussions, we hoped to achieve a rounded and balanced understanding of power in relation to contraceptive choice. Despite this, our interviews were all conducted in England and with women in receipt, or recent receipt, of ORT. We did not interview any men and our participants did not always clarify whether they were discussing recent or historical experiences of contraception. Accordingly, our findings cannot automatically be generalised to other populations and settings and additional research on the views and experiences of women in other countries, as well as men who are sexual partners of women in receipt of ORT, would be valuable.

Conclusions
Our research indicates that power in relation to contraceptive choice is multi-faceted and multi-directional, operating at both individual and structural levels. It is played out in health care settings, private homes, and sex work venues and reenacted as life circumstances and personal priorities change. Even when women appear to have limited power or control, they sometimes find a way of asserting themselves via resistance and micro-powers (Foucault, 1978(Foucault, , 1982, such as secretive use of contraception and routinization of condom use. Providing clear, non-judgemental information and advice and ensuring universal and easy access to contraceptive options underpin informed decision-making. Further service and system-level commitments are, however, needed to reduce the number of unintended pregnancies, terminations and children being taken into care. These include training professionals so that they can better support women and their partners; building closer relationships between those working in the substance use and sex work sectors and those working in sexual health so they can share knowledge; and increasing visible and free access to condoms so that women do not have to risk embarrassment by asking for them. Efforts should also be made to bring women with shared experiences together so they can learn from and support each other but also potentially raise awareness of, and challenge, some of the wider societal injustices and abuses that impede their contraceptive choice.