Screening of studies

Studies identified from the searches were collated in Mendeley, de-duplicated and transferred into Rayyan for screening. One reviewer screened all articles with a second reviewer screening 20% of articles. Full texts of the potentially eligible studies were screened by two independent reviewers. All studies identified for screening were available with title and abstract in English language or French.

Data extraction

A data extraction form was developed in Microsoft Excel based on the standard JBI template. Two reviewers each extracted data from 50% of the included studies. A third reviewer independently extracted data from 10% of studies for quality assurance. Data were extracted on study publication date, country, setting, study type, aim/purpose, total sample size, sample size of participants relevant to the inclusion criteria, participant characteristics, methodology, focus of study, outcome assessed and findings. Quantitative data were summarised and topic areas within each study were coded. Once coded by topic, studies were grouped by frequency of topic and categorised accordingly.

Year of study

The oldest included study was from 1998 [3], whilst the most recently published were from 2023 [80, 107]. Over the 25 years that relevant studies were published, 62 were published within the last 10 years. The highest number of studies were published in 2014 (n = 10). Fig 2 displays the number of included studies by publication year.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Number of included studies by publication year.

https://doi.org/10.1371/journal.pone.0311554.g002

Context

Of the 140 LMIC globally, only 20 countries (14%) were represented in the included studies. Fig 3 illustrates the breadth of countries in the review according to the number of included studies from each of those countries. All studies focused on a single country, except for one study which included participants drawn from three South American countries [29]. Most included studies were conducted in the continents of Africa (n = 41) and Asia (n = 41), with n = 18 studies being set in the continent of South America. The Middle East was the setting for n = 22 of the included studies.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Number of included studies by country.

https://doi.org/10.1371/journal.pone.0311554.g003

India was the country setting for the largest number of studies (n = 17). Other countries with >10 included studies were Brazil (n = 14), South Africa (n = 13) and Turkey (n = 12). This meant that 4 of the 140 (3%) LMIC globally accounted for 56% of all included studies.

The topic of studies carried out in India were heterogenous. Eight of the 14 studies were carried out in Brazil and 11 of the 13 studies carried out in South Africa focused on HIV. In contrast, most studies set in Turkey were focused on sexuality and sexual function (n = 6) and family planning (n = 4).

The majority (n = 7) of studies in Uganda were based around HIV and HIV risk behaviours. Four of the six studies set in China were based around sexual function, three of which utilised the Arizona Sexual Experiences (ASEX) Questionnaire [59, 60, 113]. All Tunisian studies (n = 4) were focused on sexuality and all Iranian studies (n = 3) were qualitative.

Psychiatric hospital outpatient facilities were the most frequent setting for studies (n = 46), followed by psychiatric hospital inpatient settings (n = 29). In addition, a proportion of studies utilised mixed inpatient and outpatient psychiatric settings (n = 20), five of which also incorporated participants from the community. Two studies were conducted in primary care, both of which were based in China. A small number of studies were based solely in the community and not in health facilities (n = 3).

Participants

A total of 38,510 participants were included in the studies, however some studies used the same participants so the total likely overestimates the number included. Of the total number of participants, 19,987 (52%) were relevant to the study population of this scoping review. Small numbers of men and women without SMI may be included in this total in studies that were not disaggregated by diagnosis or gender.

The most represented SMI diagnosis was schizophrenia. Twenty-eight studies focused solely on schizophrenia and people with schizophrenia were included in a further 54 studies. People with a diagnosis of bipolar disorder were included in 55 studies with an additional 8 studies exclusively focused on people with bipolar disorder.

Generally, inclusion criteria of the studies restricted participation to those 18 years of age or older. Only four studies [28, 26, 89, 102] included women aged 15-years or above and one study included those aged 16 or older [18]. The oldest recorded participant included in a study was 83 years. Reporting of age ranges within studies was variable and some studies reported mean age of participants only. Where age range was reported, it was noted that only a handful of studies included women of menopausal age and older.

Participants without lived experience of SMI were included in 11 studies [19, 33, 38, 40, 50, 54, 78, 82, 94, 103, 105]. These participants included psychiatric nurses, midwives, administrators of mental health care facilities, caregivers or family members of people with SMI, government decision makers, mental healthcare providers, and people living in communities that include people with SMI.

Ethics

Included studies were searched for evidence of ethics approval, informed consent and assessment of capacity to consent. Nineteen of the 100 included studies did not report whether ethical approval had been obtained [19, 23, 25, 34–37, 43, 44, 51, 55, 65, 66, 71, 81, 90, 102, 104]. Seventy-six studies explicitly stated that informed consent had been sought from participants but of those, only 16 studies [2, 5, 31, 32, 36, 37, 42, 52, 57, 60, 67, 70, 99, 110, 111, 113] included a statement that capacity to consent had been assessed prior to participants being included in studies. This means that almost a quarter of studies did not provide information in their published papers as to whether consent had been obtained and most studies in the review (n = 79) did not include information as to whether capacity to consent had been assessed. A small number of papers (n = 5) did not require assessment of capacity to consent due to the nature of the study or the focus of the study being participants without lived experience of SMI [19, 33, 38, 40, 112].

Of the studies that included participants aged 15 or 16 years and older (n = 5), none explicitly stated that capacity to consent had been assessed [18, 28, 89, 97, 102] and one study did not document that consent had been obtained. All five studies did provide information indicating that ethics approval had been obtained.

Analysis by country of publication indicated that the greatest number of studies not reporting ethics approval were published in India (n = 10) [3, 23, 34–37, 43, 51, 65, 66] but Tunisia had the greatest proportion with three [25, 55, 81] of the four studies not reporting this. Of the eleven studies that did not report all three areas of ethics approval, informed consent and capacity to consent, n = 4 were published in India [23, 43, 51, 66], n = 3 were published in Tunisia [25, 55, 81], n = 2 were published in Brazil [71, 90] and n = 1 were each published in Nigeria [104] and Turkey [44]. Analysis did not detect any patterns by year of publication with regards to ethics approval, informed consent and capacity to consent assessment among the included studies.

Methodology

Most studies (n = 76) implemented a cross-sectional design. Other quantitative study designs were one case control study [55] and one cohort study [37]. Questionnaires were the main data collection tool used in these studies. A sizeable number of studies also used blood samples as a data source.

There were 22 qualitative studies included in the review, two of which used ethnographic methodologies [40, 90]. There were two case studies in the review [23, 43]. Most of the qualitative studies used semi-structured interviews (n = 18). The range of sample sizes for the qualitative studies was 14–200 participants.

Study focus

Included studies frequently had multiple areas of focus. Focus areas were grouped into 6 overarching categories. The largest domains were HIV prevalence, risk behaviour and knowledge (n = 32), sexual function (n = 30), contraception use and family planning (n = 25), sexual violence (n = 21), fertility, pregnancy and the post-partum period (n = 19 studies) and non-HIV STIs (n = 11). The most relevant results from each study have been reported in S2 File.

HIV (prevalence, risk-taking behaviour and knowledge)

All of the 14 studies that ascertained the prevalence of HIV reported it to be higher in women with SMI than in men with SMI [21, 30, 39, 53, 57, 69, 70, 72, 76, 77, 80, 86, 97]; however, not all of these studies were probabilistically sampled. Sixteen studies researched HIV risk behaviour [2, 3, 18, 34, 36, 40, 49, 53, 68, 75, 83, 90, 94, 102, 106]. Results were varied, although the majority reported that HIV risk behaviour was generally higher among women with SMI than men with SMI. Key findings from three qualitative studies [40, 90, 94] found mental healthcare providers 1.) were more concerned about the avoidance of pregnancy than the avoidance of HIV, 2.) perceived patients to be at high risk of HIV, and 3.) noted that patients may not be able to ‘resist’ high risk sexual behaviour and may also lack the agency to refuse unprotected sexual intercourse.

Six studies assessed the knowledge of people with SMI about HIV/AIDS [2–5, 37, 102] with inconsistent findings between the studies. Three of the studies reported that HIV knowledge in women with SMI was lower than a comparator group of men with SMI [3, 37] and a comparator group of women without SMI [102]. Women with SMI had a higher knowledge score in one study [5]. Two studies reported no significant difference in HIV knowledge score between men and women with SMI [2, 4]. Limited evidence from one study reported that a higher proportion of women with SMI attended HIV appointments compared to men with SMI [62]. A study looking at whether men and women with SMI had ever had testing for HIV reported no significant difference [99]. One qualitative study from South Africa [41] identified barriers and challenges in delivering effective HIV counselling. Examples of how the health, social and wellbeing needs of women could be better met included improved data collection on HIV prevalence among psychiatric inpatients, capacity to support health literacy and ensure test results and diagnosis are fully understood and a need to improve follow up in the community after hospital discharge, particularly where a patient returns to a rural area [41].

STIs (not including HIV)

Most studies focused on non-HIV STIs used men with SMI as a comparator, with only one study comparing to women without SMI [75]. The prevalence of a range of STIs were reported. One study showed a similar prevalence of chlamydia between men and women [30]. In the three studies that reported the prevalence of syphilis, one showed no difference [79] whilst the other two showed prevalence to be higher in men [30, 112]. A qualitative study noted the perception of mental healthcare providers that rates of syphilis were higher in their hospital population compared to the general population [40]. Two studies reported hepatitis B prevalence, both showed a higher prevalence of hepatitis B in men with SMI than women with SMI. Similarly, a study of hepatitis C prevalence showed a higher prevalence in men [30, 32]. Prevalence of history of STI was reported by five studies [18, 32, 42, 45, 75]. Generally, studies showed a similar prevalence for history of STI between women and men but a study that compared women with SMI to women without SMI showed a higher prevalence of STI in women with SMI [75].

The two qualitative studies included utilised healthcare provider staff participants only or in combination with people with SMI [40, 90]. One qualitative study drawing on interviews with mental healthcare providers in Brazil indicated that SMI patients’ decision making in relation to risk and sexual behaviour was impaired by their mental health illness, which led to unprotected sex and in turn a higher risk of STI than the general population [90]. Dialogue with patient focus groups echoed this, indicating there may be greater sexual health needs for this population in terms of risk perception, decision making and acquisition of STIs [90].

Sexuality and sexual function

A total of 30 studies focused on sexuality and sexual function. The most used measure of sexual function was the Arizona Sexual Experiences Scale (ASEX) which was used in eight of the included studies [47, 58–61, 64, 81, 113]. Other measures of sexual function included The Changes in Sexual Functioning Questionnaire (CSFQ) [27, 48, 65] and Female Sexual Functioning Index (FSFI) [55, 85, 96], the Sexual Behaviour Questionnaire (SBQ) [25], the Golombok Rust Inventory of Sexual Satisfaction (GRISS) [56, 64], the Psychotropic-Related Sexual Dysfunction Questionnaire (PRSexDQ) [98, 101], the Sexual Risk Behaviour Assessment Schedule (SERBAS) [106], the Schizophrenia Quality of Life Questionnaire (SQoL18) questionnaire, specifically the “sentimental life score” component (measuring satisfaction with love life) [29]. One study used a bespoke scale and found more women with SMI than men with SMI reported a problem in their sexual life [44].

Of the 18 studies that used standardised measures of sexual dysfunction and compared women with SMI to men with SMI, ten studies reported higher levels of sexual dysfunction in women with SMI than men with SMI [25, 47, 59–61, 81, 85, 98, 106, 107]. Five studies reported no significant difference between women with SMI and men with SMI [27, 48, 56, 58, 63]. Only one study reported higher levels of sexual dysfunction in men with SMI [101] and another reported lower levels of sexual satisfaction in men with SMI [29]. Limited evidence from one study indicated that self-reported levels of sexual dysfunction varied by gender when type of medication was taken into consideration [64].

Of those studies that did not use a male comparator, one study comparing groups of women with SMI taking risperidone or olanzapine medication for SMI found both groups to have high levels of sexual dysfunction [65], another study reported women with SMI to have higher levels of sexual dysfunction compared to women without SMI [55]. A third study reported 70% of women with SMI had high levels of sexual dysfunction [96] although no comparator was used.

Eight qualitative studies relating to sexuality were identified [40, 67, 78, 90, 92, 103, 105, 107]. Stigmatisation was consistently reported by women with SMI and that this was at times perpetuated by healthcare providers. Examples given were perceptions that women with SMI lacked sexual needs and the promotion of lifelong abstinence. Studies including healthcare providers reported that staff caring for women with SMI agreed they have the same sexual desires as women without SMI. Both healthcare providers and people with SMI felt that sexuality could be affected by mental illness, with both increases and decreases in sexual activity described. Women with SMI described positive experiences of sexuality to be beneficial to their mental health but also negative sexual experiences to be a result of their illness. There was limited research indicating women with SMI were keen to learn more about the interaction between SMI and their sexual fulfilment. No studies reported sexuality or sexual function needs and experiences of lesbian, gay, bisexual, transgender, intersex, queer, asexual or other sexually or gender diverse LGBTQIA+ women with SMI.

Contraception and family planning

Twenty-five studies were based around contraception and family planning. The majority of these (n = 13) focused on contraception use [2, 28, 40, 43, 51, 53, 71, 74, 75, 90, 94, 105, 106]. Four studies, which did not include a comparison, reported large proportions of women with SMI had never used contraception, had not used contraception during their most recent sexual experiences and had no intention of using contraception during sexual intercourse [71, 87, 93, 109].

Of those studies that included a comparator, most studies used a comparator group of women with SMI rather than men with SMI. Contraception use compared to women without SMI was mixed with two studies stating no difference in contraceptive use [74, 75], two studies stating women with SMI were more likely to use contraception [28, 46] (in particular ‘traditional’ methods such as withdrawal and calendar methods) and one stating that women with SMI were less likely to use contraception [89]. There were similar results for studies comparing women with SMI to men with SMI; two studies found contraception use to be lower in women [2, 51] whilst one study found it to be lower in men [41].

Four qualitative studies were also identified on the topic of contraception [94, 103, 105, 106]. Healthcare professionals perceived that women with SMI were commonly engaged in unprotected sex and therefore prescribed reversible contraceptive methods for them. Women with SMI felt they may be perceived negatively by partners for using condoms. When asked about knowledge of different types of contraception, coitus interruptus, condoms and intrauterine devices (IUDs) were generally the most cited.

A small number of studies looked at informed choices and knowledge when making contraception decisions. One study assessed whether there were differences in participants being informed about contraception based on gender and diagnosis of either schizophrenia or bipolar disorder [41]. A further study looked at the proportion of women with SMI referred to a perinatal psychiatric clinic and reported that health care providers had discussed contraception issues with them [43]. A qualitative study in Uganda looked at the autonomy of women with SMI in making contraception choices and noted that this was often absent, with consent taken from family members to force a woman to use contraception if she refused [103].

Seven studies focused on family planning advice from professionals [33, 41, 82, 87, 93, 110, 104]. The focus of included studies was varied. Most studies looked at the quantity of provision or receipt of family planning advice from the perspective of either healthcare professionals or patients. None of the included studies assessed the quality of family planning advice provided. One study included a comparison and found more women with SMI received family planning advice compared to men with SMI [41]. In one study from Turkey, 73.7% of nurses stated that patients should receive family planning education however only 23.5% stated giving family planning advice to patients [33]. In another, 21.9% of people with SMI stated they received family planning advice from professionals [87]. Two further studies concluded there was a lack of family planning advice being provided to patients from healthcare professionals [93, 104]. In a qualitative study, caregivers and postpartum women with SMI felt that family planning education should be added to an existing psychoeducation programme [82]. Two studies reported that women with SMI felt that they would prefer to receive information about family planning from mental healthcare providers [104, 110] as they are considered knowledgeable about the impact of their health conditions on family planning decision making.

Two studies (from India and Turkey) provided evidence of healthcare professionals making family planning decisions for patients [24, 33], with forced abortion being considered necessary by over 39% of health professionals working with people who have schizophrenia in one study [33]. Use of sterilisation and abortion for failed contraception was reported in a qualitative study carried out in South Africa, which looked at the perceptions and experiences of mental health care staff [40]. Another study found that family planning considerations affected desire to take medication and were influenced by medication, with some women with SMI not wanting to take sodium valproate due to teratogenicity risk [95].

Sexual violence

Six studies [2, 22, 42, 61, 68, 73] compared experiences of sexual violence in women with SMI to men with SMI or women without SMI. Most studies reported lifetime experience of sexual violence. Two studies reported experience of sexual violence from a partner or non-partner[68, 87]. All but one study found women with SMI to be more likely to have experienced sexual violence. An Egyptian study reported men to be more likely to report sexual violence [22]. Six studies reported the prevalence of sexual violence in populations of women with SMI, and this ranged from 10%-100% [20, 34, 35, 87, 102, 111].

There were nine qualitative studies [19, 38, 40, 50, 54, 67, 91, 103, 110] focusing on the theme of sexual violence. Two thirds of included studies utilised non-patient participants either wholly or in combination with people who have lived experience. Only three of the nine qualitative studies included exclusively people with lived experience of SMI [67, 91, 110].

All nine studies reported the vulnerability of women with SMI to sexual violence. This included women being raped by male patients, whilst inpatients in psychiatric facilities. In addition to reports of sexual violence, participants reported difficulties accessing medical care after rape for homeless women; rape resulting in unintended pregnancies and delayed discharge from medical care facilities in women who experienced sexual violence at home.

Fertility, pregnancy and postpartum

Five studies were based around fertility [26, 51, 74, 100, 108]. These were varied in terms of the outcome being studied. Three studies looked at conception and childlessness and found similar rates between women with SMI and men with SMI or women without SMI [26, 51, 74]. A study carried out in Brazil found lower fertility rates and fecundity in women with SMI compared to women in the general population [100]. A qualitative study described difficulties in the fertility process for women with SMI, alongside parenting being described as a rewarding process that aids recovery [108].

Fourteen studies focused on pregnancy [23, 24, 41, 43, 46, 66, 87, 88, 91, 92, 95, 108–110]. Notably, the only two case studies included in this review were based around decision-making during pregnancy in the context of medication and abortion [23, 43]. Seven studies mentioned the impact of medication on pregnancy and/or the effect of withholding medication during pregnancy. One qualitative study mentioned a higher risk perception related to impacts of medication on a foetus compared to risk of unmedicated mental illness during pregnancy [110]. One Turkish study included a comparator and found that women with SMI were less likely to have experienced antenatal care during pregnancy, had a greater frequency of Caesarean birth and were subject to more birth related trauma [88]. Two studies reported that high numbers of women with SMI experience unplanned pregnancy (51.4% of women in one study [87] and 87.8% of women who had experienced pregnancy in another [109]. A study that compared women with schizophrenia to bipolar disorder [41] found those with bipolar disorder to have a lower age of first pregnancy (22 years for bipolar and 32 years for schizophrenia). Two studies reported pregnancy being discouraged for women with SMI, with 69% of Ethiopian community members felt that women with SMI should not give birth in one study [110], whilst in a study from India, there were reports of women being forced to abort or coerced not to have children [66]. A comment in another study from India reported instances of pregnancy loss due to domestic violence [91]. One study found no significant difference in the number of pregnancies or number of lost or interrupted pregnancies between women with bipolar disorder compared to controls but that a higher proportion of women with bipolar disorder had unplanned pregnancies [46].

Three studies were based in the postpartum period [82, 88, 108]. One study compared women with SMI to women without SMI and found that women with SMI were less likely to breastfeed, more likely to express concerns about infant care, experience childbirth related trauma and rely on others to care for their babies [88]. Two studies mentioned breastfeeding: a Ugandan study discussed instances of women with SMI being told not to breastfeed as mental illness could be passed onto an infant through breastmilk [82] and some women with schizophrenia in a Chinese study stated that breastfeeding could lead to a loss of routine for women with SMI that could lead to an exacerbation of illness [108].

Comparison to previous literature

Hughes et al. [115] completed a systematic review and meta-analysis of HIV, hepatitis B and hepatitis C prevalence in people with SMI globally in 2016 [70]. The review found a higher prevalence of these viral infections in people with SMI in countries with a low prevalence in the general population and similar prevalence in countries with a high prevalence in the general population. The Hughes et al. [115] systematic review included many studies that were also included in the current review. The authors identified high levels of bias present in studies focusing on HIV prevalence in LMIC. They also commented on a general theme of higher HIV prevalence in women compared to men in both high and low prevalence countries, which is in keeping with findings from this scoping review. A 2016 systematic review [116] found that 10% of women with SMI experienced sexual violence, and that this was significantly higher than men with SMI and the general population. That review mainly included studies from HIC but also had similar findings to this review.

What this review adds

Throughout the results, the burden of sexual and reproductive health needs for women with SMI in LMIC is clearly shown. Rates of STIs, HIV and risky sexual behaviour were high. A small number of studies on educational programmes for HIV showed positive outcomes. The need for more education on STIs and unwanted pregnancy prevention was identified. Studies based around pregnancy were often focused on the use of medication in pregnancy, but a small amount of research on pregnancy did state advantages (such as providing a purpose in life) and disadvantages (exacerbation of mental illness) of pregnancy for women with SMI.

The link between sexual dysfunction and SMI was found to be complex with multiple factors such as psychotropic medication, self-esteem and positive and negative symptoms of SMI to be involved. The overwhelming majority of studies about sexual violence found that women with SMI were at risk of sexual violence, including whilst admitted onto mixed inpatient wards, in the home environment or when homeless.

Limitations

Limitations were present in this scoping review. Many LMIC were not covered in this review, and with the wide variation in contexts, cultures and geography, it is difficult to generalise these findings to all LMIC. Most studies were of a cross-sectional design, therefore causality cannot be implied for a majority of papers. As this is a scoping review, bias was not assessed. There may be high levels of bias in some studies, which could skew the conclusions of this review. Many studies had small sample sizes and are considered alongside larger studies, although an effort was made to note studies that were particularly small. Grey literature was excluded in this review due to the large amount of peer reviewed literature available. Capturing unpublished literature may ensure inclusion of more LMICs.

Strengths

The strengths of this scoping review include a rigorous methodology with an iterative approach that resulted in early decisions to refine the inclusion criteria. Initially, only studies that provided disaggregated data by gender and diagnosis were included. It was felt that some useful data was being excluded based on these criteria and therefore, the inclusion criteria was refined to include aggregated data where at least 50% of participants had SMI or 80% were women. A broad definition of sexual and reproductive health needs ensured that multiple areas of the review topic were included and ensured a broad overview of this multifaced area. Double coding was utilised throughout the review process to increase robustness of findings.

Implications for a subsequent review

This scoping review systematically identified literature on the sexual and reproductive health needs of women in LMIC in relation to study type, focus of study, population, setting and outcome. The nature of scoping review methodology is limited to descriptive or simple content analysis. To move to more interpretative thematic analysis, subsequent synthesis of evidence from one or more of the focus areas from this review is planned. To progress this, a mixed methods approach to evidence synthesis would enable interrogation and interpretation of data gathered from both qualitative and quantitative studies and facilitate better understanding of this complex area of health for women living with SMI in LMIC. This would enable the literature identified to be utilised to inform decision-making in policy and practice.

This review highlighted that only a small proportion of LMIC countries are represented and a future review will likely need to consider how applicable findings are for the culture and diversity of other LMIC countries and consider the use of grey literature.

Implications for research and practice

There is a need for future research to better represent the breadth of LMIC countries and for high quality research to be carried out. Other key gaps in the literature were identified by this review. This included a need for greater focus on inclusion of people with lived experience as part of studies looking at sexual violence, as well as ensuring that ethical and consensual aspects are fully considered and reported by researchers. These are both areas underscored by the recent work of ‘The Lancet Commission on ending Stigma and Discrimination in Mental Health’, which outlined the increased effectiveness of interventions developed in co-production with people who have lived experience [117]. The Lancet Psychiatry Journal has also recently specified that research submitted for publication will be required to provide information on how people with lived experience have shaped research priorities, the design of the study and interpretation and write up of findings with authors expected to acknowledge this as a limitation of their work where it has not been carried out [118].

Marginalisation and social exclusion not only exacerbate access to mental and physical health care but are often the catalyst that leads to basic human rights of people with mental health conditions being contravened [117]. Future research would benefit from focussing on the effectiveness of interventions that integrate physical and mental health, most notably those aimed at reducing stigma and discrimination around mental health and sexual health needs, as well as culturally appropriate adaptations [117].