IPV exposure in HIV+ pregnant women poses a substantial health burden, especially in SSA. This is the first systematic review examining the effects of IPV on this group. There was a wide variety of reported IPV prevalence in the included studies of this review, with rates ranging from 19.6% [61, 62] to 56.3% . This is likely due to the various challenges involved in determining IPV prevalence globally and more specifically in SSA [7, 20]. These are discussed in more detail in the limitations section.
The review’s findings demonstrate five key issues. Firstly, the relationship between IPV and the mental health outcomes of HIV+ pregnant women. Secondly, emotional/psychological violence was reported as the most prevalent form of IPV. Third, unplanned pregnancy was highlighted as a crucial factor in increased depressive symptoms in HIV+ pregnant women experiencing IPV. Fourth, stigma was shown to have a tremendous role to play within this relationship and finally, IPV increases HIV risk behaviours in this group, such as reducing condom use and reducing adherence to PMTCT medication. These findings align with the model from Hatcher et al.  that illustrates the interaction between IPV and HIV seropositivity.
Overall, the review describes a high prevalence, and thus a substantial burden, of IPV in HIV+ pregnant women in SSA, especially emotional/psychological abuse. It is crucial to note that it is usually more difficult to measure this form of IPV, since emotional/psychological abuse can be difficult to conceptualise [22, 27]. The nature of this type of abuse, however, could go some way in explaining its high prevalence of IPV in HIV+ pregnant women in SSA. Potentially, perpetrators and others may consider these behaviours more acceptable  and consequently, they may not view them as abuse. Further research in this area is necessary to explore emotional and psychological abuse in pregnant women in more depth.
Most studies focussed on the mental health of HIV+ pregnant women exposed to IPV, which is important as the search strategy did not target mental health explicitly. They report high levels of depressive symptoms, psychological distress, and suicidal ideation because of IPV exposure. Furthermore, HIV+ pregnant women exposed to IPV who are younger or who have unplanned pregnancies were at higher risk of increased depressive symptoms, compared to women who were not exposed to IPV. Ashaba et al.  corroborate this in their qualitative study. It demonstrated that increased verbal and emotional abuse was associated with unplanned pregnancy in women living with HIV, particularly when informing their partners about the unplanned pregnancy. This has implications for determining which groups to target within healthcare settings, and the role that sexual and reproductive health (SRH) services can play in IPV interventions. Additionally, Hatcher et al.  discuss how maternal mental health associated with IPV is often overlooked during maternal contact with health professionals, which could be an important point of entry for this group.
The double burden of adversity
HIV-related stigma was highlighted as a crucial contributor to the IPV experiences of pregnant women in our study. This review revealed that stigma was associated with all forms of IPV  and depressive symptoms  within this population. The findings imply that it is not just the physical experience of HIV, IPV or both that causes difficulties, but also, the associated feelings of shame, which can be both internalised and externally imposed.
While external stigma is important, and worth highlighting, the effects of internalised stigma need emphasis as these can be particularly insidious. The negative impact of both IPV and HIV on a person’s self-identity needs to be specifically focused on, as it relates to their individual coping strategies. This can result in harmful consequences, for instance limiting help-seeking behaviour or treatment adherence [35,36,37]. This can be aggravated when these exposures co-occur, as shown by two of the studies included in this review [52, 56].
Labels such as ‘victim’ or ‘HIV-infected’ can affect how a person sees themselves and how society perceives them [33, 38, 39]. Even if IPV or HIV are not disclosed, internalised stigma can still impact a person’s ability to seek help, therefore jeopardising opportunities for diagnosis and management [38, 39], or worsening their mental health [38, 65, 66]. The presence of both HIV and IPV stigma in this review have been described to leave expectant and current mothers feeling isolated [67, 68], creating a barrier to both HIV-related and pregnancy-related care . A crucial step in IPV management, after these services have been made available, is recognising the need for them. Unfortunately, internalised stigma can hamper this tremendously. Moreover, it is a difficult aspect of this relationship to target as it requires changing both social norms and personal beliefs.
As discussed, HIV- and IPV-related stigma can influence medication adherence. Others have demonstrated a complex relationship between IPV, mental health and HIV medication adherence [40, 67]. Hatcher et al.  highlight that poor adherence can be used as a means of self-harm, as women experiencing both of these conditions during pregnancy may feel overwhelmed and therefore have a desire to end their lives. Additionally, ever present thoughts and experiences of violence can result in forgetfulness – either forgetting to take their medication or forgetting to pick them up. This effect can be aggravated by the memory impairment and concentration problems characteristic of PTSD and depression. Both these mental health conditions are strongly associated with IPV and trauma [1, 2, 69], despite PTSD not being one of the reported outcomes in the studies included in this review. For some women, however, motherhood can provide resilience, allowing them to focus on their child’s wellbeing . This may allow women to continue taking their medication, however, it does not necessarily reduce their exposure to IPV nor its devastating effects.
IPV, HIV and fear
This review provided a better understanding of the relationship between HIV and IPV with fear acting as a potential mediating factor. The findings suggest that interactions between partner disclosure, PMTCT uptake, and relationship control are key in understanding the association between HIV and IPV during pregnancy [50, 57]. This is corroborated by previous qualitative research . IPV and fear of IPV can result in women not feeling confident enough to request condom use , placing them at increased risk of HIV and other STIs [40, 68, 70]. Furthermore, Hampanda  demonstrates how IPV can reduce PMTCT medication adherence. A systematic review by Hatcher et al. , supports this as they report that the presence of IPV reduces adherence to antiretroviral therapy for HIV+ women.
There are various potential reasons for reduced medication adherence. Firstly, IPV can contribute to reduced adherence through direct partner control over access to HIV treatment [64, 67]. Second, fear of HIV disclosure can result in women avoiding taking their medication to prevent their partners finding out. Many studies have reported that this fear stems from a worry that their partners will consider them unfaithful, resulting in increased emotional or physical violence [40, 64, 70, 72,73,74]. While a study in this review  reported no association between IPV during pregnancy and HIV status disclosure, other studies have emphasised the importance of this relationship [40, 64, 72, 75, 76]. The potential impact of internalised stigma should also be highlighted here. This is because it can result in low self-esteem that means women do not feel capable of asserting safe sex practices or medication adherence. All in all, this review has demonstrated the variety of ways IPV complicates the experiences of HIV+ pregnant women.
Strengths and limitations
This review has key strengths. First, it was a systematic review, using PRISMA guidelines, to study effects of IPV on HIV+ pregnant women in SSA. Second, the search was conducted in multiple databases resulting in a more comprehensive review. Additionally, the inclusion of African Journals Online ensured that studies from SSA were not unintentionally excluded due to publication bias. Lastly, the exclusion of studies not reporting p values or data collection methods and the use of standardized quality assessment  of the included studies ensured avoidance of low-quality studies, which strengthen the reliance on the study findings. However, the explicit focus on studies published in English, the last 10 years and, within the medical literature may have excluded relevant literature.
Despite the valuable information garnered from the 14 included studies, the studies have several limitations. Firstly, all studies relied heavily on self-reporting for their key outcomes which is liable to social desirability bias, recall bias and under-reporting, especially considering the stigma associated with both conditions. The studies that used ACASI went some way in overcoming this as its use can contribute to reduced social desirability bias, and the audio function makes it more appropriate in low-literacy populations, but only if used correctly . Secondly, depression symptoms were measured using the Edinburgh Postnatal Depression Scale, without diagnostic review. Therefore, it does not provide a definitive diagnosis of depression as defined by either the Diagnostic and Statistical Manual of Mental Disorders or the International Classification of Disease-10. This may have led to some discrepancies between studies.
As mentioned previously, prevalence of IPV varied greatly amongst the included studies. This is likely to be because what constitutes as abuse is not standardised between and within countries. In addition, it is heavily reliant on self-reporting  and women recognising what they experience as IPV. Secondly, the methodology of data collection in studies differ, including questionnaires or face to face interviews. These have their own inherent biases and can also result in different responses . Furthermore, psychological/emotional forms of abuse and controlling behaviours may not be recognised as IPV, which may contribute to underreporting [7, 20]. Challenges in studying IPV and HIV are therefore likely to have impacted the studies included in this systematic review.
In addition to this, many of the studies used non-proportional sampling, meaning that some individuals had no chance of being selected. Subsequently, this inhibited the ability to estimate the effects of sample error resulting in a non-representative sample and potentially non-generalisable results. Importantly, most studies (n = 13) recruited clinical samples and only one study offered population estimates on IPV. As a result, prevalence estimates cannot be generalised. Moreover, only 3 studies included HIV-negative women in their sample, so it is difficult to draw conclusions based on this evidence. There is a need for further comparative studies to better understand this relationship.
The review identified gaps in the existing literature. While physical violence and psychological/emotional violence were often reported in the included studies, there was limited information on sexual violence and male controlling behaviours. Furthermore, the studies that reported on mental health outcomes only discussed depressive symptoms and suicidal ideation. However, anxiety, PTSD and substance abuse are crucial mental health conditions that are likely to affect this group [1, 2, 69], but this was not captured in this review. Two studies reported on substance abuse  and alcohol use , but they did not explore how IPV affects these outcomes in HIV+ pregnant women. There is therefore a need to broaden the scope of the mental health outcomes while researching this group.
Finally, there are the limitations associated with the cross-sectional study design used in a large proportion of the studies. This hinders interpretation of causal links between IPV and HIV among pregnant women. The few longitudinal studies identified in the review opted for short follow-up periods determined by pregnancy status (pre- and post-partum), independent of changes to HIV status or IPV exposure. Finally, the geographical distribution of studies was limited, with all studies conducted in English speaking countries in SSA, and most of the studies being conducted in South Africa (n = 10). Nevertheless, this review provides a starting point for enhanced research into the effects of IPV exposure on HIV+ pregnant women in SSA.
Implications of findings
The results of this systematic review build on previous literature on IPV and HIV [43, 78,79,80]. This review offers new insights into the severe problem of IPV amongst HIV+ pregnant women in SSA, with devastating consequences on their mental and physical health. These findings have important implications for further research, policy, and practice.
The review has identified important research gaps. The presence of IPV during pregnancy complicates maternal mental health extensively amongst HIV+ women [49, 51, 52, 54]. Hence, further research is required into better diagnosis and management of mental health issues related to these two stigmatising conditions, especially looking at anxiety, PTSD, and substance abuse. Despite a search strategy aiming to identify sexual violence and controlling behaviours by intimate partners as well, little information was found, which could be both a cause and a consequence of stigma. It is also important to highlight that women who may not seek healthcare services will likely be underrepresented in this research, as most of the studies focused on healthcare settings. Subsequently, more efforts are needed to gather data on the needs of this specific group, and, how best to provide IPV services for them.
Only one included study reported on PMTCT adherence, however, this is noteworthy considering that one of the main ways that HIV spreads in SSA is vertical transmission from mother to child; Hampanda  approximates that it accounts for 15% of the total global incidence. Therefore, further qualitative and quantitative research on the effects of IPV during pregnancy on PMTCT adherence is vital to inform policymaking and advocacy. This also further corroborates the value of adding discussions on IPV into sexual reproductive health and rights discourse.
More research is also required comparing HIV-negative and HIV+ women’s experiences of IPV, perhaps with case-control studies, as many of the studies only examined HIV+ women. Furthermore, all longitudinal studies identified in the review included women enrolled during pregnancy with short term follow-up pre- and post-partum. Forthcoming studies should follow women through an extended period of their reproductive life and/or include comparison groups. This could contribute to understanding potential causal links between pregnancy, IPV and HIV status in the longer-term.
The findings provide guidance for policy and practice. The WHO guidelines on responding to IPV  suggest the use of antenatal care for increased IPV screening and intervention opportunities. This review highlights the importance of developing and implementing guidelines that recognize and target the concerns of HIV+ pregnant women exposed to IPV, particularly by Ministries of Health in SSA countries. This also includes increasing resources for training health workers to overcome the barriers of stigma related to these two conditions, as ensuring that healthcare workers do not retraumatise their patients is essential . Due to the highly stigmatising nature of both HIV and IPV, including IPV management in already existing services and structures can help to capture women in this vulnerable group.
It could be argued that policies in the WHO AFRO region and at country level should consider IPV screening for pregnant women attending HIV clinics. The results suggest that exploration of outreach activities or targeted interventions, sensitised to stigma, should be considered. Our findings have illustrated how the isolation associated with HIV and IPV amongst pregnant women in SSA can affect their health-seeking behaviour. Hence, ensuring IPV screening and management is a part of antenatal, HIV and SRH services could be of tremendous value to these women.