Human immunodeficiency virus‐positive women's perspectives on breastfeeding with antiretrovirals: A qualitative evidence synthesis

Abstract Human immunodeficiency virus (HIV)‐positive women can breastfeed with minimal risk of mother‐to‐child transmission if taking antiretrovirals. Guidelines surrounding infant feeding for HIV‐positive women have evolved several times over the last two decades. Our review aimed to explore perspectives of breastfeeding with antiretrovirals from HIV‐positive women since the World Health Organization (2010) infant feeding and antiretroviral guidelines. HIV‐positive pregnant and postnatal women from all countries/settings were eligible. HIV‐positive women were either on an antiretroviral regimen at the time of the study, previously on an antiretroviral regimen, not initiated on a regimen yet, or enrolled in prevention of mother‐to‐child transmission (PMTCT) care. Quality assessment of all included studies were conducted. Four databases (CINAHL, EMBASE, MEDLINE and PsycINFO) were searched for studies conducted from January 2010 to October 2020. Nine papers were included in the review, of which two presented findings from the same study. Five analytical themes were developed via thematic synthesis: (1) awareness of breastfeeding with antiretrovirals, (2) turmoil of emotions, (3) coping mechanisms, (4) the intertwining of secret, stigma and support and (5) support needed. Support from family and health care professionals and coping approaches were important to overcome stigma and the emotional challenges of breastfeeding with antiretrovirals. Health care professionals should be familiar with the most updated national and local guidance surrounding infant feeding and antiretrovirals. Further research into interventions to encourage HIV‐positive women to adhere and commit to lifelong antiretroviral treatment (Option B+) for breastfeeding is required.


| INTRODUCTION
Breastfeeding offers many positive health benefits to the infant and breastfeeding woman, such as providing infants with immunity to disease, reduced risk of childhood obesity, and promoting motherinfant bonding, and reduced risk of breast and ovarian cancers in women (World Health Organization [WHO], 2018). However, there are differences in infant feeding advice for HIV-positive women between low-and high-resource settings. Currently, the WHO and United Nations Children's Fund (UNICEF) (2016) promotes HIV-positive women in low-resource settings to exclusively breastfeed for 6 months whilst on lifelong antiretrovirals, introducing appropriate solids and non-breast milk liquids from 6 months (complementary feeding) while continuing to breastfeed for up to 24 months and beyond, until a safe diet without breast milk can be provided. If the woman decides to stop breastfeeding completely, a 1-month period of introducing solids and non-breast milk liquids, with a gradual reduction of breast milk, is recommended. In high-resource countries, where relevant resources (clean water, formula milk) are readily available, formula feeding is recommended instead to eliminate the risk of HIV mother-to-child transmission (MTCT) entirely (American Academy of Pediatrics, 2013; British HIV Association (BHIVA), 2020; World Health Organization, 2010). However, this is not to convey that breastfeeding with antiretroviral regimens have not occurred in high-resource settings. For example, BHIVA (2020) recognises HIV-positive women's free will in the United Kingdom to be able to decide to breastfeed, if specific criteria are met. Researchers have advocated for high-resource settings to permit HIV-positive women to breastfeed with antiretrovirals, in acknowledgement of people's capacity to understand the implications of doing so (Gamell, 2018;Kahlert et al., 2018).
Antiretrovirals enable HIV-positive women to breastfeed with reduced risk of HIV MTCT (White et al., 2014) by lowering the risk of HIV MTCT during pregnancy and breastfeeding via suppressing the virus to below the detectable threshold (<50 copies per ml). Antiretroviral therapy (ART) involves lifelong use of three or more antiretrovirals (WHO & UNICEF, 2016). However, antiretroviral treatments do not fully eliminate the risk of HIV transmission via breast milk (Blumental et al., 2014), and it is apparent that antiretroviral regimens have different rates of HIV MTCT risk. For instance, Shapiro et al. (2014) found a 1.1% risk of HIV MTCT via exclusive breastfeeding (EBF) with maternal ART and a 1.7% risk for EBF with infant ART. Various antiretroviral regimens (Option A, B and B+see Table 1) exist for HIV-positive women to breastfeed. HIV MTCT risk may also be elicited by breast infections, such as mastitis and cracked nipples (Zadrozny et al., 2017). Mixed feeding is the introduction of solids and other liquids with breast milk before the age of 6 months (WHO & UNICEF, 2016). Before 2010, mixed feeding was linked to an increased risk of HIV MTCT (Coovadia et al., 2007). Since 2010, research has indicated no difference in HIV MTCT risk between EBF and mixed feeding (Bispo et al., 2017). However, mixed feeding for HIV-positive women is not an official recommendation and only a guiding practicing statement as it poses the risk of undermining EBF (WHO & UNICEF, 2016). With the absence of antiretrovirals, there is an HIV MTCT risk of 16% by breastfeeding (Nduati et al., 2000).
In the last two decades, guidelines surrounding infant feeding practices for HIV-positive women have been inconsistent due to conflicting research into HIV and prevention of mother-to-child transmission (PMTCT) care. HIV-positive women breastfeeding with antiretrovirals has been a somewhat controversial area for policymakers, leading to multiple WHO guidelines and revisions of their recommendations over the last two decades (Dunkley et al., 2018) (see Appendix A). The uptake of guidelines has been varied across low-resource settings due to dependence on local implementations and ongoing studies (WHO & UNICEF, 2016).
A range of barriers may deter HIV-positive women on antiretrovirals to breastfeed and adhere to the antiretroviral regimen while breastfeeding, including HIV-related stigma in community and health care settings (Flax et al., 2017), psychosocial challenges (Kaida et al., 2014), and inconsistent policies causing confusion (Nabwera et al., 2017). Research has found low continuity amongst postpartum HIV-positive women to adhere to antiretroviral protocol, particularly postnatally (Nachega et al., 2012). Given this evidence, there are recommendations for further research into interventions to increase retention care, such as continuous engagement in PMTCT care, and adherence to antiretroviral regimens (Rollins et al., 2014;WHO & UNICEF, 2016).
Given the increasing accessibility of antiretrovirals for HIV-positive women to breastfeed and extensive evolvement of infant feeding guidelines for HIV-positive women, an understanding of the perspectives of breastfeeding with antiretrovirals from HIV-positive women is important to inform health care practice by increasing awareness of specific care needs for this group of women and provide holistic support. The majority of clinical trials, studies and reviews focus on quantitative aspects, such as the efficacy of antiretrovirals, thus the need for this qualitative review.

Key messages
• HIV-positive women have a unique and challenging infant feeding experience, which requires tailored care and support.
• Health care professionals require updated knowledge of the most recent national and local infant feeding and antiretroviral guidelines for HIV-positive women and need to communicate their advice effectively and consistently to HIV-positive women.
• To increase social support for HIV-positive women to breastfeed with antiretrovirals, community education and interventions are essential in low-resource settings.
• Further research into HIV-positive women's perspectives of breastfeeding with antiretrovirals in high-resource countries is required.

| METHODS
The review followed the guidelines provided by "Enhancing transparency in reporting the synthesis of qualitative research (ENTREQ)." The review was registered on the PROSPERO international prospective register of systematic reviews, with registration number: CRD42019136548. The review question was as follows: "What are HIV-positive women's perspectives on breastfeeding with antiretrovirals?"   If an HIV-positive woman's CD4 count >350 cells/mm 3 , a short-term antiretroviral prophylaxis: Azidothymidine (AZT), also known as zidovudine (ZDV), twice daily from 14 weeks' gestational age (GA) Single dose of nevirapine (NVP) from the onset of labour with the initiation of daily AZT twice daily with lamivudine (3TC) for 1 week postnatally • Breastfeeding infant receives daily infant prophylaxis of NVP, initiated within 6-12 h from delivery, until 1 week after the end of the breastfeeding period. Infants who are not breastfeeding take NVP for 4-6 weeks. • WHO (2013) retracted its recommendation for Option A.

WHO (2010)
Option B • If an HIV-positive woman's CD4 count is ≤350 cells/mm 3 , antiretroviral therapy (ART) is initiated for life, due to the HIV-positive woman requiring treatment for her own health. At the end of the breastfeeding period, the HIV-positive woman can cease ART if CD4 count >350 cells/mm 3 , with continual monitoring for when CD4 count falls ≤350 cells/mm 3 as an indicator to re-start ART. • If an HIV-positive woman's CD4 count is >350 cells/mm 3 , a triple antiretroviral prophylaxis is initiated from 14 weeks' GA, throughout delivery until 1 week after the end of the breastfeeding period. • Infant receives daily NVP or twice daily AZT from 6 to 12 h of delivery for 4-6 weeks, regardless of infant feeding method.
WHO ( Figure 1 shows an example of search history from one database. The electronic search was restricted from 2010. The search was updated on 10th October 2020. Only studies published in English language were included. Reference lists of previous relevant reviews (Laar & Govender, 2013;Nyoni et al., 2019;Tuthill et al., 2014) and included studies were hand searched. KMCL and Y-SC performed a title-abstract screen of the studies. The full text of the relevant studies was retrieved and screened for inclusion independently by KMCL and Y-SC. Differences between screening results were resolved by discussion between KMCL and Y-SC.

| Quality assessment
Quality assessment for each included study was conducted independently by KMCL and Y-SC using an adapted Critical Appraisal Skills Programme (CASP) (2018) 10-item checklist for qualitative research (see Table 3). CASP question 10 was adapted from "How valuable was the research?" to "Is the research valuable?" to enable scoring. Each CASP question was answered via "Yes," "No" or "Can't tell." A "Yes" scored a value of 1. "No" and "Can't tell" scored 0. A maximum score of 10 could be achieved per paper.
Differences between reviewers' assessments were resolved through discussion between KMCL and Y-SC.

| Data extraction and synthesis
Data were extracted on the study setting, aim, antiretroviral regimen HIV-positive women received, infant feeding regimen in place at the time of the study, including whether free formula milk was provided to participants, the infant feeding method used by the participants, study design, data collection methods, data analysis and key findings. Data were synthesised using Thomas and Harden's (2008) three stages of thematic synthesis: (a) free lineby-line inductive coding, (b) development of descriptive themes from organising free codes, and (c) development of analytical themes. First, free line-by-line coding was carried out to form initial codes. Each finding was analysed and coded to reflect its content closely to the study's finding. Each finding could have more than one initial code applied. Afterwards, initial codes were compared with one another for similarities and differences, and grouped accordingly. Descriptive themes were formed to convey the meanings of groups of initial codes. Analytical themes were formed inductively to examine beyond the study's content. Data extraction and synthesis was conducted by KMCL, Y-SC and KYCL.
Final themes were agreed on by all authors.

| Ethical considerations
This is a qualitative evidence synthesis and does not require an ethical approval. No funding was received to assist with the preparation of this article.
HIV-positive women experienced various stages of antiretroviral treatments and breastfeeding methods in relation to national and local infant feeding protocols and recommendations at the time of the study (see Table 2). All studies involved qualitative methodologies.
Quality assessment of included papers (see Table 3) results in only one study scoring a total of 10 (Dunkley et al., 2018). Matovu et al. (2014) scored the lowest with 6. Many studies failed to address the relationship between the researchers and participants, which reduced the credibility of these studies. Some studies did not provide sufficient evidence in their methodology for a rigorous data analysis and recruitment strategy.

| Themes
Five analytical themes were generated from the included papers: (1) awareness of breastfeeding with antiretrovirals, (2)  HIV-positive women knew they were recommended to exclusively breastfeed for 6 months with no introduction of solids and non-breast milk liquids (Horwood et al., 2018;Matovu et al., 2014). Few HIV-positive women knew the importance of adherence to antiretrovirals while breastfeeding to reduce the risk of HIV MTCT (Flax et al., 2017;Horwood et al., 2018;West et al., 2019). On the other hand, a minority were unaware of updated infant feeding guidelines for HIV-positive women, such as not knowing HIV-positive women could breastfeed with antiretrovirals as it was not previously promoted (Dunkley et al., 2018;Horwood et al., 2018). For example, one woman said: "No, I didn't know I was going to breastfeed as I thought only [HIV] negative people were allowed to breastfeed, until I visited the clinic" (Horwood et al., 2018, p. 4).

Risks of breastfeeding with antiretrovirals
Many HIV-positive women understood the risk of HIV MTCT from breastfeeding with Option B+ but were unable to explain how HIV transmission occurs (Dunkley et al., 2018). Few HIV-positive women were aware of factors that could increase the risk of HIV MTCT while breastfeeding with antiretrovirals, such as breast sores (Horwood et al., 2018) and infant thrush (Horwood et al., 2018).
Although mixed feeding has been found to have no difference in HIV MTCT compared with EBF, some women were only promoted EBF for 6 months. These women thought mixed feeding before 6 months could increase HIV MTCT (Horwood et  • 32 lost to follow up from ART (Option B+).
• Various statuses of breastfeeding.
• Malawi followed WHO (2010)  in HIV-positive women in rural and urban areas of South Africa.
• (2) Triple ARV regimen during pregnancy and continue until end of breastfeeding.
(3) Initiation of lifelong triple ARV drugs regardless of CD4 count during pregnancy after known HIV status.
• Recommendation of EBF for first 6 months of life.

Benefits of breastfeeding
HIV-positive women's awareness of breastfeeding benefits influenced their decisions to continue breastfeeding as they desired their infant to acquire the benefits of breastfeeding (Flax et al., 2017;Horwood et al., 2018;Phakisi & Mathibe-Neke, 2019).
A South African woman reported, "He gets healthy; he is protected from diseases. Breast milk is healthy" (Horwood et al., 2018, p. 6).
Another HIV-positive woman from Malawi continued breastfeeding after stopping antiretrovirals despite the recommended infant feeding policy (Flax et al., 2017).

| Turmoil of emotions
The decision making and experiences of breastfeeding antiretrovirals were challenging for HIV-positive women, leading to an overwhelming weight of emotions conveyed by the women.

Striving for motherhood with influences of personal experiences and intuition
HIV-positive women wanted to breastfeed in order to fulfil their innate desire to achieve motherhood (Horwood et al., 2018). Furthermore, the decision to breastfeed or not was reinforced by pre-    , 2018, p. 7). Mixed feeding was instigated upon women from family and communities due to perceptions of breast milk alone being inadequate to meet the infant's feeding demands (Horwood et al., 2018;Phakisi & Mathibe-Neke, 2019).

Fear of HIV MTCT
Some HIV-positive women were reluctant to breastfeed due to fear of HIV MTCT from breastfeeding with antiretrovirals and mixed feeding before 6 months. Many HIV-positive women decided to formula feed to avoid breastfeeding entirely (Dunkley et al., 2018;Horwood et al., 2018;West et al., 2019).

Financial and lifestyle stress
Many HIV-positive women struggled to afford formula milk and had no choice but to breastfeed (Dunkley et al., 2018;Horwood et al., 2018;West et al., 2019). As a woman reported, "I chose to breastfeed because I am not working and sometimes you may find that I would not have money to buy formula as the father also doesn't have a good job" (West et al., 2019, p. 5). This stress persisted during the period of introducing solids and other nonbreast milk liquids while gradually stopping breastfeeding when trying to afford replacement feeding (Flax et al., 2017;Horwood et al., 2018). A Ugandan woman's partner begged her to breastfeed due to financial constraints (Dunkley et al., 2018). In contrast, some HIV-positive women decided against breastfeeding or stopped breastfeeding to work/return to work or seek employment for financial stability (Dunkley et al., 2018;Horwood et al., 2018;Phakisi & Mathibe-Neke, 2019;West et al., 2019). Dunkley et al. (2018) reported that HIV-positive women felt frustrated with the need for regular infant HIV testing if they continued to breastfeed with Option B+.
However, HIV-positive women undergoing Option B+ did not experience the severity of side effects as much (Chadambuka et al., 2018). For example, one lactating woman from Zimbabwe reported: I wanted to say that we accept it well because lifelong antiretroviral therapy keeps us healthy all the time and we are able to do our daily chores without being sick all the time since we will be taking our medication well. (Chadambuka et al., 2018, p. 5)

| Coping mechanisms
A turmoil of emotions mentioned above over their infant's risk of HIV transmission and infant feeding experience led to coping behaviours exhibited by HIV-positive women to withstand their challenging infant feeding experiences.

Emotional resilience
HIV-positive women commented on the need for emotional resilience while breastfeeding with antiretrovirals and making infant feeding decisions independently (Horwood et al., 2018). For example, "No, I haven't discussed anything because I am a strong person, I've told myself that I am going to breastfeed so there is no one that I fought with or say something" (Horwood et al., 2018, p. 9). Emotional resilience was often reported by HIV-positive women in other instances, such as when stopping breastfeeding to deal with the emotionally challenging experience from being unable to provide breast milk (Horwood et al., 2018), adhering to antiretroviral medications while breastfeeding (Chadambuka et al., 2018) and disregarding family members pressuring them to mixed feed (Horwood et al., 2018).

Turning to religious beliefs
A minority of HIV-positive women resorted to religion as a method to cope with their emotional struggles and fear over the risk HIV MCT while breastfeeding: "I used to pray to God to be the one to protect him … So that he does not get infected with HIV … because of breastfeeding...." (Dunkley et al., 2018, p. 7

| The intertwining of secret, stigma and support
Many HIV-positive women concealed their infant feeding method to conceal their HIV-positive status due to a strong social stigma prevalent in communities towards infant feeding and HIV. Having a strong support network, including family and health care professionals, provided HIV-positive women a better infant feeding experience.

Interlinks between (non)disclosure of HIV status and social support
The support received by HIV-positive women varied and was dependent on whether they disclosed their HIV status to their family. Sociocultural expectations from peers and family members to follow community infant feeding practices to avoid disclosing their HIV status were reported by HIV-positive women who were pressured to breastfeed (Dunkley et al., 2018;West et al., 2019) or continued breastfeeding longer than they wanted to (Dunkley et al., 2018;Horwood et al., 2018)  South African woman to deceive her family by giving "insufficient breast milk" as a reason to commence formula feeding (West et al., 2019). HIV-positive women on Option B+ were reported to experience less stigma because they could breastfeed beyond 6 months and assimilate to cultural breastfeeding norms like noninfected individuals (Chadambuka et al., 2018;Flax et al., 2017;Katirayi et al., 2016), as this quote illustrated: "… nowadays no one can tell the difference between an HIV-positive and an HIV-negative person. We all look the same..." (Chadambuka et al., 2018, p. 5).
In contrast, some HIV-positive women who disclosed their HIV status received support from their partners and family to breastfeed and follow their local infant feeding recommendations at the time, such as avoiding mixed feeding (Dunkley et al., 2018;Horwood et al., 2018). One South African HIV-positive woman openly shared her HIV-status to family and strangers and explained how she was exclusively breastfeeding with antiretrovirals (Phakisi & Mathibe-Neke, 2019). "I disclosed my status to everyone. I talk a lot about it, even to strangers. Some people used to comment about my breastfeeding while I was HIV-positive. I would tell them that my child was safe because I was taking antiretrovirals and not mixed feeding" (Phakisi & Mathibe-Neke, 2019, p. 30). Another South African woman believed her choice to exclusively breastfeed has nothing to do with her family (Horwood et al., 2018).

The value of health care professionals
Some HIV-positive women relied on health care professionals to recommend how they should feed their infants or wanted to make decisions with health care professionals (Croffut et al., 2018;Horwood et al., 2018). For instance, HIV-positive women actively sought infant feeding advice from health care professionals when they were unsure of when to stop breastfeeding: "I'm waiting for Monday to go for 6 months [clinic] visit, then they will tell me what to do" (Horwood et al., 2018, p. 4). HIV-positive women gained awareness of current infant feeding protocol for HIV-positive women and the importance of adherence to antiretroviral regimen from health care professionals at the time of the study (Flax et al., 2017;Horwood et al., 2018;West et al., 2019). Some HIV-positive women only became aware they could breastfeed via antiretrovirals after consulting with health care professionals (Horwood et al., 2018). HIV-positive women also received general infant feeding advice, such as health and bonding benefits of breastfeeding (Horwood et al., 2018), expressing and storing breast milk (West et al., 2019) and avoidance of mixed feeding as per local recommendations (Horwood et al., 2018;Phakisi & Mathibe-Neke, 2019). Some HIV-positive women were motivated by health care professionals to adhere to antiretrovirals to reduce the risk of HIV MTCT while breastfeeding (Horwood et al., 2018).

| Support needed
To neutralise HIV-positive women's challenging experience of breastfeeding with antiretrovirals, two areas of support needs were described by the women.

Role models
HIV-positive women wanted role models of other HIV-positive women breastfeeding with antiretrovirals to share their experiences and support, to encourage uptake of antiretrovirals, particularly for HIV-positive women who opted for formula feeding (Chadambuka et al., 2018;West et al., 2019).
Yes, it encourages, because if you hear from their experiences, you learn from what they would have gone through, how it helped them, where they came from and where they are going. (Chadambuka et al., 2018, p. 4) Enhancing education provision to women and their support network The need for further education and consolidation of education offered by health care professionals was highlighted by HIV-positive women. Confusion amongst HIV-positive women regarding infant feeding advice they received was frequently reported, such as the conflict between breast is best and the HIV MTCT risk associated with breastfeeding with Option B+ (Dunkley et al., 2018). Updated infant feeding recommendations in line with current national and local guidelines are required by HIV-positive women as there were inconsistencies surrounding advice on breastfeeding cessation and introducing complementary foods from 6 months (Dunkley et al., 2018;Horwood et al., 2018;Phakisi & Mathibe-Neke, 2019;West et al., 2019), and a few HIV-positive women received outdated advice (Horwood et al., 2018;Phakisi & Mathibe-Neke, 2019 (Katirayi et al., 2016, p. 5).
Women wanted information on formula or replacement feeding, not just breastfeeding advice alone (Horwood et al., 2018). Community education to inform antiretroviral use for breastfeeding so partners and families can facilitate their infant feeding experience was suggested by one HIV-positive woman in order to reduce societal stigma (Chadambuka et al., 2018). were collected; however, some women had been misinformed by health care professionals. Misinformation from health care professionals could reflect rapidly changing guidance on infant feeding advice (Nkwo, 2012). This is reflected by Chinkonde et al. (2010), whose in-depth interviews with health care professionals on PMTCT care, found they struggled to advise breastfeeding with antiretrovirals and following national and local guideline changes to infant feeding in women to administer infant antiretrovirals were a deterrent for women to adhere to optimal breastfeeding protocol with antiretroviral regimens. This correlates with previous quantitative studies finding reduced antiretroviral adherence while breastfeeding compared to pregnancy (Haas et al., 2016;Nachega et al., 2012).

| DISCUSSION
This aligns with the need for interventions to adequately prepare women for Option B+, a lifelong commitment, as addressed by WHO and UNICEF (2016).
Managing to cope with breastfeeding with antiretrovirals

| Strengths and limitations
We conducted a robust search to identify all relevant existing literature, including all studies from any country and health care setting. All included studies were critically appraised by two researchers to improve rigour. A limiting factor is the ambiguity in deciphering whether studies explored breastfeeding in the context of antiretrovirals, as studies that did not mention antiretroviral use were not included in our review. All included studies were conducted in Africa, where HIV prevalence is high and limited resources for replacement

CONFLICTS OF INTEREST
The authors declare that they have no conflicts of interest.

DATA AVAILABILITY STATEMENT
As this is a qualitative evidence synthesis, the original data are available from published articles included in the review. The references of the included studies are in the reference list.