“I am scared, I do not want to lie”: Exploring the impacts of COVID-19 on engagement in care, perceived health, relationship dynamics, and parenting among postpartum women with HIV in South Africa

Background COVID-19 and efforts to manage widespread infection may compromise HIV care engagement. The COVID-19-related factors linked to reduced HIV engagement have not been assessed among postpartum women with HIV, who are at heightened risk of attrition under non-pandemic circumstances. To mitigate the effects of the pandemic on care engagement and to prepare for future public health crises, it is critical to understand how COVID-19 has impacted (1) engagement in care and (2) factors that may act as barriers to care engagement. Methods A quantitative assessment of COVID-19-related experiences was added to a longitudinal cohort study assessing predictors of postpartum attrition from HIV care among women in South Africa. Participants (N = 266) completed the assessment at 6, 12, 18, or 24 months postpartum between June and November of 2020. Those who endorsed one or more challenge related to engagement in care (making or keeping HIV care appointments, procuring HIV medications, procuring contraception, and/or accessing immunization services for infants; n = 55) were invited to complete a brief qualitative interview, which explored the specific factors driving these challenges, as well as other impacts of COVID-19 on care engagement. Within this subset, 53 participants completed an interview; qualitative data were analyzed via rapid analysis. Results Participants described key challenges that reduced their engagement in HIV care and identified four other domains of COVID-19-related impacts: physical health, mental health, relationship with a partner or with the father of the baby, and motherhood/caring for the new baby. Within these domains, specific themes and subthemes emerged, with some positive impacts of COVID-19 also reported (e.g., increased quality time, improved communication with partner, HIV disclosure). Coping strategies for COVID-19-related challenges (e.g., acceptance, spirituality, distraction) were also discussed. Conclusions About one in five participants reported challenges accessing HIV care, medications, or services, and they faced complex, multilayered barriers to remaining engaged. Physical health, mental health, relationships with partners, and ability to care for their infant were also affected. Given the dynamic nature of the pandemic and general uncertainty about its course, ongoing assessment of pandemic-related challenges among postpartum women is needed to avoid HIV care disruptions and to support wellbeing.


Introduction
As of January 2023, over 4 million COVID-19 infections and over 100,000 COVID-19-related deaths have been reported in South Africa (SA), gures which likely underestimate the spread of the virus and associated mortality. 1 Current estimates indicate that just over 32% of the SA population has completed the initial vaccination protocol, leaving the majority of individuals in SA unprotected from severe disease. 1 Of relevance to public health in this setting, speci c populations are at higher risk for severe morbidity from COVID-19, including individuals with HIV 2,3 and adults with compromised immune systems. 4 Increasing the complexity of COVID-19 risk management, SA is home to the largest HIV epidemic in the world, with 7.8 million adults and children living with HIV. 5 HIV prevalence rates are higher among women aged 45 and younger (24.7%) compared to men (13.5%) 5 , indicating that women of reproductive age bear the greatest HIV burden and are therefore at high risk for detrimental COVID-19related outcomes.
In the context of COVID-19 and repeated periods of lockdown implemented by the SA government, there was and continues to be a strong possibility that people with HIV (PWH) might lose access to their HIV care (even for a temporary period), with particularly negative implications for postpartum women. To optimize maternal health and decrease perinatal transmission, engagement in HIV care is critical during the postpartum period. In what is commonly referred to as the prevention of mother-to-child transmission (PMTCT) cascade, women must complete a series of steps (e.g., administer antiretroviral therapy (ART) to the infant, ensure that infants are tested for HIV, adhere to breastfeeding recommendations of six to 24 months) 6 to reduce the likelihood of HIV transmission during pregnancy, labor, and breastfeeding. With PMTCT interventions, the average rate of perinatal transmission is approximately 3.5%; 7 to achieve this rate, women must effectively navigate adherence barriers to avoid falling off the treatment cascade "cliff." 8 Maintaining engagement in HIV care can be di cult under the best of circumstances. Some of the challenges to remaining in care that are unrelated to COVID-19 include stigma 9 (i.e., from providers as well as internalized stigma and shame associated with HIV), structural barriers 10,11 (e.g., the need to transfer to a new clinic after delivery), and decreased motivation to remain adherent to ART after delivering an HIV-negative infant. 12 In addition to remaining engaged in HIV/PMTCT care, postpartum women with HIV (WWH) also need to manage infant care appointments, including immunization visits, and adjust to the stressors of new parenthood and/or the complexities of caring for other children while also meeting the needs of a newborn. Then, during multiple COVID-19-related national lockdowns, postpartum WWH had to balance the need to visit clinics to obtain care for themselves and their infants, facing all of the challenges described above, with additional risk of (a) the spread of a highly infectious, possibly fatal respiratory disease, and (b) limited access to usual medical care if lockdown regulations were not followed. 13 In addition to these unique engagement in care challenges, the mental health and relational stressors already faced by perinatal WWH in SA may have been exacerbated by the COVID-19 pandemic. In sub-Saharan countries with high HIV prevalence rates, perinatal WWH have elevated rates of depression 14 (with rates approaching 50% in SA), 15 reduced social support, 16 and fears about the physical wellbeing of their children [17][18][19] given the possibility of HIV transmission. Moreover, both pregnant and nonpregnant WWH experience high rates of intimate partner violence (IPV) and sexual trauma, contributing to rates of depression and posttraumatic stress disorder. 20 Such experiences may lead to the development of avoidance-based coping strategies, 21,22 which, together with COVID-19 restrictions, may have further compromised engagement in care and limited opportunities to seek social support from individuals outside of their households. In addition, among WWH experiencing IPV prior to the emergence of COVID- 19, lockdown periods may have negatively impacted their safety and the safety of their children. Notably, for perinatal WWH who had not disclosed their HIV status to their abusive partners, accessing HIV care and PMTCT services could place them at increased risk for violence. 13 Economic stressors that are common among perinatal WWH during non-pandemic conditions-unemployment or underemployment, inability to buy food for themselves and their infants, limited funds for transportation to the clinic-may have also intensi ed during the most restrictive levels of the SA government's COVID response, further compromising efforts to remain engaged in HIV care.
With the COVID-19 pandemic now in its third year, long-term management of risk and negative health outcomes will be critical in all contexts but particularly complex in resource limited settings like SA. Here, speci c subpopulations, including postpartum WWH, will likely continue to experience COVID-19-related physical and psychological health disparities for decades to come. To actively mitigate these disparities, we must explore the multifaceted impacts of the pandemic on engagement in HIV-and infant-related care, as well as on the domains that in uence care engagement, including mental health, intimate relationships or partnerships, and associated coping strategies. In this qualitative sub-study, we leveraged an existing cohort to speak directly to postpartum WWH about these intersecting challenges, with the ultimate goal of reducing care disruptions over the course of the current pandemic and during future public health crises.

Parent study procedures
The parent trial was a prospective longitudinal cohort study based in Durban, KwaZulu-Natal, SA that had two main goals: (1) to estimate the rate of attrition from HIV care during the postpartum period and (2) to identify factors associated with attrition. Data collection was initiated in February 2018 and completed in June 2022, and the primary data analyses are currently underway.
The aim and procedures of the parent study are described in detail elsewhere. 23 Brie y, the primary quantitative outcomes of the study are HIV RNA (i.e., viral suppression) and self-reported number of visits to any HIV care provider (i.e., retention in care). Four-hundred seventy-two women with HIV between the ages of 18 and 45 were recruited at 28 weeks of pregnancy or greater. Other inclusion criteria included (1) currently taking antiretroviral therapy, (2) uent in English or isiZulu, (3) access to a phone and willing to be contacted for repeated assessments, and (4) ability to provide informed consent. Women were excluded from the parent study if (1) they had an active or untreated major mental illness that would interfere with participation (e.g., untreated psychosis, bipolar disorder, active suicidality) or (2) participation would compromise their safety, complicate the interpretation of study ndings, or otherwise interfere with study objectives. The study included ve assessment timepoints: at baseline (during pregnancy) and follow-up at six, 12, 18, and 24 months post-baseline.

Sub-study procedures
A brief quantitative assessment and a qualitative interview exploring COVID-19-related challenges to engagement in care and other associated concerns were added to the assessment battery of the parent study at all follow-up time points (six, 12, 18, and 24 months). At the end of follow-up assessments conducted between 1 June and 30 November 2020 (all telephonic), study staff asked all participants if they were willing to answer some additional questions about their experiences during the ongoing COVID-19 pandemic. Participants who agreed to do so remained on the phone to complete both the quantitative assessment and the brief qualitative interview. For the present study, we analyzed only the qualitative interviews of participants who reported at least one engagement in care challenge in the quantitative assessment.
Importantly, SA follows a ve-level COVID-19 alert system based on epidemiological trends, health system capacity to respond to disease burden, and any other factors that might in uence rates of infection, hospitalization, and mortality. The system ranges from level 1 ("low COVID-19 spread with high health system readiness") to level 5 ("high COVID-19 spread with high health system readiness"). 24 From June to November 2020, alert levels ranged from level 1 to level 3 (level 3 from June 1 to August 17, level 2 from August 18 to September 20, and level 1 from September 21 through the end of November). Notably, the country went into full lockdown (level 5) from March 26 to April 30, 2020, which was then decreased to level 4 on May 1.
Data for the sub-study were collected across two COVID-19 waves, de ned by the SA government as the period from which COVID-19 weekly incidence is equal to or greater than 30 cases per 100,000 persons until the weekly incidence is equal or below 30 cases per 100,000 persons. 24

Quantitative assessment
We utilized a brief quantitative assessment that included items adapted from the N2 COVID-19 Check-in Survey Items, as well as from the Adolescent Trials Network COVID Questionnaire Draft, 25 and questions generated by our team. The assessment was used to identify women who faced challenges (1) making or keeping their HIV care appointments ("Have you had trouble making or keeping your HIV care appointments with your health care provider during the COVID-19 pandemic?"), (2) procuring their HIV medications ("Have you had trouble getting your HIV medication from your health care provider during the COVID-19 pandemic?"), (3) procuring contraception ("Have you had trouble getting your contraceptive method during the COVID-19 pandemic?", "Have you had trouble accessing condoms during the COVID-19 pandemic?"), and (4) accessing immunization services for their infants ("Have you had trouble accessing immunization services for your baby during the COVID-19 pandemic?"). Participants were provided dichotomous (yes/no) options to the four questions above; if "yes" was endorsed, participants were then prompted to select from a list of options to identify the speci c factors that contributed to these challenges (e.g., for the rst question, sample options included limited transportation, limited appointment availability, limited availability of services, lack of childcare). To characterize the sample with respect to COVID-19-related care, participants were also asked about COVID-19 diagnoses and treatment (e.g., whether they had been screened/tested for COVID-19, received a COVID-19 diagnosis, received associated treatment).
In some cases, participants completed the quantitative assessment more than once (i.e., at more than one follow-up assessment); on these occasions, their rst assessment was used to identify whether or not their qualitative data would be included in this analysis.

Qualitative assessment
The qualitative assessment explored factors associated with the challenges identi ed in the quantitative survey, as well as participants' perceptions of the impact of the pandemic on perceived health risks, mental health, intimate partnerships, and ability to care for their infants. Individual qualitative interviews were conducted telephonically using a semi-structured interview guide that followed best practice procedures established by Huberman and Miles 26 and Strauss and Corbin. 27 Some items were adapted from the Bennett and Elliott Qualitative Interview Guide, as well as from the Gwadz Qualitative Interview Guide, both of which were accessed in April 2020 via an open-sourced document developed by the Center for Drug Use and HIV Research 25 in the early stages of the COVID-19 pandemic. Items probing aspects of partnerships and experiences as a new mother were added. The broad content areas of the qualitative interview, as well as sample questions and problems, are provided in Table 1. • Have you been separated from your partner during the COVID-19 pandemic? If yes, how did the separation affect your relationship?
• Please described any positive experiences that you have had with your partner during the COVID-19 pandemic.

Experiences as a new mother
• How has the COVID-19 pandemic impacted your nancial support? Social support?
• How has the COVID-19 pandemic impacted your family's ability to help you with the baby?
• How do these experiences or feelings affect your ability to manage your health?

Coping
• What has helped you the most to cope during the COVID-19 pandemic?
• What has helped you the most with...taking your ARVs, caring for your child(ren), demand for food and other needs, your family and partner?

Analyses
Overall, 266 postpartum WWH completed the COVID-19 quantitative assessment across all follow-up time points; 34% completed the assessment at the six-month follow-up visit, 24.5% completed the questions at the 12-month visit, 32.1% at the 18-month visit, and 9.4% at the 24-month visit. Of the 266 participants who completed the COVID-19 quantitative assessment, 55 (20.7%) endorsed at least one COVID-19related engagement in care issue. Among this subset, challenges making and keeping HIV care appointments were reported by 8.2% (n = 22) of participants, 4.9% (n = 13) reported di culties accessing HIV medications, 7.5% (n = 20) noted problems securing contraception, and 5.2% (n = 14) indicated challenges accessing infant immunization services. Fifty-three of the 55 participants (96.4%) who quantitatively reported challenges agreed to complete a brief, targeted qualitative interview, and descriptive statistics were calculated only among these 53 participants.
All interviews were digitally recorded, transcribed, and translated from isiZulu to English by SA team members (NM, KS) who identify as Black South Africans and have both expertise and lived experiences with the isiZulu culture. To facilitate quick dissemination of study ndings locally among stakeholders in Durban, we used a rapid analysis approach. 28 Following this approach, four study team members (AMS, APB, GRG, RV) trained in qualitative methods independently reviewed the same subset of transcripts using a template of neutral domain names, which were selected to correspond with each interview question (e.g., mental health, motherhood, coping; see Table 1). The reviewers reduced and summarized the data using the template, then organized the summaries into themes and subthemes. All four team members met on an ongoing basis to compare the text that they had organized under each domain, their associated themes, and sample text that they had selected to highlight those themes for the subset of interviews that they all coded. During these meetings, they resolved any domain, theme, and subtheme discrepancies; then, they were each assigned a batch of interviews to code independently. After all interviewers were coded, the team met one additional time to collate the data, discuss all themes and subthemes, reach consensus on nal themes, and review extracted example quotes to present in the tables.

Characteristics of sample
The average age of the sample (n = 53) was 29.4 years (SD = 5.3), and 58.5% (n = 31) of participants were diagnosed with HIV more than ve years ago. Almost all participants identi ed as Black South African (98.1%, n = 52). Most (77.4%, n = 41) had completed up to Standard 10 at school (the equivalent of 11th or 12th grade in secondary school), 77.4% (n = 41) were unemployed, and 90.6% (n = 48) had a monthly income of less than 4000 South African Rand (ZAR) or roughly $330 USD. Approximately 17% (n = 9) of the sample reported living with a romantic partner, and 86.8% (n = 46) of participants had a long-term partner but were not married. The average number of living children across the sample was 1.7 (range: 1-4). Full sociodemographic characteristics are provided in Table 2.

Qualitative ndings
The qualitative interviews yielded numerous themes and subthemes within ve key domains: engagement in HIV care, physical health, mental health, relationship with partner/father of the baby, and motherhood and caring for the new baby. See Table 3 for sample extracts from the interviews, categorized by domain, theme, and subtheme. Missed ART doses "Ja It got nished and I stayed for two weeks without. I kept going to clinic thinking maybe if I try my luck and go maybe I will nd someone who will understand… But they all refused, and I went to the sick people thinking that I will meet someone, but it is still the same. They refused." (Age 23) Used others' extra ART pills "It does not happen, I make sure that I collect them even if they get nished, many of us take it, so there is even extra, you see. I have never missed my treatment ever since the lock down started… There are three people who are taking pills. They have extra pills, you see, which they have not used, you see. They give me from their pills." (Age 22) Challenges picking up ART due to: Clinic closures "It has affected me in that sometimes the clinics close, when you go to the clinic to…collect your pills and you would nd that they are not working, maybe they found someone who is sick, who has COVID-19. I have been to gateway and they were closed, and I waited for a couple of days and then I heard from others that they are open, and I went back and got my pills." (Age 21) Limits on timing of public transportation "So, we were really disturbed at that time, you will nd that you take transport to collect your pills you stand there until it is time for you to go back and then you rush back again because at 10:00 the transport nishes and then the transport operates again at 16:00 and that is when they close." (Age 37) Challenges accessing immunization services for baby due to COVID-19 closures "They kept turning us back and turning us back and they said they are not doing any immunizations. Like, since my child is six months I have gone there. When I arrived, they asked why I haven't come in to bring the baby, so I went back to the farm because I had shingles and I could not get assistance. Yes, I left because I was not assisted at the clinic, so I ended up going to the farm. When I went to immunize the baby, I told them that they turned me back when Covid-19 started, so I was not advised accordingly from the clinic." (Age 33) Concerns about breastfeeding while living with HIV in the context of COVID-19 "What is important right now and something that I am always cautious about is that I am taking my ARVs accordingly and because the baby is breastfeeding, I always make sure that I am doing it correctly. As well as I how I feed myself. That is what I need to ensure as well as to take care of my family." (Age 40) Misinformation about newborn health and COVID-19 "I would say the COVID-19 had a negative impact in my life because have you heard….do you remember when the COVID-19 arrived, it was everywhere on the social media, they were saying we should not immunise the babies, they are going to kill them, you see. There are those things that spread which are not true. You do not know whether you should take your baby to the clinic or not." (Age 22) Engagement in HIV care. The impacts of the COVID-19 pandemic and efforts to manage the spread of the virus in SA impacted engagement in HIV care primarily through reduced access to ART. Often, participants described missing ART doses, either for brief periods or for month-long stretches, because they left the city, went back to rural areas with their families, and did not have access to their typical clinic, where their records are kept. Providers sometimes refused to provide medication if participants did not have their "books" (i.e., their records, which document current medications and doses) with them. Some participants resorted to sharing medications within social networks to help each other. Other challenges to picking up ART were speci c to clinic closures, reduced public transportation, and lack of funds for transportation.
Physical health. Some participants reported contracting COVID-19 and/or described heightened risk for acquiring COVID-19, which had either direct or potential downstream effects on engagement in care. Most participants were acutely aware that their HIV-positive status increased their risk for negative COVID-19 outcomes. Participants were concerned about possible COVID-19 exposure at work, in public spaces where social distancing was not feasible, and at home, particularly among family members who did not take COVID-19-related precautions seriously. Moreover, participants expressed fear and uncertainty around possible COVID-19 exposure at the clinic, especially when collecting their medications.
Mental health. Themes speci c to general decreases in mood, increased anger, and increased anxiety also emerged from the data. Participants expressed the high toll that the pandemic and associated lockdowns have taken on their emotions, increasing the intensity of expressed emotions and, for example, complicating efforts to hide sadness, anger, or fear from their children. Many factors were cited as catalysts for increased anxiety, including fear of contracting COVID-19, concerns that others (especially children) might cope poorly with COVID-19, and lack of nancial resources, as many participants lost their jobs. Women also expressed multiple levels of uncertainty about the future, including potential vulnerabilities due to the interaction of COVID-19 and HIV, impacts on educational attainment, and the impacts of continued isolation.
Relationship with partner/father of the baby. When asked about the effects of the pandemic on partnerships and/or the relationship with the father of the baby, participants described both negative and positive impacts, with descriptions that often con icted. The majority of the sample noted decreased, rather than increased, interaction with their intimate partners, as many were restricted from traveling and had partners who were working in other cities or outside of the country. Some women described the ways in which decreased in-person interaction led to concerns that their partners would leave them, forget about their children, and fail to provide nancial support, which was a major concern. Increased distance was also linked to decreased sexual intimacy with partners, an increase in suspected relationships between male partners and other women, verbal abuse, and increased alcohol use among male partners.
Other participants, however, noted that increased distance improved their relationships. A minority of women experienced lockdown with their partners; some of these participants reported improved communication and joint planning for the future, disclosure of HIV status, and increased assistance with childcare.
Motherhood and caring for the new baby. Themes associated with childcare challenges were characterized by lack of support, fears that infants would acquire COVID-19, and concerns about breastfeeding when ART access is limited. Participants described consistent lack of support throughout the different lockdown periods, both speci c to childcare, which prohibited work, and to nances, which made it almost impossible to meet infants' needs. As women were unable to secure immunizations for their infants, they feared that their babies would be more susceptible to COVID-19. Similarly, with decreased consistency in ART use, again, due to decreased access, participants worried about HIV transmission via breastfeeding but did not report speci c changes in breastfeeding behavior.
Misinformation on social media about infant immunization in the early phases of the pandemic made it di cult for some participants to assess whether taking their babies to the clinic posed more or less risk than avoiding immunizations all together.
Coping. When asked how they coped with these COVID-19-related challenges, participants elucidated several strategies. See Table 4 for descriptions of each of the coping strategies that participants employed. Participants described their use of speci c coping strategies to decrease the distress associated with reduced engagement in HIV care, negative mood, evolving relationships with their partners, and challenges meeting their infants' needs. These strategies were also used to cope with nancial limitations and associated distress around meeting their family's basic needs-especially the baby's needs. Some women intentionally sought social support from family, friends, and new partners, whereas others relied heavily on spirituality and prayer. Many participants described taking active steps toward acceptance, recognizing that they had little to no control over the course of the pandemic and expressing gratitude for what they do have, rather than focusing on that which they have lost. Some found solace in adherence to COVID-19-related protocols, and still other participants coped via adaptive distractions, such as gardening.

Discussion
This analysis of multi-method data collected from postpartum WWH who were enrolled in a longitudinal cohort study revealed that about 20% of the cohort faced COVID-19-related challenges accessing HIV care, medications, or infant-related services. Brief qualitative interviews were conducted with the participants who endorsed those challenges to more thoroughly understand their experiences. With respect to direct impacts on care engagement, participants described challenges with ART use, including missed doses and shared pills with friends, as well as numerous logistical barriers to accessing their medications, often because they moved away from the vicinity of their home clinic and were refused medications from other clinics. Participants also described COVID-19-related impacts on factors that may indirectly in uence care engagement, including physical and mental health, intimate relationships or partnerships, and new parenthood. With awareness of the association between HIV and severe COVID-19related outcomes, many participants were concerned about their health and possible COVID-19 exposure, especially at the clinic. Negative effects on mental health included increased sadness, anger, and fear, with intersecting uncertainties-unknown interactions between COVID-19 and HIV, unknown impacts of continued isolation, unknown employment or nancial prospects-leading to a lack of hope for the future.
Effects on relationships with partners were mixed, with some participants reporting strong negative consequences of decreased interactions with their partners (e.g., feared or con rmed in delity, lack of nancial support for the infant, verbal abuse) and others expressing positive experiences when cohabitating with partners during lockdowns (e.g., increased communication, increased intimacy, HIV status disclosure). Finally, effects on parenting the new infant mostly centered on lack of nancial resources, which limited participants' ability to provide food and clothing, decreased access to immunizations due to COVID-19-related closures or restrictions on travel, and concerns about transmitting COVID-19 via breastmilk. Finally, when asked how they coped with these COVID-19-related challenges, they described highly adaptive strategies, including seeking social support, moving toward acceptance, and spirituality.
The engagement in care ndings validate early concerns and recent ndings on the impacts of the COVID-19 pandemic and associated public health efforts to manage disease transmission on the HIV treatment cascade in resource-limited settings. Within the rst few months of the pandemic, researchers and clinicians based in sub-Saharan Africa and other regions with high HIV prevalence rates signaled the alarm, drawing attention to the potential for severe disruptions at each phase of the HIV care continuumfrom HIV testing to ART access and interrupted ART supply to attrition from care and HIV-related deaths. [29][30][31] Several commentaries described the ways in which the pandemic might exacerbate structural inequities and HIV disease burden, particularly among women, who are more likely to be living with HIV compared to men and who typically serve as frontline workers in low-and middle-income countries. 13,32 Indeed, reductions in HIV testing at rst antenatal visit and reductions in HIV treatment access during pregnancy were documented across 17 countries and 15 countries, respectively. 33 Recent quantitative data from 65 primary care clinics in KwaZulu-Natal, where the data for the current analysis were collected, provides a broad-scale view of the impacts of the 2020 national lockdowns on HIV testing and treatment. In the rst week of lockdown (March 30, 2020 to April 5, 2020), there was an estimated 47.6% decrease in HIV testing and a 46.2% decrease in ART initiation, whereas ART collection visits decreased only slightly and missed ART collection visits increased for a short time. 34 Though these data suggest that ART provision was largely maintained during this speci c time period, it is unclear how these patterns changed over subsequent periods. Moreover, the degree to which speci c subpopulations that face additional barriers to retention in care under normal circumstances (including postpartum WWH) 8 were more severely impacted than others was not examined. In our sample of postpartum WWH, we did nd that ART access was compromised.
Our ndings also reinforce the importance of attending to and providing resources to address the mental health of women with HIV during public health crises. General decreases in mood may have downstream effects on engagement in HIV care, with potential for heightened risk of dropout among postpartum WWH, a population that faced signi cant mental health challenges pre-COVID-19. For example, in a review and meta-analysis published in early 2020, the pooled prevalence of postpartum depression in Africa was 16.8%, 35 with pre-pandemic rates speci c to SA hovering between 35 and 47%. 36 Depression is a known barrier to engagement in HIV care, 37 and depression during the postpartum period-both related and unrelated to COVID-19-has strong negative implications for decreased ART adherence and potential perinatal transmission. 38 In addition to decreased mood, women in our sample understandably also expressed speci c worries and concerns about potential HIV/COVID-19 co-infection, nances, and their ability to nish their education as well as plan for their child's future. These worries and more general anxiety, though normative in the early phases of the pandemic, may eventually contribute to patterns of behavioral avoidance, leading to further reduced access to HIV-related care. In addition, the combined effects of intimate partner violence and sexual trauma, which are common among WWH in SA, 39,40 as well as rekindled memories of apartheid-era restrictions may also compromise care engagement. 13 Though we did inquire about physical violence from a partner during the lockdown period, we did not explore the degree to which other forms of violence, pre-COVID-19 traumas, or posttraumatic stress may have contributed to or exacerbated di culties accessing care. Nonetheless, it is evident that the pandemic contributed to poor mental health among postpartum WWH.
Although some participants highlighted positive effects of the COVID-19 lockdowns on their current romantic relationships, most women emphasized negative consequences, nancial stressors resulting in part from greater physical distance from partners, and a strong need for additional childcare support. For a minority of participants, increased quality time with partners facilitated improved communication, increased intimacy, and HIV status disclosure. These bene ts have not been widely reported in the existing literature, which has primarily focused on the ways in which the pandemic has jeopardized relationship quality and stability, 41,42 nor have they been discussed among sub-populations at heighted risk for negative COVID-19 outcomes. In some cases, COVID-19 may have offered couples the opportunity to join together against an external threat, 41 especially if partners were locked down together. But, for the most part, women in this sample described the negative effects of decreased interaction (e.g., partners initiating sexual relationships with other women, verbal abuse, increased alcohol use) as the majority of participants were locked down separately from their partners. Similarly, in a Kenyan sample of adolescent girls and young women who had romantic partners during the pandemic, reduced time with partners was the strongest predictor of decreased relationship quality. 43 For many participants, decreased interaction with partners also translated into lack of shared childcare responsibilities and decreased nancial support for the infant, exacerbating existing gender inequalities, particularly in the unpaid (care) economy. Not only did women in SA experience two thirds of the net job losses between February and April 2020, they also took on a disproportionate share of additional childcare following school closures. 44 Other notable negative effects of the pandemic on parenting during the COVID-19 lockdowns were signi cant confusion around and reduced access to infant immunization, concerns about missed ART doses in the context of breastfeeding, and fears about infecting the baby as well as other children.
Finding balance between guarding against the spread of COVID-19 and controlling well known preventable diseases has proved challenging, particularly in low resource settings, with recent modeling predicting that not maintaining routine infant immunization will lead to more deaths than deaths related to COVID-19 exposures at vaccination clinics. 45 In a survey of members of the Immunizing Pregnant Women and Infants Network (IMPRINT), over 75% of whom were based in low-and middle-income countries, 50% reported broad challenges accessing immunizations, including logistical barriers, provider issues, and not attending appointments due to COVID-19 fear. 46 Although participants did not describe changes to their breastfeeding behaviors, there were clear concerns that limited access to ART in the context of the pandemic would render their breastmilk unsafe, and with decreased nancial resources to purchase formula, for example, the health of their infants could be at risk.
Importantly, the range of coping strategies that participants described demonstrate incredible sources of strength and resilience in spite of limited access to infant care services as well as reduced access to HIVrelated care, physical and mental health concerns, and changing relationship dynamics. A combination of social support, acceptance, strength through prayer, steadfast adherence to COVID-19 guidelines, and intentional distraction with other activities (e.g., cooking, gardening, cleaning) enabled participants to navigate through some of the darker periods of the lockdowns. Similar strategies were used by survivors of the 2014-2016 Ebola outbreak in West Africa, which primarily impacted Guinea, Liberia, and Sierra Leone. 47 During this period, individuals affected by Ebola maintained active involvement in community prevention efforts, participated in prayer and bible study, and sought social support from both family members and non-governmental organizations. 48 Among survivors of the SARS epidemic in Hong Kong, socialization through activities like Tai Chi helped restore a sense of meaning to their lives. 49 Participants who identify what is important to them and what makes them feel good, even when confronted by a situation that they cannot control, may enable them to pursue meaningful goals and activities under extremely trying circumstances. 50,51 Several limitations of the current analyses should be noted. Small sample sizes are typical of qualitative work, but the size of our sample (speci cally the small size of our COVID-19 wave 2 sample) limited our ability to separate the data by lockdown or by phase of the COVID-19 pandemic. Therefore, we could not draw conclusions about engagement in HIV care and contributing psychosocial challenges that may have been unique to speci c time points over the ongoing pandemic's duration. Similarly, based on the timing of their enrollment in the parent study, participants completed the COVID-19 assessment and corresponding interview at different points in their postpartum experiences (6, 12, 18, and 24 months), with too few interviews at each timepoint to explore relationships between early vs. late postpartum and COVID-19-related barriers to care. We had fewer participants complete assessments at the 24-month follow-up period (n = 5) relative to the other follow-up assessments (18, 13, and 17, respectively), for example, because we had hopes of conducting this nal assessment in-person. Importantly, participants who were not included in this qualitative sub-study either endorsed none of the four engagement in care challenges (making or keeping their HIV care appointments, procuring their HIV medications, procuring contraception, or accessing immunization services for their infants) or did not complete the parent study assessment to which the COVID-19 questions were added. It is possible that participants had challenges with other aspects of HIV-or infant-related care or follow-up that were not included on the list and therefore were not documented. It is also possible that participants who did not attend their parent study assessment may have had different or worse experiences than those who completed the assessments.
Therefore, even though we selectively identi ed participants who did report di culties remaining in care or accessing treatment, there may have been a selection bias toward a more resilient sample, such that postpartum WWH who were lost to follow-up in the parent study may have had worse COVID-19-related outcomes.
In conclusion, a signi cant portion of postpartum WWH have faced challenges making or keeping their HIV care appointments, procuring their HIV medications and/or contraception, and accessing immunization services for their infants during the early waves of the COVID-19 pandemic in SA. COVID-19-related effects on physical and mental health, relationship with partners, and parenting/childcare are both important to address in their own right and may have critical implications for both retention in HIV care and prevention of perinatal transmission. It is also important to highlight the degree to which the pandemic affected the nancial wellbeing of this sample, rendering it ever more challenging to meet basic needs. As the pandemic continues, providers who serve this population (who were also under tremendous pressure during this time 52,53 and are therefore in need of continued support) and public health o cials who set HIV care policy should proactively address these concerns at individual and systems-levels to avoid disruption of services, especially essential services for populations already at risk of attrition from HIV care.