“You must leave but I didn’t want to leave”: Qualitative evaluation of the integration of ART into postnatal maternal and child health services in Cape Town, South Africa

Abstract

Postpartum HIV care retention rates are well below retention rates of the general adult population. The Maternal-Child Health Antiretroviral Therapy (MCH-ART) trial tested the benefit of integrating postpartum maternal ART and pediatric care through the end of breastfeeding compared to the standard of care of immediate postpartum referral of mother and infant to separate services. After the trial, twenty-one participants completed in-depth interviews to understand the acceptability of the service integration and the potentially differing ‘lived’ experiences of the women randomized to the two conditions. Key findings include: 1) the MCH-ART integrated service was found to be acceptable, 2) women in the intervention condition expressed more negative feelings around the need to be transferred to general ART services and 3) women in the intervention condition perceived that they had more influence in selecting the clinic to which they would be transferred compared to those in the control group, although there was no actual difference by study design. Future work should more directly evaluate the impact of shared decision-making and long-term relationships with clinic staff on patient engagement and retention in HIV care.

Trial Registration:

ClinicalTrials.gov

Introduction

Retention in HIV care during the postpartum period is a major concern. In South Africa, one study found that 49% of women living with HIV missed an HIV care appointment or disengaged from care by six months postpartum (Phillips et al., 2014). Risks of disengagement from HIV care and suboptimal adherence during the postpartum period include viral rebound, drug resistance, and vertical and horizontal transmission (Hosseinipour, Gupta, Van Zyl, Eron, & Nachega, 2013; Sha et al., 2011).

In South Africa, HIV care services are integrated into antenatal care (ANC), which leads to the need to transfer from these integrated services to general adult HIV services post-delivery, which may be a particularly vulnerable time (Luzuriaga & Mofenson, 2016). Furthermore, mothers take their infants to separate child health services, creating further constraints on time, transportation, and other resources. Navigating the process of transferring a mother’s own care to an adult HIV clinic may not be prioritized during this time, leading to poor retention (Phillips, McNairy, Zerbe, Myer, & Abrams, 2015). Thus, interventions to improve postpartum retention in HIV care are needed.

One proposed strategy to increase postpartum maternal and child engagement in HIV care is the integration of ART services within postpartum maternal and child health (MCH) services (Bhardwaj, Carter, Aarons, & Chi, 2015). Integration of HIV services into ANC has improved retention in HIV care during pregnancy (Suthar et al., 2013, Turan et al., 2015), but limited research has tested the integration of services during the postpartum period, largely due to the lack of postpartum services for women more generally. For example, in South Africa, there is no structured programme of postpartum care. The integration of HIV services into expanded postpartum MCH services could delay the need to transfer care to a general HIV clinic post-delivery to a time that is potentially more stable for mothers.

A recent randomized controlled trial (RCT) in South Africa tested this strategy by integrating concurrent and co-located maternal ART and pediatric care into the MCH clinic through the end of breastfeeding (the Maternal-Child Health Antiretroviral Therapy [MCH-ART] service) compared to the standard of care of immediate postpartum referral of mother and infant to separate services (Myer et al., 2016). This study found that women in the integrated services condition had improved retention in care and viral suppression at 12 months postpartum compared to participants in the control condition (Myer et al., 2018). We do not, however, understand the acceptability of this service integration and the potentially differing ‘lived’ experiences of the women each condition, including the process of transferring out of the ANC setting, how supported women felt during the transfer process, and the similarities and differences between clinic settings. In this sub-study of the MCH-ART trial, we conducted in-depth interviews to elucidate and compare the experiences of participants in the intervention and control conditions.

Materials and Methods

Study setting

The study was conducted at a public sector Midwife Obstetric Unit (MOU) in Gugulethu, a former African township community in Cape Town, South Africa, with high rates of antenatal HIV infection.

MCH-ART Intervention

In the MCH-ART trial, all participants received integrated MCH and ART services during pregnancy, following the local standard of care. Participants were randomized to receive either: 1) the MCH-ART intervention condition which kept postpartum women and their infants in the MCH clinic to receive integrated concurrent and co-located maternal ART and pediatric care until cessation of breastfeeding or a maximum of 1 year or; 2) the standard of care control, in which women were transferred out of the MCH clinic after the first or second postpartum visit (typically no more than 1 week postpartum) to attend general adult ART services and their infants were referred to separate routine child health services. The MCH-ART trial was a service-level intervention and did not incorporate individual counseling or alter the quality of the service between the two conditions. All participants received routine ANC with a nurse midwife as the primary healthcare worker providing integrated ANC and HIV services. More detailed information about the MCH-ART study’s design and primary results have been reported elsewhere (Myer et al., 2016; 2018).

Current Study and Participants

As part of the MCH-ART trial, a qualitative sub-study was conducted. Participants from each arm of the trial were randomly selected. Additionally, we purposefully oversampled women who were employed and/or who had stopped ART to include those that may have greater barriers to successful engagement in HIV care postpartum. Two sets of interviews were conducted: one at 9 months postpartum and one at 18 months postpartum. Forty-seven women completed the 9-month interview and 21 (55.3%) were retained and interviewed at 18 months postpartum. The analyses presented here includes only the 18 months postpartum interviews, which occurred after all intervention procedures were completed and all women had been transferred out of the MCH integrated services. The Human Research Ethics Committee of the University of Cape Town Faculty of Health Sciences and the Institutional Review Board of Columbia University Medical Center approved this study and all participants provided informed consent.

Data collection

Demographic and clinical data was collected through the main MCH-ART study. Two female bilingual (English and isiXhosa) qualitative interviewers were trained in the semi-structured interview agenda and received on-going supervision to ensure adherence to the interview agenda, assess thematic saturation, and discuss any issues that arose. The interview agenda included reflections of experiences within the antenatal MOU setting, perceptions of the transfer process postpartum, experiences within adult HIV care clinics, and barriers and facilitators to retention in HIV care postpartum. Interviews were conducted in a private research office. Participants were reimbursed for their time and travel costs. Interviews were translated into English and transcribed. An independent isiXhosa speaking research assistant quality checked the interviews to ensure accuracy of the translation/transcription process.

Data analysis

An initial thematic coding structure was developed based on the interview guides. Two members of the study team (JP, AW) both reviewed a subset of the transcripts (33%, n=7) to develop definitions of thematic codes and ensure consistency across the two coders. The full analysis team also reviewed the initial codebook. For the remainder of the transcripts, the two coders independently applied the coding framework, meeting regularly to discuss the framework and any additional emergent themes throughout the process. NVivo 11 was used to organize the data and queries were run to determine any differences by intervention condition using case classifications. The full analysis team reviewed the initial results and interrogated the findings, which resulted in additional queries and validity checks of the analysis.

Results

Participants in this sub-sample were on average 27 years and a minority had matriculated high school (Table 1). Very few were primigravid and some were married or cohabitating. A few participants were newly diagnosed with HIV in the intervention condition, whereas over half of those in the control were newly diagnosed. Median CD4 T-cell counts indicate that participants were fairly healthy at antenatal booking and the majority had undetectable HIV RNA viral loads at randomization postpartum. Participants in the intervention condition had a longer mean duration of exclusive breastfeeding (Mean=5.3 months, SD=1.9) than those in the control condition (M=3.9, SD=2.3). By study design, all participants in the control condition (n=12) in the qualitative study sub-sample were transferred out of MCH-ART services before 3 months postpartum. Most participants in the intervention (n=9) transferred between 9-12 months postpartum.

Table 1:

Participant characteristics of the qualitative sub-sample of the MCH-ART randomized controlled trial

	Intervention: MCH- ART service	Control: standard of care
	(n=9)	(n=12)
Mean age (SD), years	27.9 (5.3)	27.8 (5.6)
Completed secondary/any tertiary education	1 (11%)	2 (17%)
Primigravid	1 (11%)	1 (8%)
Married/cohabitating	5 (56%)	4 (33%)
Newly diagnosed HIV+ in this pregnancy	2 (22%)	8 (66%)
Median CD4 count at antenatal booking	478	571
Undetectable HIV VL (40 copies/mL) at randomisation	5 (56%)	11 (92%)
Mean duration of exclusive breastfeeding in months	5.3 (1.9)	3.9 (2.3)
Postpartum time in MCH-ART service prior to transfer
Less than 3 months	0 (0%)	12 (100%)
3 - 6 months	2 (22%)	0 (0%)
6 - 9 months	0 (0%)	0 (0%)
9 – 12 months	7 (78%)	0 (0%)

Participants were interviewed about (1) perceptions and experiences in the MCH-ART service (regardless of time spent within these services), (2) experiences and perceptions of the transfer process, and (3) experiences of the new clinic to which the participant was transferred to continue her HIV care. Differences by intervention condition are noted.

1). MCH and ART integrated services

Participants in both the intervention and control conditions reported largely positive experiences within the MCH-ART service during pregnancy and during postpartum for those in the intervention. Participants from both conditions reported feeling welcome and comfortable within this service.

Participant (P): It is a good space to be in, even when you leave you leave free-spirited you never leave with something that upsets you. (23 years old, control condition)

Interviewer (I): What did you like the most?

P: support and privacy. Privacy. You are not seen by other people, no matter what happens, you didn’t have any concerns. You were motivated to come to the clinic. (28 years old, intervention condition)

Additionally, participants from both conditions reported that the staff were approachable, generally supportive of mothers living with HIV, and tended to women quickly. Participants also noted the personal importance of being able to trust the staff ensure confidentiality.

P: I liked them because they kept our conversation confidential. Their health care was the best. They treated me very well up to the point where I felt I was the only one at the clinic. (20 years old, control condition)

Although the majority of participants reported good service, there were some instances of mistreatment in the labor ward among women in both conditions. Participants indicated that these experiences were limited to labor and delivery and did not extend to the rest of the MCH and ART integrated care services.

While most participants reported positive experiences within the MCH and ART integrated service, very few women discussed their perceptions of the actual integration of ART and MCH services, compared to receiving ART separate from baby services postpartum.

2). Transfer process from MCH and ART integrated services to adult HIV care

Participants expressed mixed feelings about the transfer process. Some participants in both conditions were happy with the transfer: several participants stated that the new clinic was closer to home and/or would be the same clinic that her baby attended. However, nearly all of the intervention participants reported some negative feelings about the need to be transferred including feeling sad, hurt, or unhappy. Only three women in the control group expressed these negative feelings.

P: […] If there was an option of me staying here (Gugulethu MOU) I would have stayed because here, they take care of people more than the clinics we are transferred to. (34 years old, intervention condition)

Furthermore, there were some differences between the conditions regarding the participant’s perceived choice in selecting the clinic to which they would be transferred. Those in the intervention group perceived that they had an influence on the decision, whereas, some of those in the control group did not feel that they had any influence on the decision.

I: Did you decide on your own?

P: I was only told to go to the clinic close to my place. I was never given a chance. (24 years old, control)

Additionally, some participants in the intervention condition reported receiving help from staff at the MCH and ART integrated services to decide which clinic to attend, whereas almost all participants in the control condition reported not receiving any help.

P: They asked me to choose the clinic I wanted to be referred to. I wanted to find out from them if they can assist me to choose the good clinic to go to. They said there is one here in KTC. They mentioned Vuyani clinic as well. I told them that I want to be transferred to Vuyani clinic. (26 years old, intervention)

Among women in both conditions, influential factors in deciding which clinic to transfer to included crowding, good reputation, and the perception of confidentiality among staff members at particular clinics. Location was also commonly cited either because some women chose clinics that were close for convenience whereas other women chose clinics that were far away because they did not want to be seen at the local clinic for fear of stigma and lack of privacy.

P: […] Someone helped me when I told her that I wanted to be transferred out to Delft. I told her the reasons why I chose to be transferred out to Delft. She made me aware that anything might happen even in Delft. She said I can run but I cannot hide.

I: How did you feel about that?

P: I felt she was telling me the truth but I found it better than to go to the clinic around my community. (22 years old, intervention condition)

With regards to the actual transfer and first appointment process, most participants in both conditions reported a fairly easy process of transferring, while a few reported feeling anxious or scared. Several participants in the control condition noted that in clinics with baby and adult care, the baby is seen at 6 weeks and the mother is scheduled the same day, making the transfer process easier.

3). Experiences post-transfer at new HIV care clinic

Participants had mixed experiences at their new HIV care clinics. Some participants had negative experiences, such as staff members being mean and judging them because of their HIV status. However, most of the participants in both conditions reported positive perceptions of the new clinic and staff.

P: They are caring, that’s what I like about that clinic, there is nothing I don’t like about them. (23 years old, control)

With regards to clinic atmosphere, a few participants reported overcrowding and staff shortages in the adult HIV clinics while other participants reported that the clinics were in good standard and that they received adequate, timely service.

P: There is a shortage of staff. You will find that only two people are serving a clinic full of people. (34 years, control)

P: There is a lot happening at the clinic. I normally go there to attend my appointments then get my ART and go back home. I don’t spend much time at the clinic. (24 years, control)

There were mixed responses from participants in both conditions with regards to privacy issues within these clinics. Some were deeply concerned about being identified as HIV positive within the clinic setting, whereas others felt that the adult HIV clinics were private.

P: … I am relaxed there. I have no stress at all. I don’t care how many times I go there because no one knows. There, everyone has the same card, same waiting room, same doctor, and same pharmacy. I am comfortable there. (28 years, intervention)

Discussion

As the MCH-ART trial was a service-level intervention and did not incorporate individual counseling or alter the quality of the service by condition, we did not expect there to be many differences in experiences of the service integration, other than length of time spent in the service. Indeed, we found similar and largely positive experiences of the MCH and ART integrated care services among those randomized to both conditions, regardless of how long women stayed within these services. We did, however, find some differences in the participants’ feelings around the need to transfer as well as the experience of the transfer process itself.

With regards to the transfer process, qualitative differences between the conditions may partially explain the main trial results of improved retention in care and viral suppression in the intervention condition. Although participants from both conditions reported liking the integrated services, women in the intervention condition more often reported feeling sad, hurt, or unhappy about the need to transfer out of these services. We believe this may have been due to the extra time spent within the MCH-ART integrated services by those in the intervention condition who likely developed closer relationships with the staff demonstrated by these negative feelings when told they needed to transfer out. These close relationships may also have played a role in engagement in HIV care among those in the intervention condition (Wood et al., 2018).

There were also differences in the perceptions of the participants’ ability to decide which clinic she would be transferred to. Women in the intervention may have felt more involved in this process because they had longer relationships with the providers in the MOU and may have perceived having more choice with regards to where they would be transferred for care. There is some evidence in the literature to suggest that having a choice, perceived or real, may have some benefits in gynecological and obstetric care (Edmonds, 2014), however, current evidence determining the impact of shared decision-making interventions on behavioral and clinical outcomes is limited (Shay & Lafata, 2015). Within the context of postpartum HIV care transitions, future research should investigate the independent effects of shared decision-making on postpartum retention in HIV care.

There are other potential explanations not captured by these qualitative findings that may also explain the significant trial results. Transferring clinics during the postpartum period creates some level of disruption in daily life. Within the context of this trial, the timing of this disruption may have been a key factor in improved patient outcomes. Women randomized to the control condition had this disruption occur when they were potentially more vulnerable to its impact. Early postpartum can be a difficult time due to managing the care of a small infant, healing from labor, and lack of sleep. Adding the disruption of a clinic transfer during this time may be particularly problematic for retention. Because women only discussed their personal experiences with the transfer process, we cannot qualitatively determine the extent to which the timing of this disruption may have contributed to intervention outcomes.

Limitations

There are several limitations to this study. A substantial number of women who completed the 9-month interview could not be reached to complete the 18-month interview (26/47; 55.3%). Women who completed the 18-month interview may be different from those who did not complete this interview in ways that may limit generalizability. Additionally, by study design, participants’ choice of new clinic was not manipulated, however, there may have been objective differences in how much choice women actually had which may have influenced participants’ perception of choice.

Conclusions

The current study, in conjunction with the main MCH-ART trial findings, indicate that the integration of MCH and ART services during the postpartum period is acceptable and positively impacts women’s retention in HIV care postpartum. Future research should investigate the implementation of this type of intervention in other settings.