Between March and September 2021, 139 perinatal women participated in key informant interviews. The majority were pregnant [n=125 (90%)], in their third trimester of pregnancy [n=74 (53%)] and spoke IsiXhosa [n=78 (56%)]. Eleven (n=11) key informant interviews were conducted with facility-based healthcare workers between March and May 2021. Two MOU Managers, seven ANC nurses and two health promotion officers were interviewed. Between June and August 2021, nine focus group discussions were conducted with ANC nurses (n=3) and OTLs (n=6).
Training
Master trainers received their training in January 2020 and were tasked with cascading the training to the relevant healthcare workers in February 2020 (Table 2). As a result of the COVID-19 pandemic in March 2020, many of the healthcare workers who had received training, were no longer available in 2021 when delivery of the HSS components were assessed. Several new and refresher trainings were held with all healthcare workers involved between March and June 2021.
Health promotion and awareness raising
Healthcare workers reported that their one-day training in 2020 was adequate and had equipped them with the skills needed to deliver the talks. One healthcare worker explained that the training had taught her to identify “the person who is behaving out of normal character” [G006H], which helped her identify women who needed to be referred to the nurse.
Detection and referral
Facility PACK trainers were first trained as master trainers. However, as a result of the pandemic, the facility PACK trainers were no longer available in 2021 and ASSET team members had to provide refresher training to some ANC nurses and completely new training to others. The modules that were intended to be delivered weekly over a four-week period, had to be condensed and delivered in a four-hour training session, with minimal time for exercises. Some ANC nurses reported that the training was very informative while others explained that they did not use the PACK guide in the antenatal clinic. Instead, nurses were trained to follow the BANC guidelines. At one MOU, the nurses reported that the training was too long, as they had been screening pregnant women (only those attending the healthcare facility for their first antenatal care visit) for symptoms of CMDs using the mental health screening questionnaire (released by the DoH after completion of the pre-implementation phase of the study, in an updated version of the MCR).
Counselling and supervision
In addition to the original counselling and supervision training in 2020, several additional trainings were delivered by the master trainers who were originally trained. While some OTLs reported that the training was adequate and had equipped the CHWs with the necessary skills, many felt that they needed more time for role-playing and practicing how to deliver the three sessions. One OTL reported that the CHWs she supervised “… didn’t have the confidence. They didn’t feel adequately trained” [R02FGD]. She suggested that the training, which was delivered over three days, would work better if it was delivered over a longer period, with time in between where CHWs could practice the skills they were taught before learning the next skill.
HSS programme components
The results are reported for each of the HSS programme components using the following implementation outcomes: acceptability, appropriateness, feasibility, adoption, fidelity of delivery, and fidelity of receipt (Table 3).
Insert Table 3
HSS programme component 1: Health promotion and awareness raising
This health promotion and awareness raising component involved health promotion officers, or other lay healthcare workers, delivering a 5–7-minute presentation to small groups of women in the waiting areas, using a flipchart as a visual aid. The presentation included information on the signs, symptoms, risk factors and consequences of depression, anxiety and experiences of domestic violence as well as to inform women about the mental health screening delivered by the ANC nurses and the home-based counselling delivered by CHWs for women with symptoms of CMDs or domestic violence. As very few talks were delivered towards the end of the study when the qualitative interviews took place, we were unable to assess the acceptability, appropriateness and feasibility from a user perspective.
Acceptability – satisfaction with the content and delivery of the health promotion and awareness raising component
Healthcare workers considered the provision of the health promotion and awareness raising talks to be acceptable. An ANC nurse explained that the pregnant women “enjoyed the talks” [R01FGD]. In addition, the content of the talks played an important role in creating awareness of the symptoms of CMD. An ANC nurse described how women would disclose their feelings “as soon as they come in (to the consultation room)” by saying “I heard the talk and I do feel like that Sister” [R01FGD].
Appropriateness – perceived fit, usefulness and relevance of the health promotion and awareness raising component
Healthcare workers felt that the talks were especially useful in alerting pregnant women to the mental health screening questions that would be asked, as well as the availability of counselling by CHWs for those who screened positive. A healthcare worker explained that the talks were “helpful to the patients, because the patients now they are willing to talk to us” [G006H]. She elaborated by saying that “they (pregnant women) know where to get help if they have a problem, they know who they can talk to” and that the talks should be “introduced at other clinics (healthcare facilities) that see pregnant women”. A health promotion officer explained that, before delivery of the talks, pregnant women “didn’t know that when you come to the clinic, you come for many things (including mental health counselling) … they thought you only come for pregnancy” [G001H].
Feasibility – suitability of the health promotion and awareness raising component for everyday use
While the content of the talks was considered suitable, provider-related issues impacted its suitability for everyday use. Following delivery of the talks, some women would seek out the healthcare worker who had delivered the talk to discuss the problems they were experiencing, leaving the healthcare worker feeling distressed and in need of support. A health promotion officer reflected on her experiences by saying “after you speak to the client, you sit down and reflect on that problem. It sometimes leaves you feeling depressed” [G001H].
Adoption – uptake, utilization and initial implementation of the health promotion and awareness raising component
The talks were not delivered regularly. At facilities without a dedicated health promotion officer, those who were trained reported that their other responsibilities were often time sensitive, which made it difficult to deliver the talks in the morning when the perinatal women were waiting for their consultations with the midwives. Healthcare workers also rotated into different departments, which resulted in those who were trained to deliver the talks, being unavailable a few months after their training. At the BANC clinic, the breastfeeding counsellor initially delivered a talk 3-4 days per week, while the enrolled nurse at the MOU delivered the talks daily. At both facilities the healthcare workers were unable to provide the talks from August when they were rotated to a different department. At the MOU with a dedicated health promotion officer, the talks were delivered approximately twice per week. On the other days, the health promotion officer chose to present other health-related information such as the stages of pregnancy, HIV testing, hygiene, and breastfeeding.
Fidelity of delivery – delivery of the health promotion and awareness raising component as intended
The talks were not always delivered as intended. Only one facility had a dedicated health promotion officer. At the other facilities, other lay healthcare workers were identified to provide the talks. At the BANC clinic, a breastfeeding counsellor was trained to provide the talks. She never used the flipchart, and only spoke about depression and domestic violence. At one of the MOUs, an enrolled nurse always used the flipchart and delivered the talk as intended. The second MOU had a dedicated health promotion officer who was trained. She chose not to use the flipchart, and in most cases spoke only about depression.
Fidelity of receipt – receival of the health promotion and awareness raising component improved knowledge and practices as intended
Five hundred and forty-five (n=545) women completed the knowledge and practices survey in the two weeks prior to the delivery of the HSS programme (baseline) and 650 women completed the survey after an awareness raising talk (follow-up). We found an improvement in health seeking behaviour after the talks (Table 4). Significantly more women were prepared to disclose feelings of depression or anxiety to family members (83% vs. 77%; p=0.020); and to disclose feelings of depression (82% vs. 76%; p=0.011) and anxiety (83% vs. 77%; p=0.009) to healthcare workers after the talks, compared to before the HSS programme was delivered. There was no significant change in the proportion of women who were prepared to disclose experiences of domestic violence to a family member (p=0.504) or healthcare worker (p=0.642) after the talks, compared to before the HSS programme was delivered.
Insert Table 4
HSS programme component 2: Detection
Detecting pregnant women with CMD and domestic violence involved an ANC nurse administering a brief screening questionnaire (available in the MCR [18]), using a patient-centred approach, to all pregnant women during their routine consultations. Women who screened positive underwent further assessment using the Practical Approach to Care Kit (PACK) - Primary Care Guide for the Adult [19].
Acceptability – satisfaction with the content and delivery of the detection component
While many of the pregnant women interviewed expressed acceptability of the screening process, some women were reticent about revealing how they felt to an ANC nurse. One pregnant woman explained that the ANC nurse “did not ask anything that could start the conversation, she concentrated on the examination” [N045P], while another woman explained that “it’s just not easy talking to a nurse” [N047P]. An MOU manager explained that “Unfortunately, we as nurses have a bad name out there. Not everybody feels comfortable speaking up – to voice out what they want to say, they will choose certain people” [G004H].
Other women were concerned about the stigma of disclosing domestic violence to a nurse. A pregnant woman who was experiencing domestic violence explained that she felt ambivalent about disclosing her experiences because “when you tell the truth, people might judge you” [G045P].
Appropriateness – perceived fit, usefulness and relevance of the detection component
All healthcare workers interviewed reported that the screening process was useful for detecting women with symptoms of CMD and domestic violence. One healthcare worker explained that the screening was “something that we’ve needed for quite some time in the maternal aspect, because it’s certain things that can be easily ignored” [G005H]. One ANC nurse commented on the importance of making eye contact (which formed part of the detection component) with a patient when administering the mental health questionnaire by saying “you have to make time to look … in the eyes. If the patient doesn't make eye contact, you know there is something going on” [RGF01]. Many pregnant women confirmed that the mental health questionnaire administered by the ANC nurses gave them the opportunity to express how they were feeling, which they would not have disclosed if they had not been specifically asked.
Feasibility – suitability of the detection component for everyday use
There were several challenges observed related to the feasibility of the detection process. The proportion of women who were detected with symptoms of CMD when a research team fieldworker did the screening was significantly more than when the ANC nurses did the screening [474/1198 (28%) vs. 190/9000 (2%)], even though the fieldworkers screened fewer women than the nurses each day (each facility had one fieldworker versus several nurses). An ANC nurse explained that when she screens patients they would say “I’m fine sister, I don’t have any problems, I’m fine” [GFG01], yet the same patient would screen positive when the fieldworker did the screening (using the same questions).
One ANC nurse explained that patients do not want to disclose their experiences in the consultation rooms as the cubicles offer little privacy – “it’s only the curtains so the patient is sitting here, she can hear what is going on, on the other side” [RFG01].
Several ANC nurses reported that patients’ disclosures during the screening process were sometimes difficult to listen to. One ANC nurse said that “it’s not that we don’t want to do these things (screening), we are doing it, but sometimes it gets too much” [G004H]. Another ANC nurse elaborated by saying “we are unable to cope because we are facing our own problems” [GFG01]. A third ANC nurse described how a patients’ disclosure had affected her by saying “the patient was very emotional, and I was seeing to the patient, so I also broke down” [GFG01].
Many healthcare workers involved in the screening process were unhappy about the increased workload. Administering the screening questions, completing additional documents and calming patients who became teary during the screening process added additional time to their already busy schedules. One healthcare worker explained that “we are already ticking a lot of things… and filling in a lot of forms. And it’s taking up a lot of our time” [G004H].
Adoption – uptake, utilization and initial implementation of the detection component
The screening component of the HSS programme was delivered inconsistently. ANC nurses at one facility chose to only screen patients at their first antenatal clinic visit, while ANC nurses at the other healthcare facilities skipped the screening process during busy times. One ANC nurses admitted that “we don’t do it (screening) routinely” [G002H]. As a result, nurses reported only being able to screen 30-40% of the women attending antenatal clinics each day.
Fidelity of delivery – delivery of the detection component as intended
The majority of nurses referred women for counselling after screening positive on the brief screening questionnaire instead of using the PACK guide to assess the severity of symptoms or exclude other possible reasons for their symptoms (such as bereavement, medication side effects, or substance abuse) before making the referral. ANC nurses reported that they did not feel comfortable or confident using the PACK guide as only a few of them had received prior PACK training. An OTL described how high-risk cases were inappropriately referred to the NPO by saying “there are cases where a mom is referred to us but their situation is beyond our capabilities… There is a lot of cases that needs more of the professional help than what we can give, abuse is one to mention” [GFG02].
Fidelity of receipt – receival of the detection component improved detection rates as intended
A review of the mental health screening questionnaire in the MCR (completed by ANC nurses) was used to assess changes in detection rates (Table 5). During the delivery of the HSS programme, both the proportion of women screened [611 (75%) vs. 496 (60%); p<0.001], as well as the proportion of women detected with CMD [42 (7%) vs. 17 (3%); p=0.011] improved significantly.
Insert Table 5
HSS programme component 3: Referral
Linkage to care involved: (1) ANC nurses completing a referral form for women who screened positive and agreed to the counselling; (2) emailing the referral form to the community-based services coordinator, who reviewed and forwarded the referral form to the relevant NPOs or when the physical referral form was collected from the healthcare facility and delivered to the NPO office; (3) assigning the patient to an OTL and CHW based on the woman’s address (OTLs and CHWS serviced a pre-defined area);and (4) an OTL or CHW making telephonic or physical contact with the referred woman to arrange counselling.
Acceptability – satisfaction with the content and delivery of the referral component
Many pregnant women who were interviewed at the healthcare facilities indicated that they would be happy to be referred to a CHW, social worker or mental health nurse if they required support. However, several women did not want anyone to visit them at home and indicated that unless the counselling took place at the healthcare facility, they would not accept the referral. Some CHWs had difficulty articulating the reason for the visit following a referral, especially when using one of the local languages (Afrikaans and IsiXhosa), which resulted in women being dissatisfied with the referral. One pregnant woman described the CHWs IsiXhosa reason for the visit by saying “someone did come to my house, but she said they were helping crazy people. So I told them I’m not crazy… so they never came again” [N047P].
Some women provided incorrect contact information because (1) they feared being turned away from their facility of choice because they did not reside in the immediate vicinity, or (2) they agreed to the counselling to please the nurse and didn’t fully understand what they were agreeing to. One OTL recounted her experience with a pregnant woman who had been referred by saying “Sometimes they will tell you - I just signed that paper there to please the sister at the MOU just to get her off my back” [RFG02].
Appropriateness – perceived fit, usefulness and relevance of the referral component
Healthcare workers based at the MOUs were happy to email their referrals to the community-based service office for dissemination to the NPOs. The process was familiar to all involved as the same process was used to refer other service users requiring community or home-based support. Since the BANC clinic only had one supporting NPO, those involved found the collection of physical forms a more suitable method.
Feasibility – suitability of the referral component for everyday use
While ANC nurses were happy to complete the referral forms, administrators and managers who were tasked with ensuring that the forms were emailed to the community-based service office reported that the increased workload was sometimes quite difficult to manage. One manager explained her dilemma by saying “sometimes I do forget to scan the referrals, especially Fridays because I am not here Saturday and Sunday, … but my aim was to always try and push and make sure that people are being seen as soon as possible” [G004H].
Adoption – uptake, utilization and initial implementation of the referral component
Many women returned to the facility weeks after being referred for counselling, without anyone having made contact with them. In some instances, by the time CHWs were able to contact a patient, their feelings of distress had already been resolved, or they had accessed support elsewhere. One ANC nurse explained that patients returned to the facility saying “I am still having a problem and they didn’t get hold of me” [NFG01]. The delay or absence in contacting women occurred for several reasons such as: (1) incomplete or illegible referral forms; (2) incorrect contact details on referral forms; (3) delays in emailing the referral forms to the community-based services coordinator;(4) delays in a community-based services coordinator accessing the referral forms and forwarding them to the relevant NPO;and (5) delays in OTLs contacting patients. One OTL explained that “sometimes it’s difficult as some patients give wrong addresses, phone numbers that don’t exist or they will say they don’t have a number, it’s not easy to trace them” [GFG02].
Fidelity of delivery – delivery of the referral component as intended
While a referral form was always completed for patients who required a referral, the forms often contained incorrect contact details, making it difficult to find the women. Many women did not own a cell phone, so were not able to provide their contact details or provide the contact number of a friend, relative or neighbour. In addition, seventeen (9%) of the women were incorrectly referred for home-based counselling as they were high risk cases that needed specialised support. These women were referred back to the facility, where they received a referral to see a mental health nurse or social worker for further support.
Fidelity of receipt – receival of the referral component as improved referral rates as intended
The referral system was partially effective at linking patients to care. Of the 198 women who were referred for counselling, 59% were contactable. Illegible, incorrect or outdated contact details were responsible for being unable to locate 41% of those referred.
HSS programme component 4: Treatment
The treatment component consisted of three structured counselling sessions, using a problem-solving approach, delivered by CHWs in patients’ homes or at the healthcare facility, while being supervised and supported by OTLs [21].
Acceptability – satisfaction with the content and delivery of the counselling component
All women who experienced one or more counselling sessions (n=39) reported being satisfied with the counselling and that it had helped them cope when they were feeling distressed. One pregnant woman explained how counselling had helped her by saying “before I attended the counselling, I was crying a lot. Everything was hard for me. But since I spoke to them (CHWs), at least it went down, it was good to talk to them” [G044P]. Another pregnant woman described her feeling of comfort while being counselled by saying “when they are with me, it’s like I was with my sisters” [G043P]. While some CHWs were happy to deliver the counselling sessions, others were uncomfortable supporting women that came from the same communities that they lived in and were experiencing the same problems they were experiencing. One OTL explained that some of the CHWs “don’t think that this is where they want to be – to give counselling…so we can’t force them” [R02FGD].
Appropriateness – perceived fit, usefulness and relevance of the counselling component
Several OTLs and CHWs reported on the usefulness of the counselling content. Some mentioned that they were able to use the counselling skills for their other patients, not just the pregnant women who were referred for counselling. One OTL commented on how it had “helped us look at the mental state and not just going to see the mums and babies” [R004H]. Many of the women who received at least one counselling session expressed how useful the counselling had been. One pregnant woman described how burdened she felt before she received counselling, and that the counselling “was actually good” [G045P] . She explained that after she started the counselling, she felt “at ease a bit – not totally, but I felt good. I felt like I can talk to them (the CHW and OTL) freely without them judging me. They were very welcoming, very supportive and very understanding at the same time”.
Feasibility – suitability of the counselling component for everyday use
Several issues were raised regarding the feasibility of the counselling component. Both patients and healthcare workers felt that CHWs were not the right cadre of staff to provide counselling. As CHWs lived in the same community in which they worked, patients were concerned about confidentiality. In addition, OTLs were concerned about the capability of CHWs as they were lay healthcare workers with no previous mental health training or experience. An OTL explained that CHWs “have never dealt with anything mental before… they’ve got no experience; even their home-based care training, some of them it’s so minimal … there are huge problems there with using CHWs in the communities” [RFG02].
Using patients’ homes for counselling was also problematic, as many women lived in small homes or informal settlements with large household numbers. It was often difficult to find a private space to deliver the counselling. One pregnant woman explained that “We live in a shack, so when you talk in your house, someone can hear you from next door, so it’s not private enough” [G045P]. In addition, the some women were concerned that other household members and neighbours would be curious about the CHW visits, as these visits were historically connected to patients with HIV or TB. Some women voiced their concern that others would assume that they were receiving treatment for HIV or TB.
Both CHWs and OTLs reported feeling distressed after hearing patients’ experiences of distress, especially when they were describing domestic violence. One OTL described her difficulty by saying “I struggled to shut the visuals in my mind. I couldn’t sleep that night” [GFG02].
Adoption – uptake, utilization and initial implementation of the counselling component
Between April and July 2021, 198 pregnant women were referred for home-based counselling (Table 6). Only 20 (10%) of those referred completed all three counselling sessions and reported feeling better able to cope. An additional 19 (10%) felt better after one or two sessions while 13 (7%) reported feeling better without counselling. Many women were either not found [82(41%)] or no longer interested in receiving counselling [35 (18%)] once they knew what it entailed. Some women misunderstood the referral and thought that the CHWs would be bringing them money, food or clothes for the baby.
Insert Table 6
Fidelity of delivery – delivery of the counselling component as intended
Many CHWs had difficulty delivering the counselling sessions as intended. Even though OTLs spent time preparing the CHW for a counselling session, they often became overwhelmed during the counselling session and did not follow the structured process. An OTL described how CHWs were “allowing patients to go off-topic and load it with all sorts of things, but not following the steps” [RFG02]. A pregnant woman who received counselling confirmed the OTLs’ views when she explained that the CHW tasked with providing the counselling had “asked me to pray about the situation whenever I feel like I’m alone or I don’t have someone to talk to. I should just talk to myself about the whole thing as if I’m with someone” [G047P]. Another pregnant woman explained that the CHW had given her a book to “read to forget” and had advised her to speak to her pastor to “help her cope” [R039P]. Yet, the few OTLs who evaluated the CHWs fidelity to the counselling steps and structure using a checklist, reported that the steps were done partially or done well. Further comments on the evaluation forms referred to the CHWs interest in helping the patients, their good listening skills and their efforts to reassure patients of the confidentiality of the counselling sessions.
Fidelity of receipt – receival of the counselling component decreased symptoms of distress as intended
Even though the counselling sessions were not delivered as intended, all the pregnant women who received counselling reported feeling better after speaking to a CHW. A 20-year-old who was pregnant with twins explained that she “could open up to someone I don’t really know and they made me feel comfortable” [R037P]. Many women appreciated having someone that checked on them and cared about how they were feeling. One woman explained that “they just come to see what I’m doing … and to ask me about my pregnancy and when I’m due” [R034P], while another woman explained that she found it helpful to “talk to someone else that can relate or have an understanding” [R033P].