Peripartum common mental disorders (CMD) are highly prevalent in low- and middle-income countries (LMIC) such as South Africa. With limited public mental health resources, task sharing approaches to treatment are showing promise. However, little is known about the feasibility and acceptability of, as well as responses associated with problem-solving therapy (PST) for the treatment of prepartum CMD symptoms in South African public health settings.
To investigate participants’ preliminary responses to a task sharing PST intervention, and to evaluate the feasibility and acceptability of the intervention.
A Midwife and Obstetrics Unit attached to a Community Health Centre in a Western Cape district.
Using mixed methods, 38 participants’ responses to a PST intervention, and their perceptions of its feasibility and acceptability, were explored. Primary outcomes included psychological distress (Self Reporting Questionnaire; SRQ-20) and depression symptoms (Edinborough Postnatal Depression Scale; EPDS). Semi-structured interviews were conducted three after the last session. Six stakeholders were also interviewed.
Significant reductions were seen on EPDS (Cohen’s
Results support task sharing PST to Registered Counsellors to treat antenatal CMDs in perinatal primary health care settings. Research is needed on how such programmes might be integrated into public health settings, incorporating other non-specialists.
Perinatal common mental disorders (CMDs), such as depression and anxiety, are highly prevalent in low-and middle-income countries (LMICs)
Several systematic reviews have investigated the effectiveness of task-sharing mental health interventions in LMICs.
As one of the World Health Organization’s (WHO) Mental health Gap Action Programme (mhGAP)-recommended treatments,
Given the comparatively recent recognition of the burden associated with maternal mental illness in LMICs,
Data were collected at a MOU that serves a large district in the Western Cape province of South Africa, with a primarily low-income population of more than 300 000 people.
We employed a mixed-methods design comprising two phases. In phase 1, quantitative data were collected to measure participants’ preliminary responses to the intervention, whilst qualitative data were collected to explore women’s perceptions of its feasibility and acceptability. This phase addressed the study’s first two aims. In phase 2, qualitative methods were again used to collect data concerning the MOU personnel’s (stakeholders) perceptions of the barriers to and facilitators of intervention delivery. This phase addressed the study’s third aim. Given the small sample sizes in both phases, qualitative methods for examining feasibility and acceptability were deemed most appropriate.
Recruitment procedures and intervention process.
The PST intervention used in this study was adapted from Sorsdahl et al.
Description of the problem-solving therapy sessions’ content and procedures.
Session 1 | Session 2 | Session 3 |
---|---|---|
In addition to orienting the participant to the PST model, this session involved helping her identify what is most important to her life. Problems and worries would then be listed and categorised in one of three groups: problems that are not important (group A), problems that are important but unsolvable (group B) and problems that are important and solvable (group C). A problem from the third group would then be selected and together the registered counsellor and participant would develop a step-by-step plan to solve the problem. | The participant would be reminded of the PST model and a list of adaptive coping strategies would be discussed. A problem from the ‘not important’ category would then be selected and the ways in which coping strategies might be applied to these would be discussed. In addition, a problem from the third group would be selected and together the registered counsellor and the participant would develop a step-by-step plan to solve the problem. | Again, the participant would be reminded of the model and then more coping strategies would be discussed. Thereafter, a problem from the ‘important but unsolvable’ category would be selected from the list made in the first session, and ways of coping with this problem would then be discussed. Again, a problem from the third group would also be selected and together registered counsellor and the participant would develop a step-by-step plan to solve the problem. |
Intake nurses were trained to use the EPDS, incorporate it into standard assessment procedures and make referrals to the registered counsellor. In addition to her formal Bachelor of Psychology degree training, the registered counsellor also received a 3-day training course in maternal mental health and 18 h of training in the PST model and manual. She also received at least 1 h of clinical supervision per week from the first author, a registered clinical psychologist. Procedural matters, case management and fidelity to the therapeutic protocol were addressed in supervision. A random review of recorded sessions did not reveal any protocol drift.
The primary outcome of this study was psychological distress. The secondary outcomes included perinatal depression, functional impairment, substance use involvement, perceived stress and perceived social support.
Stakeholders’ roles in the project.
Title/designation | Role in the project |
---|---|
Health worker | Screening and referral |
Senior nursing assistant | Screening and referral |
Midwife, nursing sister | Screening and referral |
Midwife, nursing sister, acting head of MOU | Referral source, primary liaison |
Social worker | Referral source and resource |
Registered counsellor | Collected baseline data and delivered PST intervention |
MOU, Midwife and Obstetrics Unit; PST, problem-solving therapy.
Quantitative data were analysed using SPSS version 23.0. Participants’ socio-demographic data were analysed with descriptive statistics. We used both paired
Ethical approval to conduct the study was obtained from the Faculty of Health Sciences Research Ethics Committee (HREC) at the University of Cape Town. Permission to collect data at the midwife and obstetrics unit was also obtained from the Western Cape Department of Health as well as the facility management. Written informed consent was obtained from all the participants. In order to protect the identities of the stakeholders, identification codes were omitted from quotes included in this article.
Of the 38 women who participated in the study (see
The demographic characteristics of the problem-solving therapy intervention sample (
Variable | Total sample ( |
% of sample |
---|---|---|
18–24 years | 11 | 28.9 |
≥ 25 years | 27 | 71.1 |
Partnered | 21 | 55.3 |
Unpartnered | 17 | 44.7 |
Primary school | 24 | 63.2 |
High school | 11 | 28.9 |
Tertiary qualification | 3 | 7.9 |
Black African | 8 | 21.1 |
Coloured |
30 | 78.9 |
Islam | 8 | 21.1 |
Christianity | 30 | 78.9 |
English | 14 | 36.8 |
Afrikaans | 16 | 42.1 |
isiXhosa | 3 | 7.9 |
Other indigenous South African languages | 3 | 7.9 |
Other languages | 2 | 5.3 |
Unemployed | 17 | 44.7 |
Unemployed by choice (student, homemaker) | 8 | 21.1 |
Employed (full-time or part-time) | 13 | 34.2 |
None received | 17 | 44.7 |
Childcare grant recipient | 21 | 55.3 |
< R1000/month (< ±US$74) | 22 | 57.9 |
R1000 – R5000/month (±US$74 – US$370) | 13 | 34.2 |
R5000 – R10000/month (±US$370 – US$740) | 3 | 7.9 |
, The authors are cognizant of the deeply and historically problematic use of this classification. However, given South Africa’s troubled socio-political history, the use of these markers allows for the monitoring of improvements in health and socio-economic disparities that originated within such a classification system.
Addressing the study’s first aim, which sought to investigate women’s preliminary responses to the intervention, several significant gains were seen on both primary and secondary outcome measures, as reflected in
Participant pre-post differences in outcome measures imputing for missing data (
Outcomes | Pre-intervention |
Post-intervention |
Comparison |
|||||||
---|---|---|---|---|---|---|---|---|---|---|
Mean | s.d. | Mean | s.d. | Mean diff | s.e. | Correlation | Hedges’ |
Cohen’s |
||
19.4 | 3.6 | 16.2 | 6.3 | 3.2 | 1.1 | 0.222 | 0.60 | 0.61 | < 0.01 | |
14.9 | 3.6 | 11.3 | 6.1 | 6.3 | 1.2 | 0.484 | 0.67 | 0.68 | < 0.01 | |
Work | 5.9 | 3.2 | 4.5 | 3.5 | 2.5 | 0.9 | 0.44 | 0.41 | 0.42 | 0.02 |
Social life | 7.2 | 3.1 | 4.6 | 4.1 | 6.8 | 0.5 | 0.386 | 0.69 | 0.71 | < 0.01 |
Family life | 6.5 | 3.5 | 4.9 | 3.9 | 5.6 | 1.0 | 0.456 | 0.42 | 0.43 | 0.01 |
30.68 | 5.5 | 26.0 | 8.7 | 9.1 | 2.0 | 0.306 | 0.62 | 0.63 | < 0.01 | |
Tobacco involvement score | 14.0 | 14.2 | 14.5 | 13.2 | 0.7 | 1.6 | 0.912 | 0.04 | 0.04 | 0.66 |
Alcohol involvement score | 6.7 | 9.7 | 6.9 | 10.8 | −0.3 | 2.0 | 0.763 | 0.02 | 0.02 | 0.87 |
Significant other | 21.3 | 6.8 | 22.3 | 6.4 | −1.8 | 1.1 | 0.836 | 0.15 | 0.15 | 0.11 |
Family | 15.5 | 8.1 | 16.1 | 8.6 | −1.1 | 1.9 | 0.709 | 0.07 | 0.07 | 0.56 |
Friends | 15.9 | 8.3 | 16.8 | 8.1 | −1.7 | 2.2 | 0.594 | 0.11 | 0.11 | 0.47 |
Overall social support | 52.7 | 17.7 | 55.2 | 19.2 | −4.6 | 3.5 | 0.789 | 0.13 | 0.13 | 0.21 |
CMD, common mental disorders; s.d., standard deviation; s.e., standard error.
Data from interviews with women who participated in the intervention (participants) and staff members involved in the delivery of the project (stakeholders) highlighted several emergent themes, addressing the study’s second and third aims regarding the intervention’s feasibility and acceptability.
Most of the participants felt that they derived some benefit from the intervention, with nearly all participants reporting that they would recommend such a programme to friends. The opportunity to hear another perspective, to talk about past experiences or to have time for themselves was critical. The opportunity to confide in a non-judgmental person, who was not previously known, was deemed particularly helpful:
‘I found out I was pregnant. I didn’t want the baby, all of that and – it was so painful for me but – really after talking to her it – just to speak to someone that’s not family, man, someone you don’t know, really helped, it’s almost like it’s just a burden off your shoulder. Since the first session we had, I could see a light again and I could actually feel that this is my baby …’ (Participant #26, aged 28)
From these descriptions, PST-specific factors of the intervention may be less important to some participants than having the registered counsellor’s impartial and empathic ear. Other participants valued the problem-solving approach itself. Some participants expressed appreciation for its pragmatism, whilst others referred directly to the problem-solving aspects of the intervention as useful, such as the development of better coping mechanisms:
‘I learned to control the problems I have and how to solve it and what to do and so on. That is how it helped me.’ (Participant #24, aged 23)
‘How she taught me how to sit and think and not allow my thoughts to run through my mind but to allow it to let it stop and so on. Yes, think more about the positive things.’ (Participant #24, aged 23)
Reports on improved self-efficacy that in turn led to more positive feelings about themselves, such as ‘being stronger’ and feeling like ‘a better person’, were described. For some participants, these benefits were linked to improved relationships, whilst others found that they were more able to seek out social experiences than before:
‘… [
A few participants spoke about using what they had learnt from the intervention to help others, going so far as to make copies of the booklet for friends. One participant even arranged to meet with her friends at the time that her weekly appointment with the registered counsellor had been due, to teach them the PST techniques.
On the other hand, two participants reported that they did not find the intervention helpful at all (one of them still attended all three sessions). Whilst both seemed to suggest that their objections were related to the registered counsellor’s style, it is possible that the PST model was a poor fit for them, as evident from the following comment: ‘there wasn’t really space for me to talk, we were just reading out of the booklet’. Despite one participant’s appreciation for the ways in which the model helped her better manage distressing thoughts, she also made an important observation about the limits of counselling interventions for someone who lives in poverty. In this instance, this participant highlighted the tension between the need to think about how to find money and the anxiety and stress that these thoughts generated for her:
‘I can think because I’ve got children, you see, I’ve got no job, I must make something for my children, but everything it’s use the money, you must have the money, you just think and keep thinking “where must I get the money? Where must I do that?” But the only thing is to stop thinking. I can’t stop thinking. I must think.’ (Participant #18, aged 29)
From the stakeholders’ perspective, most seemed to feel that the programme lightened their own workloads, as it gave them a resource to refer distressed patients to, instead of having to manage the patients themselves. Several stakeholders described how interactions with distressed patients could be burdensome and stressful for the staff, in that containing the patient took time and energy from their own limited resources, as highlighted in the following extract:
‘And I’m just asking [
All the stakeholders talked about the ways in which having a counsellor at the MOU relieved some of the demands that distressed patients represented. This seemed to be the most significant and meaningful contribution that the programme made to staff:
‘I think how it impacted on my work is that a lot of the clients that was actually screened to be seen by [
Participants who had missed appointments with the registered counsellor or had prematurely terminated their participation provided several reasons for doing so. Although a few participants highlighted how stigma associated with attending counselling sessions at the MOU might be a barrier, such as ‘people will think something is wrong with you, structural barriers, mainly related to financial constraints and lack of transport, were the most commonly cited. Childcare and work commitments also presented obstacles to attending sessions:
‘Okay all her available times, then I was busy – it was either work or I had to be by my child’s school.’ (Participant #24, aged 23)
The stakeholders highlighted several challenges associated with the programme, many of which appeared to be related to an overburdened system. As referrals increased, the registered counsellor had less time to immediately see all patients referred to her and in some instances she would need to arrange an appointment on another day instead. When asked about what she thought was problematic about the programme, one stakeholder had this to say:
‘The amount of patients that was referred … Because I don’t think [
The role that the programme played in relieving the staff members seemed to be echoed in this way and perhaps highlighted a sense of inadequacy or anxiety about having the capacity to manage a distressed patient. One stakeholder alluded to this:
‘But even those patients really need help so if – I mean [
Every stakeholder made reference to the overburdened and understaffed state of the system and the consequent demands placed on staff members. Interestingly, whilst the value of the programme for the staff appeared to be in the relief it offered from the demands of distressed patients, feeling overburdened may have, in and of itself, represented a barrier to the acceptability of the programme. In this vein, two stakeholders expressed frustration at their colleagues for their unwillingness to participate in projects in general: ‘I think people work in little squares and they’re just concerned about what happens in their little square.
‘I know, it also has to do with the lack of staff and the amount of clients that need to be seen. So I also understand that perhaps people feel overwhelmed and so [
The registered counsellor also reported that the intake nurses felt that the screening and referral process was burdensome:
‘Those that were involved – let’s just say the intake nurses found it to be a “las” (burden) – it’s extra work for them, [
The lack of space at the MOU was another obstacle to the acceptability of the programme, as stated by one stakeholder: ‘The only thing here is the space thing you understand?’ However, one stakeholder felt that the lack of space was sometimes used as a reason to prevent new programmes from being adopted, as new programmes often represent additional work. In this way, physical space may well have represented staff members’ capacity – in terms of time and energy – to accommodate the additional duties that programmes often bring with them:
‘It’s difficult because they agree to a lot of things and then when it needs to happen, there’s no space available, people don’t want to share their space.’ (Stakeholder #2)
Given that a majority of participants stated that the intervention was acceptable the way that it was, few recommendations for improvements were made. Of those who provided recommendations, many stated that group sessions would be beneficial in providing support and that the intervention should be made available at other locations. This recommendation was to address the practical difficulties in accessing the clinic, or to protect participants from the stigma associated with receiving mental health services:
‘Ja [
Several participants stated that the number of sessions was inadequate and that more or longer sessions would improve the intervention:
‘Longer sessions because you just deal with this and now you get to, not a breaking point but you know, you get to a point where you think, where you feel there’s still a lot for you to resolve but the sessions is too
Despite some ambivalence, all stakeholders stated that having a counsellor at the MOU was essential. Stakeholders stated that they needed someone who would attend specifically to patients’ mental health, worrying that in the meantime, as the programme has terminated, staff might not detect problems:
‘We need our own counsellor. We are not picking up depression, we are not picking up postnatal depression because we’re not looking for it. As it is now there is a lot of people slipping through our hands that need help … And we are [
Three stakeholders reported that increasing both patient and staff awareness about mental health and counselling would improve the service and retention rates:
‘I think that if we can get to a point where we actually get people to understand that there’s more to wellness than just physical health um – we would have done a lot – maybe we can do a lot better then.’ (Stakeholder #2)
Two stakeholders stated that stigma associated with mental illness needed to be addressed in order to improve the programme. Both felt that being seen by other patients to use the service made the women feel self-conscious and therefore less inclined to take up counselling:
‘Maybe they don’t want to be seen – I think – when people don’t want to be seen – maybe they thought “this one knows me and they know I’m coming for this and I’m coming for that” – see them in a certain time or maybe give them appointments to come.’ (Stakeholder #4)
Despite the concern about space, one solution offered by several stakeholders to ensure that all patients are seen on the day was to have more counsellors available so that a walk-in service could be made possible. However, one stakeholder stated that a walk-in service would mean that counselling is treated as a ‘crisis service’. She opined that this would send the wrong message to women about taking care of their mental health:
‘I don’t know how effective that is also um – because it also creates the wrong perception with the client in terms of intervention and what can be done – and again in my opinion I think it would be better to say to people you know, mental health is a thing that you should pay attention to continuously and not only when you are in crisis or when there’s a problem.’ (Stakeholder #2)
This is the first study in South Africa to investigate the feasibility, acceptability and preliminary responses to an adapted PST intervention for psychologically distressed pregnant women. Quantitative data provide initial support for the potential benefits of the intervention for reducing symptoms of psychological distress and improving functioning. In line with findings from other studies,
Retention rates of almost 40% for the full intervention, whilst not high, appear to be in line with prepartum mental health interventions conducted in other settings.
For participants, the intervention’s acceptability seemed to lie primarily in the opportunity to talk confidentially to a non-judgemental and empathic person about their problems. Whilst the PST model seems to have had an influence on many participants’ thinking, this appears to have been a secondary benefit for some. This is consistent with evidence from studies to show that task-sharing PST interventions are generally acceptable and feasible to intervention participants in primary care.
For stakeholders, the programme was generally perceived as expanding and improving the quality of services provided by the facility. Having a professional resource to refer to seemed to relieve them of the pressures of managing distressed patients during the course of routine care. Mental health problems appeared to add to the burden of care experienced by MOU staff who reported not having the time, capacity or skills to manage psychologically distressed patients. In this respect, the intervention was widely deemed to be acceptable by stakeholders. To our knowledge, other studies have not found this. However, the overburdened state of primary healthcare systems might in and of itself represent a barrier to the successful integration of programmes that rely on staff members’ participation. Similar South African studies have shown that stakeholders experience the inclusion of new interventions into usual care as generating additional burden, and that staff buy-in is central to the success of programmes.
There are some limitations of this study. The main limitations are the small sample size and the absence of a control group, restricting our ability to comment on the effect of the intervention. Furthermore, it is possible that participants might have experienced spontaneous remission of symptoms and the study’s positive outcomes simply reflect that. However, these findings suggest that a scaled-up randomised controlled trial of a task-sharing PST intervention to reduce psychological distress amongst pregnant women might have positive outcomes. In addition, both women who participated in the intervention and stakeholders involved in its delivery generally found value in the programme.
Despite the study’s limitations, in combination with the qualitative data, the outcome data from this study support the feasibility and acceptability of this task shifting brief intervention as well as its potential to effect positive outcomes in the treatment of prepartum psychological distress and CMD symptoms. Perhaps most significantly, the results of this study suggest that integrating mental healthcare interventions into primary care services may improve the mental health of services users, in addition to reducing the burden that patient’s psychological distress may represent for healthcare providers.
The authors have declared that no competing interests exist.
M.S. participated in the conceptualisation of the study, conducted the data collection, analysed the data and wrote the article. D.J.S. participated in the conceptualisation of the study, assisted with data analysis and made substantial contributions to the review and editing of the article. T.N. participated in the conceptualisation of the study and made substantial contributions to the review and editing of the article. B.M. assisted with data analysis and made substantial contributions to the review and editing of the article. P.C. assisted in the development of the intervention protocol and data analysis and made substantial contributions to the review and editing of the article. K.R.S. participated in the conceptualisation of the study, assisted with data analysis and made substantial contributions to the review and editing of the article.
Maxine Spedding is supported by the National Health Scholarship Programme of the Medical Research Council of South Africa and the Hendrik Vrouwes Scholarship. Dan Stein is supported by the South African Medical Research Council.
Data for the study will be made available upon written request, with the permission of all authors.
The views and opinions expressed in this article are those of the authors and do not necessarily reflect the official policy or position of any affiliated agency of the authors.
The authors recognise the deeply and historically problematic nature of this; however, the capacity to monitor developments in health and socio-economic disparities, which originated from such a classification system, is made possible by the continued use of these markers in South Africa
The term ‘participants’ is used to denote women who received the PST intervention, whilst the term ‘stakeholders’ represents staff who were involved in the delivery of the intervention.