Background
In the context of HIV service delivery, mentor mothers (MMs), also known as expert mothers, are HIV-positive women with first-hand experience as exemplary clients in the prevention of mother-to-child transmission of HIV (PMTCT) who provide experiential guidance for other women living with HIV [
1‐
5]. Beyond their personal experience, MMs often receive training to enhance their peer support services and work alongside formal healthcare workers (HCWs) [
1,
4,
5]. In this regard, MMs can be considered lay health workers: lay people who have been trained for short periods to assist formal HCWs and take over certain tasks [
6‐
9]. MMs and similar lay HIV health workers often do not have specific qualifications other than being persons living with HIV [
4,
10‐
12]. They work in health facilities, in clients’ homes, and in the larger community and ultimately act as a link between health facilities and communities [
4,
6,
10,
11].
Depending on the setting, MM roles at health facilities include HIV testing, pre- and post-test counseling, enrolling clients into PMTCT/HIV care, booking client appointments, assisting in drug dispensing and adherence counseling, and tracking clients who have missed appointments or dropped out of care [
12‐
16]. These MM roles are played largely in the framework of task shifting, which the World Health Organization (WHO) defines as “the rational redistribution of tasks among health workforce teams” [
17]. In HIV care, task shifting may occur from doctors to nurses or other formal HCW cadres [
18,
19] and from doctors, nurses, and other formal HCW cadres to lay health workers [
4,
11,
19,
20]. The professional-to-layperson model of task shifting has been formally or informally adopted to various degrees in several sub-Saharan African countries to facilitate scale-up of HIV (including PMTCT) services, reach and retain clients, reduce disease burden, and ultimately improve treatment and prevention outcomes [
4,
10,
11,
19,
20]. Task shifting is particularly helpful in low-resource settings where there is a shortage of human resources for health with concomitant high burden of disease.
At 3.2 million, Nigeria has the second largest population of people living with HIV globally, after South Africa [
21]. Furthermore, Nigeria has a large PMTCT burden, along with wide program gaps: only 30% of approximately 200 000 HIV-positive and pregnant Nigerian women receive antiretroviral drugs annually, and only 9% of HIV-exposed infants receive timely early infant diagnosis testing [
22,
23]. In Nigeria, structured MM peer support has been shown to improve maternal retention and viral suppression [
24] as well as timely infant presentation for HIV testing [
25]. Similar findings on the positive impact of peer support on PMTCT outcomes have been reported from other African countries [
1,
26‐
30].
Despite the well-established benefits of peer support in PMTCT, many high-burden countries—including Nigeria—have not formally adopted these interventions at the national level. For example, in 2005, Nigeria introduced expanded roles for lay voluntary workers including people living with HIV as expert patients in pilot projects such as the Integrated Management of Adult and Adolescent Illness [
31]; this was however not fully implemented nationwide. Both the 2009 national decentralization program for HIV treatment scale-up to primary healthcare centers and the 2014 national task shifting/task sharing guidelines [
32] formalized policies to only support expanded roles of non-physician health workers already in the Nigerian civil service structure (e.g. nurses, midwives, community health officers, community health extension workers and pharmacy attendants). MMs and other HIV-positive treatment supporters are not part of the current civil service structure and are typically supported by externally funded implementing partners.
Furthermore, challenges exist in terms of defining MMs’ roles/niche in PMTCT/HIV programs in particular and the formal health system in general [
7,
9,
13,
14,
33]. Under these circumstances, professional relationships between MMs (as HIV-positive lay health workers) and the formal HCWs who supervise them may be complicated. Placed in this hierarchical environment, MMs, with often poorly defined roles, little or no education or professional credentials, low wage-earning capacity, and known to be HIV-positive, may be highly vulnerable to stigma, discrimination, marginalization, non-supportive supervision, or other negative experiences at the healthcare facility [
10,
12‐
14,
34].
The nature of MM’s working conditions, especially HCW-MM working relationships, have so far not been well-characterized in Nigeria, a country which stands to gain significantly from the scale-up of peer support interventions in its challenging PMTCT program. This paper explores the nature of the working environment for MMs at primary healthcare centers in rural North-Central Nigeria. Specifically, we aim to describe, from the perspective of MMs, how interactions with healthcare workers shape MMs’ working conditions and influence their performance.
Methods
Study design
This qualitative study was nested within a larger PMTCT implementation research project, the MoMent (
Mother
Mentor) study, in North-Central Nigeria. MoMent was a prospective cohort study that compared a standardized, closely supervised MM program with the less-structured, less-supervised routine MM program at primary healthcare centers in rural North-Central Nigeria [
35]. Main outcomes included postpartum maternal retention and viral suppression and timely uptake of early infant diagnosis [
35]. This article draws its findings from focus group discussions (FGDs) conducted towards the end of the prospective study follow-up, to capture the experiences and opinions of all MoMent MMs (intervention and control) regarding their roles and working conditions during study implementation.
Study setting and population
This study was conducted in rural communities of the Federal Capital Territory and Nasarawa State in North-Central Nigeria. Study participants were MMs engaged at all 20 (10 intervention and 10 control) primary healthcare centers that served as MoMent study sites. Table
1 compares training, supervision and scope of work for all MoMent MMs working in both intervention and control arms [
24,
35]. MoMent MMs in both arms were chosen from communities surrounding the primary healthcare centers they were assigned to. All of these women had completed the PMTCT cascade at least once and were expected to guide other women in navigating and being compliant with PMTCT services. MMs in both arms were expected to work at both facility and community-level and were provided the same stipend amount. The major differences between the two MM groups were in the supervision and structure built into the intervention arm: intervention MMs received baseline training via a standard curriculum with daily, hands-on supportive supervision from a study-designated MM supervisor. Furthermore, all intervention MMs utilized standardized logbooks for documenting client calls and visits. Random quarterly performance audits were conducted via client feedback, in order to improve and/or maintain MM work performance in the intervention arm [
36].
Table 1
MoMent peer support program description, roles, and responsibilities
Program description and requirements |
HIV-positive woman | Yes | Yes |
Experiential PMTCT knowledge | Yes | Yes |
Standard pre-engagement selection criteriaa | No | Yes |
Standard baseline training and curriculum with certificationb | No | Yes |
Stipend for work activities | Yes (~ 50 USD monthly) | Yes (~ 50 USD monthly) |
Schedule for client interactionc | No | Yes |
Scope of client-related work formally communicatedd | No | Yes |
Facility or community-based activities | Both | Both |
Linked to HIV+ clients at ANC clinic | Yes | Yes |
Activity documentation | Weak, not standard, not consistently linked to PMTCT outcomes, often not reviewed | Emphasized, standard logbook provided, PMTCT outcomes-specific, reviewed by MM supervisor |
Supportive supervision | Weak support and supervision, responsibility of facility staff-in-charge | Emphasized, responsibility of supportive MM supervisor who reports to staff-in-charge |
Audits of client interaction activitiese | No | Yes |
Formal performance evaluations | No | Yes, based on client feedback |
Re-orientation or disengagement based on performance | No | Yes |
Peer support activities |
Document client interactions | Yes, inconsistent (in non-standardized notebooks) | Yes, required (in standardized, outcomes-specific logbook) |
Conduct HIV tests for ANC clinic clients | No | No |
Immediately flag clients with missed appointments | No | Yes |
Track clients with missed appointments | Yes, after several weeks | Yes, required within 3 days |
Daily feedback sessions with supervisor | No | Yes |
Participant recruitment
Over its 5-year implementation period, MoMent engaged a total of 38 MMs across both intervention (structured peer support) and control (unstructured peer support) sites. The number of MMs assigned per site was guided by a ratio of 1:10–15 between MMs and pregnant or postpartum clients (up to 18–24 months post-delivery) [
35]. Ultimately, intervention arm MMs had an average ratio of 1:12 while control MMs averaged 1:14 clients [
24].
All MoMent ever-engaged MMs were eligible to participate in the MM FGDs. Towards the end of the study, all 38 MMs (regardless of whether still actively engaged or not) were contacted by telephone, by research officers who had been stationed at each MoMent site during the study. MMs were not contacted or recruited by healthcare workers for these FGDs. The research officers briefed each of the 38 MMs about the FGDs; possible dates and the MMs’ availability on those dates were discussed. A total of 36 out of the 38 MMs indicated their interest and availability to participate in the FGDs. These 36 interested and available MoMent MMs were provided information on the date and location of their specific FGD. Two MMs, both from the control arm, were interested; however, they were unable to participate in the FGDs on the scheduled dates—one was recovering from surgery while the other had to travel out of town.
All 36 successfully recruited MMs presented for the FGDs, which were conducted on non-clinic days in private rooms at study primary healthcare centers within the communities where the MMs worked, or at a private venue within their work catchment area. This was done to preserve confidentiality and to encourage discussions on this topic without fear of victimization by facility healthcare workers. Written informed consent was obtained from all MMs before the FGDs. Snacks and transport reimbursement (depending on distance traveled) were provided on the day of the FGD for all participants. Healthcare workers neither participated in nor observed the FGDs.
Data collection
Seven FGDs (four among intervention MMs and three with control MMs) were conducted between September and November 2016, which marked the end of the 6-month postpartum follow-up for all participants for the prospective study’s primary outcomes. Each of the seven FGDs conducted had four to six MM participants, ultimately representing all 20 MoMent sites.
Prior to each FGD, an interviewer-administered form was used to capture participants’ socio-demographic information including educational attainment, marital status, religious affiliation, parity and duration of engagement as MMs.
Each discussion was audio-recorded and guided by a trained bilingual (English and Hausa) facilitator, with or without a co-facilitator, and an observer. During each FGD, at least one observer took notes on non-verbal cues, which were used to assist in data analysis and interpretation. All facilitators and observers were study staff familiar to participants, due to their interactions with these MMs at study sites during MoMent data collection. All study FGD staff had a minimum of 2 years working experience with the MoMent study and had had at least 1 year experience in conducting qualitative interviews. None of the FGD facilitators or observers worked as facility staff nor had any supervisory role over MMs participating in their respective FGDs.
The FGD guide explored MMs’ opinions on their workload and stipends, terms of engagement, scope of work, and relationships with healthcare workers. Each FGD lasted for 60–90 min.
Since the number of FGDs to be conducted was limited by the specific number of MoMent MMs engaged and available, data saturation was not a consideration. However, after transcription and analysis of the initial seven FGDs, we conducted a “member check” to gain participant feedback on the initial findings and to validate the collected data and its interpretation [
37]. A cross-section (
n = 4) of the 36 original participants were recruited for the member check; these participants were selected in order to equitably represent religion, high and low levels of education, and intervention and control arms. The member check was performed in September 2017 as a group discussion, where the key findings from qualitative data analysis were presented to participants for confirmation, correction, and additional commentary. Ultimately, the member check was in agreement with initial findings. The member check was conducted by the same facilitators and observers who implemented the initial FGDs.
Data transcription and analysis
Audio-recorded FGDs were transcribed verbatim in English or transcribed and translated from Hausa to English were necessary. Manual transcription and analysis were performed by the same facilitators and observers who conducted the FGDs. For data analysis, we adopted the constant comparative method in a grounded theory approach [
38]. In this approach, inductive methodology is used to systematically generate theory from the data collected. We selected a series of code words to develop themes and sub-themes from the qualitative data. Preset codes were related to the general themes in our FGD guides and served as the root of our coding tree. The root pre-set code words for our coding tree were “work relationships,” “stigma/discrimination,” “working conditions and pay,” and “roles/responsibilities.” Eight paired analysts independently coded and analyzed the data. This was followed by group review, triangulation, and content analysis by iteration until a final consensus on patterns and categorizations was achieved. The research consultant (AO) additionally independently analyzed and coded data with Nvivo 11 (QSR International, Victoria, Australia) and compared the findings to emerging themes identified by the paired researchers.
Discussion
Our study provides detailed insight into the working conditions of mentor mothers and their professional relationships with healthcare workers in North-Central Nigeria. We have highlighted a mixture of positive and negative examples that on one hand demonstrate supportive working environments in some instances; however, there were also other instances where MMs’ working environments were less than conducive, largely due to tenuous relationships with HCWs. Some of the issues emerging were stigma and discrimination on the basis of MMs’ HIV-positive and non-formal work status, unclear scope of work at the facility level, and assignment of non-relevant tasks by HCWs.
Our findings support those of prior studies that have reported issues with lack of recognition, complementarity, and integration of lay HIV health worker roles vis-à-vis HCW roles in sub-Saharan Africa [
4,
10,
11,
13]. The non-integrated, poorly structured programs in which MMs and similar lay health workers operate may actually limit their impact in the roles for which they were engaged. In our study, MMs discussed being alienated by HCWs on the basis of their non-formal work status. For example, the lack of training certificates and identity cards made MM vulnerable to dismissive treatment by some HCWs. Given the continued threat of vertical transmission to the HIV/AIDS elimination agenda [
22], the opportunity to capitalize on the gains from maternal peer support in PMTCT cannot be taken for granted.
A poorly defined scope of work—especially at the facility level—was a major complaint from MMs in our study. Vagueness in job descriptions for lay health workers in HIV and PMTCT have been reported among MMs as well as other lay health workers in HIV programs [
4,
11,
14,
16]. The non-formal work status of MMs and other lay health workers in HIV may contribute to the persistence of poorly defined scopes of work. This may also perpetuate HCWs’ practice of assigning non-relevant tasks to MMs, which distract them from core duties. Similar experiences have been reported among HIV peer educators in Ghana [
12] and expert mothers in Malawi and Zimbabwe [
4]. However, as demonstrated by the MoMent study, providing structure can significantly improve the impact of lay peer support on maternal-infant outcomes in PMTCT [
24,
25]. However, MM scope of work at both facility and community-level must be well-defined, along with oriented input and buy-in from both experienced and new HCWs. The introduction of structure and standards can improve the quality and sustainability of peer support while harnessing the unique motivation of people living with HIV.
In addition to their non-formal work status, MMs’ HIV-positive status also factored into some HCWs’ attitudes towards to them. MMs reported experiencing HIV-related stigma and discrimination from the HCWs they were working with. Stigma and discrimination from HCWs towards clients is well-documented [
39‐
42], but our study additionally highlights that directed from HCWs towards MMs. Health systems that engage MMs and other people living with HIV should be aware of this and make provisions for HCW sensitization and advocacy/protections against workplace stigma and discrimination.
In our study, payment for services was noted to be critical to MMs’ motivation to work. While stipends were provided to all MoMent study MMs for the purpose of client home visits and phone calls [
4,
35], these funds were also used for MMs’ livelihood. Both paid and unpaid models of peer support have been implemented in HIV programs, all having different degrees of impact [
4,
11]; however, head-to-head comparisons of paid and unpaid peer support models within the same study setting are lacking. In their analysis of lay HIV health worker programs in sub-Saharan Africa, Herman et al. report that adequate remuneration in the setting of quality supervision and continuous training is critical for quality and sustainability [
10]. Cataldo et al. report similar findings from their synthesis of expert mother studies (including MoMent) in Malawi, Nigeria, and Zimbabwe, noting that adequate remuneration and training are likely to maximize the impact of these interventions in PMTCT [
4]. That said, instances, however rare, of MM stipend “garnishing” by HCWs are unacceptable and unethical, and avenues for reporting and addressing these phenomena need to be available to MMs. During the MoMent study, the opportunity for this type of corruption to occur was minimized by paying all MM through their bank accounts and not by cash (via HCWs/clinic administrators), creating a safe avenue for all MMs to report such cases with minimal retaliation, and within the routine PMTCT program, involving local chapters of the Network of People Living with HIV/AIDS in Nigeria in the MM program structure. These local chapters also serve as potential pathways for MMs to report issues at work that may be taken up to funding implementing partners and/or the local health authorities.
While we have discussed MM-HCW tensions in the workplace, it is noteworthy that MMs are engaged to complement and not supplant HCWs’ jobs. As other studies have reported, much of the tensions between HCWs and lay health workers stem from poorly defined lay worker roles and HCWs’ fear of their roles being usurped from “encroachment of territory” [
10,
12‐
14]. Thus, training and empowering MMs may work against them in their relationships with HCWs at the facility level. It is interesting to note that in our study, MMs mentioned little of tensions with HCWs or poorly defined scopes of work with regard to community-level MM activities. We suggest positioning MMs as clearly defined task shifting resources at the facility level, while protecting time for MMs to perform their community-level duties. Data on the costs and cost-effectiveness of PMTCT peer support programs have been encouraging [
43] and further support the call for their standardization, integration, and scale-up.
The potential impact of MoMent’s MM program structure on MM working conditions should also be mentioned. Part of the intervention package included supportive supervision from knowledgeable, PMTCT-trained staff who acted as advocates for their assigned MMs. While not reported here, intervention arm MMs have noted how collegiate support from their study-assigned MM supervisors made them feel valued and helped them cope with job-related stress [
4]. The supportive element of the supervision may very well have contributed to better MM client outcomes in the prospective MoMent study by way of higher-quality, more impactful MM counseling [
24].
Our study is limited in that only the views and experiences of MMs are presented here. Our approach to the MM FGDs was to gather information on their experiences and working conditions during the MoMent study. While we interviewed HCWs (among many stakeholders) to assess acceptability of MMs as part of the formative aspect of MoMent [
35,
44], we did not interview HCWs for the latter FGDs—they were limited to MMs only. Obtaining HCWs’ views on the issues explored here may have yielded additional perspectives on MM working conditions and roles. Additionally, exploration of community-level challenges faced by MMs can potentially fill in prevailing gaps in understanding their working conditions; this was not addressed in this paper. Lastly, our study was not designed to explore HCW-related experiences of MMs with shorter versus longer-term engagements; it is thus difficult to tell whether HCW-MM relationships have changed over time, for instance before MoMent and during/after MoMent. However, other reports published before this paper point to similar prevailing conditions for lay workers in HIV, albeit not in Nigeria. It appears not much has changed, likely because not much attention has been paid to developing and implementing solutions.
Acknowledgements
The authors would like to thank all the mentor mothers who volunteered to participate in this study and shared their experiences. We also appreciate the contributions of research staff who assisted with data collection, transcription, and analysis. Lastly, we thank the World Health Organization and Global Affairs Canada for providing financial support for this study.