Role of trust in sustaining provision and uptake of maternal and child healthcare: Evidence from a national programme in Nigeria

Despite increasing attention to implementation research in global health, evidence from low- and middle-income countries (LMICs) using realist evaluations, in understanding how complex health programmes work remains limited. This paper contributes to bridging this knowledge gap by reporting how, why and in what circumstances, the implementation and subsequent termination of a maternal and child health programme affected the trust of service users and healthcare providers in Nigeria. Key documents were reviewed, and initial programme theories of how context triggers mechanisms to produce intended and unintended outcomes were developed. These were tested, consolidated and refined through iterative cycles of data collection and analysis. Testing and validation of the trust theory utilized eight in-depth interviews with health workers, four focus group discussions with service users and a household survey of 713 pregnant women and analysed retroductively. The conceptual framework adopted Hurley’s perspective on ‘decision to trust’ and Straten et al.‘s framework on public trust and social capital theory. Incentives offered by the programme triggered confidence and satisfaction among service users, contributing to their trust in healthcare providers, increased service uptake, motivated healthcare providers to have a positive attitude to work, and facilitated their trust in the health system. Termination of the programme led to most service users’ dissatisfaction, and distrust reflected in the reduction in utilization of MCH services, increased staff workloads leading to their decreased performance although residual trust remained. Understanding the role of trust in a programme’s short and long-term outcomes can help policymakers and other key actors in the planning and implementation of sustainable and effective health programmes. We call for more theory-driven approaches such as realist evaluation to advance understanding of the implementation of health programmes in LMICs.

This paper report results from a component of a broader mixed-methods theory-driven study (Mirzoev et al., 2016), that adopted a RE approach to examine the relationships between contexts, mechanisms and outcomes of community health workers (CHW) programme in Anambra State, Nigeria. It complements our previous reports of focused theoretical lessons (Ebenso et al., 2019; Mirzoev et al., 2020) and empirical results on advocacy for maternal and child health in Nigeria (Okeke et al., 2021; Uzochukwu et al., 2020), security in the provision and utilization of maternal healthcare (Etiaba et al., 2020) and health worker motivation (Ebenso et al., 2020). RE is a theory-driven approach that guides the implementation of complex interventions through iterative theory development, testing and refinement (Pawson and Tilley, 1997; Robert et al., 2012; Wilson and McCormack, 2006; Wong et al., 2017). Programme theories developed within realist studies explore which contexts trigger which mechanisms that produce intended or unintended outcomes in different contexts. This enables a clear understanding of the ‘whys’ and the ‘hows’ of programme outcomes within a particular context that is well suited to evaluating programmes implemented at diverse levels of the health system investigations in low-resource settings (Marchal et al., 2012). In RE, data extraction proceeds from baseline enquiries and development of programme theory to testing/refinement and consolidation of the programme theory, using empirical data (Dalkin et al., 2015). The study was conducted in three phases, corresponding to the building of initial programme theories (IPTs), testing/validation and consolidation/refinement (Manzano, 2016; Pawson and Manzano-Santaella, 2012; Pawson and Tilley, 1997) (see Table 1 in supplementary file). In phase 1, we reviewed key SURE-P/MCH programme documents and relevant MCH Federal and state policies, between June and September 2015, to understand the programme architecture and design (Ebenso et al., 2019). Initial qualitative interviews were held (May–November 2016) with purposefully identified 96 stakeholders comprising in-depth interviews (IDIs) with 10 policymakers, 11 programme officers, 16 health workers/PHC staff, and 15 facility managers at federal and state levels. Focus Group discussions (FGDs) were held with 8 Village Health Workers (VHWs), 12 WDCs, 12 service users (pregnant women) and 12 family members of service users. Different numbers of interviewees reflect the three phases of our research and different engagements with the Trust theory. The selection of all these different respondents was done to explore their views and experiences of the SURE-P/MCH programme. These data facilitated the development of 8 programme theories for the main study (Mirzoev et al., 2016). In phase 2, which entailed data testing and validation of trust theories, qualitative and quantitative methods were utilized. Data were collected in 12 PHCs and three general hospitals purposively selected to reflect the implementation of the SURE-P/MCH programme in Anambra state, Nigeria. These facilities were clustered into three, each cluster comprised one general hospital and four PHCs. The focus on the clusters reflects the setup of the SURE-P/MCH programme. Two of the clusters benefitted from the SURE-P/MCH intervention, while the third cluster was used as a control. This was relevant to enable us to determine if there were any differences in MCH service utilization in the clusters that benefited from the intervention compared to the control cluster. For the qualitative methods, 8 IDIs with health workers and 4 FGDs with service users were conducted (August–September 2018). The IDIs and FGDs were guided by a semi-structured question guide designed around the different versions of initial programme theories and included questions for testing and validating the different components of the programme theories for the main study. The FGD interviews were conducted in the Igbo language, while the IDIs were conducted in both Igbo and English languages depending on respondents’ preference. All IDIs and FGDs were conducted face-to-face and were audio-recorded with respondents’ consent, transcribed and translated into English as necessary. Female researchers (NE, (Sociologist), UE, (Health Economist), UO and EE (Medical doctors) trained in realist interviewing undertook the data collection while NE, UE, UO, TE (male Health economist), EE and AM (female Sociologist) were involved in data analysis. Qualitative data collection was complemented by data from a quantitative household survey. The survey was based on a community listing of all households in the project cluster areas that had a birth in the last 6 years; covering a period before, during and after the SURE-P/MCH programme. A stratified random sample of 713 women was selected for quantitative interviews across the three project areas. A questionnaire was administered, which collected information on maternal health-seeking behaviour to the care given and socioeconomic information on the household between May and June 2018. In phase 3, we refined and modelled the complex relations between the actors, contexts, intervention processes and mechanisms, and its outcomes (December 2018). Using the Context-Mechanism-Outcome (C-M-O) configuration, we examined the emerging data on trust to make inferences about the relationships between contexts, mechanisms and outcomes (Fig. 1). We examined the quantitative data critically to explore the effect of the intervention on various sub-groups of women and to identify sub-period variation in outcome relative to the period before, during and after the withdrawal of the programme. Patterns across data sets were identified by accumulation (the same factor was present within a set and across sets) and causal relationships were established with further support of the theoretical literature and our qualitative data set. This enabled us to refine and consolidate our programme theory on trust which states as follow: In the context of improved staff attitude, upgraded health facilities and functioning WDCs achieved during the implementation of the SURE-P/MCH programme, pregnant women who receive sustained financial and non-financial incentives to use MCH services (Context), are likely to develop and maintain a sense of improved trust (including confidence and satisfaction) with health facilities and staff (Mechanism), ultimately leading to the improved likelihood of repeated and regular utilization of MCH services from these health facilities (Outcome). Fig. 1: CMO template visualizing the causal linkages among contexts (Cs), Mechanisms (Ms) and Outcomes (Os) (See Fig. 1 in supplementary file). To explain how trust works, we drew upon Hurley’s (2006) perspective on the ‘decision to trust’ and Straten et al. (2002) framework of factors that influence public trust in healthcare systems. We also utilized social capital theory (Agampodi et al., 2015; Bourdieu, 1986; De Silva et al., 2007). in our interpretation of the sustainability of trust during the SURE-P/MCH Programme and the existence of residual trust by service users after the withdrawal of the programme. Elements of Hurley’s (2006) perspective relevant to explaining trust in LMIC context include security; the number of similarities between the trustee and truster; if the trustee shows benevolent concerns, trustee’s capability to do their work. Straten et al. (2002) provide a useful framework in the explanation of factors influencing public trust at micro, meso and macro levels. They specify that at the micro-level, people are more concerned about the behaviour of the healthcare providers, whether they will listen to them and handle their problems appropriately. At the meso level, peoples’ concern is whether the health providers are cooperating among themselves; at the macro level, people are worried about impacts of interventions accompanying the development process in the society on their access to, as well as the quality of healthcare. This framework is relevant in the analysis of trust in LMICs. Peters and Youssef (2016) indicate that at the micro-level, the interaction between the doctor and patient can become more effective based on trust and consequently will enhance the patient’s satisfaction and compliance with treatment while at the macro level, the importance of trust is expressed in the impact it makes in society by influencing efforts being made to meet societal expectations. Studies on trust in Nigeria (Amuta-Onukagha et al., 2017, Fagbamigbe and Idemudia, 2015; Ugboaja et al., 2018), reinforce the relevance of articulating factors at the micro, meso, and macro levels in the analysis of trust in health systems. Social capital is conceptualized regarding entitlements to resources including information, financial benefits, favours and services individuals get through membership to a community and participation in networks. It also implies an expectation of reciprocity among members of the networks (Bourdieu, 1986; De Silva et al., 2007). Hence social capital is perceived as ‘tangible’ and ‘intangible’ resources that members of a group have access to on account of their membership to the group (De Silva et al., 2007). Three types of ties namely ‘bonding’, ‘bridging’, and ‘linking’ social capital are identified. Bonding social capital refers to relationships of trust and cooperation, with strong ties among people who have shared identity such as ethnicity, social class, age and place of residence. The bonding ties serve as means through which individuals seek help and support from members of the network (Erickson, 2011). Bridging social capital derives from respect and mutual relationships in networks that are not homogeneous (Erickson, 2011; Putnam, 2000), while linking social capital, on the other hand, refers to “vertical” ties existing among people who belong to different levels of power in the society (Erickson, 2011). Three dimensions of social capital are structural, cognitive and relational social capital. Structural capital refers to the existence of social networks through which people have access to resources, people, roles, rules and procedures (Bourdieu, 1986). Cognitive social capital refers to people’s perceptions and interpretations of the shared relationships in the networks. The relational social capital is concerned with the nature of personal relationships existing among people through interactions in the social system as well as feelings of trust in the network (Claridge, 2018, Harpham and Grant, 2002). Although the social capital theory has its origin in developed countries, it has been successfully used in the analysis of health behaviour and outcomes in some LMICs including Nigeria (Agampodi et al., 2017; Lau et al., 2020, Ozawa and Walker, 2011; Semali et al., 2015; Ware et al., 2009). The social capital theory demonstrates that health outcomes are dependent on income inequality levels with a greater impact on communities where inequality is higher and safety nets lower (Rodgers et al., 2019; Vincens et al., 2018) in most LMICs. Therefore, social capital plays a vital role in increasing the levels of trust in the analysis of health behaviour and outcomes in some LMICs and is relevant to our study in Nigeria.

• ezumah, nkoli
• manzano, ana
• ezenwaka, uchenna
• obi, uche
• ensor, tim
• etiaba, enyi
• onwujekwe, obinna
• ebenso, bassey
• uzochukwu, benjamin
• huss, reinhard
• mirzoev, tolib