Influences on policy-formulation, decision-making, organisation and management for maternal, newborn and child health in Bangladesh, Ethiopia, Malawi and Uganda: the roles and legitimacy of a multi-country network

Collective Decisions

Collective decisions, evidenced by a Memorandum of Understanding (MoU) or joint statement for example, should be artifacts of any multi-stakeholder network [15] and both bolster and reflect legitimacy of the network. Collective decisions were observed at the global level as the network emerged, where all partners had an agreement with the WHO-based QCN secretariat regarding the formation of the network. This suggests QCN legitimacy is linked to legitimacy of the WHO.

At the national level, partners of all case countries working for QCN implementation, including those in the private sector, have contracts or longstanding bilateral arrangements with the MoH and with each other. The prior performance of partner organisations in each country –their history of contributing to quality improvement efforts– influenced their contribution to collective decisions.

Implementing partners in each country usually co-produced knowledge, guidance and gave statements on key issues with the MoH but were keen for the MoH to be seen as taking the lead and were in support roles of the ministry’s strategic and operational direction.

“In new districts, at first we have one to one interaction with leadership where we give some overview. After that, we organized an inception meeting, all of the leaders attended that meeting, and through that process, we make them oriented as well as engaged with our activities.”

(Technical and Implementing Partner – National level – Bangladesh Round 3)

In Bangladesh, QIS and other partner organizations attended the follow-up meeting with WHO QCN secretariat together and sent a narrative report and working plan to the WHO QCN secretariat, which was a good example of joint work. Another example of a collective decision was observed in Ethiopia in identifying partners and selecting learning sites by MoH and WHO in country, at the initial stage of QCN.

Autonomous decision-making persists along with coordination and collective decisions. For instance, in Uganda, partners took the autonomous decision during the selection of sites for scaling up. Bangladesh also experienced independent decision-making (e.g. selecting scaling up areas) and influence on MoH (e.g. joining the network, running capacity building activities by NIPSOM) in decision-making by the partners. The partners in Malawi also seem to have a lot of autonomy concerning decisions of which activities they will support and where. Similarly, Ethiopia also has the experience of autonomous decision-making on site selection since the selection of facilities was made based on partners existing support or pre-existing support by another project.

Collaborative efforts

Collaborative efforts include co-organizing events, co-coordinating activities, or co-authoring publications. Partners across global, national and local levels, and especially key national level partners in all four case study countries displayed strong collaborations. Cohesive participation in developing the forthcoming National Health Quality Strategy (NHQS) by all key partners at the national level in Bangladesh was observed for instance, though USAID was steering the strategy and communicating with different directorates as well as the WHO global network.

Developing QCN roadmaps in Ethiopia, Malawi and Uganda were the result of collaboration between QCN partners and in Uganda, this collaboration at the national level increased after MoH’s leadership transition which also made clear roles and responsibilities of the different units. Many respondents expressed their realization of the importance and benefit of QCN setting-up opportunities for collaboration.

Co-organizing different events, workshops and training were common across all study countries though the events were led by different partners depending on the topics. For example, in Bangladesh, national events were mostly led by government; they led in agenda setting and decision-making, and development partners supported technically and/or financially, so they could push the government to organize such types of collaborative events. Such financial and technical support to proceedings provided development partners authority to strategically influence QCN, though, as in Bangladesh, government often set the agenda and made the decisions on what was implemented and by which organizations. This was not always the case though. Two perspectives were observed regarding funding and decision-making in Malawi - influence of development partners in implementation and conditional funding, which is experienced through UNICEF, and reliance on partners for direction.

“…..we are hoping that partners like UNICEF, GIZ would come and say; ‘okay, what will be our direction?’ This is because other than WHO, we need to engage other partners”

(Government - National level - Malawi Round 3)

Sharing resources

At the global level, BMGF and USAID were the primary funding partners. BMGF primarily supported through funding the WHO-based QCN secretariat and UNICEF for national implementation. Later, BMGF did not fund network activities beyond the global secretariat which shrunk the implementation activities in UNICEF funded countries and spaces though they overcame this quickly through alternative funds. For instance, in Bangladesh, BMGF was the primary funder for the Kurigram project, while they also worked in 5-6 other districts with funding from Global Affairs Canada and the UN Emergency Fund for the Rohingya in Cox’s

Bazaar. USAID funded QI efforts in-country via the MoMENTUM award, a global initiative covering 30 countries including Bangladesh and Malawi (but not Ethiopia as MoMENTUM award was not running there and Uganda because of their own system) that will continue some of the QCN activities after QCN funding stops in 2022-2023 [26], with funding going directly to implementing partners rather than MoHs. Involvement of most funding partners in all four case countries was either via the continuation of previous QI efforts or MNCH activities, or via alignment with QCN activities (e.g., Mamoni project in Bangladesh). QI was also incorporated into pre-existing work, for example, World Bank/GFF supported funding for QI activities in Uganda, but they didn’t really align themselves with QCN activities.

Funding or sharing resources at the national and local levels is catalytic. In addition to receiving direct budget support from donors, it was common for partners in all four case study countries to mobilise their own resources, though the amount of resources varied across countries. Since there was no specific budget for QCN in Ethiopia, partners were using their own budgets earmarked for other similar QI activities to prepare learning sessions and support facilities. Big partners like USAID, UNICEF, WHO, Save the Children also provided human resources support through providing direct funding to the government or through implementing partners or by themselves at the national and/or facility level in all countries.

“We provided funding to the government directly to hire additional human resources like officers for/located at the regional hospitals”.

(Technical and Implementing partner – National level - Uganda Round 1)

Except human resources support, partners also provided logistics and financial support and additionally, in some cases, support in the reconstruction and renovation of different facilities for MNH services across our four case countries. Pooling resources depended on necessity and on the ministry’s request and/or facility manager’s request. For example, in Ethiopia, rather than focusing solely on quality, all activities were considered such as project design, implementation planning, and so on and when a need developed, such as when the government requests assistance or when gaps exist, donors such as UNICEF helped to fill those gaps, either in kind or cash. In Bangladesh and Malawi, partners also directly provided resources (e.g., equipment) to learning sites. In Uganda, partners pool their resources at the national level for QCN activities but previous funding experiences and prevailing implementing partner working arrangements with MOH, including tight timelines and targets, led funders to prefer directly funding their own QI initiatives whilst ensuring alignment and reporting to MOH priorities.

Domestic resources have been used for QCN activities in Ethiopia and Bangladesh in MoH or development partner’s initiatives. In Ethiopia, MoH eventually shared some budget with the regions. However, having financial autonomy, some of the regions used the money for other purposes. On the other hand, one partner in Bangladesh succeeded to convince the National Institute of Local Government (NILG) to properly utilize their budget for MNH. Our meeting observation and interview showed that Save the Children succeeded to do it by engaging NLGI’s in different events, like, advocacy meetings, establishing functional linkage between health and family planning departments and NLGIs, engaging them in data-driven decentralized planning and regular communication and follow-up. This helped to develop ownership by local government and contributed to sustainability. Such type of devolved funding was not observed in Malawi or Uganda. This suggests that the organizational and political culture of the MoH –their willingness to allow devolution or flexibility of funding– influences allocation of resources for QCN activities and reflects the extent to which the local level of QCN implementation was seen as legitimate by governments in each country.

However, our interviews and observations indicate that resource allocation was not sufficient in most facilities to run QCN activities smoothly. In Uganda, there was a common theme that the global level of QCN did not consider the physical constraints or level of rapid resourcing needed from the government to be able to achieve the goals and meet the standards, particularly around experience of care. A similar claim regarding organizational and structural capacity, came from respondents in Bangladesh.

Normative Interaction

The commitment, norms and principles of all QCN actors of all countries overlapped with national goals and were consistent with previous works that have been discussed in the background section. However, all agreed that network activities gave them more impetus to act and be accountable at global level. All the actors sought synergy though they didn’t always experience synergy in terms of joint working on implementation and achievement of goals.

The governments of participating countries were leading the QCN through their commitment for improving quality of care and adaptation of the Network’s strategic objectives to their country contexts through growing the partnership with the different organizations. All case study countries also adopted and adapted the QoC standards considering their own country context, led by MoH or the responsible department of MoH [11]. Each country had exposure to QI activities through partners and mostly previous partners were working on network implementation. This was supposed to be beneficial, but this was not always the case. For instance, in Uganda, new coordination of actors was required as initial lack of clear coordination led to many participants reporting a lack of awareness or sense of cohesion. While participants at the national level were aware of and unified behind the QCN’s goals, there was often a sense that each actor was continuing to operate in its predetermined silo at sub-national level. Another challenge in design for Uganda’s in-country approach was that facility-level stakeholders, especially frontline workers were not fully oriented to the QCN separately; but rather some network activities were mainstreamed into other existing standard operating procedures, practice guidance or QI initiatives implemented by other implementing partners. A lack of a standardized implementation plan, including a clear set of timed targets, made it difficult to ensure alignment and cohesion around QI at all levels of the network in Uganda. One MoH participant reflected that having clear, time-bound and measurable commitments would have increased motivation and momentum by encouraging accountability. However, this situation was not static: our last round of data indicated better coordination and leadership from the centre.

Ethiopia experienced similar challenges throughout the entire implementation period. At the start of QCN, most of the actors in Ethiopia perceived network activities as WHO’s work until the MoH announced it was their flagship initiative. In Ethiopia there was some disconnect between the federal and regional levels. The federal level and regional level blamed each other. The regional level complained that the central (federal) level didn’t share anything or give clear direction, nor assign them with responsibilities properly, and that made them see the work as the federal level’s project. On the contrary, the federal government was complaining about the regional level’s lack of commitment whilst the regional level associated this with lack of capacity. In addition, regional level informants also mentioned they had a minimal sense of ownership.

“…..the MoH announced the QCN program in their own; they simply ask us to send them one or two participants in a meeting.”

(Government-Regional level-Ethiopia Round 1)

In the initial phases of QCN emergence, lack of coordination was also observed between the quality improvement and MCH departments of MOH. However, this was to improve later after a clear allocation of roles and responsibilities was undertaken. Lack of clear direction to the sub-national level was observed as a reason for slow inception of activities at local level in Malawi. Less concentration and poor political commitment of central level gradually made the RHD less active. Besides this, involvement of numerous partners, and persistence of discrepancies between the partners’ and local objectives, created difficulties to coordinate and consolidate various efforts.

Evidence showed that implementation activities in Bangladesh were mostly ‘DP [Development Partner] centric’ despite the government chairing the QCN. Like the other three case study countries, awareness was mostly at central level, and the network was not known to most of the sub-national level actors who perceived the activities as partner’s work. This perception was, however, transformed progressively though not entirely, through meetings, workshops, and training organized by implementing partners, and when health workers at facilities started to comprehend the benefits.

The above findings indicate that the beliefs of network actors about the purpose of QCN influence its legitimacy. This is exemplified at the local level, where the network was less well known by many health workers, and consequently had less authority.

Exchange of information

Learning and sharing occurred between and within countries and at various levels: global, regional, national, sub-national, district, sub-district and facility level. Here, sub-national, district, and sub-district levels are identified as local level (Figure 2). To continue the learning within country, district level learning networks were established in Bangladesh (in 2020) and Ethiopia (in 2019) [2]. A national level learning hub was also established in Bangladesh. QCN had a learning platform at global level and other methods of learning and sharing included: i) regular calls between partner countries and the QCN Technical Working Group (TWG) where countries share implementation progress and challenges, ii) a topical webinar series co-organized with partners, with a focus on sharing national level experience and know-how, and iii) in-person global level meetings where all global partners and network countries send delegations of eight to ten people [12].

• Download figure
• Open in new tab

Fig 2. QCN learning and sharing methods at global, regional, national and local levels

Exchange of information between countries was executed by all case study countries through regular calls and international meetings. From Bangladesh, TWG representatives attend the meetings. The QCN secretariat usually communicated with the focal person of development partners, and they coordinated and joined meetings along with the MoH. The MoH and development partners then shared updates on implementation activities in country with the QCN secretariat. In Uganda, the Ministry of Health and the WHO country office had weekly calls with the QCN global leadership and participate in network meetings, and their interaction increased and improved over our study period.

Of our four case study countries, Bangladesh and Malawi participated and shared experiences in different webinars initiated by the WHO QCN secretariat. MoH of Ethiopia and Uganda were initially reluctant to share at the global level though they agreed subsequently. In Bangladesh, development partners mostly coordinated it and included participants from local level also identified by them, with the consensus of the

MoH. However, the webinars were mostly “DP centric” as government stakeholders hardly attended. A national respondent in Bangladesh noted:

“Yeah! It’s good but the government people don’t have the time to attend/participate in this meeting. They don’t want to talk, understand on this issue. So I think these are DP centric.”

(Technical and Donor partner-National level-Bangladesh Round 4)

Though the learning platform is appreciated by most stakeholders, global participants mostly benefited and few stakeholders at national or local level attended.

All the partners including MoH, attended the global meetings, held during 2017–2019 before the COVID-19 pandemic, and shared experiences [12]. Key development partners of Bangladesh also attended other global or regional meetings and webinars to share their methods, experiences and learning. Uganda also shared with neighbouring countries at the initial stage of QCN in their country.

Information exchange within countries was experienced by all countries in diverse ways and at both national and local levels. However, these meetings are mostly initiated and financed by implementing or technical partners, and dependent on their support. For example, facility level meetings in Malawi were initiated and financed by implementing partners, so stopped functioning for lack of resources when partners stopped supporting them. Shortage of budget and mistrust between the ministry and the regions affected regional level meetings in Ethiopia. In all four countries, the MoH also conducted national level sharing and learning meetings including large collaborative learning sessions, and implementing partners undertook district collaborative learning sessions and facility-based learning visits to learning facilities within and between districts.

QCN was structured in Bangladesh and Malawi using already established national, subnational and facility level committees. Many of these were inactive though and started to be functional (mostly at the facility level) due to support and continuous focus by implementing partners.

Transfer of concepts and methods

Some respondents talked about adapting lessons from Bangladesh by other countries, though there was no evidence of transfer of concepts or methods by the study sites, i.e., neither the case study countries adopted learning from other countries nor other countries adopted learning from these countries.

Within countries, no one discussed implementation or adaptation of learning from other partner’s quality improvement methods; rather all the partners’ implemented QCN activities following their own approaches, and often in separate geographical silos. In Uganda, different partners operated in different regions and used different specific tools dictated by their different funding mechanisms. In Malawi, partners continued to support the kind of activities they had been doing before the launch of the QCN, working in similar areas. In Ethiopia, all partners used their own approaches, for example, one partner provided coaching every month, and another provided district-based coaching every quarter. Another one went to the districts from the center.

“Partners have their own interests; they all have different approaches that they follow. For example, we say learning collaborative should be prepared in three months. Some do it within six months. Some conduct coaching every month, the others do it quarterly. Therefore, it lacks uniformity.”

(Government - National level - Ethiopia Round 1)

The two key partners in Bangladesh were following two different approaches. However, at the initial stage of the study, one stakeholder of one development partner (UNICEF), mentioned about a cross learning process, where they have learned 5s-CQI-TQM from the other partner (Save the Children)’s previous planned piloting area which was originally implemented by JICA Bangladesh. Similarly, the USAID team visited Kurigram as Kurigram was already established as a model district. A national respondent from Bangladesh noted:

“Many [stakeholders] from different districts came to Kurigram including the whole USAID team…….they visited Kurigram before they started the Project. They observed the measurement system including other good system”

(Technical, Implementing and Donor partner - National level - Bangladesh Round 1)

However, as our study progressed, no stakeholder talked about this type of visiting or learning. This may be due to a lack of embedding of such cross-learning in organizational cultures of institutions involved in QCN including the units of government ministries of health leading QCN in each country. Political instability, e.g., in Ethiopia, and the COVID-19 pandemic –in all countries– may also have contributed to QCN being unable to achieve significant cross-learning.