Across countries and types of shock, effective coordination and partnership were important influences on shock preparedness and response, either as a supporting factor when coordination was strong, or a limiting factor when coordination was weak. Based on country examples of both effective coordination and gaps, our synthesis identified key issues for effective coordination and partnership related to i) the value of inclusive and representative coordination and partnership among different actors, ii) structural aspects of coordination bodies, iii) adequate capacities for coordination, including learning, and iv) political enablers. We consider each of these areas in turn.
Adequate capacity of coordination bodies and learning from previous shocks
Country experience showed the influence of organisational capacity on coordination, including technical and political skills, sufficient staff, infrastructure and funding.
The influence of human resource capacity in national coordination bodies was seen in several countries. In Nepal, NRA staff were seconded from line ministries and all positions were temporary. This secondment structure brought several difficulties that reduced human resource capacity: high turnover, which affected institutional memory; a lack of coherence and complementarity in skills; lack of rewarding career paths within the NRA to attract high calibre staff; and hierarchical issues related to individuals having less senior positions in the NRA decision making hierarchy compared to in their parent ministry, with consequent demotion. Previous experience in Nepal also indicated the importance of political skills among the leaders of coordination bodies, and sufficient credibility of these leaders with the wider government bureaucracy, to ensure they can win support and cooperation from staff and other ministries [
61]. A further consideration from experience in Nepal was flexibility of institutional capacity, with a permanent core structure that can expand as needed for disaster response, including through additional staffing [
61].
Further human resource challenges for coordination bodies were indicated by experience at the NDRMC in Ethiopia. One issue was the overall availability of skilled staff: the Commission had skilled technical staff, but some experienced staff were leaving, and there were insufficient skilled personnel to provide technical support on disaster management mainstreaming to line ministries or to monitor policy implementation. NDRMC staff capacity to support coordination was also strained by multiple concurrent emergencies; for example. early NDRMC engagement in the COVID-19 response was limited by simultaneous work on a desert locust plague, major floods, and conflict-induced displacement [
40,
60]. Commission experience also highlighted the need for specific technical skills to manage different types of shock: the NDRMC’s historical strength has been with drought response, but a changing humanitarian environment, particularly increased conflict and displacement and more recently COVID-19, required new skills [
40].
Country experience also showed the influence of human resource capacity on subnational coordination. In Nepal, local governments were relatively newly established and still building organisational capacity, including for disaster management. Many local authorities had established or begun developing EOCs, but a lack of human resources contributed to limited EOC functionality and ability to ensure effective coordination [
50]. Elsewhere, the influence of human resource capacities on subnational coordination was evident in relation to COVID-19. In Pakistan, insufficient technical expertise for emergency preparedness, response and coordination reduced the capacity of provincial coordination bodies [
41]. In Ethiopia, subnational capacity varied between regions and levels. For example, in Oromia, availability of trained staff assigned to emergency response at regional level supported early activation of the EOC, communication across levels, and development of plans to support coordination. However, insufficient staffing and experience at zone and woreda levels contributed to lack of sub-regional EOCs and hindered coordination [
45]. In Sidama, a new regional administration, limited training meant insufficient technical skills and experience in emergency coordination at regional level, and there were also insufficient dedicated staff at sub-regional levels. This reduced the functionality of regional and woreda Emergency Coordination task forces and EOCs for the COVID-19 response, hindered coordination, and contributed to delays and insufficient harmonisation of response activities [
44]. Insufficient skilled staff also hindered coordination between federal and subnational government [
39,
51].
Experience with subnational coordination capacity for the COVID-19 vaccine rollout in Ethiopia highlighted the influence of ongoing health systems human resource issues: long-standing shortages of government and immunization programme staff contributed to limited time capacity to participate in coordination fora, which reduced the functionality of some regional and zonal working groups for COVID-19 vaccination. Weak functioning of subnational working groups in turn reduced their capacity to coordinate with national structures, delayed sharing of information, and hindered vaccine rollout in some regions [
57].
As well as staff, subnational capacity was affected by funding and communications infrastructure. For example, in Pakistan, availability of accurate and consistent information to support coordinated action has been limited by insufficient technology and connectivity at lower health system levels, as well as by insufficient administrative capacity, unfamiliarity with information systems, and incentives related to self-reporting [
41]. In Nepal, provincial EOC capacity is limited by shortages of equipment and funding, as well as staffing [
50]. In Ethiopia, phone outages and subnational gaps in internet access and information technology hindered emergency coordination and information sharing between regional, woreda and federal government and with regional development agencies, including during COVID-19 [
51]. Where communications equipment was available, for example in the Southern Nations Nationalities and People's Region and Oromia, this enabled information sharing and coordination [
45,
46]. For example, in Oromia weekly virtual meetings with leadership in all zones and towns helped to identify supply gaps and distribution needs [
45]. In South Sudan, unreliable internet was further compounded by disruption to road networks, limiting sub-regional attendance at physical meetings to discuss the COVID-19 response [
57].
Adequate funding also affected capacity for coordination, partly due to allowances associated with participation in coordination meetings. In Kenya, strained county health budgets hinder coordination mechanisms, for example when nutrition coordination meetings cannot be hosted due to lack of funds [
43]. Lack of budget has also affected development agency coordination with local government for the COVID-19 response in South Sudan, due to the costs associated with workshops [
57].
Capacity is needed not just within the health sector, but across sectors given the importance of multi-sector coordination and action. In Ethiopia, experience during COVID-19 showed that several government ministries and departments lacked staff with training and experience in public health emergency management and coordination, and did not have established structures to support disaster management. This contributed to difficulties in cross-sector coordination for the COVID-19 response, and increased reliance on the health sector [
39].
Development agencies can support the capacity of coordinating structures; for example, they supported communications infrastructure in some regions of Ethiopia [
51]. However, short term support has limited the effectiveness of capacity development. In Ethiopia, development agency support to the NDRMC has often involved short-term technical assistance. This temporarily boosted capacity, but technical assistance staff sometimes lacked personal investment in the NDRMC, and their short terms of appointment hindered sustainable improvement and institutional learning [
40]. Similar sustainability issues were seen in Pakistan, where development agencies provided capacity building for the Province Disaster Management Agencies and Health Departments, but with insufficient ownership and gaps after donor funding ended [
41].
Using previous learning to support effective coordination
Country experience showed that learning from previous shocks could support effective coordination, by demonstrating the value of partnership and so encouraging coordination efforts, and by developing mechanisms for coordination or lessons on effective coordination approaches.
In Kerala, experience with a series of shocks (such as floods and the 2018 Nipah virus outbreak) showed government the value of citizen engagement, and helped it to develop mechanisms to convene other government departments and to collaborate with external stakeholders. This in turn supported coordination for the COVID-19 response [
52]. In Sierra Leone, learning from Ebola contributed to active community engagement and involvement of traditional leaders in the COVID-19 response, and enabled swift activation of coordination structures and systems for COVID-19, such as use of the Ebola Emergency Operations Committee [
37,
38]. Similarly, previous experience of emergency coordination using an incident management system in some regions of Ethiopia facilitated establishment of the regional PHEOC and task forces for COVID-19 coordination, which in turn supported engagement with partners and other stakeholders [
44,
46]. Regional governments have also used learning from past droughts to support coordination during nutrition emergencies, including holding more frequent coordination meetings and creating different sector task forces [
40].
While learning can support effective coordination, country experience also indicated several factors that can limit identification and use of lessons from previous emergencies. In Ethiopia, the NDRMC had used lessons from previous years to strengthen coordination of drought relief, but more extensive learning was limited by issues such as lack of opportunity to reflect as crisis rapidly follow one another; limited opportunity to absorb lessons from evaluations or reviews; coordination structures often becoming dormant after emergency response is concluded; and a lack of leadership to act on learning [
40]. The latter issues also hindered learning in line ministries: task forces and committee meetings only took place during emergency response, reducing opportunities to discuss and reflect on learning after crises end [
59]. In addition, sector focal points were temporarily assigned for emergencies and then returned to their usual duties, and they were not directly accountable and evaluated for their involvement in emergency management [
40], hindering incentives and opportunities for learning. At subnational level, staff turnover and poor documentation were further issues hindering learning for development of effective coordination systems [
40].
Political considerations and effective government leadership
The final area identified in our analysis was the role of political considerations, including leadership across different government levels, sectors and stakeholders, and incentives for coordination.
Experience during COVID-19 indicated the value of national government leadership for coordination between government sectors and levels, and with development agencies. In Rwanda, government led coordination for vaccine rollout, and asked development agencies for assistance in specific areas, which helped to harmonise their support [
57]. In Pakistan, the Prime Minister’s political ownership of the National Command Operation Centre (NCOC) established for COVID-19 supported the NCOC’s effectiveness and contributed to engagement of other government sectors and departments [
41]. Similarly in Ethiopia, the National Ministerial Committee established for COVID-19 was accountable to the Prime Minister, the Minister of Health chaired the overall COVID-19 coordination group, and senior government leadership provided close follow-up and support for activities [
39,
57]. This high-level leadership and engagement promoted collaboration across different national government ministries, and coordination with development agencies and subnational levels, including through early activation of the PHEOC and creation of multi-stakeholder coordination fora [
39]. High-level national leadership also supported coordination at subnational levels, particularly via material and political support for the functionality of regional PHEOCs [
44‐
47].
Leadership at subnational levels also affected the strength of coordination during COVID-19. In Pakistan, effective leadership by Chief Ministers in some provinces contributed to cross-sector coordination, such as involvement of the district administration, police, and government departments on health, education, water, sanitation and others in quarantine facilities [
41]. In Kerala, leadership from the Chief Minister to establish a platform for discussion across ministries helped to ensure engagement of other state ministries, with all departments (not just the Ministry of Health) instructed by senior government to focus on COVID-19 [
52]. In Ethiopia, where regional political leaders and senior regional government were actively engaged, this facilitated early establishment of the EOC, availability of staffing for emergency coordination platforms, coordination across sectors, links between the regional EOC and lower levels, and engagement with stakeholders such as health professional associations [
44,
45] Where political leaders were less engaged, this made information sharing, communication and coordination of the response less systematic or streamlined [
39,
45,
46]. For example, in Oromia, political engagement declined as growing political instability took leaders’ attention. The resulting leadership gap contributed to reduced engagement by departments beyond health, interruption of coordination fora, insufficient resources for the response, and duplication of some activities [
45].
While high-level political leadership supported engagement of different government sectors, effective multi-sector coordination also required sufficient leadership within different government sectors, including ministries beyond health. In Ethiopia, national policy designates lead institutions to implement disaster risk management for different hazards; for example, the Ministry of Agriculture for agriculture-related emergency management, the Ministry of Water, Irrigation and Environment for floods, the Ministry of Peace for conflict-related emergencies, and the MoH for health-related emergencies. However, commitment to disaster management from ministers in other sectors has been limited, particularly beyond the MoH, with a tendency to rely on the NDRMC secretariat. This lack of leadership contributed to limited development of sector disaster management plans and structures, which in turn reduced of focus on and capacity for disaster management, and hindered coordination. For example, lack of clear responsibility and structure in other ministries weakened effective participation in multi-sector coordination fora and accountability for action [
60].
Experience during COVID-19 indicated the need to balance high-level political or government leadership with sufficient technical input, from within and outside government. In Ethiopia, health professional associations, development agencies, technical government staff and other technical experts were part of coordination structures [
39]. For example, coordination of vaccine rollout was initially managed by MoH, but government later brought in the wider, existing immunisation structures to provide technical input (e.g. the National Immunization Technical Advisory Group and the immunisation TWG) [
57]. In contrast, in South Sudan, senior government officials decided to change the vaccine distribution strategy previously agreed by government and development agencies, without consultation. This led to delays in implementation, reduced population access, and ultimately to vaccine doses being unused and returned to COVAX. A later change in government structures brought more collaboration with development agencies and use of technical guidance, leading to a revised and more effective distribution strategy [
57]. A final example from Sierra Leone showed the importance of balancing political leadership with engagement of technical government staff: MoHS staff were concerned that decisions made by political leaders with insufficient input from technical and medical experts reduced the effectiveness of COVID-19 response strategies [
38].
Beyond leadership, other political considerations were also evident in country experience. A particular issue was political tension between national and subnational government, which hindered coordination across levels. In Pakistan, there was friction between national and provincial governments in the COVID-19 response, with open disagreement regarding some provincial policies (such as a more complete lockdown in Sindh). This tension partly reflected wider political systems, with different political parties in government in different provinces and at federal level [
54]. In Nepal, the District Disaster Management Committee could support national-local coordination as the committee chair is a federal government official (the Chief District Officer, or CDO), and members are politically-elected local government leaders. However, the Committee’s effectiveness is reduced by concern among local leaders that CDOs exert central government power and side-line subnational government [
50].
A further political consideration indicated by country experience related to incentives and openness regarding information sharing. Country experience showed that timely, open and accurate information sharing and unified data systems can support coordinated action, as well as underpinning other key elements of shock response such as identification of affected populations [
41,
61]. However, political incentives sometimes reduced availability of accurate data. In Ethiopia, early warning information underpins the annual Humanitarian Requirements Document and Response Plan (the basis for government and agency work and for coordinated action). However, early warning information is collected by different ministries and government agencies, and sometimes withheld or manipulated, partly because this early warning data plays a role in determining budgets and resources. For example, tension over early warning information between the NDRMC and Ministry of Agriculture (which manages the Productive Safety Net Programme, PSNP) reflected concern about the size of the PSNP caseload and affordability of safety nets. Resources to the regions are also affected by population size, and associated under- or over-estimates of population figures reduce the accuracy of needs assessments. Political concerns over a perception of Ethiopia as an aid dependent ‘famine country’ have also led to late publication of early warning findings. One example came from the 2017–19 Gedeo/West Guji displacement crisis, where IDP numbers were sometimes inflated to attract additional resources, or understated to protect political image. These tensions reduced the effectiveness of early warning information, and both reflected and contributed to weaknesses in coordination [
40]. Similar issues have affected more recent NGO response to nutrition emergencies:, gaps and delays in provision of information from subregional to regional level, delays in reporting of crises from regional government to the federal level and other stakeholders, and discrepancies between woreda and regional data regarding the number of people in need meant early warning information was late and uncertain, which hindered timely and coordinated NGO response [
58].
The role of transparency in information sharing was seen with COVID-19 vaccine rollout in Rwanda and South Africa. In both countries, there have been gaps in information from government, including difficulty in accessing health management information system data in Rwanda (for example, on coverage of priority groups), lack of budgetary information on resources provided and funding gaps in both countries, and in South Africa, limited information on vaccine purchasing agreements. This lack of shared information on progress, costs, and resources hindered effective coordination and prioritisation of development agency support [
57].
Discussion
This paper has highlighted the importance of effective coordination and partnership for emergency preparedness and response in fragile and shock-prone settings. Based on empirical examples, the paper identified a set of issues that characterise and enable effective coordination. Overarching issues involve inclusive coordination across government sectors and levels and with other stakeholders, such as development agencies, NGOs, the private sector and research institutes; structural issues, including the availability of coordination fora, ongoing coordination structures that function before and after shocks, the mandate and authority of coordinating bodies, and streamlined information systems; the capacity of organisations with a role in coordination, including staff, skills and infrastructure, across relevant sectors and at different geographic levels; and political considerations and incentives, including government leadership.
The synthesis findings suggest key elements that can enable effective coordination, both for preparedness and response to shocks. Table
1 summarises key findings and recommendations.
Table 1
Summary of findings on key elements of effective coordination
Inclusivity of coordination | • Engagement and input from all relevant government sectors (multisectoral coordination), including sufficient technical input on health but also coordination with areas such as water and sanitation, social protection, agriculture, disaster management, and finance • Effective cooperation and communication between national and local governments (vertical coordination), including consistent and aligned guidance to local levels, clarity on roles, clear systems for reporting information upwards, and national government responsiveness to district needs • Government engagement with other stakeholders, including development agencies, local leaders and civil society, religious leaders, research institutions and the private sector • Development and humanitarian agencies working through agreed coordinating structures and funding systems, and approaching work collaboratively • Coordination and unified contact points among development agencies on the one side, and among government stakeholders on the other, to facilitate communication between multiple development agencies and government bodies • Gender equity in representation and involvement in coordination fora, across government and other stakeholders |
Structural features of coordination mechanisms | • Availability of functioning coordination structures, including regular meetings at national and sub-national levels • Clear roles for each coordination structure and organisation responsible for coordination, including at different government levels, to avoid overlapping remits. Tools such as accountability matrices can help to avoid duplication and clarify responsibilities • Clear relationships and sets of responsibilities between disaster risk management coordination authorities and sectoral ministries • Mandates for coordination bodies that are sufficiently wide to support responses to the range of relevant shocks • Sufficient authority for coordination bodies to convene relevant actors and ensure the implementation of agreed plans. Positioning directly under the president or prime minister (rather than in a ministry) can support this authority • Coordination structures that function on an ongoing basis—before shocks occur to enable preparedness and anticipatory planning, and after shocks to support learning and recovery • Using existing structures can support coordination during shocks, by providing established organisational arrangements, roles, relationships and ways of working; new structures may require additional support |
Adequate capacity of coordination bodies and use of learning from previous shocks | • Adequate numbers of staff with political and technical expertise, including expertise for all relevant types of shock, within coordination bodies at different levels • Career paths that reward expertise and provide stability of employment for at least core staff, to support skills, motivation and accountability • Adequate funding and communications infrastructure for organisations responsible for coordination, across relevant government sectors and levels • Support for sustained and ongoing capacity, rather than short term technical assistance or capacity only during emergencies • Learning from previous shocks, enabled by time for reflection, leadership to act on learning, accountability for emergency response as part of core staff roles, and the retention and exchange of organisational learning, including through standing (rather than ad hoc) coordination structures |
Political considerations and effective government leadership | • Effective political leadership, at national and subnational levels and in different government sectors, balanced with technical input from government and other stakeholders • Political, organisational, and individual incentives to support coordination, including in relation to transparent and accurate information sharing • Regular communication and reporting across government levels to promote effective subnational leadership and accountability |
The synthesis demonstrates the complexities of coordination, resulting from issues such as myriad actors, requirements for multiple capacities and insufficient resources, power imbalances, the influence of pre-existing health system conditions, and the role of factors both within the health system and beyond, such as wider governance, national infrastructure and support from other sectors. The findings also demonstrate the interactions between different components and feedback loops, for example with human resource systems affecting capacity to learn from previous shocks, and scope for learning then affecting availability of adequate expertise. Similarly, high-level political support can affect the authority of coordinating bodies, in turn influencing engagement and information sharing by different sectors. These interactions reinforce the emphasis on systems approaches in the health system strengthening and resilience literature [
64,
65].
Our findings build on existing analyses and strategies that emphasise the importance of coordination and partnership across sectors, levels and stakeholders. Some specific enablers and characteristics of effective coordination identified in our synthesis have also been identified in other contexts. For example, a recent discussion of experience with coordination during COVID-19 indicated the importance of responsive leadership and strong political will, clear roles and responsibilities, and effective information systems, as well as the need for coordination across sectors, levels, and with community and religious leaders [
66]. Weak leadership, shown in our findings as affecting engagement from different government sectors and levels, has also been identified as hindering coordination for shock response in other reviews [
14,
67]. In humanitarian contexts, the role of collaborative relationships, development agency commitment to alignment, capacity for coordination, and effective information management have been identified as enabling coordination [
68].
Some characteristics and enablers identified in our synthesis are also stipulated in existing guidance on coordination for emergency management. In particular, the WHO toolkit for assessing crisis management capacity indicates the need for a high-level multisectoral committee with regular meetings, a clearly defined mandate and set of responsibilities, clear authority, and sufficient staff, equipment, funding and support systems to fulfil its mandate [
21], in line with our findings on the structural requirements for coordination bodies and the influence of capacity on coordination. This toolkit also emphasises the need for coordination across government sectors and with the private sector and civil society.
Several other characteristics and enablers of effective coordination from our findings have been discussed as key requirements for broader shock response, health systems resilience or health systems strengthening, rather than in direct connection to coordination and partnership. For example, the roles of political leadership, equitable gender representation, functioning information systems, learning, and technical skills and staffing for emergency management are widely highlighted as important for broader health system strengthening, emergency management and resilience [
69‐
72]. Our work shows the specific relationship of these requirements to effective coordination for shock preparedness and response, illustrating interactions between coordination and other key health systems and resilience capacities.
This analysis was restricted to information from reports developed or supported by OPM, and therefore did not aim to provide a comprehensive assessment of all aspects of coordination and partnership, or a full review of available evidence. Synthesising findings from OPM reports brings new evidence to the discussion, as some reports were not publicly available, and none had been published in the academic literature, but a wider review of all available research would provide additional evidence. The reports included in this review had different areas of focus and research questions, and as such, they did not provide directly comparable information across countries. Due to varied focus and purpose of the original reports, some provide in-depth discussion and detail, whereas others provided only summary information on coordination and partnership. Reports were not evenly spread across countries; for example, many reports were from Ethiopia due to a substantial OPM programme on resilience. However, the reports did cover varied country contexts, which helped to identify variations in the strength and approach to coordination.
Conclusion
COVID-19 and other shocks have highlighted the importance of effective coordination and partnership across government and with other stakeholders. We identified four key areas that characterise and enable effective coordination for emergency management, involving inclusive approaches, and the structures, capacities, and political incentives for coordination. Within this, the synthesis identifies a set of issues to consider in strengthening coordination for health systems resilience and shock response, such as political skills, organisational authority, retention of learning, communication platforms, and alignment by development agencies. Interactions between these components and with the wider health system and governance architecture indicate the need to consider coordination as part of a complex adaptive system, so requiring attention to interdependencies between different components.
The paper raises several areas where further evidence could support understanding of effective coordination systems. These include the political economy of coordination, including the roles and membership of coordination structures and incentives for engagement among different ministries; effective structures and formats for coordination fora, such as optimal constitution and size; gender equity in coordination, including representation of women in decision making structures, and strategies for coordination that have supported an inclusive response; the influence of funding systems on coordination, including use of pooled funding; and variation in coordination between different types of shock, including acute or protracted crises and whether shocks are framed as ‘health’ emergencies, and how this affects representation of health and other sectors.