Health systems governance, shocks and resilience: a scoping review of key concepts and theories

Abstract

Introduction

It is often argued that health system resilience to shocks is influenced by governance. Multiple theoretical perspectives seemed to have emerged in the literature, but there has been limited analysis of their foundation and implications so far. To address this gap, the aim of this paper is to carry out a scoping review of conceptualisation and theories on the definition and relationship between governance and health system resilience in the context of shocks.

Methods

Six electronic databases were searched using the following keywords: resilience, health system, framework and governance. 835 abstracts were screened, and 47 full texts were extracted and analysed in MAXQDA using inductively generated codes. We analysed the conceptualisation of health system resilience, governance and shock and theories on the relationship between them.

Results

Although there are no collectively agreed definitions, health system resilience is typically defined as the capacity of a system to continue to perform its fundamental functions in the setting of acute disruptive events (shocks). Existing conceptualisations and theories focus on normative descriptions of the ‘capacities’ that should be developed to promote health system resilience. They usually emphasise the bidirectional relationship between governance and health system resilience, but underlying mechanisms remain undertheorised. While many authors stress the value of a forward-looking approach that anticipates scenarios to build resilience, there is no consensus on whether a centralised ‘top-down’ is more effective than a decentralised ‘bottom-up’ approach to respond to shocks.

Conclusions

Overall, the literature suggests that governance is critical to understanding health system resilience in response to shocks, yet these relationships are considered complex and potentially bidirectional and remain undertheorised. Future research should develop more nuanced theories that capture the dynamics of these relationships and consider how the health system relates to broader (non-health) systems that have the potential to foster health system resilience.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• The COVID-19 pandemic highlighted the importance of health system resilience. Multiple conceptualisations and theoretical frameworks have emerged over the past years. There is a growing recognition that health systems should be resilient against a variety of shocks that differ in nature, temporality, severity and frequency. Governance is generally considered vital for health system resilience, but how governance can foster resilience is not well understood, nor is how the nature of shocks can affect this relationship.

WHAT THIS STUDY ADDS

• This study reviews and contrasts the conceptualisation of shock, governance and health systems resilience used in the literature. It identifies the key points to consider when studying health system resilience in each specific context, reviews theories on the dynamic relationship between the three core concepts and identifies literature gaps. Theories highlight the fact that shocks are context-dependent and therefore require context-specific governance approaches to foster health system resilience.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• Our findings highlight the need for a consolidated theoretical framework and for developing sound theories on how governance can improve the resilience of health systems to shocks. This will support empirical research on health systems resilience and in turn help design strategies to strengthen how we govern systemic crises such as the COVID-19 pandemic.

Introduction

Health systems resilience (HSR) has gained increased attention in global health, particularly following the COVID-19 pandemic.1,4 HSR generally refers to the capacity of health systems to prepare for, manage and learn from health crises and shocks, while maintaining core functions and services.15,7 HSR is considered key for global health⁸ as it may contribute to protect human life,⁵ strengthen health systems,⁹ achieve universal health coverage (UHC) and health security¹⁰ and guarantee high health system performance.^{11 12} A growing number of studies have evaluated the drivers and determinants of HSR.713,15 Among those drivers, governance has been identified as a key determinant influencing how health systems prepare for, respond to and recover from shocks.^{16 17} Yet, there is no widely accepted theoretical framework by which we can understand the role of governance in fostering HSR.¹⁸ In addition, existing frameworks focus on specific shocks such as the COVID-19 pandemic, but there may be differences in the way different shocks influence resilience, as well as the role of governance in this process. The aim of this paper is to carry out a scoping review of the literature to better understand the definitions and theories that conceptualise the relationship among HSR, governance and shocks and to identify significant gaps that may require theory development.

Methods

We conducted a scoping review in accordance with the methodology outlined by the Joanna Briggs Institute Guidance¹⁹ to address the following research question: how are the concepts of governance, shock and resilience defined in the international literature and how do studies theorise the relationship between these concepts?

The search strategy was developed in collaboration with an expert librarian, to be comprehensive and to identify both published and unpublished (grey literature) evidence on the study topic. The search strategy was based on four keywords: resilience, health system, framework and governance. For each concept, the librarian found the freetext terms, the emtree and the mesh. The validated search strategy was adapted for each included database: Medline OVID SP, Embase.com, Central – Cochrane Library Wiley, Web of Science – Core collection and SocIndex. In addition, Google Scholar was employed to search the grey literature, retaining the first 200 references. The search included only papers written in English, with no limitation imposed on the publication year. The main search was carried out on 25 May 2022 by the expert librarian. The research strategy for PubMed is reported in the supplementary materials.

The librarian removed duplicates and exported the final research into EndNote X8.1. All the publications were then imported into Rayyan²⁰ to facilitate the screening of the titles and abstracts by two independent reviewers (GB, CP). The two reviewers independently assessed the abstracts and discussed throughout the literature research any disagreements. Full texts of retained abstracts were read, and the agreement of the two reviewers was necessary to exclude an article. The reasons for exclusion were recorded and reported in the review process flowchart (figure 1).

Figure 1. PRISMA-flow diagram of study selection.

Only theoretical articles on HSR were included. Theoretical articles were defined as papers that discussed concepts, theories, strategies, models or frameworks on HSR, but did not directly involve data collection. They included reviews, expert opinions, policy briefs, reports of non-governmental organisations and commentaries. We excluded papers written in any other language than English, articles not focused on HSR or not centred on the health system globally (ie, articles focused on a single WHO’s health system building block²¹ or on one health system facility or a single clinical unit) and empirical articles. Academic theses and full textbooks were also excluded. The qualitative data analysis tool MAXQDA was used for data extraction, as it allowed us to label and code full texts.²² The codebook was established using inductive coding methods. For each included article, all text sections concerning governance, resilience and shocks and their conceptual definitions were coded by at least one author (GB and/or CP). The relationships between ‘resilience and shock’, ‘governance and shock’ and ‘governance and resilience’ were extracted by analysing paragraphs where both concepts were coded.

Patient and public involvement

Patients and the public were not involved in the design, conduct or reporting of this research.

Results

The search strategy identified 1350 studies from the selected electronic databases (figure 1). After removing duplicates, 835 studies remained. 761 studies were excluded by title and abstract screening. Four books were also excluded. 70 papers were read in full text to be assessed for eligibility: two studies were excluded because of their empirical design and 21 because they did not focus on HSR. 47 articles were retained for the review: 14 articles, 13 reviews, eight non-governmental organisations’ publications, five commentaries, four editorials, one book chapter, one meeting report and one round table abstract (online supplemental table S1). The main conceptual frameworks used in the literature were the WHO building blocks and functions,²¹ complex adaptive systems, organisational resilience, network analysis and social-ecological models. Although not formally theoretical models, some articles used environmental science frameworks, the Sustainable Development Goals framework and health policy frameworks, such as UHC and Health Security. Several articles did not use a well-defined conceptual framework, but instead were based on experts’ opinion and lessons from the COVID-19 pandemic or other acute shocks (online supplemental table S1).

Definition of ‘shock’

Eight formal definitions of ‘shock’ were identified in the literature, although additional terms were used which referred to synonyms of ‘shock’ (ie, challenge, stress, crisis, threat and disaster) (table 1). Six key features of ‘shocks’ were identified. First, all articles referred to duration, with definitions of ‘shock’ including attributes linked to the acute nature (short duration) of the event, that is, ‘sudden/immediate’, ‘transient’ and ‘time-bound’. Second, studies referred to the intensity of events, that is, ‘extreme’, ‘surprising’ and ‘atypical’. Third, some studies referred to shocks as unexpected events (although a few studies made reference to ‘recognisable’ events). Fourth, studies conceptualised two durations of effects: one definition included a medium-term perspective of the shock, while another definition made reference to the longer term. Fifth, studies also made reference to the disruptive nature of shocks, with two definitions defining ‘shock’ as an interruption to the normal pattern of development or an additional burden to the health system. Sixth, most papers discussed health-related shocks (ie, disease outbreaks), but some studies discussed how non-health societal shocks could potentially have an impact on HSR (eg, insecurity, economic crisis, natural disasters and conflicts) (table 1). Finally, Rogers et al proposed a classification of shocks to a health system based on four dimensions: time, expansion, origin and the target of its impact,²³ while Thomas and Sagan viewed the shock in a dynamic four-stage cycle.^{2 6 24}

Table 1. Shock definitions.

Definition	Concept defined	Features	Causal factors	Impact	Examples
‘Shocks are classified as acute if they are sudden in occurrence and transient in nature. These included disease outbreaks, insecurity, economic crisis, unspecified natural and/or manmade disasters and rapid policy reforms’. (Barasa et al11)	Acute shock	Sudden in occurrence Transient in nature			Disease outbreaks, insecurity, economic crisis, unspecified natural disasters, manmade disasters and rapid policy reforms
‘…Bloom et al proposed to classify events or crises that affect systems into two different groups: (1) shocks (transient disruptions) and (1) stresses (enduring pressures). Shocks are atypical and short-term interruptions to the ‘normal’ pattern of development (eg, natural disasters, wars, debt), whereas…’. (Blanchet et al, based on Bloom G, Edström J, Leach M, Lucas H, MacGregor H, Standing H, Waldman L. Health in a Dynamic World: STEPS Working Paper 5. Brighton: STEPS Centre; 2007)	Shock	Transient disruptions, atypical, short-term		Interruptions to the ‘normal’ pattern of development	Natural disasters, wars and debts
‘A sudden and often surprising event that causes an additional burden to the health system, most often for a short period of time. Pandemics such as Ebola or natural disasters caused by climate change were the commonly used examples of a shock’. (Fridell et al40)	Shock	Sudden, often surprising event		It causes an additional burden to the health system in a short period of time.	Pandemics (Ebola) and natural disasters caused by climate change
‘We use the term shocks to mean stresses and extreme challenges to the system caused by external events. These can be immediate and time-bound, such as a tsunami or a flood affecting a health system or can unfold over a period of time—such as a financial crisis’. (Hanefeld et al30)	Shock	Immediate, time-bound, unfold over a period of time	Caused by external events		Tsunami, flood and financial crisis
‘The term ‘shock’ rather than crisis was chosen as the types of events examined were comparatively short- to medium-term in nature’. (Hanefeld et al30)	Shock	Short-term and medium-term in nature
‘There is no consensus definition of system shocks in the literature on health system resilience, but two main areas of research can be identified. On the one hand are studies concerned with immediately recognisable shocks such as pandemics, natural disasters, national or international financial crises and armed conflict’. (Ismail et al46)	Shocks	Immediate Recognisable			Pandemics, natural disasters, national or international financial crises and armed conflict
‘Shock is a sudden and extreme change which impacts on a health system, and is thus different from the predictable and enduring health system stresses, such as population ageing’. (Thomas et al6 and Sagan et al41)	Shock	Sudden, extreme change		Impacts on a health system
‘We understand shocks to be sudden and extreme disturbances, such as epidemics, natural and other disasters, and financial crises. We think of a shock in a dynamic way—a cycle that consists of four stages, with interlinkages between the recovery from a shock and preparedness for the next shock cycle, as we go through the loop again’. (Sagan et al24)	Shock	Sudden, extreme disturbances			Epidemics, natural and other disasters and financial crises
‘Only two papers focused on the resilience of health systems to chronic, everyday challenges. Challenges are described as chronic if they are persistent and recurrent over long periods of time’. (Barasa et al11)	Chronic everyday challenges	Chronic, everyday persistent, recurrent over long periods of time
‘…resilience, where the context is not a shock but a chronic challenge that occurs every day over a long period of time and that cannot always be predicted’. (Fridell et al40)	Chronic challenge	Occurs every day over a long period of time. It cannot always be predicted
‘We consider shocks to be sudden and extreme disturbances, such as epidemics, distinguishing them from more chronic health system strains, such as those caused by population ageing, and relate key aspects of resilience to the different stages of the shock cycle’. (Thomas et al6)	Chronic health strains/stress	Chronic			Population ageing
‘On the other hand, are studies concerned with the effects of chronic, largely internal stressors (eg, workforce shortages, payment delays or policy changes), drawing primarily on insights from local or regional health systems particularly in sub-Saharan Africa’. (Ismail et al46)	Internal stressors	Chronic, largely internal			Workforce shortages, payment delays and policy changes
‘…Bloom et al proposed to classify events or crises that affect systems into two different groups: (1) shocks (transient disruptions) and (2) stresses (enduring pressures).… whereas stresses are effects that play out over a far longer time span than established policies can generally cover’. (Blanchet et al25 based on Bloom G, Edström J, Leach M, Lucas H, MacGregor H, Standing H, Waldman L. Health in a Dynamic World: STEPS Working Paper 5. Brighton: STEPS Centre; 2007)	Stress	Enduring pressures play out over a far longer time span than established policies can generally cover		Over a far longer time span
‘We recognise that there are many longer-term crises, for example, epidemics of chronic disease or underfunding over a period of years which affect and equally require health systems resilience, however these require further analysis. (Gilson et al).' ‘Similarly, climate change is an ongoing process which is likely to affect disease patterns and food security (hence nutritional illnesses) over a long period of time (Mayhew and Hanefeld 2014)—we focus …’. (Hanefeld et al.30)	Longer-term crisis	Over a long period of time		Affect and equally require health systems resilience. Climate change is likely to affect disease patterns and food security.	Epidemics of chronic disease underfunding over a period of years; climate change.
A crisis is ‘a disruption that physically affects a system as a whole and threatens its basic assumptions, its subjective sense of self, its existential core’ (Brodie et al31 based on Pauchant and Mitroff, 1992).	Crisis			A disruption that physically affects a system as a whole and threatens its basic assumptions, its subjective sense of self and its existential core.
WHO describes public health threats as: ‘new or newly emerging diseases, the accidental release or deliberate use of biological, chemical or radio-nuclear agents, natural disasters, human-made disasters, complex emergencies, conflicts and other events with a potentially catastrophic impact on human health’. (Bayntun et al32 from Final Report A Multi-State Study to Develop and Test a Benchmark Protocol for Assessment of Health Emergency Preparedness in Eastern Mediterranean Region Title)	Public health threat			Events with catastrophic impacts on human health.	New or newly emerging diseases, the accidental release or deliberate use of biological, chemical or radio-nuclear agents, natural disasters, human-made disasters, complex emergencies, conflicts and other events
‘A Disaster is a serious disruption of the functioning of a community or a society involving widespread human, material, economic or environmental losses and impacts, which exceeds the ability of the affected community or society to cope using its own resources’. (WHO 201533)	Disaster	Serious		Disruption of the functioning of a community or a society involving widespread human, material, economic or environmental losses and impacts, which exceeds the ability of the affected community or society to cope using its own resources

The main shock features mentioned in the literature were the nature, the occurrence (acute/sudden, chronic or slow moving), the severity, the frequency and its internal (endogenous) or external (exogenous) origin. Some examples of ‘external’ acute shocks were disease outbreaks, natural and manmade disasters, conflicts and economic crisis.^{7 11 12 25 26} (Chronic) everyday shocks were often considered as internal (endogenous) to the governance system and included the instability in governance structures, payment delays, dysfunctional policies, limited human resources and high levels of disease burden.711 27,29 A significant shortcoming in the public health literature is the tendency to conceive shocks as exogenous, ignoring the fact that they may in fact relate to inherent weaknesses in healthcare system capacities, planning and preparedness or governance.

Finally, the terms ‘challenge’ and ‘stress‘ were primarily employed to denote chronic everyday disturbances (the main attributes were ‘chronic’, ‘every day‘, ‘persistent’, ‘recurrent’ and ‘long period’) such as population ageing, workforce shortages, payment delays and policy changes, climate change and epidemics of chronic diseases (table 1). The two definitions of crisis emphasised its impact on HSR and on the system as a whole.^{30 31} A public health threat was defined as an event with catastrophic impacts on human health (eg, newly emerging diseases³²). Lastly, the WHO defined a disaster as a disruption of the functioning of a community/society which exceeds its ability to cope using its own resources.³³ Overall, this literature is useful to understand the variety of shocks affecting health systems.

Definition of health system resilience

39 definitions of HSR were identified (online supplemental table S2). 77% of HSR definitions identified it as a capacity/ability, while other studies conceptualised it as the amount of change a system can undergo, an emergent property, a system feature, a response to shocks or an active process. All except two definitions implied that resilience is observed in the context of a disruptive event such as a shock or a crisis. All definitions identified at least a prescriptive resilience action/capability before, during or after the disruptive event (online supplemental table S2).

The main capabilities attributed to HSR were the capability to adapt (51% of included articles), to respond/deal with (44%), to absorb (41%), to transform (36%) and to maintain core functions and/or structures (28%) (figure 2). Health system performance, the maintenance of the core functions and the improvement/strengthening of health systems including health needs and good health outcomes were the most common objectives of resilience mentioned in the literature. Lastly, three definitions of everyday resilience were found, referring to health system’s capacity (or flexibility) to continue delivering good quality services and to react and respond to chronic challenges or immediate stressors.^{7 29 34} The analysis of the concept of resilience in general, in contrast to the more specific definition of HSR, is shown in online supplemental table S3.

Figure 2. Frequency of health system resilience Actions and capabilities (A) and the purposes of health system resilience (B) in HSR definitions.

Definition of Governance

13 definitions of governance were identified, three of which highlighted the difficulty in defining and measuring governance and the lack of a universally accepted definition/framework35,37 (table 2).

Table 2. Governance definitions.

Article	Definition
Afolabi et al42	‘Leadership and governance ensure that health authorities take responsibility for steering the entire health sector, while providing effective regulation through a combination of guidelines backed up by legal measures and enforcement mechanisms.’
Barasa et al11	‘Governance is used here to mean the rules and processes that guide operations and affairs of organisations.’
Blanchet25	‘The governance of health systems (ie, how health systems function and perform and how their actors interact with each other)’
Blanchet et al16	‘Governance relates to the implicit and explicit rules and institutions that shape power, relationships between actors, and the actions of these actors.’
Fridell et al40	‘The focus on leadership and governance in both the published literature and by experts is not surprising, given that governance is concerned with how a health system and its actors function and perform.’
Kluge et al39	‘Governance is about making and implementing collective decisions. It is therefore vitally important to health policy and implementation and is a pivotal, yet often underestimated, enabler for leading a health system in times of emergencies, preventing them from becoming a crisis.’
Mfutso-Bengo et al38	‘Leadership and governance involves ensuring strategic policy frameworks exist and are combined with effective oversight, coalition building, regulation, attention to system design and accountability.’
Rogers23	‘Governance and leadership help to ensure that everyone works towards a common goal, including cooperation across health systems.’
Sagan et al24	‘…, we mean here governance in the broader sense, that is, going beyond the governance of the health system alone. ……Governance has also been identified as ‘the mortar that binds all other components together’, rather than a standalone function. It creates trust in the system. As such, it enables the other functions to work properly and contributes to the strengthening of the system as a whole.’
Sagan et al35	‘Governance – the way decisions are made and implemented – enables the financing, resource generation, and service delivery functions to operate as intended and in coordination with the rest of the system to achieve maximum overall system performance and, by extension, resilience.’‘Governance is the most important enabler of health system functioning. It provides a foundation and lever for resource generation, financing and service delivery and ensures they operate well and in coordination with the rest of the system. It also extends beyond the health system through interactions between levels and actors.’‘As a result, there is no single concept of health system governance with a unanimously accepted framework.’
Sagan et al41	‘It is perhaps the most complex area to unpack, not least because there is no universally agreed concept of health system governance, but also because of its enabling role for all the other health system functions and the various linkages and feedback loops this creates.’
Saulnier et al18	‘Governance aims to drive the behaviour, priorities, interactions, participation, accountability and decisions of system actors; power influences the ways in which actors can participate in and contribute to governance.’
Huckel Schneider37	‘Governance and Leadership’ is frequently cited as a key pillar of strong health systems—although it has proven difficult to define and even harder to measure.’ ‘Health system governance can be defined as the structures and institutions that determine the roles and responsibilities of and relationships between various health system actors in taking and enacting policy decisions.’‘Framing health governance as the organisation of relationships between systems parts helps guide the construction of an empirical basis for health system improvement—and ultimately the further development of the Universal Health Coverage agenda.’

Five articles defined health system governance, five provided a general governance definition^{11 16 18 24 38} and three focused on both general and health system governance.^{23 35 39} All definitions conceptualised governance broadly as encompassing rules and processes,^{11 16} structures and institutions^{16 37} and how a health system functions and performs,^{25 40} while some viewed it as the enabler of health system functioning.^{24 35 41} In addition, several definitions of governance were related to leadership.^{23 37 38 40 42}

The main governance functions included in these definitions were decision making and implementation,^{18 35 37 39} the determination of roles and responsibilities of different actors and the interactions between them,^{16 18 25 35 37 38} the enablement of other components to work^{24 35 41} and their coordination^{23 24 35 41} (table 2). Moreover, governance was considered important to improved system performance and resilience,³⁵ to strengthen the system,^{24 37} to prevent emergencies from becoming a crisis³⁹ and to achieve UHC.³⁷

Relationship between health system resilience and shock

HSR was almost always applied to acute shocks, sudden in occurrence and transient in nature such as the Ebola disease outbreak in West Africa and the COVID-19 pandemic.47 11 12 16 18 23 26,30 34 41 43 However, resilience to everyday ongoing chronic challenges, enduring and recurrent over long periods of time, called ‘everyday resilience’, was also related to HSR. Several authors argued that developing ‘everyday resilience’ may promote the system’s ability to face acute shocks.47 11,13 26 30 40 43 50 52 Everyday resilience is described, just like resilience, as arising from a combination of absorptive, adaptive and transformative strategies.^{7 11 43} Yet, strategies proposed to build HSR differed depending on whether the shock is acute and unanticipated. It is often argued that variations in shocks, contexts and health systems mean that there is not a ‘one-size-fits-all’ strategy to build resilience. However, an adaptable ‘all-hazard’ approach may work due to common vulnerabilities.^{32 51} For example, a standardised methodology to test resilience to various types of shocks has been proposed by the WHO,²³ which proposes a flexible HSR toolkit that is adjustable to the national and local context.³³ A shock may affect either the whole system or only some of its components and produce cascading effects across different systems.⁴⁶ As a result, it is argued that adopting a ‘system-of-systems’ approach is useful, instead of analyses limited to single levels of health systems (such as the building blocks approach).⁴⁶ It is also argued that health system resilience should be assessed ‘holistically’, as all components are interconnected and crucial.⁵³

In the literature, building HSR requires several core capacities. In addition to reactive capacity, a key resilience-building strategy is a proactive vision, focused on anticipating and preparing for potential shocks.26 7 11,13 16 18 23 The occurrence of a shock is not necessary for a health system to be considered resilient: a resilient health system is governed to understand the range of potential shocks and seeks to reduce the risks if they happen.^{6 24 27 41} Resilience approaches based on ‘what if’ adverse scenarios were examples of the operationalisation of this forward-looking approach.^{13 23} This highlights the essential characteristics of foresight and anticipation in the conceptualisation of governance of health systems.

It is argued that features of the shock, such as its nature, stage, scale and impact, influence the type of resilience strategy to be employed.^{6 7 12 53} These theories hypothesise that the severity of the shock influences how much resilience is required to effectively respond, while the scale and duration of the shock mostly impact health system performance.⁶ The intensity of the shock is believed to affect the amount of change that the health system needs to introduce.^{6 30 44} While absorptive strategies are often implemented to face small shocks like the recurrence of influenza each year, adaptive strategies are used in response to more intensive shocks not manageable with absorptive strategies. Finally, transformative capacities are for major shocks where the system would not survive without a change of the core functions and goals.^{2 6 12} The type of adaptation is influenced by the different causal factors or drivers of the shock.⁴³

Although HSR is a key element to reduce system vulnerability to shocks,⁷ maintaining health system performance may not always be possible in case of severe shocks.³³ Understanding the type of shock which affects the health system is important to evaluate the scope of potential shock effects⁴⁶: the recovery trajectory of health system performance is influenced by the nature, severity and duration of a shock, the prior resilience status of the system and its underlying structures and processes including elements amplifying or dampening shock effects.⁴⁶ From a complex systems perspective, the effects of a shock depend also on the phase of the adapting cycle theory (ie, growth conservation, collapse, renewal or reorganisation) in the system, and it is greater if the system is unstable.⁴⁶ Resilience is not only shaped by the nature of the shock but also by the state of the system. Some routinely collected measures can help evaluate the trajectory of resilience after a shock, in terms of the extent and speed to return or exceed the baseline.¹³

Overall, a recent shift has happened from a static equilibrium conceptualisation of resilience to a more dynamic understanding of the concept. In other words, HSR is not simply ‘bouncing back’ from a shock to the state before the shock, but it is instead an active and complex process within a dynamic health system that evolves and transforms, ideally into something better.^{7 11 24 27 30 35 40 41 52}

Relationship between shock and governance

The literature often conceptualises governance of health systems as essential for an effective and appropriate response to acute shocks,^{11 30} as it affects all other health system functions.²¹ However, its role is often overlooked during a crisis.^{24 30} Chronic shocks could also be favoured by weak and unstable governance and leadership and, in some cases, by intentional choices made by international, national or local actors.^{4 43} Although changes in governance may result in desired and/or undesired consequences, they are rarely identified as disturbances in the health system literature. A possible explanation is that their principal features (ie, less sudden and more structured) do not fit the common usage of the term shock.⁴³ For example, institutional collapse might lead to large-scale disruptions of the system as seen after the mortality crisis that followed the collapse of the Soviet Union.⁵⁹

Several ‘good’ governance principles have been identified. The loss of trust in institutions affects governance during times of crisis,³⁰ with transparency considered particularly important.¹¹ However, during an acute shock, such as the COVID-19 pandemic, which requires fast responses, it may be challenging to maintain governance attributes that characterise normal times such as transparency, legitimacy and public accountability.^{11 35 41} The same challenge applies to maintaining equity in access to health services, which may be diminished during a shock.⁴⁰ Emergency interventions are considered as opportunities for increased corruption and abuse.^{35 41} These risks can be mitigated by anticorruption institutions, which ensure the transparency of procurement changes.^{35 41} This is important as a system of governance may be severely affected during a crisis.⁶⁰ These conceptualisations emphasise the fact that institutions may need to be designed to govern in ways that resilience may be maintained during crisis.

Relationship between governance and resilience

Mutual influence between governance and HSR

It is argued that governance and HSR influence one another⁴⁰: governance is vital to a resilient response to both acute and chronic shocks^{11 24 30 35 40 41 61} and in turn governance may be affected by the lack of HSR.^{18 40} Governance enables other health system building blocks (eg, financing, resources generation and service delivery) to function as intended and in coordination with the rest of the system.^{2 23 24 35 40 41} Indeed, HSR is mainly a ‘function of the actions and decisions taken by individuals, networks and groups managing these systems’,¹⁶ who need to manage the three aspects of resilience (ie, to absorb, to adapt, to transform) and to decide when to shift from one to another.⁶ While there is consensus on the importance of governance strategies for building HSR, how governance can foster resilience across different types of shocks remains undertheorised.³⁰ The lack of a single accepted governance framework,^{2 35 37 41} the difficulty in evaluating governance performance,^{24 37 41} the continuous evolution of governance over time² and the unpredictability of shocks³⁰ may all contribute to this gap in understanding. Overall, investments related to governance for enhancing resilience are not easy to identify.^{2 30 41}

Mechanisms linking governance to health system resilience

The literature proposes several mechanisms by which governance may influence HSR. A proposed framework identified knowledge, the ability to anticipate and cope with uncertainties, the capacity to manage interdependence, and legitimacy as the main interlinked dimensions.^{16 25} Alternatively, interlinked elements to achieve HSR were leadership, ethics and governance.³⁸ The lack of any of these elements cannot be easily compensated by the other and in turn affects the other two elements.³⁸ As discussed above, governance affects all other system dimensions, and good governance is essential to enable success in actions across all building blocks.³⁰ Managing the whole shock cycle requires key governance functions such as effective and participatory leadership, coordination and effective information systems and flow, while some specific skills may only be relevant to a specific stage.⁶ These differences are sometimes captured by ‘social capital’ as the central driver of resilience.⁴

Most of the included papers discussed governance in a broader way since it is difficult to isolate health system governance. In fact, ‘the governance function goes beyond the health system through interactions with other sectors, since population health is largely determined by actions outside of the health sector’.³⁵ Additionally, the management of large acute shocks often falls outside the control of the health system itself. Consequently, ‘the health system and public health leaderships have to find ways to interface with other systems and the wider political context’.⁴¹ Health systems are embedded within other complex systems (ie, political, economic, social systems).^{16 25} This means that shocks originating from a sector (ie, economic or environmental sector) can impact another sector’s outcomes (ie, mental health).³⁰ In consequence, ‘health actors must engage at higher levels of governance and must accept that health may not be the lead in these processes’.³⁰

Governance structure and HSR

A resilient health system establishes a clear governance structure prior to a shock,^{12 35 38 39 48 49 51} and it should be able to adapt in a flexible way to ongoing circumstances.²³ Critically, resilient health systems shift their governance structure in favour of a common goal as a response to a shock.⁴ Different governance approaches could be adopted according to the context. For example, during an acute, sudden uncontrolled and rapidly expanding shock (ie, Ebola outbreak), centralised, hierarchical, top-down, military-style governance is often adopted, at least in the early stages.^{30 35 37} This solution is often legitimised by declaring a state of emergency and/or enacting emergency legislation⁶ which allows the creation of special committees.³⁵ Centralisation could increase the efficiency at the onset of a crisis by ensuring the coordination at all system levels and across different sectors.^{6 30 41} However, it could limit the inclusion of experiences and values of community and people involved in the response, which may affect effectiveness in the long run.³⁰

Decentralisation of health systems and HSR

As happened in many countries during the COVID-19 pandemic, in later stages of the acute shock, there is often a shift to decentralised governance with the preservation of the coordination role of central governments.³⁵ Decentralisation may promote resilience in making the system responsive to evolving situations and changing needs, as a distribution is territorially sensitive,⁴¹ control empowers local actors¹¹ and ‘provides the necessary flexibility that facilitated timely responses to everyday challenges and in times of crisis’.^{4 11} However, decentralisation may also contribute to the confusion of roles and responsibilities with consequent inefficiencies due to coordination failures and efforts duplication.⁴ There is currently no consensus on whether a centralised ‘top-down’ or decentralised ‘bottom-up’ approach is the best, with two approaches ‘having merits at particular points in a crisis’.^{30 35 41}

Other relevant forms of governance encompass polycentric governance, a system of decision-making where multiple governing bodies interact to make and enforce rules within a specific policy area or location. Polycentric governance may foster resilience and spread the impact of a shock.⁴⁰ Intersectional governance is key to building health system resilience against shocks.³⁰ A common health-related authority through international agreements would improve global coordination and enhance cross-border resilience.^{39 62 63} Finally, appropriate and effective multilevel governance, conducted across levels (at each health system function, at overall health system and beyond health system), is considered important³⁵ for a resilient response, and it is considered more important than technical capacity to respond.²⁴ In fact, an effective governance design can safeguard against poor political leadership by establishing clear health policies and limiting politicians’ scope of actions.³⁵ Health systems governance needs reshaping and strengthening to account for diverse contexts and incorporate multilevel and multi-stakeholder approaches^{37 43} ensure flexible structures that adapt to changes, promote innovation and prevent abuses; and build and reinforce relationships within and outside the health system.^{35 43}

Legislation is often considered a crucial instrument of governance to enhance resilience, as an effective response relies on the transparency of pre-existing legal authority.^{6 32 35 51} Emergency legislation outlining rapidly deployable emergency powers is crucial.^{32 35 41} Laws enacted during stable periods are believed to allow time for consensus building.^{12 41} However, a legal framework alone is considered insufficient; a robust judicial system is essential to ensure officials operate within established laws.³⁵

Governance functions that promote HSR

The main features of governance functions to promote HSR are shown in online supplemental table S4.

Important responsibilities of governance and leadership cited were accountability,^{16 23 30 33 35 38 40 41 58 61} transparency,611 12 30 35 38,41 legitimacy,^{26 35 41 46} equity,1238,40 responsiveness¹² and integrity,⁴¹ as well as the ability to monitor and evaluate other system components (eg, financing, resource generation, service delivery).^{12 27 38} Governance practices of resilient systems should be inclusiveness,¹¹ flexibility^{11 12 41 48} and trust in the system.^{4 26 41} Another important element to HSR against acute and chronic shocks is effective leadership capacity (see online supplemental appendix 1).

Discussion

This review analysed conceptualisation and theories on how governance influences HSR to shocks and considered the different ways in which these terms are defined in the literature. Many of the analysed articles used the WHO building blocks and functions, as well as complex adaptive systems, as their foundation. Our review shows that there is a diversity of views and theoretical approaches to understand HSR. Yet, there are sets of important features of existing conceptualisation and theories that arise from our analysis. First, existing scholarship typically adopts a ‘prescriptive’ approach—that is, they focus on normative descriptions of the ‘capacities’ that should be developed to promote HSR. Many authors, for example, highlight the need to adopt a ‘system-of-systems’ approach (as opposed to analysing isolated health system components). Second, there is generally consensus that the impact of governance on HSR to shocks depends heavily on the context; the extent to which the shock is acute/chronic and unanticipated/expected; and on the phase of the adapting cycle in which the shock takes place. Third, most conceptualisations/theories emphasise the value of a forward-looking approach that anticipates different scenarios to maintain resilience. Yet, these theories also emphasise how ‘good’ principles of governance in normal times may not always be easy to maintain and effective to respond to acute shocks. They emphasise, therefore, the need to develop institutions that are prepared to respond to shocks while at the same time being able to perform regular functions in normal times. Fourth, the literature emphasises the bidirectional relationship between governance and HSR: while governance can foster resilience, resilience can also affect governance. However, the mechanisms underlying this bidirectional relationship remain undertheorised. Fifth, some theories argue that the interaction between multiple government bodies, as in polycentric governance, may foster resilience to shocks. Finally, there is no consensus on whether a centralised ‘top-down’ is more effective than a decentralised ‘bottom-up’ approach to respond to shocks. While decentralisation may promote resilience, it may also contribute to role diffusion during crises.

Shock, governance and HSR relationships analysis

The relationship between resilience and governance is often conceptualised as a three-way relationship where the shock cannot be removed from the picture. Theories highlight the need for a more detailed articulation of this triad in empirical research. Conceptualisation/theories suggest that pre-established strategies to HSR need to consider context: that is, the features of the health system, the ‘larger’ systems in which the health system is nested, the type and intensity of shock that occurs and the forms of governance applied in normal times and during emergencies. Our review does not identify a single form of governance to foster HSR; instead, ‘perspective’ theories emphasise the importance of considering the context, for example, type of shock and the different phases of the shock cycle.

The literature has started to articulate how resilience depends on the type of shock, yet this question remains undertheorised. For example, some theories suggest that absorption is the first line of defence against shocks, while adaptation and transformation allow for broader and long-term change in the system. An important question is whether a shock can also accelerate change that was required but not possible because of the usually gridlocked political processes in normal times.

While many theories emphasise the importance of building HSR against chronic shocks (everyday resilience), there is a lack of theorisation on this question, as most conceptualisations focus on acute shocks. Everyday resilience may prevent those small chronic shocks from becoming severe and promote resilience against acute shocks,¹¹ but (governance) strategies against chronic shocks have not received much attention in the literature. Even more, there is a need to understand how governance, including the different meanings of the governance of global health, could contribute to HSR in situations when it is itself the source of the chronic disturbance, which negatively affects the system and its resilience.⁶⁴

The literature in our review highlights the importance of governance to enhance HSR. However, there are different views on the boundaries of governance: only a few articles focus on governance and leadership beyond the scope of the health system. This could be related to the common adoption of the WHO building blocks as a conceptual framework in many of the articles reviewed.²¹ However, some similarities can be identified for shocks and HSR concepts: the word ‘shock’ is usually used to describe acute, sudden disruptive events while the terms ‘challenge’ and ‘stress’ are employed for chronic ones; most of HSR definitions agree in considering it as a system capacity and 95% of them include the concept of shock (acute or chronic) in their formulation suggesting that a (potential) shock is a requirement to talk about resilience. Indeed, actions to build resilience can be adopted before, during or after a shock. Governance was heterogeneously defined, which may reflect different disciplinary backgrounds about governance and suggests a potential lack of engagement of the public health literature with the governance literature and, more specifically, the complex and sociohistorically situated construction of political systems. However, governance definitions put the focus on the role assigned to it in decision making and implementation and, in the identification and coordination of the actors. Understanding the relationship between resilience and governance requires looking beyond public health to incorporate insights from political science, economics and social sciences, where governance has been extensively theorised and conceptualised.⁶⁵

Overall, the examination of the definitions highlights the importance of defining the terms. More attention should be devoted to the context of the shock, the understanding of the governance system and the resilience of what. As a result, there is an overly strong focus on leadership within the health system, which somewhat disregards the complex links between health and other (non-health related) systems.

Two key objectives of health system resilience considered in the literature are improving health system performance and strengthening the health system. These are also the objectives of health system governance (figure 2 and online supplemental table S2). The following question, therefore, arises: do HSR and governance share common objectives, and is governance a necessary condition for resilience to achieve these objectives? A potential implication would be that HSR cannot be conceptualised without considering governance. On the other hand, this may also reflect a lack of rigorous language and definitions of these terms in the literature.

As shocks often transcend physical and sectorial boundaries and health governance functions extend beyond the health system, it has been argued that conceptualising HSR requires a ‘multisystemic’ (ie, systems of systems) resilience approach.⁶⁰ The fundamental question is how the resilience of a system depends on the resilience of its parts (or one of its parts) versus results from interaction between the parts. Most frameworks in public health place the primary responsibility on the state (public health authorities at the local or national levels), but the COVID-19 pandemic highlighted the need for coordination at the regional and global levels. Further research should approach resilience from both regional and global perspectives, while incorporating additional governance questions, such as the influence of governance on the financialisation of the health system, the dynamics of public-private partnerships and its role in promoting or undermining equity within the health system and the society and the role of cross-national institutions that shape the ability of national state governments to respond to shocks. Finally, our review also highlights that resilience is influenced by the timing of implementation and cessation of measures, as early or delayed actions can influence a system’s ability to absorb, adapt and recover from disturbances. For example, during the COVID-19 pandemic, measures were often adopted too late and then maintained for too long. A critical question is therefore how to translate the dynamic understanding of resilience in complex systems into resilience capacities.

Strengths and limitations

Our literature review covers a wide literature, ranging from international publications to grey literature, including publications from non-governmental organisations. Our results support the need for further research aimed at defining comprehensive approaches/models for resilience and governance that take into account the context where they act and the potential shocks that affect a given health system.

However, some limitations should be considered. First, given that only English papers were included in this review, we may have excluded valuable theoretical papers, particularly those from low-and-middle income countries (LMICs), which may not be published in the international literature. Overall, our review highlights the fact that the literature on HSR theorisation has taken place in high-income countries and emphasises the need for more theory development emerging and focusing on LMICs. Second, our review primarily included articles from the public health literature, which may explain the emphasis on ‘prescriptions’ and solutions, rather than rigorous theorisations of concepts from a disciplinary standpoint. Third, most of the literature focused on public health goals, ignoring the potential tension with other societal goals (which may ultimately undermine or delay the public health response).⁶⁶ For example, most theories did not consider the broader set of actors and institutions outside the health system that have the power to impact health system resilience.⁶⁴ Health systems and policy are inherently political, competing for attention and resources. This political economy of health remains largely undertheorised in the field of public health.⁶⁷

Conclusion

There is an extensive literature conceptualising the link between governance and HSR to shocks. Most of these theories adopt a solution-oriented approach that identifies features of governance that may be ‘prescribed’ during shocks to enhance resilience. Our review highlights the need for more theorisation on the complex mechanisms underlying the bidirectional relationship between HSR and governance, drawing on diverse perspectives and exploring avenues rooted in political science, political economy and public policy theories. First, political economy approaches should expand beyond nation-states to examine transnational actors, resource dependency and power dynamics within health systems. There is also a need to extend political economy theories to incorporate the role of global health actors (eg, WHO, World Bank) and their influence on HSR during shocks. Second, theories should explore how environmental and ecological dimensions, such as climate-related shocks and planetary health, can highlight interdependencies with broader systems. Theories such as adaptive governance and transformability, which are related to the study of resilience in social-ecological systems, have proved useful to understand intertwined ‘natural’ and social ‘issues’.^{68 69} Theories that capture the dynamic and complex relations among resilience governance and shocks, considering not only health systems but also broader (non-health) systems have the potential to foster a more integrated view of resilience to interconnected shocks. Third, a range of theories from the constructivist approach (and political sociology) can provide more insights into how resilience is framed and operationalised as a result of norms, values and political ideologies. Fourth, more theory development from LMICs is essential to deepen understanding of local adaptations and the unique dynamics of postshock recovery trajectories in LMICs and contribute to decolonise global health. Ultimately, such efforts at theorisation will not only support empirical research but also adapt our governance framework to address future global health crises.

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information.