Leading under uncertainty: a qualitative case study of crisis leadership in Canada’s health system

Abstract

Background

Prolonged public crises challenge leaders to make high-stakes decisions under conditions of uncertainty while sustaining organisational performance and public trust. Understanding how leaders navigate these complex demands is essential for strengthening health system preparedness and resilience.

Objectives

The goal was to explore and unpack lived leadership experiences during a public crisis situation and to unpack how contextual factors shape their behaviours under uncertainty and prolonged disruption.

Design

This study was conducted in Canada during the middle phases of the COVID-19 pandemic (October 2020–March 2021). Canada’s healthcare system operates within a federal-provincial framework, with public health responsibilities spanning national, provincial and local levels. 16 senior leaders with provincial or national responsibilities across healthcare delivery, government, academia and research participated in the study. Purposive sampling was used to achieve variation across organtional and geographic regions. A qualitative case study design was employed, with semistructured interviews serving as the primary data collection method and thematic analysis used to analyse the data. The analysis was informed by Sriharan et al’s Crisis Leadership Framework and reported in accordance with the Standards for Reporting Qualitative Research.

Methods

Videoconference interviews (40–50 min) explored leadership challenges, decision-making under uncertainty, communication, and contextual enablers and constraints. Transcripts were analysed using Braun and Clarke’s six-phase approach with triangulation and reflexive documentation to enhance rigour.

Results

Leaders described multifaceted roles that spanned clinical, administrative and advisory functions. Six leadership domains emerged from the analysis: adaptive leadership, learning-oriented leadership, systems thinking, psychological stewardship, implementation agility and strategic communication. Leadership enactment was shaped by structural integration and centralisation, which enabled real-time situational awareness and coordinated resource deployment. Rapid cycle decision-making was supported by emergency legislative flexibility, multidisciplinary evidence synthesis structures and peer learning networks. Constraints included systemic fragmentation, bureaucratic inertia, data sharing limitations and sustained workforce fatigue, all of which influenced leaders’ ability to enact these domains under pressure.

Conclusions

Effective crisis leadership depended less on static competencies and more on dynamic enactment shaped by structural, operational and relational mechanisms. The findings extend the crisis leadership framework towards Dynamic Capability Leadership for Learning Health Systems, positioning leadership as a strategic capacity to sense emerging challenges, seize opportunities through rapid and accountable decision pathways and transform processes to sustain performance under uncertainty. These capabilities were enacted through adaptive and learning-oriented leadership, implementation agility, systems thinking, psychological stewardship and strategic communication, all amplified by structural integration and evidence-synthesis infrastructures. Overall, the findings highlight crisis leadership as a system property shaped by context, coordination and the ability to translate evolving knowledge into timely action.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• This study drew on purposive elite sampling of 16 senior leaders across five Canadian provinces, enabling depth of insight but limiting transferability to a single national context.

• Data triangulation, analyst triangulation and reflexive documentation strengthened analytic rigour, but the study’s exclusive focus on senior leaders excluded the perspectives of frontline staff and junior management.

• Interviews were conducted during the middle phase of the pandemic, which may have introduced recall bias given the chaotic nature of the earlier response period.

• Time constraints during some interviews, conducted between emergency briefings, may have limited the depth of exploration for certain topics.

• Direct quotations were limited and presented as paraphrased summaries to protect the confidentiality of a small, identifiable group of elite participants, consistent with established strategies for elite interviewing.

Introduction

Extended public crisis such as the COVID-19 pandemic requires leaders to balance immediate operational demands with long-term organisational sustainability within complex adaptive health systems, moving beyond linear, event-based models of crisis management towards more dynamic, relational and system-oriented approaches.^1–5 Foundational crisis management and crisis communication research emphasise anticipatory planning, structured response processes and reputational protection,^{6 7} whereas work on public leadership under pressure highlights how crises are inherently political, contested and value-laden, requiring leaders to navigate scrutiny, blame and legitimacy alongside technical problem-solving.¹

Crisis leadership reviews in healthcare identify task-oriented competencies (rapid decision-making, coordination, resource allocation and procedural agility); people-oriented competencies (psychological safety, relational coordination and emotional stewardship) and adaptive competencies (experimentation with incomplete data, humility and systems thinking), all shaped by political, structural and cultural contexts.^{8 9} However, these studies catalogue what leaders should do but rarely unpack how they enact competencies under sustained uncertainty.⁹

This study responds to that gap by examining lived experiences of senior leaders during COVID-19’s middle phases and unpack the competencies and enactment of competencies. In this study, ‘leadership competency’ refers to underlying capacities as defined by Boyatzis,¹⁰ who conceptualised competencies as stable, underlying characteristics that predict superior performance across situations. ‘Enactment’ describes the observable behaviours through which these capacities are expressed in context.¹¹

Methods

Study design

We used a qualitative single-case study methodology to unpack the lived experiences of health systems leaders during the COVID-19 crisis in response to evolving system pressures, uncertainty and organisational complexity.¹² The case study design is well suited to examining complex social phenomena in real-world contexts and elucidating how and why leaders enact behaviours amid contextual constraints.¹² The analysis was informed by the Sriharan et al Crisis Leadership Framework,⁹ which served as a sensitising lens to guide interpretation of task, people and adaptive leadership within political, structural and cultural contexts while allowing new themes and explanatory mechanisms to emerge from participants’ accounts.¹³

Study context

This study was conducted in Canada, where the healthcare system is constitutionally divided between federal and provincial jurisdictions. This division means that the provinces are primarily responsible for healthcare delivery and public health functions, whereas the federal government oversees areas such as border control and inter-provincial coordination. This structural context shaped the leadership experiences and system mechanisms observed in the study. Data collection occurred from October 2020 to March 2021 during the middle phases of the pandemic, a period after the initial emergency response and before vaccine scale-up, to capture sustained leadership under prolonged uncertainty. This timing aligns with calls for mid-crisis research to examine adaptive mechanisms beyond initial sensemaking.¹³

Sampling and recruitment

We used purposive elite sampling to recruit senior healthcare leaders.¹⁴ This approach targets information-rich cases with deep system knowledge, optimal for revealing leadership mechanisms in complex policy contexts.^{15 16} Eligible participants held senior leadership positions in healthcare delivery organisations (hospitals or regional health authorities), government policy settings (ministries of health or public health agencies) or academic institutions contributing to pandemic decision-making.

Potential participants were identified through publicly available information regarding COVID-19 leadership roles, snowball sampling from initial interviews and the research team’s professional networks. We contacted 24 individuals and received 16 positive responses (a 66.7% participation rate). Reasons for non-participation included scheduling conflicts (n=6) and non-response (n=2).

16 senior healthcare executives were interviewed, representing a purposive sample sufficient for thematic saturation in a single-case study of elite informants within a bounded system. This aligns with empirical guidelines for research involving homogeneous, high-status participants where data depth exceeds breadth.¹⁶

Data collection

Semistructured interviews were conducted via videoconference because of pandemic restrictions. Two experienced qualitative researchers (AS and TS) conducted all interviews through Microsoft Teams or Zoom using a semistructured guide (online supplemental file 1). The guide was developed by AS and TS in parallel with, and informed by, an earlier review of the medical and business literature on crisis leadership during pandemics,⁹ which identified the task, people and adaptive competency domains that shaped the interview questions; it was not formally pilot tested. The guide explored leadership challenges, strategies used during the pandemic, decision-making under uncertainty, communication and team management, contextual enablers and barriers, and reflections on effective leadership practices. Interviews lasted 40–50 min and were audio recorded, transcribed verbatim and anonymised.

Data analysis

Data analysis followed Braun and Clarke’s reflexive thematic analysis,¹⁷ with findings from our parallel knowledge synthesis of crisis leadership frameworks⁹ serving as a sensitising lens rather than a deductive coding framework. This approach enabled identification of mechanisms, which are the processes linking competencies to enactment, while prioritising data-driven patterns across participants’ accounts. The first author (AS) led the analysis, including development of the initial codebook, primary coding and iterative theme construction. The second and third authors reviewed the codebook, examined coded excerpts and provided feedback on theme refinement to ensure coherence, credibility and alignment with the analytic aims.

Mechanisms were identified through iterative pattern analysis, comparing coded segments across participants and mapping how leadership behaviours interacted with contextual features of the Canadian health system. Rigour was strengthened through analyst triangulation (multiple researchers reviewing interpretations), data triangulation (comparing patterns transcripts) and reflexive documentation throughout the analytic process.

Themes were refined through repeated comparison with the dataset, and an audit trail documented analytic decisions. Analysis adhered to Standards for Reporting Qualitative Research guidelines.¹⁸

Researcher reflexivity

Given the small number of senior healthcare executives interviewed (n=16) and their elite status within a specific regional system, direct quotations were limited to protect participant confidentiality. Instead, findings were presented through paraphrased summaries that preserved meaning while obscuring identifiable phrasing or context, consistent with established strategies for elite interviewing.^{14 19}

Results

Participant characteristics

The final sample (n=16) consisted of healthcare organisational leaders (n=5), government policy leaders (n=5) and academic or research leaders (n=6). Participants represented five provinces, included 11 men and 5 women and averaged more than 15 years of leadership experience. 11 participants were trained health professionals (physicians or nurses) and 5 were policy or administrative leaders. Leaders frequently described performing multiple concurrent roles (‘double duty’), including balancing clinical or research and administrative responsibilities, chairing provincial and federal advisory tables and shifting from organisational leadership to regional or provincial incident management roles. This multiplicity of responsibilities reflected the breadth and intensity of leadership demands during the pandemic.

Key findings

Participants described a set of core competencies (table 1) that shaped their leadership during the crisis. Each competency was enacted through specific behaviours that supported decision-making, coordination and team stability under conditions of sustained uncertainty.

Table 1. Crisis leadership competencies in action.

Competency	Context	Enactment	Outcome
Ambiguity tolerance	Prolonged uncertainty	Decisive action under uncertainty	Timely decisions
Intellectual humility	Evolving evidence	Transparent acknowledgement of uncertainty	Adaptive decisions
Structural knowledge	System fragmentation	Dot-connecting across silos	Coordinated response
Emotional anchoring	Sustained emotional strain	Visible stability	Team resilience
Bureaucratic agility	Administrative constraints	Strategic process navigation	Rapid implementation
Evidence sensemaking	Information overload	Strategic filtering and interpretation	Actionable guidance

Online supplemental file provides participant paraphrased quotes.

Ambiguity tolerance: decisive action under uncertainty

Participants reported operating in an environment defined by incomplete information and rapidly shifting conditions. Ambiguity tolerance enabled them to act decisively despite the absence of reliable data. Leaders described publicly acknowledging uncertainty while maintaining forward momentum. As one participant noted, “You have to be able to tolerate a little bit of ambiguity… and accept it publicly” (SP 1). This stance supported timely decision-making, even when evidence was insufficient. Another participant described the need for “comfort in decision making without adequate data” (SP 9). Leaders also emphasised helping others manage uncertainty (SP 8) and adopting a wartime posture that prioritised speed over consensus. As one senior leader stated, “You have to be decisive… we don’t have time for debate” (SP 15).

Intellectual humility: transparent acknowledgement of uncertainty

Participants described intellectual humility as essential for navigating evolving evidence and frequent shifts in guidance. Leaders emphasised the importance of openly acknowledging what was unknown and correcting course as new information emerged. One participant described humility as “remaining open enough to say we don’t know everything” (SP 5). Another noted the inevitability of error, stating, “Be humble from the outset because you’re going to be wrong a lot” (SP 9). Leaders described communicating uncertainties directly, such as “I don’t have an answer that’s perfect… here’s where the uncertainties are” (SP 1). This stance also required self-compassion, reflected in the comment, “We need to forgive ourselves… this is a hard time” (SP 8). Intellectual humility supported trust building and enabled ongoing course correction.

Structural knowledge: dot-connecting stewardship

Participants highlighted the importance of deep structural knowledge of the health system and its interdependencies. This competency enabled leaders to connect fragmented components across hospitals, public health, long-term care and government. One participant described their role as “helping people connect dots across departments and levels of government” (SP 15). Leaders emphasised that actions in one part of the system inevitably affected others, noting, “Nothing happens without impacting other parts… systems that don’t talk can’t plan” (SP 16). Several participants described the system’s fragmentation, including the observation that “We are very fractured. System isn’t really a system at all and really can’t work as a system in the way that it’s been constructed” (SP 10). Structural knowledge allowed leaders to bridge silos and align responses across jurisdictions.

Emotional anchoring: visible stability

Participants described the need to project emotional stability to support teams during prolonged crisis conditions. Leaders recognised that their own emotional state influenced organisational morale. One participant explained, “I cannot be down… the whole senior team will go that way” (SP 2). Others described fostering collective efficacy during shared adversity, such as “Facing a shared calamity… collective efficacy” (SP 8). Leaders emphasised the importance of being a steady presence, with one noting the value of “being that calm presence that helped coordinate the response” (SP 3). Emotional anchoring served as a stabilising force that helped prevent organisational fatigue and maintain team cohesion.

Bureaucratic agility: strategic process navigation

Participants described using their deep understanding of bureaucratic structures to accelerate action in a system typically slowed by procedural complexity. This agility enabled leaders to streamline or temporarily suspend processes that would normally impede change. One participant noted that “Virtual care moved in ten days after ten years of no progress” (SP 13). Another described rapid operational restructuring, including “Moving entire clinical units in twenty four hours… we learned quickly” (SP 4). Bureaucratic agility allowed leaders to mobilise coordinated action across municipal, provincial and federal systems during the crisis.

Evidence sensemaking: strategic filtering and interpretation

Participants described managing overwhelming volumes of emerging scientific information while facing public and political scrutiny. Leaders emphasised the challenge of filtering contradictory and undigested evidence, noting “Too much contradictory, undigested raw evidence” (SP 15). Effective sensemaking required disciplined interpretation and clear communication. One participant described presenting information neutrally to support sound decisions, stating, “Presenting things dispassionately helps get to the right decisions” (SP 1). Others emphasised integrating diverse inputs and articulating a coherent direction, reflected in “Listening, synthesizing, and communicating a decision or vision” (SP 3). Evidence sensemaking enabled leaders to make informed choices during periods of scientific volatility.

Enabling mechanisms

Participants identified several system-level features that supported leadership enactment. Structural integration and centralisation enabled real-time situational awareness and facilitated coordinated resource allocation. Legislative flexibility and emergency authorities allowed leaders to bypass standard procurement and administrative processes, enabling rapid implementation of innovations such as virtual care. Expert collaboration structures, including scientific advisory tables and modelling groups, provided shared interpretations of evolving evidence and supported consistent decision-making. Peer networks further enabled real-time learning and validation of emerging strategies in the absence of established playbooks.

Constraining mechanisms

Leaders also described system-level constraints that impeded their ability to enact crisis leadership competencies. Systemic fragmentation resulted in siloed operations and hindered coordinated action across sectors. Bureaucratic inertia and longstanding risk-averse cultures slowed adaptation to the pace required by the crisis. Data-sharing limitations produced incomplete or inconsistent information, complicating planning and response efforts. Sustained high-intensity workloads contributed to leadership fatigue and reduced stamina over time.

Discussion

This study advances understanding of crisis leadership by showing that effectiveness in prolonged public health emergencies is not defined by static traits or competencies but by the dynamic enactment of those competencies within complex adaptive systems. The findings extend the Sriharan et al⁹ Crisis Leadership Framework, originally organised around task, people and adaptive domains, by demonstrating that these operate as capabilities that expand or contract depending on structural integration, culture and context. Crisis leadership is therefore best understood as an ongoing process of implementation, sensemaking and adaptation rather than a checklist of traits.

From competencies to enactments

First, the study clarifies the distinction between leadership competencies and leadership enactments. Existing work has long differentiated underlying skills or competencies from observable leadership behaviour, emphasising that skills matter only when translated into performance in context.¹¹ Consistent with this insight, the findings show that latent skills such as ambiguity tolerance, intellectual humility, structural knowledge and bureaucratic literacy became meaningful only when enacted as urgent decisiveness, public admissions of uncertainty, dot-connecting stewardship or strategic process redesign. Leadership effectiveness in these data resided less in the possession of traits and more in the situated enactment of those capacities under conditions of uncertainty and constraint, echoing behavioural and context-sensitive perspectives on leadership.^{9 11}

Adaptive and learning-oriented leadership

Second, the study extends adaptive leadership theory by detailing how leaders enacted adaptation in practice. Adaptive leadership emphasises mobilising people to tackle complex problems by experimenting and working with uncertainty rather than relying solely on technical solutions.²⁰ Leaders described rapid cycle experimentation, high-stakes decisions with incomplete data, course corrections as evidence evolved, and transparency about uncertainty. These patterns align with adaptive leadership and implementation science principles such as iterative testing, feedback loops and adaptive redesign.²¹ They also reflect a learning-oriented leadership stance, in which public admission of uncertainty and real-time course correction position humility as a strategic competency that enables collective learning rather than a purely personal characteristic.

Psychological stewardship

Third, the analysis suggests that existing people-focused leadership constructs can be reframed as psychological stewardship. Rather than centering solely on communication or team management, leaders in this study acted as emotional anchors, intentionally regulating their presence to prevent organisational despair and sustain collective efficacy. This aligns with scholarship on psychological safety and relational coordination, which highlights how leaders’ relational behaviours and high-quality communication support team learning, resilience and performance.^{22 23} Framing this work as psychological stewardship underscores that the emotional labour required to maintain stability, especially amid virtual fatigue, personal attacks and information overload, is a core leadership function in learning health systems rather than a discretionary interpersonal skill.

Bureaucratic agility

Fourth, the study reframes task-oriented leadership behaviours as bureaucratic agility: the ability to strategically bypass, compress or reconfigure administrative processes to accelerate implementation. Crisis management research shows that leaders frequently rework or suspend normal procedures to enable timely action under pressure.¹ In these data, leaders leveraged the pandemic as a catalyst to deploy long-planned innovations, such as virtual care, that had remained stalled under routine bureaucratic conditions. These practices mirror implementation strategies such as facilitation, de-implementation of low-value processes, and adaptive redesign,²¹ highlighting bureaucratic agility as a critical determinant of whether decisions could be translated into action at the pace required by the pandemic.

Structural integration as an implementation multiplier

Finally, the study identifies structural integration as a determinant of leadership effectiveness. Learning health system frameworks emphasise interoperable data infrastructures, cross-boundary collaboration and continuous feedback as prerequisites for system learning.²⁴ Consistent with this work, integrated systems in this study enabled real-time data sharing, coordinated resource allocation and rapid deployment of interventions, whereas fragmented systems forced leaders into ad hoc committee building, duplication of effort and reactive problem solving. Structural integration functioned as an implementation multiplier that amplified or constrained leaders’ ability to enact their competencies, showing that leadership enactment is coproduced by people, processes and system design.^{1 9 24}

Dynamic capability lens for learning health systems

Building on these findings, we interpret crisis leadership in learning health systems through a dynamic capability lens. Dynamic capabilities refer to an organisation’s ability to sense emerging opportunities and threats, seize them through timely strategic action and reconfigure resources to remain effective in rapidly changing environments.²⁵

Figure 1 synthesises the empirical findings and interprets them through a dynamic capability lens. The findings suggest that crisis leadership effectiveness emerged not simply from possessing competencies, but from their situated enactment under conditions that enabled or constrained action. Adaptive and learning-oriented enactments supported sensemaking under uncertainty; bureaucratic agility enabled emerging knowledge and decisions to be translated into timely action; and psychological stewardship and systems-oriented leadership supported coordination and workforce stability during sustained disruption. Structural integration, data sharing, legislative flexibility, collaborative networks, organisational culture and system fragmentation further shaped leaders’ capacity to enact these capabilities. Viewed through a dynamic capability lens, these findings can be understood as interconnected processes of sensing, seizing and transforming, through which leadership and system conditions jointly shape the capacity of learning health systems to adapt during periods of sustained uncertainty.

Figure 1. Dynamic capability leadership for learning health systems.

Strengths and limitations of this study

This study provides mechanism-focused insights from senior leaders across diverse jurisdictions and roles, capturing how context shaped leadership enactment under pressure. Several strengths enhance the credibility and richness of the findings. The interviews included leaders from integrated provincial systems, fragmented provincial systems and the federal policy arena, offering a comparative view of how structural context influenced leadership behaviour. The timing of the interviews, conducted as leaders were transitioning from the first to the second wave of the pandemic, enabled real-time reflection that preserved the immediacy of crisis decision-making. Leaders spoke candidly about personal strain, emotional labour and the limits of their knowledge, providing unusually transparent accounts of leadership under uncertainty. The dataset also surfaced structural levers that are often invisible in leadership studies, such as the role of centralised data systems in enabling rapid procurement and coordinated action.

The study also has limitations. Several leaders described the early months of the pandemic as a blur, which introduces the possibility of recall bias and smeared timelines. The sample consisted exclusively of senior executives and clinical leaders, meaning the perspectives of frontline workers and junior managers are not represented. Many interviews occurred under significant time pressure, sometimes between emergency briefings or back-to-back virtual meetings, which may have constrained the depth of exploration for some topics. The data also reflect a specific moment in the pandemic, before leaders had the benefit of hindsight regarding long-term outcomes such as delayed cancer care or the durability of virtual care adoption. These constraints shape interpretation but do not diminish the value of the insights generated.

Implications for organisations, leadership development, policy and research

Organisations could strengthen their capacity to sense emerging challenges by establishing cross-functional evidence synthesis hubs and real-time dashboards that monitor scientific signals, operational pressures and workforce well-being. They could also create decisive and accountable pathways for rapid decision-making and resource mobilisation, supported by pre-authorised emergency rules for procurement and staffing. Psychological stewardship could be treated as a core operating capability, with routines that maintain team psychological safety, relational trust and cross-boundary coordination. Mechanisms that connect hospitals, public health, primary care and social services around shared goals and shared knowledge could be reinforced.

Leadership development should shift from trait-based curricula to practice-based capability building. Leadership curricula should be modernized to promote scenario planning and simulation exercises to cultivate adaptive capacity, psychological stewardship and systems thinking literacy to enable leaders to lead under uncertainty.

At the policy and system levels, governments could legislate flexible emergency governance pathways that authorise rapid cycle decision-making while safeguarding transparency. Investments in structural integration, including interoperable data platforms, coordinated supply chains and shared situational awareness, are essential. Workforce resilience indicators could be embedded in preparedness standards.

Future research should focus on unpacking dynamic capability for health systems and test leadership enactment under different crisis scenarios through qualitative and quantitative approaches.

Conclusion

This study shows that crisis leadership is not defined by the possession of individual traits or competencies but by the dynamic enactment of those competencies within complex and evolving system conditions. Leaders were required to navigate uncertainty, synthesise rapidly emerging evidence, stabilise teams under psychological strain and implement innovations at unprecedented speed. Their effectiveness depended not only on internal capacities but also on structural and cultural environments that enabled or constrained their ability to act.

The findings extend the Sriharan et al Crisis Leadership Framework by reframing task, people and adaptive competencies as dynamic capabilities that unfold in real time. These capabilities operated through practices such as rapid cycle experimentation, psychological stewardship, bureaucratic agility and integrative sensemaking. Leadership was enacted through mechanisms shaped by system integration, information flows, team capacity and governance structures, which positions crisis leadership as a system property rather than an individual attribute.

Interpreting these findings through a dynamic capability lens positions crisis leadership as a strategic capacity to sense emerging challenges, seize opportunities through rapid and accountable decision pathways and transform processes to sustain performance under uncertainty. Embedding these principles into leadership development and system design can help cultivate resilient and agile learning health systems capable of addressing crisis scenarios.

Data availability statement

Data are available upon reasonable request.