Transformative Learning Through Implementation: A Realist Case Study of Frostbite Care Protocol Implementation

Abstract

Background

Learning Health Systems (LHS) emphasize continuous improvement through cycles of evidence generation, implementation, and evaluation, yet relatively little attention has been paid to the educational processes through which clinicians and organizations learn during implementation. This study examined how implementation of a frostbite treatment protocol functioned as a process of transformative learning within a LHS.

Methods

We conducted a realist case study of implementation of a frostbite treatment protocol across four tertiary-care hospitals within a Canadian health system. Semi-structured realist interviews were conducted with nine interest holders representing six professional groups and analyzed alongside implementation documents, including clinical protocols, electronic health record order sets, and educational materials. Data were analyzed iteratively using Context–Mechanism–Outcome Configurations (CMOCs), which were synthesized into program theories explaining how implementation generated learning and organizational change.

Results

Eighty-five CMOCs were synthesized into five program theories describing an iterative learning process through which interest holders: (1) recognized deficiencies in existing frostbite care, (2) overcame organizational barriers, (3) mobilized collective action, (4) institutionalized new clinical practice, and (5) reinforced and sustained transformation through improved patient outcomes. A chance clinical encounter functioned as a disorienting dilemma, challenging assumptions about appropriate frostbite care and motivating change. Trust, professional credibility, multidisciplinary collaboration, and distributed leadership enabled interest holders to reinterpret uncertainty, navigate organizational barriers, and develop new clinical pathways. Rather than occurring before implementation, transformative learning emerged through participation in implementation itself. Improved patient outcomes subsequently reinforced confidence, facilitated dissemination, and stimulated continued organizational learning.

Conclusions

This study begins to extend transformative learning beyond formal educational settings by demonstrating how implementation can function as a collective learning process within healthcare organizations. By integrating realist inquiry, transformative learning theory, and LHS, we propose implementation as a mechanism through which clinicians and organizations collectively question assumptions, reinterpret evidence, and sustain system transformation.

Plain Language Summary

Healthcare organizations are increasingly adopting Learning Health Systems, which improve patient care by continuously generating evidence, implementing new practices, and learning from the results. However, researchers know much less about how healthcare teams learn and change while they introduce new ways of working.

In this study, we examined how healthcare teams implemented a new frostbite treatment protocol across four hospitals in Alberta, Canada. We interviewed doctors, nurses, pharmacists, and other healthcare professionals and reviewed clinical protocols, educational materials, and other implementation documents. We used a research approach called realist evaluation to understand what helped the implementation succeed, for whom, and under what circumstances.

We found that healthcare teams did more than introduce a new clinical protocol—they learned together throughout the implementation process. They recognized gaps in existing frostbite care, overcame organizational barriers, worked collaboratively to develop new approaches, and integrated the protocol into everyday practice. An unexpected patient encounter challenged long-held assumptions about frostbite treatment and prompted clinicians to rethink their approach. Trust, collaboration, shared leadership, and early improvements in patient outcomes built confidence and encouraged wider adoption of the protocol.

Our findings show that implementation can help clinicians and healthcare organizations learn, adapt, and improve together. Rather than treating implementation and learning as separate activities, healthcare organizations can use implementation as an opportunity to question existing practices, build new knowledge and capabilities, strengthen teamwork, and achieve lasting improvements in patient care.

Introduction

Health systems worldwide are increasingly expected to learn and adapt in response to rapidly evolving evidence, technological innovation, and changing patient needs.^1,2 Learning Health Systems (LHS) have emerged as a model for continuous improvement by integrating data generation, evidence synthesis, implementation, and evaluation into routine clinical care. Their central premise is that every patient encounter can contribute to a cycle of learning that improves subsequent care.³ While this vision has influenced health system redesign, much of the LHS literature has emphasized the technical architecture required to support learning—including data infrastructure, feedback loops, performance measurement, and implementation processes—rather than the educational processes through which clinicians, teams, and organizations actually change.⁴

Transformative learning theory offers a complementary perspective by focusing on how individuals fundamentally revise their assumptions, beliefs, and ways of understanding the world.⁵ Originally conceptualized by Mezirow, transformative learning proposes that disorienting experiences stimulate critical reflection, dialogue, and ultimately perspective transformation.⁶ Although transformative learning has become an influential framework within health professions education (HPE), its application has largely focused on individual learners in formal educational settings.⁵ Comparatively little attention has been paid to how transformative learning unfolds collectively within healthcare organizations as clinicians negotiate uncertainty, evolve clinical practice, and participate in system change.

Implementation of novel clinical care protocols represents an opportunity for transformative learning. Introducing new models of care requires clinicians to move beyond simply acquiring knowledge to collectively interpreting emerging evidence, challenging established routines, negotiating new professional roles, collaborating across disciplinary boundaries, and adapting care pathways to local contexts.⁷ In this way, implementation becomes not merely a process of knowledge translation, but a catalyst for individual, collective, and organizational learning. Learning occurs not before implementation, but through participation in implementation itself.⁸ These experiences have the potential to reshape professional identities, relationships, and patterns of clinical work, yet they are rarely conceptualized as educational processes. As a result, implementation science and HPE have developed largely in parallel, with limited integration of theories explaining how learning drives sustainable system change.

Realist evaluation provides a useful methodological bridge between these fields. Rather than asking simply whether an intervention works, realist inquiry seeks to explain what works, for whom, under what circumstances, and through which underlying mechanisms.⁹ By examining the interaction between contexts, mechanisms, and outcomes, realist evaluation can illuminate the generative processes through which implementation activities influence both practice and learning.¹⁰ This approach is particularly well suited to understanding transformation within complex adaptive systems, where outcomes emerge through interactions among individuals, teams, organizational structures, and local environments.¹¹

We propose that implementation should be understood not merely as a process of knowledge translation, but as a transformative learning environment in which clinicians and organizations co-construct new ways of thinking and working. From this perspective, successful implementation reflects more than protocol adoption; it represents the development of new collective capabilities, relationships, and patterns of practice that become embedded within the health system.

To explore this proposition, we conducted a realist case study of a frostbite protocol implementation. Frostbite is an uncommon but potentially limb-threatening condition requiring timely, coordinated, multidisciplinary care. It is also a condition that disproportionately affects vulnerable patient populations in urban settings.¹² Although implementation of the treatment protocol, leveraging a special access medication previously only provided in critical care environments, was associated with improved access to advanced therapies and reduced rates of frostbite-related digit amputation,^13,14 these outcomes alone cannot explain why implementation succeeded. We sought to explain how contextual conditions and underlying mechanisms enabled learning, adaptation, and transformation. In doing so, we argue that this case extends beyond frostbite care to offer a broader conceptualization of implementation as a transformative learning process within a LHS.

Materials and Methods

Study Design

This study employed a realist case study design¹⁵ to examine how implementation of a frostbite treatment protocol functioned as a learning process. Realist methodology was selected because understanding implementation requires explanation of the mechanisms through which change occurs rather than simply determining whether an intervention was successful.⁹ Consistent with realist philosophy, the study sought to identify how contextual conditions influenced interest holders’ reasoning and actions to produce implementation outcomes.

The case comprised development and implementation of a frostbite treatment protocol across four urban hospitals within a single provincial health system. The protocol, implemented in February 2019, was adapted from an established model implemented in northern Canada, introduced standardized pathways for the recognition and grading of frostbite injury, specialist consultation, and delivery of therapies to improve timely access to care and reduce frostbite-related morbidity.¹⁶ Rather than evaluating the clinical effectiveness of the protocol, this study sought to explain how recognition of deficiencies in existing frostbite care, the development of a novel treatment pathway, and participation in its implementation collectively generated transformative learning and sustained changes in clinical practice.

The study is reported in accordance with the RAMESES II reporting standards for realist evaluations.¹⁷

Theoretical Framework

The study was informed by LHS principles and transformative learning theory. LHS served as a sensitizing framework during study design by conceptualizing healthcare organizations as continuously learning systems that evolve through iterative cycles of evidence generation, implementation, evaluation, and adaptation.¹⁸ We drew on three LHS frameworks, each applied to varying degrees across our institution, as sensitizing concepts to ensure the interviews captured a broad range of perspectives.^19–21 The three frameworks included the 1) Learning health systems: pathways to progress from the Alliance for Health Policy and Systems Research; 2) The Alberta Strategy fo Patient-Oriented Research Support Unit Learning Health Systems Framework; and 3) Learning Health System Action Framework from the Institute for Better Health. Transformative learning theory, used as an interpretive lens, emphasizes changes in assumptions, perspectives, and professional practice that occur through engagement with challenging experiences.⁶

Realist evaluation provided the analytic lens and explanatory framework linking these perspectives by examining how contextual conditions activated mechanisms that generated individual, team, and organizational learning during protocol implementation.⁹

Setting

The protocol was implemented across four tertiary-care hospitals within an urban Canadian healthcare system. Frostbite is a seasonal, potentially limb-threatening condition. Historically, frostbite received relatively little clinical and organizational attention, and management was largely limited to supportive wound care while awaiting tissue demarcation, with severe injuries frequently culminating in delayed digit amputation.²² This was a substantial shift for several reasons. Iloprost for the treatment of frostbite is a Special Access medication, and its use is supported primarily by physiological rationale and emerging clinical evidence rather than large, randomized trials. As a potent vasodilator with the potential to cause significant haemodynamic effects, iloprost administration had historically been restricted to the intensive care unit (ICU). Implementing this protocol therefore required not only the introduction of a novel therapy but also changes to clinical workflows, interprofessional collaboration, and institutional processes to safely deliver treatment outside of the ICU.

Participant Recruitment

Purposive sampling was used to recruit participants from emergency medicine, intensive care, plastic surgery, nursing, pharmacists, and health system leaders who played key roles in recognizing the need for change, developing, implementing, delivering, and sustaining the frostbite protocol. Participants were selected based on their ability to contribute insights into implementation processes across multiple organizational perspectives. Recruitment continued until sufficient explanatory depth had been achieved to refine program theories rather than until thematic saturation.

Data Collection

Data were collected over a 3-month period using semi-structured realist interviews (see Supplemental Material 1 for interview guide) and document analysis.

The interviews explored participants’ experiences of existing frostbite care, developing and implementing the protocol, critical implementation events, interprofessional interactions, and perceived changes in clinical and organizational practice. Interview questions were informed by initial program theories (see Supplemental Material 2) derived from LHS literature.

Throughout data collection, emerging findings were used to iteratively refine interview questions in keeping with realist methodology.

Interviews explored not only what occurred during protocol development and implementation, but also why participants believed particular implementation strategies succeeded or failed within their local contexts. Particular attention was paid to examples illustrating changes in professional reasoning, interprofessional collaboration, and organizational practices. Interviews were audio-recorded and transcribed verbatim prior to analysis. Transcripts were subsequently edited for readability by removing verbal fillers (eg, “um”, “uh”), false starts, and repetitive speech, and by correcting obvious grammatical errors where these did not alter the meaning or intent of participants’ responses.

Document analysis was undertaken to supplement and contextualize participants’ accounts and to trace the development and implementation of the protocol over time. Documents included versions of the clinical protocol, order sets, and educational materials.

Data Analysis

Analysis followed an iterative realist approach centred on the development and refinement of Context–Mechanism–Outcome Configurations (CMOCs).⁹

Data were coded to identify contextual conditions, resources and opportunities introduced through protocol development and implementation, interest holders’ reasoning and responses, and resulting outcomes. Interview data were particularly useful for examining participants’ interpretations, reasoning, and responses, while documents provided evidence of the historical and organizational context and intended implementation processes. Data sources were analyzed iteratively and compared to identify convergence, divergence, and complementary evidence.

Analysis was iterative and abductive, seeking the most plausible explanations for observed patterns through an ongoing dialogue between the data and relevant theory.²³ Emerging CMOCs were continually compared across participants, stakeholder groups, and documentary evidence to refine, reject, or elaborate the initial program theories. Through this process, explanatory accounts were progressively developed that better explained how contextual conditions influenced stakeholder reasoning and generated implementation outcomes. Individual CMOCs were subsequently synthesized into program theories explaining how recognition of deficiencies in care and participation in protocol development and implementation generated learning, adaptation, and sustained changes in clinical practice.

Results

Nine individuals representing six stakeholder groups participated in realist interviews. Interviews were analyzed alongside key implementation documents, including successive versions of the clinical protocol, electronic medical record order sets, and educational materials developed throughout implementation. Together, these data sources yielded 85 Context–Mechanism–Outcome Configurations (CMOCs) (Supplemental Material 3), which were synthesized into five program theories (Figure 1). The program theories describe an iterative learning process through which interest holders (1) recognized deficiencies in existing care, (2) overcame organizational barriers, (3) mobilized collective action, (4) institutionalized the protocol through organizational learning, and (5) reinforced and sustained transformation through improved patient outcomes.

Figure 1.

Five program theories (PT) describing the iterative learning process through which stakeholders recognized deficiencies in existing care, overcame organizational barriers, mobilized collective action, institutionalized the frostbite protocol through organizational learning, and sustained transformation through improved patient outcomes. The program theories were synthesized from 85 Context–Mechanism–Outcome Configurations (CMOCs).

Program Theory 1. Recognizing the Need for Change

Implementation was catalyzed when one participant, through a chance clinical encounter, became aware of an alternative approach to frostbite management being used successfully elsewhere. At the time, the available evidence consisted primarily of observational studies, case series, and local experience—evidence that would not typically justify widespread adoption of a novel therapy. However, within the context of a rare, limb-threatening condition with limited treatment options, this exposure prompted the participant to reconsider whether existing approaches to frostbite care remained acceptable. The realization that a tertiary referral centre could not provide treatment already available in a smaller northern hospital challenged assumptions about what constituted appropriate frostbite care and triggered critical reflection regarding professional responsibility to patients. Rather than dismissing the limited evidence, participants interpreted it as sufficiently credible to warrant further exploration of advanced therapies despite ongoing uncertainty (Quote 1, Participant 1, Table 1).

Table 1.

Representative Participant Quotations Supporting the Program Theories

Quote Number	Program Theory	Quote
1	Program Theory 1	“we brought [a patient with frostbite] to Calgary thinking that we could probably do the same thing that they’re doing in the Yukon just to find out that in fact we couldn’t …, my frustration and I think the frustration of the plastic surgeon at the time was, you know, why are they, you know, how come they can do something in Whitehorse that we cannot do in our big fancy tertiary care center? At that point we started engaging in a conversation and I engaged our clinical pharmacist in a conversation about; how do we go about doing this? Like, how do we go about getting this drug?” (Participant 1)
2	Program Theory 1	“When we initially started using iloprost, before we got the full protocol in place, there were some patients who were treated appropriately [and went to the ICU because that was the only place advanced therapies could be administered] but they were going out for smokes in the ICU and using substances which, you know, the ICU colleagues are wonderful, but they’re not that flexible on some of those things. ‬So that caused a few phone calls. That was part of the impetus. We knew right from the beginning we need to figure out a protocol for how to give it on the ward. And we knew we didn’t want patients staying in emerg for five days for their Q24 one to one iloprost infusion”. (Participant 6)
3	Program Theory 1	“… but we all know that it happens and it’s. It’s devastating, right? Like, we’ve all looked after patients whose lives have been changed because of their frostbite injury. So, there is a fertile ground for change management. Everybody was very keen to think of something that would help.‬” (Participant 7)
4	Program Theory 2	“It was more logistical barriers. Right. So it was more around, where are these patients going to go? Where should they be treated? You know, management people being like, well, I do not really want them on my unit, on my ward…”. Participant 1
5	Program Theory 2	“… when we looked at implementing this, interestingly enough, getting the drug itself was actually probably the easier part. Getting the buy in was probably the harder part.‬” (Participant 2)
6	Program Theory 2	[There was] “general acceptance that the therapy was worthwhile. And I think that was facilitated by the fact that we do not have a lot else to offer these patients …. It was like this or nothing.” (Participant 1‬)‬‬
7	Program Theory 3	“I think what happened is that we had strong voices to start off with, so we had the right people at the table to order to get this done. Dr. X from the emergency department, I believe he was the head of Foothills at the time. And then I forget who the head of plastics was also involved with and the head of ICU was also in agreement. Basically, we had a unified front from the start. And that’s really what I think allowed this to move forward and convinced the higher ups in leadership to allow this to occur.” (Participant 2‬‬)‬‬
8	Program Theory 3	“I think a few things that would be important is, is that the, the champions came to the department.‬.And they came to both educate and engage … They didn’t come with a demand for implementation. So then you’re being, you’re being approached to be a partner as opposed to simply someone that has to implement. And I think that has a very different, that creates a very different mindset amongst members of the department”. (Participant 5)
9	Program Theory 4	“… it was one of those first cases that went to the floor just because otherwise he was totally fine. There was no indication for him to go to ICU. And we were only going to use iloprost with him. That I think, was one of the first cases that prompted the change to go to the floor. They got some comfort with that. And that was some heavy, heavy lifting from the plastics team in order to convince the unit 32 that this was totally fine to do on their unit.” (Participant 2)‬‬
10	Program Theory 5	“You know, you look at these other centers who are employing these modalities and they’re, in some cases, reporting that they’re able to reduce their proximal amputation rate by 35 or 40%.‬” (Participant 8‬‬)‬‬
11	Program Theory 5	“What we didn’t account for was the amount of demand that existed despite it being limited in its evidence at the time. And so before, oh my gosh, probably within two weeks of starting our first patient on iloprost, we got requests from probably every single site within the zone. And within a month we were getting requests from [other cities] to get access to the drug in order to provide treatment and care to patients that were developing frostbite within those centers”. (Participant 2)
12	Program Theory 5	“One of the major changes we made based on our clinical experience, initially we were just treating grade three and four with iloprost and grade 4, we were adding thrombolytics. But I did notice that after rewarming there were some grade 2 that looked worse the next day.‬ And we’re actually maybe grade 3. And so we decided to be aggressive, combining that with how well people are tolerating the iloprost. We decided to treat grade 2 aggressively” (Participant 6)

Recognition of deficiencies in existing care generated interest in introducing advanced therapies locally; however, formal protocol development was overtaken by a serendipitous convergence of contextual conditions. The availability of iloprost, the simultaneous presence of clinicians familiar with the emerging treatment approach, and the presentation of a patient with severe frostbite during a cold snap culminated in a real-time decision to administer new treatments before a formal protocol had been established. The successful treatment reinforced participants’ confidence that implementation was both feasible and worthwhile, while simultaneously exposing logistical and operational barriers related to clinical workflows and coordination across services. These experiences reshaped participants’ understandingof what would be required to deliver the intervention reliably and prompted development of a standardized protocol, educational resources, and supporting organizational infrastructure (Quote 2, Participant 6, Table 1)‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬.‬

Collectively, these CMOCs suggest that recognition of unmet patient needs did not directly produce implementation. Rather, within a context of emerging evidence and limited therapeutic alternatives, exposure to a successful alternative model of care prompted clinicians to reinterpret existing practice as inadequate. This shift in reasoning generated a sense of professional responsibility, transforming dissatisfaction with existing care into motivation to pursue change (Quote 3,Participant 7, Table 1).

Program Theory 2. Overcoming Organizational Barriers

Once interest holders committed to pursuing a protocol, they encountered administrative and organizational barriers, including navigating centralized pharmacy processes, awaiting provincial authorization to implement an electronic health record order set, securing operational support, and coordinating across multiple clinical services. These contextual constraints had the potential to stall implementation, particularly given the limited evidence supporting the intervention. However, existing professional relationships with colleagues in Northern Canada, visible physician leadership, pharmacy advocacy, and support from plastic surgery provided interest holders with trusted sources of expertise and organizational legitimacy. These resources increased confidence that the proposed innovation was credible, feasible, and aligned with the shared goal of improving patient care, reducing concerns about the uncertainty associated with implementing an unfamiliar therapy (Quotes 4 and 5, Participant 1 and 2, Table 1).

Participants described collaborative discussions as creating opportunities to collectively interpret implementation challenges rather than viewing them as insurmountable barriers. Through these interactions, interest holders shifted their reasoning from emphasizing regulatory and operational risks toward identifying practical solutions that would enable safe implementation. Rather than asking whether implementation should proceed, participants increasingly focused on how it could be achieved within existing organizational constraints while maintaining patient safety.

Collectively, these CMOCs suggest that successful implementation depended not simply on the availability of organizational resources, but on how those resources shaped interest holders’ reasoning. Trusted relationships, multidisciplinary engagement, and visible leadership enabled participants to reframe uncertainty as a manageable challenge rather than a reason for inaction, generating the confidence and collective commitment required to sustain implementation despite organizational complexity (Quote 6, Participant 1, Table 1).

Program Theory 3. Mobilizing Collective Action

As regulatory barriers were resolved, implementation depended upon coordinated action across multiple professional groups. Participants consistently described implementation as a shared endeavour involving emergency medicine, intensive care, pharmacy, plastic surgery, nursing, hospital medicine, and operational leadership. Bringing these diverse perspectives together enabled interest holders to develop a shared understanding of the problem and recognize that successful implementation required coordinated contributions across clinical and operational domains rather than ownership by any single discipline.

Physician leaders who held both clinical and operational roles played an important contextual role by connecting frontline clinical priorities with organizational decision-making. Their dual responsibilities enabled them to translate clinical concerns into operational priorities while communicating organizational constraints back to clinical teams. This increased interest holders’ confidence that implementation decisions were both clinically appropriate and operationally feasible, facilitating coordinated action across professional boundaries (Quote 7, Participant 2, Table 1).

Participants also emphasized the importance of engaging departments as partners throughout protocol development rather than seeking endorsement of a completed protocol. This collaborative approach created opportunities for interest holders to contribute their expertise, address concerns, and shape implementation strategies. As participants saw their perspectives incorporated into the evolving protocol, they developed a sense of shared ownership and commitment that strengthened engagement and reduced resistance to change (Quote 8, Participant 5, Table 1).

Early implementation generated additional opportunities for learning. Contrary to initial expectations, clinicians discovered that most patients did not require ICU-level care and that implementation challenges were primarily logistical rather than clinical. These experiences prompted interest holders to reconsider their initial assumptions, adapt clinical workflows, refine patient pathways, and collaboratively develop solutions that better aligned with local resources and patient needs.‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬

Collectively, these CMOCs suggest that implementation advanced not simply because multiple disciplines participated, but because multidisciplinary engagement created opportunities for collective sensemaking. Through repeated interactions, interest holders developed shared confidence in the evolving model of care, reinterpreted emerging challenges as opportunities for adaptation, and collectively constructed implementation strategies that were both clinically credible and locally feasible.

Program Theory 4. Institutionalizing New Practice

Participants emphasized that implementation did not conclude once the protocol became operational. Instead, embedding the protocol into routine clinical practice required repeated cycles of education, experiential learning, and adaptation. As clinicians encountered additional frostbite cases and organizational processes were progressively refined, interest holders developed increasing confidence in both the protocol and their ability to deliver it safely and consistently.

Targeted educational initiatives, standardized documentation, electronic health record order sets, and ongoing support from Clinical Nurse Specialists and pharmacists provided clinicians with practical resources that reduced uncertainty surrounding a rarely encountered condition. These organizational supports enabled clinicians to view the protocol not as a novel or high-risk intervention, but as an increasingly familiar and legitimate component of routine practice.

Participants described early implementation as an iterative learning process in which each patient encounter generated new insights into clinical workflows, logistical challenges, and opportunities for improvement. Experience prompted interest holders to continually refine patient pathways, clarify professional roles, and adapt local processes to better align with clinical realities. Through these repeated cycles of action, reflection, and adaptation, confidence in the protocol increased, organizational processes became more reliable, and the new model of care became progressively normalized within everyday practice (Quote 9, Participant 2, Table 1).

Collectively, these CMOCs suggest that institutionalization occurred not simply because educational resources and organizational infrastructure were introduced, but because these resources changed how clinicians reasoned about the innovation. As uncertainty diminished and experience accumulated, interest holders increasingly viewed the protocol as the expected standard of care, enabling the transition from implementation as a discrete project to implementation as routine clinical practice.

Program Theory 5. Improved Patient Outcomes Reinforced Organizational Learning and Sustained Transformation

As implementation matured, interest holders described improved patient outcomes not simply as evidence that the protocol was effective, but as new experiences that reshaped their understanding of the intervention and its place within clinical practice. Early successful cases, the absence of major safety concerns, and subsequent reductions in digit amputations provided tangible confirmation that the considerable effort required to develop and implement the protocol had resulted in meaningful improvements in patient care. These experiences strengthened interest holders’ confidence that the new model of care was both clinically justified and operationally sustainable (Quote 10, Participant 8, Table 1).

Participants consistently described positive outcomes as reducing residual uncertainty surrounding the protocol. As confidence in the intervention increased, interest holders became more willing to invest in its continued refinement, advocate for its adoption, and support ongoing education. Rather than viewing implementation as a completed project, clinicians increasingly perceived the protocol as an evolving model of care that warranted continued adaptation and improvement.

Improved patient outcomes also influenced dissemination beyond the original implementation sites. Trusted professional relationships and informal clinical networks provided opportunities to share local experiences and outcome data with colleagues across Alberta. Seeing favourable patient outcomes locally, together with retrospective evidence demonstrating reduced digit amputations, led clinicians at other centres to view the protocol as both credible and feasible despite differences in local resources. These findings shifted reasoning from questioning whether implementation was warranted to considering how the protocol could be adapted within their own contexts (Quote 11, Participant 2, Table 1).

Collectively, these CMOCs suggest that improved patient outcomes functioned as more than implementation endpoints; they became resources for subsequent learning. Positive clinical experiences increased confidence, strengthened organizational commitment, supported dissemination, and generated further refinement of clinical practice, creating a recursive cycle in which implementation outcomes became the context for continued organizational learning and transformation (Quote 12, Participant 6, Table 1).

Discussion

This realist case study examined how implementation of a multidisciplinary frostbite treatment protocol functioned as a process of transformative learning. Rather than viewing implementation as the execution of predefined strategies or the translation of evidence into practice, our findings suggest that implementation constituted a dynamic learning environment in which clinicians collectively questioned existing assumptions, interpreted uncertain evidence, negotiated organizational barriers, adapted emerging practices, and developed new ways of working. Through iterative interactions between individuals, teams, and organizational structures, implementation became a mechanism through which new collective capabilities were created rather than simply adopted. This perspective extends conventional implementation narratives by highlighting learning not as a prerequisite for implementation, but as a process that unfolds through participation in implementation itself.

Transformative learning theory provides a useful interpretive lens for understanding this process.⁶ Mezirow describes transformation as beginning with a disorienting dilemma that challenges existing assumptions and prompts critical reflection before new perspectives are developed. In our study, this dilemma emerged through a chance clinical encounter in which one physician discovered that advanced frostbite therapies were being used successfully elsewhere. The realization that a tertiary referral centre could not provide treatment already available in a smaller northern hospital challenged participants’ assumptions about what constituted appropriate frostbite care. It also created a sense of professional discomfort and urgency, prompting clinicians to question whether the existing standard of care was acceptable and to pursue organizational change.

Importantly, transformation occurred despite the absence of the type of evidence traditionally considered sufficient to justify widespread clinical adoption. Participants recognized that the available evidence consisted largely of observational studies, case series and local experience. Rather than abandoning principles of evidence-informed practice, interest holders collectively reinterpreted the available evidence within the context of a rare, seasonal condition associated with devastating morbidity and few therapeutic alternatives. This finding suggests that transformative learning involved not only acquiring new knowledge but fundamentally reconsidering how evidence should be interpreted under conditions of uncertainty.²⁴ Such collective reinterpretation of evidence has received relatively little attention within either implementation science or HPE.^25,26

Our findings also extend transformative learning beyond its traditional focus on individual learners.^5,27 Although transformative learning has largely been conceptualized as an individual cognitive process occurring through critical reflection, the transformations observed in this study were fundamentally collective. New understandings emerged through multidisciplinary collaboration, shared problem solving, and distributed leadership rather than through isolated individual reflection. Participants jointly negotiated logistical barriers, redefined professional roles, redesigned clinical workflows, and developed organizational infrastructure that enabled new patterns of practice. In this way, transformation occurred simultaneously at individual, team, and organizational levels, suggesting that implementation can function as a collective transformative learning process within healthcare organizations.

This perspective also offers a complementary way of understanding LHSs. Existing LHS frameworks frequently emphasize continuous cycles of evidence generation, implementation, evaluation, and adaptation.^19–21 While these frameworks describe what LHSs do, they are less explicit about how clinicians and organizations actually learn. Our findings suggest that transformative learning provides one possible explanatory mechanism linking these iterative cycles. Recognition of unmet clinical needs stimulated critical reflection, implementation generated new experiences, early clinical outcomes reshaped stakeholder reasoning, and organizational adaptations created new capabilities. Rather than viewing implementation as a discrete phase occurring after learning, our findings suggest that implementation itself may represent one of the principal mechanisms through which LHSs generate organizational learning.

One of the most striking findings was that implementation was neither linear nor entirely planned. A serendipitous convergence of circumstances—including medication availability, the presence of knowledgeable clinicians, and the arrival of a patient with severe frostbite—led to treatment before formal organizational structures had been established. Rather than representing a failure of planning, this early clinical experience exposed unforeseen logistical challenges that subsequently shaped protocol development, educational resources, order sets, and new clinical pathways. This sequence suggests that implementation may often evolve through cycles of action, reflection, and adaptation rather than through strict adherence to predetermined implementation plans. Such findings resonate with contemporary views of healthcare organizations as complex adaptive systems in which learning emerges through interactions between people, context, and evolving practice.²⁸

Our program theories also highlight the central role of relationships in organizational learning. Successful implementation depended less on formal authority than on trust, professional credibility, psychological safety, and collaborative problem solving. Physician champions, pharmacy leaders, nurses, operational leaders, and specialists did not simply facilitate implementation; they created opportunities for shared sensemaking that enabled interest holders to reinterpret uncertainty, negotiate competing priorities, and collectively develop confidence in a new model of care. These findings reinforce the view that implementation is fundamentally a social process in which learning emerges through interaction rather than through dissemination of information alone.^29–32

Finally, our findings suggest that improved patient outcomes should not be viewed solely as endpoints of implementation but also as catalysts for continued learning. Early clinical successes reinforced clinicians’ confidence in the protocol, strengthened organizational commitment, encouraged dissemination to additional centres, and prompted ongoing refinement of clinical pathways. In this way, outcomes became inputs into subsequent cycles of organizational learning, illustrating the recursive nature of LHSs. Rather than terminating the implementation process, successful outcomes generated new questions, adaptations, and opportunities for further improvement.

Implications

Our findings have implications for both implementation science and HPE. For implementation scientists, they suggest that implementation strategies should be understood not only as mechanisms for changing behaviour but also as opportunities to facilitate critical reflection, collective sensemaking, and organizational learning. For health professions educators, the findings broaden transformative learning beyond formal educational settings by demonstrating how clinical practice itself can become a powerful educational environment. Supporting implementation may therefore require attending not only to technical infrastructure and educational resources but also to the relational and organizational conditions that enable clinicians to question assumptions, learn collaboratively, and adapt practice in response to emerging challenges.

Strengths and Limitations

This study is strengthened by its use of realist methodology, which enabled us to explain how contextual conditions activated mechanisms that generated learning and organizational change rather than simply describing implementation activities. Triangulation of interviews with protocol documents, educational materials, and electronic order sets strengthened the credibility of the program theories by integrating participants’ experiences with the documented evolution of the intervention. Nevertheless, this was a single case study involving 9 participants within one Canadian health system. Although implementation spanned four hospitals, these sites operated within an integrated system and shared many clinicians and implementation processes. The sample was therefore assessed in relation to the explanatory purpose of the study rather than the number of sites, and recruitment concluded when sufficient data had been obtained to refine and substantiate the program theories. While the resulting program theories are intended to be transferable rather than statistically generalizable, the identified mechanisms may operate differently in other organizational, cultural, or health system contexts. Finally, implementation was shaped by the limb-threatening nature of severe frostbite and the limited availability of alternative treatments which may constrain the relevance of the findings to less urgent innovations or those with established alternatives. Future research should examine whether similar transformative learning processes occur during implementation of other complex clinical innovations.

Conclusion

Implementation is often conceptualized as the translation of evidence into practice. Our findings suggest a different perspective. For this rare, limb-threatening condition participation in implementation created opportunities for clinicians and organizations to question assumptions, reinterpret evidence, develop new relationships, and construct new patterns of practice. Through this process, implementation became a vehicle for transformative learning rather than simply the endpoint of knowledge translation. By integrating realist evaluation, transformative learning theory, and LHSs, this study offers a conceptual framework that explains not only how clinical innovations become embedded in practice, but also how healthcare organizations might learn, adapt, and transform through the work of implementation itself. In particular, LHSs typically rely on structure continuous improvement cycles but there may be benefit from combining these structured cycles with some latitude to capitalize on unpredictable organic triggers to drive organizational change.

Funding Statement

This research received no specific funding. No sponsor or funder was involved in the study design, data collection, analysis, interpretation, manuscript preparation, or the decision to submit the manuscript for publication.

Generative AI Statement

During manuscript preparation, the authors used ChatGPT (OpenAI, GPT-5.5) to assist with language editing, improving clarity, grammar, and readability, and to support the drafting of non-scientific text. The AI tool was not used to generate or analyze study data, interpret findings, or make scientific conclusions. All content was critically reviewed, verified, and revised by the authors, who take full responsibility for the accuracy and integrity of the manuscript.

Abbreviations

LHS, Learning Health Systems; HPE, Health Professions Education; ICU, Intensive Care Unit.

Data Sharing Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request. Because the study includes qualitative interview data that may contain potentially identifiable information, the data are not publicly available in order to protect participant confidentiality. De-identified data may be shared, where appropriate, in accordance with the approved ethics protocol and institutional policies.

Ethics statement

This study was approved by the University of Calgary Conjoint Health Research Ethics Board (CHREB) (REB25-0705). Informed consent including publication of results including de-identified responses was obtained from all participants.

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

Dr Catherine Patocka reports grants or contracts from CIHR; honoraria from Royal College of Physicians and Surgeons of Canada, Canadian Medical Association; meeting/travel support from European Society of Emergency Medicine; leadership or fiduciary roles from College of Physicians and Surgeons of Alberta, Foothills Hospital Medical Staff Association, outside the submitted work. The authors report no other conflicts of interest.