Wolf Creek XVIII Part 5: cardiac arrest survivorship science

Introduction

While advances in emergency care and post-resuscitation treatment have improved survival rates, survivorship after sudden cardiac arrest involves more than just cardiac recovery. It encompasses a complex journey of physical, emotional, cognitive, and social healing. Many survivors face persistent challenges such as memory issues, fatigue, anxiety, depression, leading to poor quality-of-life (QOL) and societal participation.¹

Despite growing awareness of the complex needs of cardiac arrest survivors, outcomes are traditionally measured using largely clinician-centered rather than patient-centered metrics. Most research and clinical evaluations focus on clinical metrics such as survival rates, clinician-derived neurological scores, or hospital discharge status—outcomes that, while important, fail to capture the lived experiences of survivors.² For example, the Cerebral Performance Category (CPC), a clinician-assessed 5-point scale, is widely used to evaluate neurological outcomes after cardiac arrest. However, the common dichotomization of CPC scores (1–2 vs. 3–5) obscures meaningful distinctions in recovery and highlights the scale’s limitations in capturing cognitive, psychological, and quality-of-life outcomes. Notably, CPC 1 (good cerebral performance with minimal deficits) and CPC 2 (moderate disability with independence in daily living but inability to resume prior work or activities) are often grouped as “favorable outcomes” in resuscitation research,³ despite clinically significant differences. This practice reflects the CPC’s emphasis on survival and functional independence rather than domains such as cognition, emotional well-being, and health-related quality of life that are crucial to whole-person recovery.

Post–cardiac arrest care has historically emphasized short-term neurological outcomes and mortality, with comparatively limited attention to rehabilitation needs and recovery trajectories across the continuum-of-care. Rehabilitation after cardiac arrest presents unique challenges that are not fully addressed by existing paradigms derived from other neurologic conditions. Hypoxic–ischemic brain injury is heterogeneous, dynamic, and influenced by complex interactions between biological injury, environmental exposures, and personal factors. There is therefore a critical need for translational research frameworks that bridge molecular and systems-level biology with survivor-centered functional outcomes. Such approaches are essential to generate mechanism-informed strategies that guide prognosis, treatment selection, and recovery-focused interventions. This need has catalyzed growing interest in rehabilitation-focused, biologically informed models of survivorship that integrate multidimensional outcomes with advances in biomarker science.

Addressing multidimensional rehabilitation needs at scale will require analytic approaches capable of integrating longitudinal, high-dimensional data across biological, functional, and contextual domains. With the exponential growth of potential innovations such as Artificial Intelligence (AI) to revolutionize the way we collect, manage, and interpret longitudinal patient data across health conditions, including cardiac arrest. By leveraging data from electronic health records, wearable devices, mobile apps, and other digital sources, AI can track health trends over time, detect subtle changes, and identify early warning signs of deterioration or improvement.4, 5, 6 Based on examples in other areas of healthcare, these tools can automate and streamline data collection, reducing both patient and clinician burden while enhancing the accuracy and completeness of information and analyzing vast and complex datasets to uncover patterns that may not be apparent through traditional methods.7, 8

With this foundational knowledge in mind, this session aimed to facilitate a “what if?” discussion on survivorship science that challenges current practice and pushes the boundaries of what we think is possible.

Methods

Since its inception in 1975, the Wolf Creek Conference has a well-established tradition of providing a unique forum for robust intellectual exchange between thought leaders and scientists from academia and industry focused on advancing the science and practice of cardiac arrest resuscitation. The 50th Anniversary Wolf Creek XVIII Conference was hosted by the Max Harry Weil Institute for Critical Care Research and Innovation in Ann Arbor, Michigan, USA, on June 19–21, 2025. Meeting invitees included international academic, industry scientists, and thought leaders in the field of cardiac arrest resuscitation. All participants were required to complete conflict of interest disclosures.

Cardiac Arrest Survivorship Science topics included critical evaluation of medication effects in the post–cardiac arrest rehabilitation period (Dr. Amy Wagner), approaches to outcome measurement after cardiac arrest (Dr. Katie Dainty), emerging applications of artificial intelligence and other digital tools for longitudinal data collection (Dr. Sachin Agarwal), and integration of survivor perspectives (Mr. Matt Wood). We have summarized the content of the presentations and audience discussion and outline considerations for future directions derived from the session below.

Reframing our definition of a good outcome

In thinking about key concepts in survivorship, it is important to begin with framing the concepts of health and function and their interrelationships for this population. For this, we can use the lens of World Health Organization’s International Classification of Functioning, Disability and Health (ICF) framework.⁹ Traditionally, the biomedical model defines health as the absence of disease, and this is based on the state-of-the-science, about what we know about pathophysiology, the signs, the symptoms that go along with disease. A more careful look at health suggests it can be reflected in the ability to adequately cope with all the demands of life. Within this framing of health that we have assumptions we can make about health that are based on level of functioning which opens up the conversation on health and recovery in a totally different way. To further evolve a functional definition of health, we can consider health as a state of balance between individuals and their environment and how health reflects an internal equilibrium that allows individuals to get the most out of life despite their disease. When viewing health through this lens, we begin to see how to incorporate the values of individuals, of cultures, and their communities, and really understand the multidimensionality of health through the lens of function and the ICF model.

When we consider health through this functional framework for cardiac arrest survivors as well as co-survivors, a logical construct emerges for forging a person-centered path forward with respect to goals-of-care planning, long-term recovery trajectories, functional prognostication, as well as sustainable health living and community integration. As health care providers, operating within this function-focused framework provides the space needed to consider how to personalize treatments that target functional recovery. Including these expanded dimensions of health and our understanding of function allows more rehabilitation-relevant framework to be operationalized from a resuscitation research perspective.

The opportunity for tailored rehabilitation

The “Rehabilomics” research model aims to provide an ‘omics’-based overlay to the scientific study of rehabilitation processes, how we think about the multidimensionality of outcome, and how it links to the WHO–ICF framework.10, 11, 12 This model focuses on three domains of impairment, activities, and participation (see Fig. 1). Importantly, it also considers the moderating effects of various environmental barriers and the boundary effects of personal factors that impact this nonlinear, often bidirectional, and somewhat reciprocal framework. The -omics overlay has been used widely in traumatic brain injury (TBI) and illuminates how biomarkers can provide biological information about the environmental exposures and personal biology that individuals bring to conditions and complications that can occur as a part of the hypoxic ischemic brain injury associated with cardiac arrest.

Fig. 1.

Operationalization of the World Health Organization International Classification of Function (ICF) framework for cardiac arrest survivorship.

(A) Impairment of body functions reflects the functional relationship between the disease-based manifestations associated with cardiac arrest and their symptom burden. (B) Activities reflect the variety of individual needs that must be met in order to do what they want in the community. (C) Life participation is distinguished from activity through intent and impact. Aspects of respect and dignity are captured in concepts of self-agency and self-determination in the context of community interactions. .

Biomarkers can be used for prognosis as well as biomarker-guided clinical decision algorithms, screening and prevention programs, and stratification for randomized controlled trials and comparative effectiveness studies. Even though survivorship is a concept that has only become a part of the resuscitation conversation in recent years, this framework could nicely frame where the cardiac arrest community might want to go in terms of survivorship science. A nice example is the POST-ICECAP trial, a National Institutes of Health-funded multi-site study, involving more than 65 academic hospitals in the United States.¹³ When the investigators were delineating the outcome battery for this study, they did a thoughtful job of linking their selection of outcome measures to the various domains of function noted above, with careful attention to the types of environmental factors that may arise as barriers to the recovery trajectory, as well as personal factors. Given the considerations made during study design development to adhere to an ICF focused framework, one can see how it might to overlay a biomarker battery (see Fig. 2) on top of this evolving infrastructure, to include genetic information serum-based biomarkers, imaging parameters, as well as other physiological might provide some unique biological information that informs function as well as function focused treatment.

Fig. 2.

International Classification of Function (ICF) framework applied to cardiac arrest survivorship populations as a model for function focused precision research and integrative care. Here rehabilomics can be overlaid on this framework by incorporating biomarkers (e.g. genomic, proteomic, etc.) better understand the biological basis of function over the life course.

Many areas of medicine borrow treatment and management concepts from more established areas of practice in clinically related populations. Rehabilitation systems of practice and research tend to be focused on multidisciplinary domains and interprofessional provider care models. In addition to learning from long standing research networks like the Traumatic Brain Injury Model Systems (TBI-MS) Network,¹⁴ and the acknowledgement of TBI as a chronic condition¹⁵, multi-specialty care for post-cardiac arrest recovery research and systems of care can use this experience and guidance to uniquely consider bench-to-bedside research programs and multisystem complex care (e.g. physiatry, cardiology, neurology, psychiatry) across the life course to address the varied needs of individuals surviving cardiac arrest. Salwa et al. recently wrote a very nice narrative review on current perspectives on rehabilitation following return of spontaneous circulation after cardiac arrest, stating that “Structured interdisciplinary interventions encompassing cardiopulmonary, neuromuscular, and cognitive domains can effectively mitigate long-term disability, facilitating return to daily activities and employment.”¹⁶

As cardiac arrest survivorship becomes more prevalent, practitioners have started to borrow neurostimulation pharmacotherapy strategies from the TBI field with the expectation of similar treatment effects. For example, dopamine agonist strategies are a mainstay of TBI neurostimulant pharmacology17, 18 based on preclinical models showing functional deficits in striatal dopamine neurotransmission in cortical deafferentation models of TBI, normalization of this defect with neurostimulants like methylphenidate,19, 20, 21 and symptomatic clinical improvements in arousal, consciousness, and cognition in clinical populations TBI.¹⁶ However, striatal DA in experimental cardiac arrest models show a hyperdopaminergic neurotransmission state,²² and clinical case reports suggest that early use of alternative pharmacological strategies promoting cholinergic tone and providing GABAergic support is effective in facilitating early emergence among cardiac arrest survivors who are in a decreased level of consciousness (DoC) state. Clinically, our experience is that some patients also benefit from cholinergic therapies with relative improvements in critical illness weakness and movement disorders such as myoclonus. We highlighted a case of a 51-year-old male with out-of-hospital cardiac arrest due to ST-elevation myocardial infarction and pulmonary embolism. His hospital course was complicated by severe critical illness weakness and myoclonic activity primarily affecting his limbs. He had neurosensory hearing loss and was cortically blind for several weeks. Acute care treatment with donepezil was temporally associated with gradual but notably rapid improvements in arousal, limb strength, and ability to follow commands. Early treatment-related improvements moderated his recovery trajectory such that after 5-weeks of acute care, the patient transferred to acute inpatient neurorehabilitation and made remarkable gains in strength, cognition, communication, swallowing, balance, and activities of daily living, such that he could successfully return home with his spouse. Case examples like this demonstrate that, with structured support and appropriate early interventions, even individuals with severe deficits can achieve significant improvements in QOL.²³ The discussion emphasized that such outcomes should no longer be considered exceptional but rather attainable for a larger proportion of survivors.

Measuring what matters

As seen in the POST-ICECAP example above, research findings and, in particular, randomized trials are defined by the outcome measures that are chosen. Frequently, research teams defer to measures used in previous studies, come from a trusted source, or are easiest to obtain. But measuring the endpoints that are “always measured”, that researchers feel are important, and categorizing outcomes as “good” or “fine” using arbitrary cut points is not adequate. This point directly ties into the discussion above about defining health and the journey of survivorship and recovery, focusing on a survivor-centered approach to outcomes.

The panel critiqued the limitations of current outcome measures, such as short-term survival metrics and broad neurologic scales, which fail to capture the lived experiences of survivors and their families. We highlighted the need to develop and adopt psychometrically robust outcome measures capable of assessing cognitive, psychological, and social dimensions of recovery, as well as capturing the fluctuations that survivors experience over time. Considerable attention must be given to identifying outcomes that reflect what matters most to survivors themselves, including independence, mental health, fatigue management, and return to meaningful roles in society.

So how can this be done differently in cardiac arrest? The neurotrauma community has learned a considerable amount by simply asking patients to ask themselves the question, “What does function mean to me?”.²⁴ Many themes came out of this work, but some noteworthy topics included the idea of whether we are conducting research in areas that matter to survivors, particularly around the functional consequences of the secondary conditions that arise over time during one's recovery. Another common area of importance was adjustment to new health status and disability, as well as the ability to use biomarkers to study TBI as an evolving and dynamic chronic condition, not just a fixed insult that people move on from. And lastly, we discussed the importance of co-survivorship and the impacts and dynamics of the event for them. This included the need for education, expectation setting, and equipping them with the coping skills that they need to take on new functional roles and prepare them with the self-agency that they need to navigate what is oftentimes a complex and challenging healthcare journey.

When looking at the problem differently, using a bio-psycho-social ICF approach, utilizing biomarkers to measure the actual impacts of this injury, then the field may actually be able to move the needle on helping survivors rehabilitate their brain and decrease the impact of functional impairments on their lives. Two-thirds of cardiac arrest survivors suffer from cognitive deficits, particularly memory, planning, problem-solving, and attention, two-thirds experience symptoms of anxiety and depression, one-third develop post-traumatic stress symptoms, and only half return to their previous occupation.25, 26, 27, 28 They also report experiencing long-term fatigue and not just fatigue because their heart is recovering, but from dealing with the activities of daily living. The study by Christensen et al. in 2023 showed that at 12 months, 45% of survivors in that cohort were still on full-time sick leave.²⁹

Cardiac arrest survivorship is an iterative and non-linear journey and taking point-in-time measurements at very early stages does not tell the whole story of recovery after this life-altering event. In research and clinical settings, we tend to do routine measurements back at 3-months, 6-months, 9-months, 12-months, etc., because it is standardized, clean and matches clinical follow-up intervals, but there is a movement now to think about having people report on their outcomes when they feel they need to report, or when they feel like they need to come back for follow-up. An approach like this would consider cardiac arrest as a complex chronic condition and move to a future that uses a synergistic lens to assess improvement using both patient-centered outcomes and clinical outcomes.

Traditional research models need a complete transformation

Clinical research today operates under immense strain. Our current research models, particularly those studying recovery and function, are expensive, inefficient, and increasingly unsustainable. Long timelines, low recruitment rates, inaccurate measurement, and inconsistent data collection across sites has become the norm.³⁰ Rehabilitation and survivorship studies are especially vulnerable: they are often multi-site, dependent on human assessors, and rely on subjective or self-reported outcomes.³¹ Oversight and monitoring alone consume nearly two-thirds of total trial budgets, while administrative burden and staff turnover continue to erode efficiency.³¹ The consequences are predictable yet unacceptable. Nearly 80 percent of trials miss enrollment timelines, budgets are routinely overrun, and sites withdraw from studies altogether.³² Translating research findings into clinical practice is often delayed by years, during which patients may go without access to potentially beneficial interventions. The problem is structural: trials depend on outdated workflows, paper-based assessments, and data systems that do not communicate with one another. As treatments become more personalized and precision-based, the available patient pool shrinks, making recruitment even more difficult. Many eligible participants live in rural or community settings that lack the infrastructure to host formal trials, further widening inequities in access and representation.33, 34 In active U.S. trials seeking participants, most clustered around urban centers on the coasts. To move forward, we must re-engineer this model using technologies that lower costs, expand reach, and improve data quality while maintaining rigor and transparency.

Artificial intelligence (AI) offers practical and not-so-futuristic solutions

Remote monitoring technologies such as wearables, smartphones, and ecological momentary assessments allow researchers to collect continuous, real-world data on physical activity, sleep, and mood (see Fig. 3).³⁵ Such “quiet” data capture can document survivorship trajectories without overburdening patients or staff. From these streams, AI can derive interpretable digital phenotypes that reveal patterns like disrupted sleep or reduced mobility. In mental-health-related recovery research, sequential rule-mining algorithms can detect temporal relationships between behaviors, such as “if reduced activity, then disrupted sleep,” patterns commonly associated with depressive episodes. These interpretable rules enrich precision digital phenotypes that characterize individual behavior over time.

Fig. 3.

Wearable and digital devices used for continuous physiological and behavioral monitoring.

The figure illustrates the ecosystem of wearable and connected devices employed for remote data collection. Devices include the Zio Patch for cardiac rhythm monitoring, eCAP for medication adherence tracking, actigraphy and GENEActiv for measuring activity and sleep, the BioIntelliSense BioSticker for continuous physiological monitoring, and a blood pressure monitor for hemodynamic assessment. Data from these devices are transmitted to mobile devices and cloud-based platforms for secure storage and analysis.

An effective AI system can integrate these behavioral patterns with multiple layers of knowledge, including scientific evidence from the literature, clinical expertise from electronic health records and practitioner input, and patient values captured during counseling. The result could be a clinician dashboard that visualizes recovery, flags emerging issues, and recommends data-driven actions. In this vision, AI does not replace clinicians; it amplifies their insight, translating complex multimodal data into actionable understanding.

AI-powered functional assessments leveraging conversational agents and large language models may offer novel solutions to enhance patient evaluation and follow-up.³⁶ The goal is simple but transformative. We need to use AI to make outcomes-based research faster, smarter, and more sustainable. Imagine spending less time collecting and cleaning data because intelligent systems harmonize it automatically, flag inconsistencies, and learn to adapt. Recruitment could be streamlined by AI tools that scan electronic health records to identify eligible participants, predict potential dropouts, and personalize engagement. These same systems could automate oversight tasks, detect protocol deviations, and ensure consistency across sites, drastically reducing administrative costs.³⁷

Nowhere is this transformation more visible than in functional assessment. Traditional outcome measurement relies on structured tests administered by trained staff, a model that is labor-intensive, subjective, and geographically limited. Conversational AI chatbots change that equation. These systems can conduct structured or adaptive assessments, guide participants through tasks, and capture behavioral or linguistic markers of cognition and recovery. They are available around the clock and can speak in multiple languages. In rehabilitation research, this means replacing periodic, high-burden testing with continuous, low-friction observation.³⁸ Instead of snapshots, we can now watch recovery unfold in real time.

Our own work has begun exploring AI chatbots as “human-in-the-loop” (see Fig. 4) research assistants, audio-based agents that interact with study participants conversationally but within standardized protocols. Each interaction is scored independently, preventing context drift and ensuring reliability. Technical challenges remain. Poor transcription accuracy often leads to confusing responses or incorrect scoring, especially when the AI must interpret repetition or self-correction. Further, current transcription-based models lose the richness of intonation, pacing, and emotional tone embedded in speech. Voice-to-voice models that process raw audio promise more nuance but still struggle with logical consistency. These are solvable problems, but they highlight the need for deliberate, iterative development before widespread deployment.

Fig. 4.

Conceptual model of a human-in-the-loop framework for artificial intelligence (AI) development.

The figure illustrates the iterative interaction between humans and AI systems in a human-in-the-loop design. The AI model generates an output that is subsequently reviewed by a human expert, who confirms, rejects, or modifies the result based on contextual understanding and domain expertise. This feedback is then used to refine and improve the AI model, enabling continuous learning, performance enhancement, and alignment with clinical or operational goals.

Beyond conversational agents, large language models (LLMs) extend this potential even further. These systems can synthesize large bodies of research, generate reports, and summarize complex findings for diverse audiences. They can serve as intelligent interfaces between researchers and data, drafting study protocols, creating educational materials, or even generating participant-friendly summaries of ongoing trials. In survivorship research, they can transform educational resources into accessible formats such as podcasts or multilingual transcripts. Combined with conversational agents, LLMs can personalize information delivery, ensuring that every patient and clinician receives context-specific, evidence-based communication.

The future of AI in recovery and survivorship research after cardiac arrest

The rapid pace of generative AI adoption brings enormous promise and equally significant responsibility. Before these technologies become fully embedded in the research enterprise, the field must systematically assess both benefits and harms. Without structured evaluation and transparent sharing of experiences, positive and negative alike, we risk creating a landscape of random, unverified practice. The speed of AI implementation must be balanced with scientific rigor, ethical safeguards, and a sustained commitment to quality.

Developing intentional strategies, such as clear methodological guidelines, will be essential. Just as reporting standards like CONSORT³⁹ improved clinical trial transparency, similar frameworks for AI can enhance reproducibility and trust. A thorough mapping of the research process is a necessary first step. Only by identifying where current bottlenecks exist, data collection, participant engagement, analysis, and dissemination can we discern where AI introduces incremental or transformative value. Low-risk tools used for document generation or trial coordination may warrant minimal oversight, while high-risk applications, including data analysis, endpoint adjudication, or automated decision-making, require rigorous validation and regulatory scrutiny.

The importance of the survivor and co-survivor voice

Throughout this panel session, we highlighted the importance of including the survivor and co-survivor voice in resuscitation science. Mr. Matt Wood joined us to share his lived experience and thoughts on the future of cardiac arrest research. Like many cardiac arrest survivors, he and his partner were unprepared for the complexity of recovery after hospital discharge. Much of his care and follow-up focused on physical recovery. Both Matt (survivor) and Kenyon (co-survivor) experienced the trauma from the sudden cardiac arrest event in very different ways and were impacted at different timepoints. Within a few weeks of Matt’s discharge, Kenyon experienced recurring mental and emotional stress, which he sought to overcome through mental health therapy. Five months post-discharge, mental health symptoms presented with Matt; He said, “Once my body reached a certain point in recovery, my mind said it's my turn.” Matt cited the lack of resources and information, no coordination or referral of mental health resources, and no neurological evaluation post-discharge as very challenging and problematic. It left both Kenyon and him to seek out care on their own for conditions that neither had experienced before the cardiac arrest event.

The delayed onset of post-traumatic stress symptoms, severe depression, mood changes, and cognitive symptoms, compounded with existing anxiety, fatigue, and memory issues, came at a time when Matt was attempting to reintegrate into the workforce. His symptom severity eventually resulted in him stepping back to continue to work on the mental health side of his recovery. Over the next two years, this included multiple attempts to find a suitable anti-depressants, cognitive behavioral therapy, and Eye Movement Desensitization and Reprocessing (EMDR) after enduring a 6-month waitlist. Matt stated, “In many ways, I feel that Kenyon and I are exceptions to the majority of patients; we are actively engaged in our care. I take notes, log, and track symptoms and do research to be prepared for medical appointments. This way, we can work with my team of providers for the best possible outcome. But we have to do it all on our own.”

It is important for researchers to recognize that just like the various stories within cardiac arrest, recovery vastly differs for every individual and the families impacted. Social, cognitive, and family structure support can vary just as widely as care received across systems. This is where innovative research approaches discussed on the panel, like biomarkers, patient-centered measurement, and AI, can bring new thinking to identify care gaps and early intervention opportunities for whole-body recovery. Matt and Kenyon suggest that tools and resources be provided early from hospital social workers and discharge planning, initial follow-up visits with cardiologists and primary care via medical records, prompts in medical care, and case/care management outreach efforts provided through insurance companies. Kenyon noted, “Sudden cardiac arrest survival and release back into the world is extremely unique compared to the various other reasons people are in the hospital, and the needs of survivors and families in this situation are not being addressed as such by the hospital staff at large.”

Knowledge gaps and priorities for research and implementation

The future for understanding and developing approaches to high-quality survivorship in cardiac arrest is bright. This session highlighted the importance of cross-disciplinary learning as a key ingredient for the future. Fields such as traumatic brain injury, stroke, and oncology survivorship can provide valuable frameworks and lessons that could accelerate progress in cardiac arrest survivorship. There will be challenges, including the absence of long-term preclinical models in cardiac arrest research and systemic issues such as premature withdrawal of care, lack of structured rehabilitation, and ignoring long-term outcomes that matter to patients and families, which can create “self-fulfilling prophecies” that limit recovery potential.

It has taken the TBI field decades to understand and embrace the variable progressive and chronic conditions associated with it. In particular, the Centers for Medicare and Medicaid Services (CMS) have recently recognized TBI as a chronic condition, a designation that can facilitate the implementation of TBI survivor care as a part of their chronic special needs plans, potentially broadening access to continuous, coordinated care.¹⁵ As an emerging rehabilitation population, cardiac arrest survivors will need to leverage gains made in policy and care provision for other populations with acquired brain injury like TBI, hopefully enhancing research capacity and health care for this population to impact function.

The transformation of research through technology like AI is not about replacing the human element but amplifying it. We stand at a moment where the tools exist to capture recovery in its natural context, analyze complex patterns beyond human perception, and deliver insights instantly to those who need those most, creating a continuously learning ecosystem of care and science. But realizing that vision requires careful balance: innovation guided by rigor, speed tempered by safety, and automation grounded in empathy. The future of AI in research will belong not to those who move fastest, but to those who move wisely, building systems that are transparent, inclusive, and relentlessly focused on improving lives.

To realize the visionary ideas outlined in this session, a multi-pronged funding strategy will be essential. Federal and state/provincial health research agencies, such as the National Institutes of Health, can provide foundational support for large-scale, multi-site studies focused on patient-centered outcomes and rehabilitation models. Targeted funding opportunities through programs emphasizing digital health innovation and artificial intelligence integration can be leveraged to accelerate technology-enabled research. Partnerships with industry stakeholders—including medical technology firms, pharmaceutical companies, and insurers—offer additional avenues for co-development and scalability, particularly for AI-driven tools and remote monitoring solutions. Furthermore, philanthropic organizations and cardiac health foundations such as the American Heart Association can play a critical role in supporting pilot projects and community engagement initiatives, ensuring that survivor and co-survivor voices remain central to the research agenda. By combining public, private, and philanthropic resources, resuscitation science can build a sustainable funding ecosystem that supports innovation and implementation while maintaining equity and accessibility.

Lastly, the concept of co-survivorship was particularly emphasized, recognizing the health, psychological, and advocacy roles assumed by family members and caregivers alongside survivors. As such, we would be remiss not to highlight the need to prioritize stronger community engagement in our science. Survivors and co-survivors, like our session partners Matt Wood and his partner Kenyon, must be central in shaping the research agenda, ensuring that scientific inquiry remains aligned with the outcomes and challenges that matter most to those directly affected by cardiac arrest.

Conclusions

In conclusion, this panel called for a paradigm shift in survivorship research and care: from a narrow focus on survival metrics to a comprehensive approach that prioritizes long-term function, QoL, and meaningful participation. Achieving this vision will require interdisciplinary collaboration, investment in robust outcome measurement, development of precision models of care, and the establishment of infrastructure such as registries and standardized assessment batteries. The future of research will depend on how well we integrate AI-driven capabilities into a coherent ecosystem, and information science provides the connective tissue linking EHRs, imaging data, wearable sensors, and unstructured clinical notes into unified, learnable systems. When designed responsibly, these systems can continuously refine themselves, detecting bias, correcting errors, and learning from every iteration. The goal is not to automate human judgment but to elevate it, creating research environments that are both intelligent and compassionate. The session affirmed a collective commitment to advancing survivorship science so that meaningful recovery becomes the norm rather than the exception for individuals who experience cardiac arrest and their families.

Declaration of generative AI and AI-assisted technologies in the manuscript preparation process

During the preparation of this work, the author(s) used ChatGPT in order to make minor improvements to language and flow. After using this tool/service, the author(s) reviewed and edited the content as needed and take(s) full responsibility for the content of the published article.

CRediT authorship contribution statement

K.N. Dainty: Writing – review & editing, Writing – original draft, Validation, Project administration, Data curation, Conceptualization. A.K. Wagner: Writing – review & editing, Writing – original draft, Validation, Data curation, Conceptualization. M. Wood: Data curation, Conceptualization. S. Agarwal: Writing – review & editing, Writing – original draft, Validation, Data curation, Conceptualization.

Funding

This content, which is summarized here, was presented at the Wolf Creek XVIII Conference hosted by the Max Harry Weil Institute for Critical Care Research and Innovation in Ann Arbor, Michigan, USA on June 19–21, 2025. The authors travel to and from the conference was paid by the conference.

Declaration of competing interest

All of the authors declare no competing financial interests with the content of this manuscript. KND is a member of the Resuscitation Plus Editorial Board but was not involved in the peer-review of this article and has no access to information regarding its peer-review. Full responsibility for the editorial process for this article was delegated to another journal editor.