Life In The Slower Lane

Improving Emergency Care for Older Adults with Dementia through “Conversations‑Before‑Investigations”

nashte
45–68 minutes

Background and Rationale

Older adults with dementia or cognitive impairment represent a growing and vulnerable population in emergency departments (EDs) worldwide. In Australia alone, more than 400,000 people live with dementia, a number expected to double by 2058(1–3). International data indicate that nearly one in ten older ED patients has dementia, yet few ED care models are tailored to their complex needs(4). Conventional ED processes – rapid triage, protocol-driven investigations, and a focus on ruling out acute threats – are often misaligned with the care needs of persons with dementia. These patients frequently endure extended ED stays and distressing interventions with minimal benefit. For instance, advanced dementia is a terminal condition: a hospitalisation for pneumonia or hip fracture in end-stage dementia carries a six-month mortality of over 50%(5). Aggressive interventions in such cases may merely prolong suffering. As one critical care physician observed, many terminal patients spend their final weeks in “fruitless efforts to prolong death in the misguided hope of prolonging life”. This stark reality underscores the ethical imperative to reconsider default investigation and treatment pathways for cognitively impaired elders.

Drivers of Over-Investigation: Multiple factors contribute to a “default” ED approach of exhaustive investigation for older people with cognitive impairment, even when the likely yield is low. Clinicians often act out of fear – fear of missing a life-threatening diagnosis and fear of medicolegal repercussions. This “defensive medicine” mindset is well-documented; emergency physicians cite fear of malpractice, intolerance of uncertainty, and a sense of obligation to “do something” as key drivers of over-testing(6–10). Epistemic uncertainty looms large when a patient with dementia cannot communicate history or symptoms clearly. Faced with diagnostic ambiguity, ED staff may reflexively order a battery of blood tests and scans in an attempt to gain clarity or protect themselves from error. Such intolerance of uncertainty and risk aversion fuels the overuse of imaging and other investigations. Culturally ingrained ED norms further reinforce these behaviours: the fast-paced ED environment values rapid diagnostics and intervention. There is often implicit pressure from colleagues and the institutional culture to follow protocols that promote the efficient ordering of investigations and admissions. In busy EDs, taking the time for in-depth conversations can feel at odds with the dominant ethos of efficiency and throughput. Indeed, research shows that ED staff tend to prioritise readily “fixable” physical issues and procedural tasks, while patients with complex cognitive and social needs are de-prioritised. Clinicians may ask only brusque, binary questions to rule out immediate dangers and then proceed down routine investigative pathways, inadvertently leaving the older patient feeling invisible and uninformed about their care.

Impact on Patients with Dementia: For older adults with dementia, these default investigational pathways can be not only unhelpful but actively harmful. Many common ED interventions (e.g. repeated blood draws, intravenous lines, urinary catheters, loud imaging studies) can precipitate pain, agitation, or delirium in cognitively impaired patients. The potential harms of over-testing in such populations are well recognised, including false positives leading to further invasive procedures, overdiagnosis of clinically insignificant findings, and emotional or physical trauma. Critically, when a test is unlikely to change management or improve outcomes for a frail older person, subjecting them to it violates the bioethical principle of non-maleficence. Yet fear-driven cascades often continue “just to be safe,” even if they yield only burdens with no benefit. Additionally, a “prognostic illusion” may be at play: clinicians and families can hold an overly optimistic belief that aggressive interventions will substantially improve an older person’s trajectory when, in reality, advanced dementia runs a downhill course. This optimism bias, coupled with the difficulty of accepting the limits of medicine, delays an honest reckoning with likely outcomes. For example, families’ and clinicians’ hope can unintentionally inflate estimates of survival or recovery, leading to interventions that postpone essential end-of-life planning. In sum, fear, uncertainty, and cultural habitspush ED care toward maximal intervention, even as evidence suggests such approaches often neither prolong life nor enhance its quality for those with severe cognitive impairment.

Conversations Before Investigations – A Paradigm Shift: This research proposes an alternative approach, placing “conversations before investigations” at the heart of emergency care for patients with dementia. In practice, this means that upon an older cognitively impaired person’s ED presentation, clinicians would first pause to engage in a thoughtful conversation – with the patient (as feasible) and/or their family or caregivers – before reflexively ordering tests. The content of these conversations would include understanding the person’s baseline cognitive status and health trajectory, eliciting their values and care preferences, clarifying the goals of the ED visit, and discussing the potential benefits versus burdens of investigation or hospitalisation. Such an approach aligns with age-friendly healthcare models that emphasise identifying “what matters” to older adults and aligning care with their goals. Crucially, it counters the impulse to act immediately on every diagnostic possibility, instead encouraging clinicians to tolerate a degree of uncertainty while gathering context and building trust. By instituting a conversation step, the clinician can determine whether an investigation with the associated risk of harm is warranted or whether a more conservative, comfort-focused plan would better serve this patient. For instance, an ED physician might learn that an elderly woman with advanced dementia, brought in after a fall, resides in a nursing home. In light of her prognosis and prior expressed wishes, the clinician and family might decide to forgo a head CT for a minor unwitnessed fall, opting instead for observation and symptom management. International evidence supports the wisdom of this approach: integrating geriatric palliative care principles into dementia care has been shown to reduce unnecessary acute care utilisation significantly(11,12). These outcomes were achieved through proactive communication, advance care planning, and caregiver support – essentially, conversations and planning in lieu of crisis-driven ED trips. While conducted in the United States, this finding resonates with Australia’s push for dementia-friendly health systems and suggests that conversation-centred care can meaningfully improve outcomes.

Significance: By examining and implementing “conversations-before-investigations,” this research addresses a critical gap in emergency and aged care practice. It is grounded in the Australian context, where an ageing population and the recent National Dementia Action Plan underscore the need to improve dementia care across all settings(13). Yet the work will draw on and contribute to international knowledge, as the challenges it tackles – defensive medicine, communication barriers, and ethical dilemmas in acute care of the cognitively impaired – are global. This study is innovative in its interdisciplinary integration of perspectives from bioethics, ecological psychology, virtue ethics, Stoic philosophy, and feminist ethics of care. Such lenses will be used to interrogate the status quo and to design a more humane, values-driven model of emergency care. Ultimately, the proposed research is significant for its potential to shift ED culture from reflexive investigational “pathways” to a more reflective, conversation-led approach. In doing so, it aims to improve clinical outcomes (reducing delirium, stress, and unwanted treatments), uphold the dignity and autonomy of older patients with dementia, and support ED clinicians in practising under their ethical commitments rather than in fear of uncertainty. By centring meaningful conversation before investigation, emergency teams can better discern when high-tech intervention is genuinely in the patient’s interest – and when compassionate communication and comfort should take priority.

Research Aims and Questions

Aim 1: Identify and critically examine the factors that drive “default” investigation pathways for older ED patients with dementia or cognitive impairment. Research Question 1: What roles do fear (e.g. of missed diagnoses or litigation), epistemic uncertainty, “prognostic illusions,” and entrenched ED cultural norms play in clinicians’ decision-making for cognitively impaired older adults in emergency care? This aim will illuminate why current practice often defaults to extensive investigations by exploring clinicians’ perspectives and experiences.

Aim 2: Explore the perspectives of patients with dementia, their caregivers, and ED clinicians on the “conversations-before-investigations” approach to care. Research Question 2: How do people with dementia (and/or their families) and emergency staff experience and perceive the process of discussing goals and care preferences before medical investigations? This aim focuses on understanding the needs, values, and communication preferences of those directly involved – illuminating how ED processes could be redesigned to better centre the person with cognitive impairment. It will also identify perceived barriers and facilitators to implementing conversation-first practices in the ED context.

Aim 3: Develop an interdisciplinary, evidence-based framework for emergency care of older adults with dementia that operationalises “conversations before investigations” and evaluate its potential impact on clinical practice.Research Question 3: What would a model of emergency care look like that prioritises early ethical dialogue and relational decision-making for patients with dementia, and how might such a model affect outcomes such as investigation use, patient/caregiver satisfaction, and clinician moral comfort? Through a synthesis of theoretical insights and empirical findings (from Aims 1 and 2), this aim will produce a proposed care framework or set of guidelines. It will also consider strategies for implementation (education, workflow changes, policy) and lay the groundwork for future evaluation of the model’s effectiveness.

Research Design and Methodology

Overall Design

This study will employ a qualitative, interdisciplinary research design comprising two main empirical components – in-depth interviews and ethnographic observation – integrated with a strong conceptual analysis. The researcher’s philosophical positions are interpretivist and pragmatist, recognising that the realities of emergency dementia care are socially constructed through interactions yet aiming for practical solutions to improve care. The methodology is inherently transdisciplinary, as it interweaves empirical data with ethical and psychological theories to generate a rich understanding of the problem and potential interventions. The design is iterative and exploratory. Insights from initial observations may inform interview questions, and emerging themes from interviews will be interpreted through multiple conceptual lenses. By triangulating data from various sources and perspectives, the study will construct a comprehensive picture of the ED decision-making ecology and how it may be transformed.

Philosophical and Conceptual Frameworks

To ground the inquiry, the research intentionally draws on a nexus of five complementary frameworks:

  • Bioethics (Principlism and Beyond): The foundational bioethical principles – autonomy, beneficence, non-maleficence, and justice – provide an ethical yardstick for examining ED practices. In the context of dementia care, respect for autonomy entails finding ways to honour the patient’s values and prior wishes despite impaired decision-making capacity. This study adopts a relational interpretation of autonomy, acknowledging that decision-making for persons with cognitive disabilities is a shared process involving family and clinicians. The principle of non-maleficence(“do no harm”) is particularly salient: performing investigations that cause discomfort or harm without likely benefitting the patient violates this principle. By analysing current ED routines through a bioethical lens, the research will highlight where the practice may be ethically problematic (e.g. subjecting a frail person with dementia to a painful procedure of low yield) and where ethical duties (like communicating honestly about prognosis) are unmet. Bioethics thus motivates the need for change and informs the design of the conversation-first intervention to ensure it is ethically sound (e.g. upholding dignity and informed consent via dialogue).
  • Virtue Ethics: Virtue ethics shifts the focus from rules to the character and virtues of the clinician. This perspective asks: What virtues do emergency practitioners need to provide good care to cognitively impaired older people, and what vices or moral weaknesses might current pressures encourage? Literature on dementia care ethics argues that cultivating virtues such as courage and receptivity is crucial for ethical caregiving. In the ED, the virtue of courage is especially relevant – courage to resist the pervasive fear of uncertainty and litigation, courage to depart from aggressive protocols when they conflict with the patient’s best interests, and courage to “slow down” in a fast environment to truly listen. Similarly, virtues such as compassion, practical wisdom (phronesis), and patience are crucial in caring for vulnerable older adults. This framework will be used to interpret clinicians’ narratives about their decisions: for example, ordering every test “just in case” might be seen as an act of cowardice (a failure to face one’s fear of uncertainty or blame), whereas engaging in an uncomfortable goals-of-care conversation might exemplify moral courage. The research will also explore how ED culture might cultivate or stifle these virtues. Ultimately, the virtue ethics lens will inform educational recommendations – e.g. fostering moral courage and empathy in emergency training – as a complement to protocol-level changes.
  • Stoic Philosophy: Stoicism provides a timeless philosophical perspective on managing fear, emotion, and uncertainty – issues at the heart of this research. The Stoics advocated using reason and focusing on what is within one’s control while accepting the limits of life and one’s power. Applying Stoic principles, a clinician should strive to remain calm and rational in the face of an older patient’s complex presentation rather than panicking or being ruled by fear. Stoicism counsels that fear is a harmful passion that can cloud judgment; in an ED context, fear-driven reactions (such as reflexive over-investigation) can be seen as yielding to emotion over reason. The Stoic emphasis on acknowledging mortality and the natural course of life directly opposes the “prognostic illusion” of endless rescue – a Stoic-influenced clinician might better recognise when death is imminent and focus on a good death rather than futile intervention. For instance, Epictetus’s advice to “not rashly accept appearances” corresponds to avoiding rash action based on worst-case hypothetical diagnoses. This framework will enrich the interpretation of data by providing clinicians with a language and concepts to address their internal struggles. E.g. a doctor’s sense of anguish at notintervening aggressively could be re-framed as an opportunity to exercise Stoic virtues of acceptance and wisdom. Stoic philosophy thus supports the conversation-before-investigation ethos by encouraging practitioners to master their fear and make decisions guided by rational compassion, not defensive reflexes.
  • Feminist Ethics of Care and Relational Autonomy: A feminist ethics perspective centres relationships, context, and power dynamics in ethical decision-making. It challenges the hyper-individualistic notion of autonomy and instead promotes relational autonomy, wherein autonomy is achieved through supportive relationships and open communication. This is highly pertinent in dementia care: patients often cannot voice their wishes independently, so autonomy must be co-constructed with input from caregivers who know the person and professionals who guide decisions responsibly. The ethics of care emphasises empathy, attentiveness, and responsibility in caring relationships – qualities often undervalued in the task-driven ED setting. Using this lens, the study will examine how current ED processes may marginalise the voices of patients and families (for example, families not being consulted before a hospital transfer) and how a care ethic could rectify that. Feminist ethics also attune the research to issues of power: ED clinicians hold significant power over vulnerable patients, and this power imbalance can lead to paternalism if left unchecked. The data analysis will thus pay attention to instances where patients or surrogates are excluded from decisions or where institutional norms silence specific perspectives. By involving caregivers and (when possible) people living with dementia in the research (e.g. through interviews or co-design of materials), the study also embodies feminist research principles. Insights from this framework will directly shape the conversation protocols proposed – for example, ensuring they empower patient and family voices and train clinicians in empathetic listening and shared decision-making rather than authoritative directives. Ultimately, a care ethics approach reinforces the necessity of relationship-centred care, echoing evidence that relationship factors, such as empathy and effective communication, are core components of a dementia-friendly ED.
  • Ecological Psychology (Systems and Environment): Ecological psychology examines the dynamic interactions between individuals and their environment. This study provides a framework to analyse the ED as an environment – a fast-paced, sensory-overloading, protocol-saturated space – and how that environment influences behaviour and decision pathways. Key to ecological psychology is the concept of affordances: the environment “affords” specific actions. The ED, for example, affords rapid testing (with on-site labs and imaging available 24/7) and affords only brief communication (due to noise, time pressure, and lack of privacy). These environmental cues can unconsciously steer clinicians towards investigations and away from conversations. Using an ecological lens, the ethnographic component of the research will observe how physical layout, sounds, and workflow processes affect interactions with older patients. Does the lack of quiet rooms or the constant beeping of monitors deter family meetings? Do triage protocols encourage providers to order a standard battery of tests for every confused older person? By mapping these factors, we can identify modifications to the environment and system that would afford conversation: for instance, creating a designated low-stimulation room for cognitively impaired patients or integrating a “goals of care” prompt into the electronic order system. Additionally, an ecological approach extends to systems thinking – recognising that ED care for dementia involves a broader ecosystem (paramedics, aged care homes, primary care, and inpatient services). The research will consider how inter-system communication and resources (or lack thereof) drive ED decision-making. For example, poor integration of advance care plans into ED information systems is a known barrier to honouring patient wishes. By viewing the problem as an ecosystem issue rather than just individual behaviour, the study will be able to recommend multi-level interventions (workflow redesign, policy changes, cross-sector communication tools) that create a context more conducive to conversation and individualised care.

Together, these frameworks form a robust conceptual foundation. They will be explicitly referenced in the analysis and interpretation of findings, ensuring that the discussion moves fluidly from concrete observations (what is happening in the ED) to higher-order understanding (why it’s happening, what it means ethically) and finally to actionable recommendations (how we can change it). Notably, the use of multiple lenses also guards against one-dimensional analysis: for example, if interviews reveal a clinician ordering a CT on an agitated nursing home resident, a bioethics lens might critique it as non-beneficial (violating non-maleficence), a virtue ethics lens might see it as lack of courage to choose comfort, a stoic lens might identify fear driving the choice, a care ethics lens would note the absence of caregiver input, and an ecological lens might blame a protocol or environment that made that CT the path of least resistance. These interpretations, taken together, provide a deep, nuanced understanding that will inform a more effective intervention model.

Qualitative Interviews

Participants: The study will conduct qualitative interviews with three stakeholder groups: 

  • ED Clinicians – including doctors (consultants and trainees) and nurses working in Australian emergency departments, particularly those with significant experience treating older patients. 
  • Older Patients with Dementia or Cognitive Impairment – specifically, those with mild to moderate dementia who have the capacity to share their recent ED experiences (or dyads of a person with dementia and their family caregiver to support communication); and
  • Family Caregivers/Decision-Makers of older adults with more advanced dementia who have visited the ED. Including this range of perspectives ensures that the research captures the whole triadic interaction in emergency care (patient, family, clinician). 

A purposive sampling strategy will be employed to achieve diversity in participants – for clinicians; sampling will seek variation in years of experience, hospital settings, and professional roles; for patients and caregivers, diversity in the background (including representation of Indigenous and culturally diverse communities, given different cultural attitudes to dementia and end-of-life care) will be sought. The anticipated sample size is approximately 20–25 clinicians, 10–15 patient/caregiver dyads, and 10–15 caregivers of patients with advanced dementia, or until data saturation is reached within each cohort.

Interview Procedures: Interviews will be in-depth and semi-structured, lasting ~60 minutes each. A semi-structured interview guide, informed by the literature and the study’s aims, will be used flexibly to ensure key topics are covered while allowing participants to introduce what they feel is essential. For clinicians, interviews will explore their experiences and thought processes when managing older patients with dementia in the ED. Sample questions include: “Can you walk me through a recent case of an older patient with dementia in your ED? What decisions were made and why?”; “How do you feel when you’re not sure about a diagnosis – what influences your next steps?”; “Have you ever felt that you ordered a test or treatment that, in hindsight, might not have benefited the patient? What factors contributed to that decision?”; “Describe any occasions where you took time to talk with a patient’s family about goals of care in the ED. What prompted that conversation, and how did it affect the care plan?”. These prompts aim to elicit narratives about fear, uncertainty, and norms (Aim 1) as well as attitudes towards the proposed “conversation-first” approach (Aim 2). Clinicians will also be asked directly about their perceptions of the feasibility and impacts of implementing more conversation (e.g. “What do you see as obstacles to having deeper conversations before ordering tests in the ED?” and “What might help you feel more comfortable not doing a test that is unlikely to benefit a patient with severe dementia?”).

For patients with mild dementia, interviews (likely with a caregiver present to support) will focus on their recollection of ED visits: “How did the doctors and nurses talk with you when you were in the emergency department?”; “Did you feel involved in the decisions about your care? Why or why not?”; “What would you have wanted the doctors to know about you or to ask you that perhaps they didn’t?”. With family caregivers, interviews will gather their experiences and expectations: “Tell me about when you took your family member with dementia to the ED – what was that experience like for you and them?”;“Were you able to discuss your loved one’s medical history or preferences with the ED staff?”; “Do you think the tests and treatments done were in line with what your loved one would have wanted or needed?”. Caregivers will also be queried about how they’d envision an ideal ED encounter and their reaction to the idea of a doctor pausing to discuss options (e.g. “If the ED team had sat down with you on arrival to talk about what you and your loved one hope for from care, how do you think that might have changed things?”). Throughout, the interviewing approach will be empathetic, especially for caregivers who may recount stressful or traumatic experiences. If needed, shorter or multiple sessions will be arranged for patients with cognitive impairment to minimise fatigue.

Data handling: All interviews will be audio-recorded with consent and transcribed verbatim. Field notes will be written immediately after interviews to capture the interviewer’s observations (e.g. emotional tone, key points). Transcripts will be de-identified; all names and specific locations will be replaced with codes to maintain confidentiality. Participants will be given pseudonyms or codes in all records. Transcripts will be imported into qualitative data analysis software (e.g., NVivo) to facilitate the organisation and coding of the data.

Ethnographic Observation in Emergency Departments

To complement the interviews, an ethnographic component will be undertaken to observe real-time practices and interactions in the ED involving older patients with cognitive impairment. Ethnography is suitable for capturing the context, behaviours, and social norms that people may not fully articulate in interviews. It aligns with the ecological and sociocultural focus of this study, allowing for a firsthand understanding of how the ED environment and culture shape care.

Field Setting and Access: The researcher will conduct observations in one or two EDs (for example, one large metropolitan ED and one regional ED in Australia) that serve substantial numbers of older patients. Necessary permissions will be obtained from hospital administration, and ethical approvals will cover observation of staff and patients (with appropriate consent processes described below). The researcher will likely take on the role of an observer-as-participant, meaning they will not be involved in clinical care but will be a visible presence, observing and occasionally asking clarifying questions of staff during breaks (without disrupting care). The observations will focus on cases of older adults (approximately age ≥75) who exhibit dementia, delirium, or other cognitive impairment during their ED visit.

Observation Procedures: The researcher will employ shadowing and field note-taking as primary data collection techniques. They may shadow an emergency clinician (with that clinician’s consent) for specific shifts or station themselves in areas such as triage or resuscitation where older patients are commonly present. They will pay particular attention to decision points (e.g. when tests are ordered when disposition decisions are made) and the communication (or lack thereof) surrounding those decisions. Key aspects of observing include how the initial assessment is done for a confused older patient (Are family or nursing home records consulted? Is there a discussion of advance directives? Or is it straight to blood tests and IV lines?); how clinicians discuss the case among themselves (Do they express fear of missing something or pressure to follow protocol? Do they mention the patient’s dementia in decision-making or mainly focus on acute issues?); how and when family members are included (Are conversations occurring? Are families pleading for or against specific interventions?); and environmental factors (Is the patient put in a quiet room or left in a hallway bed with alarms beeping? How does that affect their behaviour and the staff’s approach?).

The researcher will unobtrusively note instances that illustrate default pathways – for example, if an elderly man with advanced dementia and multiple comorbidities is automatically sent for a full-body CT scan after a minor fall, the researcher will document what prompted that decision (was it routine policy, or did anyone question it?). Conversely, any instances of conversations before investigations in action will be golden nuggets: e.g., if a clinician takes the time to contact a patient’s GP or family to discuss goals or if they sit at the bedside to explain options to a patient – these will be documented in detail as positive deviance cases. The observational field notes will be as descriptive as possible, capturing direct quotes, timing of events, and interactions. Where relevant, the researcher may also review documentary artifacts in the ED, such as protocol checklists or forms (for example, if there’s a sepsis protocol that mandates numerous tests, how does that interplay with clinical judgment?).

Consent and Ethical Considerations in Observation: Observing patient care raises ethical issues, especially with vulnerable patients like those with dementia who cannot fully consent. The study will implement an opt-in consent processfor observing specific cases. For example, if a patient with dementia arrives accompanied by a family member, the researcher (or a designated nurse) will briefly explain their presence and ask the caregiver (or the patient, if capable) for verbal consent to observe. In cases with no family present and the patient lacks capacity, the ethics protocol may allow observation under a waiver of consent (justified by the minimal risk and high potential benefit of improving care practices), but the researcher will remain vigilant to protect privacy (not recording any identifiers) and will cease observing if anyone (patient or staff) appears uncomfortable. Staff members in the ED will be informed about the research, and their consent to be observed will be obtained collectively (e.g. through staff meetings and written information with an opt-out option). The researcher will take care to position themselves in a way that does not interfere with care and to respect confidentiality (for example, not reading medical records over others’ shoulders). All field notes will be coded and anonymised (e.g. “Patient A, an 82-year-old with dementia, accompanied by daughter, presented with X…”).

Reflexivity in the Field: Given the researcher may have a clinical background in emergency or aged care (which is likely, considering the interdisciplinary nature), reflexivity is crucial during observation. The researcher will maintain a reflexive journal to record their reactions, assumptions, and emotional responses to the situations observed. For example, witnessing a distressing over-treatment might provoke frustration or confirmation bias in the researcher; these feelings will be acknowledged and bracketed to avoid skewing the impartial recording of events. By being mindful of one’s lenses (e.g., virtue ethics), the researcher can differentiate between what is actually observed and the interpretation thereof, which will be presented later in the analysis.

The ethnographic data will ultimately provide a thick description of ED practices – the kind of granular detail that an interview alone might miss (such as a nurse’s body language when a physician orders another test or a family member’s unanswered question in the chaos). Such detail is invaluable for identifying systemic issues and for designing interventions that fit the real context.

Data Analysis Plans

The analysis will proceed in phases, corresponding to the different data sources, and then converge into an integrative interpretation that addresses the research questions. All analysis will be supported by qualitative data analysis software; however, coding and theme development will be conducted following established qualitative techniques.

Thematic Analysis: The primary analytic approach for the interview data (and much of the field note data) will be thematic analysis, using a blend of inductive and deductive coding. Initially, transcripts from interviews will be openly coded line-by-line to capture meaningful units (descriptive codes like “fear of blame,” “family not consulted,” “default protocol,” “expressed uncertainty,” “prior negative outcomes,” “value of comfort,” etc.). An inductive stance will ensure that unexpected themes (perhaps unique cultural factors or emotional nuances) are not overlooked. At the same time, some deductive codes derived from the conceptual frameworks and research questions will inform the coding schema – for instance, codes such as “malpractice fear (litigation),” “uncertainty intolerance,” “cultural norm – thoroughness,” “clinician virtue or vice,” “ethical conflict,” “examples of conversation-first,” etc. The coding process will involve multiple researchers/coders whenever possible to enhance reliability. An initial set of transcripts will be double-coded by two team members, and coding differences will be discussed to refine the codebook.

After coding, codes will be clustered into candidate themes and sub-themes. For Aim 1 (factors driving default investigations), likely themes might include “Defensive culture and fear” (with sub-themes like fear of missing something, fear of legal consequences, professional peer pressure); “Diagnostic uncertainty and cognitive bias” (sub-themes: discomfort with uncertainty, overestimation of intervention benefit, cognitive biases such as the action bias); “ED norms and time pressure” (sub-themes: rushing, task-oriented care, lack of geriatrics resources); and “Emotional burden and moral distress” (sub-themes: clinicians’ anxiety, regret in hindsight). For Aim 2 (perspectives on conversation-first), themes might include “Value of communication” (with patients/caregivers stressing the importance of being heard, and clinicians noting improved clarity when conversations happen), “Communication barriers” (e.g. noise, lack of time, lack of training, or attitudes like “families are in denial”), “Navigating goals of care” (stories of conflict or alignment between staff and family on what the plan should be), and “Trust and relationship” (how conversations – or the lack – affect trust). The interview data will yield rich narrative examples that will be used to illustrate each theme in the write-up, accompanied by participant quotations.

Narrative and Discourse Analysis: In addition to broad thematic analysis, the study will incorporate narrative analysis for selected cases to preserve the holistic story and examine the meaning-making by participants. For example, a clinician’s recounting of a specific patient case can be analysed as a narrative: what plot (sequence of events) do they construct? Do they frame themselves as a hero, victim, or bystander? How do they justify the outcome? This can reveal insights into prognostic illusions or ethical reasoning. A clinician might narrate a case where they ordered extensive tests for a 90-year-old with dementia who ultimately died – analysing their language might show an internal conflict between duty and doubt or reveal how they rationalise the default pathway (“I just had to be sure nothing was treatable”). Similarly, caregiver narratives of ED experiences can be analysed for themes of empowerment vs. alienation. The narrative analysis will thus complement thematic coding by providing in-depth vignettes that humanise the themes and highlight contradictions or tensions.

Furthermore, a discourse analysis approach will be employed to examine how language reflects power and ethics. For instance, the way clinicians describe a patient (“frequent flyer,” “no code,” “bed blocker”) might be scrutinised, as such terms carry value judgments that can affect care. The analysis will note whether paternalistic language is used (“we decided for her best” vs. relational language, “we talked and agreed…”). Attention to language and story is essential, given the philosophical frameworks; for example, is there evidence of virtue language (such as courage, duty) or fear language? Are Stoic principles echoed in anyone’s accounts (acceptance of death, focus on control vs uncontrollable)?

Ethnographic Data Analysis: Field notes from observations will undergo their thematic coding, though with a focus on process and interaction themes. The observer will code notes for events such as “investigation ordered without discussion,” “family attempts to communicate,” “staff-staff communication patterns,” “environmental stressors (noise, crowding),” and “adaptations made for patient comfort,” among others. These codes will help identify patterns (for example, one might find that during busy times, conversations drop, and test-ordering spikes, or that specific triggers, such as “possible head injury,” automatically lead to protocols overriding individual judgment). Ethnographic data will particularly inform themes related to systems and environment. For instance, a theme might be “Institutional momentum”, describing how once a patient is on a pathway (e.g. triaged to resuscitation), a cascade of tests happens almost automatically. Another might be “Missed moments for conversation”, noting when opportunities for communication arose but were lost in the shuffle. The ecological perspective will drive analysis of how space and process influenced each noted interaction.

After separate analyses, the integration phase will merge insights. A technique akin to framework analysis may be used, where the themes from each data source are mapped onto the research questions and conceptual framework components. The researcher will create a matrix with rows representing key themes and columns for data sources (clinician interview, caregiver interview, observation, and theoretical lens insights). This allows for cross-validation (triangulation); for example, if both interviews and observations highlight “time pressure” as a barrier to conversation, that finding is strengthened. If caregivers report not being heard and observations confirm that families often wait unaddressed, that triangulates the issue of communication breakdown. Discrepancies will also be noted (for example, if clinicians claim they always discuss problems with families, but observations show otherwise, that gap is itself a finding about perception versus practice).

Role of Frameworks in Analysis: The earlier-described conceptual frameworks will not only guide interpretation but can also serve as a coding framework if appropriate. For example, one might tag segments of data with codes like “beneficence vs non-maleficence conflict,” “expression of moral distress (virtue ethics),” “language of fear (Stoicism),” and “care relationship present/absent (care ethics).” This helps ensure the final analysis explicitly addresses the philosophical dimensions. During the discussion chapter, these codes will help organise how results relate to each framework (e.g. a section on “Navigating Uncertainty: Stoic and Virtue Ethics Insights” might draw together codes of fear, courage, and acceptance from across the data).

In terms of software, after coding and theme development, visual mapping tools (like mind maps or network diagrams in NVivo) will be used to visualise connections between themes (such as how “fear of missing something” connects to “protocol use” and to “lack of conversation”). This visualisation can spur more profound analytic questions: why are those connected? Is there an underlying concept, such as “professional identity,” linking fear and thoroughness?

Throughout the analysis process, peer debriefing will be employed: the researcher will regularly discuss emerging findings with advisors or a small reference group (possibly including an ED clinician, a geriatrician, and an ethicist) to challenge interpretations and ensure that alternative explanations are considered. This is especially vital given the researcher’s potential biases; external input helps to validate that themes are grounded in data, not just expectations.

Ensuring Trustworthiness

Ensuring the credibility, dependability, confirmability, and transferability of the findings (the qualitative equivalents of validity and reliability) is a priority in this study. Several strategies will be implemented:

  • Triangulation: By drawing on multiple data sources (interviews with various stakeholders, as well as direct observations) and numerous theoretical lenses, the study employs both data triangulation and theoretical triangulation. Converging evidence from interviews and observations will increase the credibility of the results. For example, if both caregivers’ words and field observations independently indicate that families are being sidelined in decisions, one can be confident in that finding. The use of different investigators or peer debriefers in the analysis also serves as analyst triangulation, thereby reducing the likelihood that one researcher’s subjective view will dominate.
  • Member Checking: Where feasible, preliminary findings or interpretive summaries will be shared with participants (primarily clinician and caregiver interviewees) to validate the researcher’s understanding. This could be done in follow-up meetings or via email summaries of themes. For instance, the researcher might share a summary like: “Many clinicians in our study noted that fear of missing a diagnosis leads them to over-test. Does this resonate with your experience?” and invite feedback. For caregivers, one might present a distilled narrative (de-identified) of a typical ED experience and ask if it feels familiar or if anything is misrepresented. Member checking will help refine the accuracy of themes and ensure the participants’ voices are correctly understood. It also enhances authenticity, giving participants a chance to correct or elaborate on the researcher’s interpretation.
  • Reflexivity: As mentioned, the researcher will engage in continuous reflexivity by maintaining a journal to record thoughts, assumptions, and emotional responses. In writing up methods and results, the researcher will state their positionality (e.g. “The researcher is an emergency physician with training in bioethics, which informed her interest in this topic and may incline her to see certain practices as problematic. She remained aware of this bias and also sought to represent clinicians’ constraints fairly.”). By being transparent about perspective, readers can judge how it might influence the research. Reflexive notes will also be used during analysis meetings to discuss how the researcher’s background may shape theme development (for example, ensuring that themes are truly evidenced by participant data rather than being imposed by preconceptions).
  • Audit Trail: The project will maintain a clear audit trail of all research activities and decisions. This includes documentation of how interview guides were developed, how codes were generated and refined (with examples of merging or splitting codes), theme evolution records, and decision points in interpreting ambiguous data. Research memos will be written throughout the analysis to capture the rationale behind analytic choices. All versions of the coding manual and analytic memos will be preserved. An external reviewer (such as a supervisor) could review portions of this audit trail to confirm that the findings are grounded in data. This approach bolsters dependability and confirmability by demonstrating that the findings are not the result of random or arbitrary processes but rather the product of traceable and logical analysis steps.
  • Rich, Thick Description: In reporting results, the study will provide detailed descriptions of the context and participants, along with numerous direct quotes from interviews and illustrative excerpts from field notes. This detailed description enables readers to assess the transferability of the insights to other settings. For example, by detailing the Australian ED context (staffing, patient demographics, etc.) and then noting, “However, similar dynamics have been observed in North American and UK EDs in the literature,” the study acknowledges both specificity and broader relevance. Transferability is further addressed by linking the findings to international literature (the study will reference global dementia care and emergency medicine practices), suggesting that while the context is Australian, many conclusions are likely to apply in other healthcare systems facing an aging population.
  • Engagement and Duration: The researcher will spend a substantial amount of time in the field (e.g., several weeks of observation across different shifts) and in interaction with participants, which lends credibility through prolonged engagement. Being present in the ED environment over time helps the researcher build trust with staff (so they act naturally) and better understand subtleties (distinguishing an unusual one-off event from a routine pattern). Persistent observation will ensure that the data collected are representative of typical ED operations and not just a snapshot that might have been anomalous.
  • Negative Case Analysis: The analysis will actively seek out and examine cases that deviate from emerging patterns. For instance, if most evidence suggests that over-investigation is the norm, but there are a few instances of minimal investigation without adverse outcomes, those will be analysed to question the rule and refine the understanding. Similarly, suppose a couple of clinicians report no fear of uncertainty (perhaps they have a high tolerance or a different approach). In that case, those narratives will be scrutinised to see if they truly contradict the main themes or if there is an explanation (such as they had special training or support). Incorporating negative cases helps ensure the conclusions are robust and nuanced rather than overly generalising.

Ethics and Researcher Reflexivity

Ethical conduct is paramount, given the involvement of vulnerable individuals (people with dementia) and the sensitive critique of clinical practices. Ethical approval will be obtained from the appropriate Human Research Ethics Committees (HRECs) covering both the academic institution and the participating hospitals. The ethics application will detail how informed consent is handled, how data are protected, and how potential risks are mitigated.

Informed Consent: For interviews, written informed consent will be obtained from all participants (or their legal guardians in the case of participants who lack capacity). The consent process for patients with dementia will be tailored to their cognitive level – those with capacity will provide consent directly (with clear, simple explanations and confirmation of understanding), whereas, for those lacking capacity, a proxy consent from a legally authorised representative (usually next of kin or power of attorney) will be obtained. Even when proxy consent is used, the patient will be involved as much as possible (through assent), with consideration given to their comfort during the process. The consent forms will emphasise that participation is voluntary, that participants may decline to answer any question or withdraw at any time, and that their decision will not affect their care.

For observations, as discussed, a combination of consent and ethical waiver will be used. ED staff effectively consent to be observed by agreeing to the study’s presence (with confidentiality assured), and patient-related observation will involve situational consent when possible. The researcher will refrain from recording any personally identifiable details in field notes.

Privacy and Confidentiality: Strict measures will be taken to protect participants’ confidentiality. Audio recordings and transcripts will be stored on password-protected devices or secure servers, accessible only to the research team. In publications or the thesis, individuals will not be identifiable – quotes will be anonymised, and any potentially identifying context details will be altered or omitted. Given the small community of emergency clinicians, extra care will be taken (for example, using generic descriptors like “a regional hospital ED nurse stated…” rather than naming the hospital or any unique role). Participants will have the option to review their transcript to ensure that any sensitive information is controlled.

Potential Risks and Safeguards: There is a possibility that interviews could cause emotional distress, particularly for caregivers recalling traumatic ED events or clinicians confronting moral distress. The researcher will conduct interviews with empathy, allowing breaks if someone becomes upset and refraining from pressing any question that clearly causes discomfort. A protocol will be in place to refer participants for support if needed. For instance, having on-hand contacts for counselling services or dementia support organisations for caregivers and clinicians will be encouraged. Additionally, the use of employee assistance programs will be promoted if the discussion raises unresolved stress. However, many participants may actually appreciate the opportunity to voice their experiences – indeed, caregivers often find meaning in contributing to research that may improve care for others.

There is also an ethical obligation to consider if, during the research, an acute ethical issue is observed (e.g. a case of overt patient harm or neglect in the ED). The researcher, being a clinician as well, will have to balance these roles carefully. The primary role in the study is the researcher, not the intervening clinician; however, per ethics and good conscience, the researcher cannot ignore a serious ethical breach. The ethics protocol will include a plan; for example, if severe distress or unsafe care is witnessed, the researcher might discreetly bring it to the attention of the treating team or an appropriate authority after the fact. In less acute situations, the data would be noted and later used (de-identified) to inform recommendations.

Researcher Reflexivity: The researcher’s background (for example, as an ED physician concerned with geriatrics and ethics) confers both insight and bias. Reflexivity has been integrated throughout the design and analysis, and it is reiterated here as a core ethical stance. The researcher continuously reflects on how their values (perhaps a bias toward less intervention and more palliative approaches) might colour interactions. They must be careful, for instance, not to leadclinician interviewees to feel defensive or to supply “right” answers. Interview questions will be phrased neutrally (e.g. “Tell me what happened” rather than “Don’t you think that was too much?”). During observation, reflexivity means acknowledging any urge to intervene or any judgment formed and consciously pulling back to remain an observer. By keeping reflexivity at the forefront, the researcher ensures the data collected are as authentic as possible and that the analysis considers multiple angles, not just the researcher’s viewpoint.

Finally, given the interdisciplinary nature, ethical reflexivity also involves consulting with experts in philosophy or ethics when interpreting ethically charged data. For example, suppose a clinician describes an ethically grey decision. In that case, the researcher may discuss (without revealing identity) with a bioethicist colleague to verify that their interpretation is reasonable and to consider alternative ethical perspectives.

Proposed Thesis Structure

The PhD thesis will be organised into chapters that reflect the research journey from context to findings to implications. A provisional structure is as follows:

  1. Introduction – This opening chapter will introduce the research problem and its significance. It will clearly state the aims and research questions and provide an overview of the thesis structure. The introduction will set the scene with a vignette or scenario that illustrates the issue (e.g., a brief story of a person with dementia in the emergency department facing a cascade of tests), hooking the reader to real-world importance. It will also outline the interdisciplinary approach and define key terms (such as “conversations-before-investigations”).
  2. Background and Literature Review – This chapter will offer a comprehensive review of relevant literature and frameworks. It will cover the epidemiology of dementia in acute care (drawing on sources about ED utilisation and outcomes for people with dementia), known challenges in emergency medicine for older patients (e.g. communication barriers, delirium, over-investigation), and current guidelines or initiatives (like Age-Friendly Health Systems, Geriatric ED models, and Australian dementia care standards). Crucially, this chapter will also delve into the conceptual literature, summarising the contributions of bioethics, virtue ethics, Stoicism, feminist care ethics, and ecological psychology to healthcare practice and identifying gaps. For instance, it may review prior studies on clinical uncertainty and defensive practice, ethical decision-making in dementia care, and models of shared decision-making. By the end of the chapter, the reader will understand the state of knowledge and how this research will extend it. The chapter will likely end by refining the research questions in light of the literature.
  3. Methodology – This chapter outlines the study design and methods, as summarised in the sections above. It will articulate the philosophical paradigm (interpretive, qualitative inquiry) and justify why a qualitative ethnographic approach is appropriate to the aims. Subsections will describe participant recruitment, data collection methods (including interviews and observations), and analytical techniques (thematic and narrative analysis), thereby expanding on the plan outlined previously. The use of the theoretical frameworks in guiding analysis will be reiterated. This chapter will also include ethical considerations, describing consent processes, risk mitigation, and reflexivity. By providing transparency about how the research was conducted, this chapter underpins the trustworthiness of the findings. (If any adjustments to methods occurred during the study, those will be explained here as well.)
  4. Findings – Factors Driving Default Investigations (Aim 1) – The first chapter of findings will present results related to Aim 1, focusing on why the current ED culture defaults to extensive investigations for older patients with dementia. It will be organised into thematic sections such as: “Fear and Defensive Practice” (with evidence like clinicians admitting fear of malpractice leads to “better safe than sorry” testing ), “Uncertainty and the Comfort of Tests” (clinicians describing discomfort with diagnostic doubt and the relief a clear scan gives them, even if it’s likely normal), “ED Norms and the Pressure to Act” (nurses and doctors describing the unwritten rules of always ‘doing something’ and feeling guilty if they don’t), and “Prognostic Illusions and Hope” (instances where staff or family had overly optimistic expectations influencing aggressive care). Each theme will be illustrated with quotations from interviews and, where relevant, observational anecdotes. For example, under “Pressure to Act,” a quote might be: “If an elderly dementia patient comes in with, say, confusion, there’s almost an expectation from everyone – the triage note, the next shift – that bloods and imaging will be done. You’d be ridiculed if you didn’t”. Observational data might add: During one observed resuscitation, an ED registrar hesitated to perform a lumbar puncture on an 85-year-old with advanced dementia and sepsis, but a senior doctor commented, ‘We have to dot all i’s,’ so the procedure went ahead.The chapter will synthesise how these factors interrelate, possibly culminating in a conceptual model (e.g., a diagram) of the “default pathway,” showing the cascade from uncertainty to fear to tests to further interventions reinforced by culture.
  5. Findings – Experiences and Perceptions of Conversations before Investigations (Aim 2) – The second findings chapter will centre on participants’ perspectives regarding communication and decision-making in the ED. It will include themes that emerged about the absence or presence of conversation and its impact. Potential sections: “Voices Unheard” (caregivers describing instances where they were not asked about the patient’s history or wishes, and the consequences of that; patients feeling confused about what was happening ), “When Conversations Happen” (positive examples where ED staff did engage in discussion – how it helped clarify goals or avoid unnecessary treatment), “Barriers to Communication” (both practical barriers like noise, and attitudinal ones like clinicians’ discomfort with end-of-life discussions), and “Perspectives on the New Approach” (how participants reacted to the idea of ‘conversations-before-investigations’ – perhaps clinicians saying it’s ideal but “not realistic in a busy ED” versus others who have found it is feasible with minor adjustments; caregivers likely endorsing it, e.g. “I wish they had asked me about Mom sooner; it would have saved her a lot of pain”). Throughout this chapter, differences or tensions between stakeholder groups will be highlighted. For instance, clinicians might believe they communicated well, but caregiver interviews may contradict that – such contrasts will be presented to underscore the need for improvement. The chapter will effectively answer how communication (or lack thereof) currently unfolds and is perceived in dementia emergencies, providing the user-centred evidence base for the proposed intervention.
  6. Discussion – This chapter will integrate the findings with the conceptual frameworks and broader literature, addressing Aim 3. It will likely be structured around key thematic discussions rather than by data source. Subsections might include: “Reframing Risk and Uncertainty in Emergency Dementia Care” – discussing how the findings about fear and uncertainty can be addressed by fostering a Stoic resilience and virtue ethics approach among clinicians (e.g., training to improve tolerance of uncertainty and encourage moral courage to deviate from overuse). Here, the study might propose, for example, reflective workshops or narratives that help clinicians recognise when doing less is actually virtuous and in line with the principle of primum non nocere. “Towards a Relationship-Centred Emergency Care” – using feminist ethics and the findings on communication to argue for systemic changes that elevate conversations: this could reference how relational autonomy can be operationalised in ED (perhaps by having a checklist that includes contacting the substitute decision-maker early or involving geriatric consultation for goals-of-care talks). The discussion will compare the results with existing international strategies, for instance, aligning with the concept of dementia-friendly EDs (as others have co-designed statements emphasising empathy and communication). It will acknowledge any surprising findings or contradictions and explain them in the context of relevant theory. “Ecological Interventions: Changing Environments and Norms” – a subsection focusing on system-level fixes, like adjusting triage protocols to trigger a palliative care consult for certain patients (inspired by how IN-PEACE proactively kept people out of ED) or redesigning ED physical space and workflow as recommended in the literature (e.g. quieter spaces, better integration of care information) The discussion will also address limitations of the study, such as generalizability (noting the context specifics) and potential biases, while arguing why the conclusions still hold merit. Overall, the chapter will propose the new framework or model – likely presented as a set of principles or steps for “conversations-before-investigations” – and justify it by showing how it answers the failings identified. For instance, it might outline: Principle 1: Elicit goals of care on arrival (addresses autonomy and reduces unnecessary intervention); Principle 2: Train for uncertainty (addresses fear); Principle 3: Adjust the environment to facilitate communication, etc., tying each back to data and ethical theory.
  7. Proposed Framework for Practice (optional chapter) – Depending on the structure, the thesis may include a distinct chapter that concisely presents the actionable output of the research: a draft protocol or guideline for implementing “conversations-before-investigations” in EDs. This could be formatted as a stand-alone chapter or as an appendix. It would synthesise all preceding analyses into a practical toolkit (for example, including conversation templates, suggested policy changes, and educational curriculum points). If the PhD includes a small pilot implementation or feasibility study (although not explicitly stated in the proposal, it could be a future direction), the results of that study could be included in this chapter. If not, the chapter serves as a bridge from research to practice, which examiners and stakeholders often appreciate.
  8. Conclusion – The final chapter will summarise the key findings and contributions of the research. It will revisit the aims and confirm how each was addressed. Importantly, it will discuss implications for clinical practice, policy, and further study. It will likely assert statements such as: “This research demonstrates that a significant portion of investigations in ED dementia care are driven by modifiable factors (fear, habit, lack of communication) rather than patient benefit. By integrating ethical conversations early, ED care can be recalibrated to serve patients’ wellbeing and dignity better.” Practical recommendations will be enumerated for clinical care, which may include introducing briefing huddles in the ED for complex geriatric cases to discuss goals or implementing decision aids for families explicitly. For policy implications, support for geriatric emergency nurse consultants or routine access to advance care plans across hospitals (such as readily checking if a patient has a “do not hospitalise” directive) may be necessary. The conclusion will also emphasise international relevance – noting that as populations age globally, the lessons from this Australian-focused research (improving emergency care through communication and ethical clarity) are broadly applicable. It may reference how the findings could inform global initiatives, such as the WHO’s dementia action plan or the dissemination of Geriatric ED guidelines. A short reflection on how the interdisciplinary approach enriched the findings may be included to encourage similarly holistic strategies in addressing healthcare challenges. Finally, the chapter will outline avenues for future research: perhaps testing the conversation-before-investigation model in a trial to measure outcomes (ED revisit rates, patient satisfaction) or exploring its applicability to other populations (e.g. adults with intellectual disabilities facing acute care, as noted in the proposal prompt to nod to cognitive disability).

Each chapter will begin with a brief introduction and conclude with a summary, maintaining clarity and logical flow through academic writing conventions. The thesis will be written in a formal, scholarly style appropriate for publication in medical and health research domains while also ensuring accessibility to the multiple disciplines it engages (medicine, ethics, psychology, etc.). Citations will be thoroughly provided throughout, as in this proposal, to ground the analysis in the existing literature and to demonstrate academic rigour.

Conclusion and Implications

In conclusion, this proposed PhD research aims to generate a deep, multidimensional understanding of how emergency care for older adults with dementia can be improved by slowing down and prioritising conversation before investigation. The driving hypothesis is that many interventions in the ED are initiated out of habit, fear, or uncertainty rather than patient-centred reasoning – and that by inserting a compassionate, ethical dialogue early in the care process, we can often avoid unnecessary tests while improving care quality. The research will produce evidence that challenges the status quo; for instance, it is expected to show that clinicians’ “worst-first” mindsets and ED culture of urgency can sometimes lead to over-treatment that is misaligned with these patients’ actual prognoses and preferences. By surfacing these issues, the study equips clinicians and policymakers with insight to inform practice change.

Practical Implications for Clinical Care: The findings of this work will have direct applicability to the ED. If successful, the research will outline clear strategies for ED teams to implement. For example, EDs might introduce a protocol that for any patient ≥80 with known dementia, the attending physician must pause after initial stabilisation to conduct a “goals-of-care huddle” – whether that’s a quick bedside talk with family or a phone call to the patient’s nursing home or GP to understand the care context. Clinicians may incorporate simple but powerful questions into their routine, such as asking caregivers, “What would your loved one consider a good outcome? What are their values and wishes?” before proceeding with high-intensity interventions. The research will likely advocate for the creation of dementia-specific guidelines or checklists in ED (analogous to trauma checklists, but tailored for cognitive impairment) that remind staff to assess issues such as baseline cognitive function, decision-making capacity, and the presence of advance directives, and to involve surrogate decision-makers early. Educational implications include training emergency personnel in communication skills and ethical decision-making: simulation exercises or workshops could be developed (inspired by this research’s findings) to practice handling uncertainty and having end-of-life conversations in the frenetic ED context. By highlighting the experiences of patients and families, the study will also sensitise ED providers to the human impact of their practices – fostering empathy and a mindset that sees talking with the patient as part of the treatment. Ultimately, the goal is a culture shift: from an ED that treats every older adult as a set of symptoms to be investigated to an ED that treats them as persons whose overall well-being and values guide clinical actions.

Policy and System-Level Implications: At a health system level, this research could influence policy in several ways. Hospitals and healthcare services may revise their policies regarding the care of the elderly and individuals with cognitive disabilities in acute settings. For instance, emergency departments might be encouraged (or required) to attain “dementia-friendly” status akin to how some pursue geriatric-friendly accreditation. This could involve physical environment modifications (dedicated quiet rooms, appropriate signage, and training security and ancillary staff in dementia communication) as well as process changes, such as allowing caregivers to be present as advocates during ED care (similar to paediatric care models that keep parents involved). Policy-makers could integrate the conversation-before-investigation principle into broader initiatives. For example, Australia’s National Dementia Action Plan and state health quality standards could explicitly recommend that acute care providers document goals-of-care discussions for patients with dementia. Additionally, the findings may support policies for better uptake of advance care planning. If ED data show that many patients arrive without documented wishes and then receive burdensome care, it bolsters the case for funding and promoting advance care directives in the community, as well as ensuring they are accessible to ED clinicians (for example, via electronic health record alerts). Another implication is in medicolegal guidelines – paradoxically, clinicians often cite legal fear as a reason for over-treatment, but clear policy statements and legal safe harbours for honouring patient preferences can empower clinicians to do less when appropriate. The research can contribute to guidance that prioritising what is right for the patient (even if that means not performing a test) is supported by ethical and legal standards, particularly when it follows from a documented conversation. Health services might also consider resource allocation changes: for example, embedding geriatric or palliative care specialists on-call for EDs, who can assist with these crucial conversations and decision-making for complex cases (thus institutionalising the conversations-first approach).

International Relevance: Although anchored in Australian healthcare, the core insights from this project will have global resonance. Many developed countries face aging demographics and have reported similar patterns of frequent ED use and potentially avoidable interventions in advanced dementia. The qualitative nuances gleaned – such as the emotional drivers of clinician behaviour or communication gaps – are likely to be present in EDs across Canada, the UK, and Japan. Thus, the proposed framework could serve as a model beyond Australia. It aligns with movements like Choosing Wisely (which globally urges reducing unnecessary tests and emphasises clinician-patient conversation) and with the push for more humane dementia care, as seen in the Alzheimer’s Association international recommendations. The research’s interdisciplinary flavour also sets a precedent: it shows the value of blending ethics and psychology into acute care improvement, which could inspire similar studies or interventions abroad. For instance, an emergency department in the US might pilot a “GERI-ED Talk” program drawing on this work, or European hospitals might incorporate Stoic resilience training in their clinician well-being programs to help staff cope with uncertainty. By disseminating the findings through academic publications and conferences in emergency medicine, gerontology, and bioethics, the work will contribute to a global conversation about the quality of acute care for individuals with cognitive impairment. It effectively advocates that the rest of the world’s EDs, too, should consider that sometimes the best care is not a new scan or blood test but a compassionate conversation.

In closing, this PhD will not only advance scholarly understanding of the nexus of emergency medicine and dementia care, but it will also offer practical, implementable solutions to improve care experiences and outcomes. It strives to demonstrate that by addressing fear with ethics and replacing reflex with reflection, we can transform emergency departments into spaces that better serve our elders with dignity and wisdom. The envisioned outcome is a virtuous cycle: improved patient and family satisfaction, reduced unnecessary investigations and hospital admissions, and enhanced clinician moral satisfaction, knowing that the care provided is truly in line with what matters to the patient. In a healthcare environment often dominated by technology and rapid action, this research reminds us of the healing power of conversation and human connection as the first “tool” in the clinician’s toolkit. By implementing conversations before investigations, emergency care for people with dementia can become safer, kinder, and more aligned with the ethical mandate to care, not just cure.

References

1.           Arvanitakis Z, Shah RC, Bennett DA. Diagnosis and Management of Dementia: A Review. JAMA [Internet]. 2019 Oct 10 [cited 2024 Jul 15];322(16):1589. Available from: /pmc/articles/PMC7462122/

2.           Harrison KL, Hunt LJ, Ritchie CS, Yaffe K. Dying with Dementia: Under-recognized and Stigmatized. J Am Geriatr Soc [Internet]. 2019 [cited 2024 Sep 20];67(8):1548. Available from: /pmc/articles/PMC6684346/

3.           Piers R, Albers G, Gilissen J, De Lepeleire J, Steyaert J, Van Mechelen W, et al. Advance care planning in dementia: Recommendations for healthcare professionals. BMC Palliat Care [Internet]. 2018 Jun 21 [cited 2024 Sep 15];17(1):1–17. Available from: https://bmcpalliatcare.biomedcentral.com/articles/10.1186/s12904-018-0332-2

4.           Adams G, Hoffman JW, Goldberg EM. It’s time for an overhaul of emergency department operations for people living with dementia. J Public Health Emerg [Internet]. 2024 Jun 25 [cited 2025 Jun 7];8(0). Available from: https://jphe.amegroups.org/article/view/9992/html

5.           Sampson EL, Leurent B, Blanchard MR, Jones L, King M. Survival of people with dementia after unplanned acute hospital admission: a prospective cohort study. Int J Geriatr Psychiatry [Internet]. 2013 Oct 1 [cited 2025 Jun 7];28(10):1015–22. Available from: /doi/pdf/10.1002/gps.3919

6.           Lamba S, Nagurka R, Zielinski A, Scott SR. Palliative care provision in the emergency department: Barriers reported by emergency physicians. J Palliat Med [Internet]. 2013 Feb 1 [cited 2025 Jun 7];16(2):143–7. Available from: https://www.liebertpub.com/doi/pdf/10.1089/jpm.2012.0402?download=true

7.           Argintaru N, Quinn KL, Chartier LB, Lee J, Hannam P, O’Connor E, et al. Perceived barriers and facilitators to goals of care discussions in the emergency department: A descriptive analysis of the views of emergency medicine physicians and residents. Canadian Journal of Emergency Medicine [Internet]. 2019 Mar 1 [cited 2025 Jun 7];21(2):211–8. Available from: https://pubmed.ncbi.nlm.nih.gov/29737962/

8.           Wu FM, Newman JM, Lasher A, Brody AA. Effects of initiating palliative care consultation in the emergency department on inpatient length of stay. J Palliat Med [Internet]. 2013 Nov 1 [cited 2025 Jun 7];16(11):1362–7. Available from: /doi/pdf/10.1089/jpm.2012.0352?download=true

9.           Grudzen CR, Emlet LL, Kuntz J, Shreves A, Zimny E, Gang M, et al. EM Talk: communication skills training for emergency medicine patients with serious illness. BMJ Support Palliat Care [Internet]. 2016 Jun 1 [cited 2025 Jun 7];6(2):219–24. Available from: https://spcare.bmj.com/content/6/2/219

10.        Cameron B, Laupacis A. Where do goals of care conversations belong? A case for the emergency department. Canadian Journal of Emergency Medicine [Internet]. 2019 Nov 1 [cited 2025 Jun 7];21(6):698–700. Available from: https://www.cambridge.org/core/journals/canadian-journal-of-emergency-medicine/article/where-do-goals-of-care-conversations-belong-a-case-for-the-emergency-department/943265E066DD7B63CCBF420C2B7DB761

11.        Popejoy LL, Vogelsmeier AA, Alexander GL, Galambos CM, Crecelius CA, Ge B, et al. Analyzing Hospital Transfers Using INTERACT Acute Care Transfer Tools: Lessons from MOQI. J Am Geriatr Soc. 2019;67(9):1953–9. 

12.        Kane RL, Huckfeldt P, Tappen R, Engstrom G, Rojido C, Newman D, et al. Effects of an Intervention to Reduce Hospitalizations From Nursing Homes: A Randomized Implementation Trial of the INTERACT Program. JAMA Intern Med. 2017;177(9):1257–64. 

13.        National Dementia Action Plan 2024–2034 | Australian Government Department of Health and Aged Care [Internet]. [cited 2025 Jun 7]. Available from: https://www.health.gov.au/our-work/national-dementia-action-plan