Article body

introduction

Recent medical practice, research, and technology advances have enhanced diagnostic and treatment capabilities. As a result, people are living longer, but at the cost of increased burden of chronic illness and reduced quality of life (1-3). Similarly, the experience of dying has changed recently: where death was once often the result of an acute event, it has increasingly become a chronic process. Many patients with chronic disease suffer recurrent complications of their disease, leading to progressive functional and cognitive decline (4). Consequently, more patients now die in institutional settings, such as hospitals or long-term care facilities, despite most expressing a preference to die at home. In Ontario, 2023 data showed that 55.3% of deaths occurred in hospitals, while 43.3% occurred in non-hospital settings, including long-term care homes, retirement homes, private residences, and other facilities (5). Given the complexity of their conditions and evolving prognoses, clear and ongoing communication by healthcare providers (HCP) is crucial (6-8).

Advance care planning (ACP) is a proactive and ongoing process where HCPs explore a capable patient’s values and healthcare priorities in light of their health condition (9). In contrast, goals of care (GOC) discussions build upon ACP by translating these wishes into more specific clinical decisions and may occur with either a capable patient or their designated substitute decision-maker (SDM) if the patient is incapable (10). A critical subset of these discussions involves documenting patients’ expressed capable wishes, which guide future healthcare decisions (1). Patients are encouraged to designate an SDM who can make healthcare decisions on their behalf in accordance with their most recently expressed wishes. Code status discussions — such as preferences around cardiopulmonary resuscitation (CPR) and intubation — are one concrete outcome of GOC discussions (10-12). These discussions may result in a decision for “full code” (chest compressions, defibrillation, intubation, and ventilation) or “do not resuscitate” (DNR), or something in between.

ACP is associated with increased patient satisfaction, enhanced patient autonomy and more efficient use of healthcare resources by reducing interventions that do not align with a patient’s expressed wishes. Despite these benefits, only 37-62% of oncology patients engaged in ACP conversations (13). Similarly, only 47.9% of hospitalized elderly patients at high risk of death within six months engage in ACP, leading to documented wishes. Of these patients, 76.3% had thought about death before hospitalization, but only 11.9% expressed a desire for life-prolonging measures. Furthermore, only 30% of documented wishes aligned with patients’ previously expressed desires (8,14,15).

On February 1, 2008, the Ontario Ministry of Health and Long-Term Care (MOHLTC) introduced the DNR Confirmation Form (DNR-CF; Appendix A) (16,17). This form was created to provide clear direction to first responders, specifically paramedics and firefighters, regarding whether to initiate resuscitative efforts in prehospital emergency settings. When a valid DNR-CF is present and signed by an authorized healthcare provider, paramedics and firefighters are instructed not to initiate basic or advanced CPR, including chest compressions, defibrillation, artificial ventilation, intubation, or the administration of resuscitation drugs (17,18). In such cases, paramedics and firefighters may provide comfort care measures, such as oxygen or medications for symptom relief, in accordance with their scope of practice (18). Before the implementation of the DNR-CF, first responders were obligated to perform CPR unless the patient met very narrow criteria for obvious death. As the interventions listed on the DNR-CF reflect the standard resuscitative actions carried out by paramedics and firefighters, signed DNR-CF serve as a directive not to initiate these procedures (18,19). The DNR-CF is bilingual and can be completed by a physician (MD), registered nurse (RN), registered nurse — extended class (RN EC), or registered practical nurse (RPN) (20).

In Canada, approaches to pre-hospital DNR documentation vary across jurisdictions. While Ontario uses a standalone DNR-CF for paramedics and firefighters, other provinces integrate resuscitation preferences within broader medical orders. For example, Alberta employs the Goals of Care Designation (GCD), a standardized medical order that categorizes treatment preferences into levels of care, rather than a binary DNR decision (21,22). British Columbia and Nova Scotia use Medical Orders for Scope of Treatment (MOST), which function similarly by incorporating resuscitation preferences into a patient’s overall care plan (23-25). Saskatchewan has developed the Saskatchewan Medical Orders for Scope of Treatment (SMOST), which also incorporates a patient’s goals and values into pre-hospital medical decision-making (25). These documents are generally accessible to paramedics and firefighters, ensuring that resuscitation efforts align with patients’ wishes in out-of-hospital settings.

A study of 96 patients presenting to the emergency department showed that only seven patients were aware of the DNR-CF, and only one patient had completed the DNR-CF. Barriers to completion included a lack of awareness and discomfort with end-of-life care discussions (16). It is assumed that most DNR-CFs are completed by primary care practitioners (PCPs). Although PCPs are essential in advance care planning, discussions of goals of care (GOC) occur frequently during inpatient admissions. Particularly for patients with chronic diseases requiring recurrent hospitalizations, the illness trajectory may lead to a different outcome from those discussed with PCPs. While many studies have explored the nuances of GOC discussions, limited literature exists on the effectiveness of the inpatient-to-outpatient translation of this communication (26-30). To date, no study has examined the DNR-CF completion rates by HCPs and the perceived barriers they face in completing these forms. Here, we aimed to identify the awareness of, rates, and perceived barriers to completing the DNR-CFs among inpatient general internists.

Methods

Setting and context

This study was conducted across two adult acute care hospitals — Hamilton General Hospital (HGH) and Juravinski Hospital (JH) — within the Hamilton Health Sciences (HHS) network in Hamilton, Ontario, Canada. These hospitals serve the population in the south-central region of Ontario and act as a major referral centre for other areas in Ontario. The JH is adjacent to the Juravinski Cancer Centre, which provides comprehensive cancer care to patients in Hamilton.

Study Population

The participants included general internists at HGH and JH in either inpatient or mixed inpatient-outpatient settings.

Ethical Consideration

This study was designed as a baseline evaluation to inform a larger quality improvement initiative. As such, our institutional research ethics board (REB) waived a full review. All participant data were anonymized, interview transcripts were de-identified, and interviewers were assigned numerical codes.

Study Design and Outcomes

A mixed-methods design was employed, comprising an online survey followed by a qualitative focus group. The primary quantitative outcomes were: 1) the proportion of general internists who had previously completed a DNR-CF, and 2) the frequency and types of healthcare provider-related and patient-related barriers identified in the survey. The qualitative outcomes included themes related to physicians’ perceptions of the DNR-CF’s utility, barriers to its completion in inpatient settings, and suggestions for improving uptake among hospital-based providers.

Survey Phase

Online surveys were distributed to 25 general internists. This survey collected demographic information (e.g., age, sex, years in practice, and practice location), along with participants’ awareness of DNR-CF, frequency of form completion, and perceived barriers to its use (Appendix B).

Focus Group Phase

A focus group was conducted with three general internists who expressed interest following the completion of the survey. A qualitative descriptive approach was used to explore participants’ experiences and perspectives regarding the DNR-CF. A semi-structured interview guide (Appendix C), informed by relevant literature and our objectives, was developed and pilot-tested internally for robustness. The 60-minute session was conducted via Zoom and moderated by a member of the research team trained in qualitative methods. The session was audio-recorded, transcribed (Appendix D) and de-identified by the focus group facilitator (Appendix E). The audio recording and the transcription were securely saved on a password-encrypted hospital computer, and audio recordings were deleted following transcription.

Quantitative Analysis

Descriptive statistics from the survey were analyzed using Google Forms with the Free Spanning Stats. Results were summarized as proportions and frequencies.

Qualitative Analysis

Braun and Clarke’s six-step framework was used to guide the thematic analysis (31). Two independent researchers (MJ and JH) read through the transcript several times to become familiar with the content, then worked separately to develop initial descriptive codes. They met regularly to compare their interpretations, refining the codes through discussion and reaching consensus along the way. Codes were then grouped into overarching themes and subthemes. Investigator triangulation, through the use of two coders, enhanced the credibility of the analysis. Discrepancies were resolved by consensus, and a finalized coding framework was applied consistently across the transcript.

Ensuring Trustworthiness

Several strategies were used to ensure that the findings were credible, dependable, confirmable, and transferable. Credibility was established through independent coding by two independent assessors, with discrepancies addressed through discussion to reach consensus. An audit trail was maintained throughout the process to keep track of how codes and themes evolved, which helped ensure dependability. To support confirmability, direct participant quotes were included in the Results section to demonstrate that interpretations were grounded in the data. Finally, detailed information was provided regarding the study setting and participant characteristics allowing readers to assess how well our findings apply to other clinical scenarios.

Results

Survey

A total of 14 general internists completed the surveys. Most participants (71%) practiced both inpatient and outpatient, whereas 29% exclusively practiced inpatient (Table 1). Although all responders were familiar with DNR-CFs, only 16.7% completed a DNR-CF on discharge. None of the participants filled out a DNR-CF during the month preceding the survey. The most frequently cited HCP-related barriers were the lack of awareness and availability of the forms (Figure 1). In contrast, the most common patient-related barriers, as perceived by physicians, included lack of access to SDM and language barriers (Figure 2). As patients were not study participants, these barriers reflect physician perceptions rather than direct patient-reported experiences.

Table 1

Demographic information

Demographic information

-> See the list of tables

Figure 1

HCP-related barriers to completing DNR-CF identified in the survey

HCP-related barriers to completing DNR-CF identified in the survey

-> See the list of figures

Figure 2

Patient-related barriers to completing DNR-CF identified in survey as perceived by physicians

Patient-related barriers to completing DNR-CF identified in survey as perceived by physicians

-> See the list of figures

Focus Group

Three general internists participated in the focus group. Four main themes were identified from the discussion: general, benefits, barriers, and suggestions.

General

All participants assumed these forms were completed solely in the community setting, usually triggered by a new nursing home or hospice admission. None of the participants have considered revisiting GOC conversations on discharge to enable the completion of DNR-CFs. They acknowledged the infrequent completion of the forms in the inpatient setting, with one participant saying: “…I have never seen anyone filling out this form in a hospital” and “I thought the forms needed to be completed in the community.”

Benefits

The General Internists agreed on the benefits of the DNR-CF when caring for an incapable patient whose SDMs are unavailable. However, it is only helpful when the patient does not want any form of resuscitation. It becomes challenging when patients may wish for some, but not all, of the treatments outlined on the form.

Barriers

We identified three key themes for HCP-related barriers to filling out DNR-CF forms:

  1. Knowledge gap: Physicians in the focus group stated they “did not think of it.” In addition, focus group physicians felt hesitant to complete the forms as they did not understand the medico-legal implications of completing a DNR-CF.

  2. Accessibility: physicians stated that “forms are not readily accessible,” and they were not sure if it was kept on wards.

  3. Outside the scope of inpatient physician: a physician in the focus group commented that completing DNR-CFs after GOC planning was “not my responsibility.”

In addition, physicians did not feel that the forms were reliable and that there was inherent redundancy in filling out the forms. The key themes we identified as patient-related barriers to having DNR-CF completed included:

  1. Systemic barriers: this was most commonly identified as “language barriers.”

  2. Lack of social support: physicians mentioned that patients who were unable to make GOC decisions often had “no available SDM,” posing a medico-legal challenge with completing the DNR-CFs.

The focus group participants reiterated some of the barriers addressed in the survey, such as lack of awareness of using these forms in the inpatient setting, form availability, limited time, and insufficient remuneration.

One participant stated:

“I try to find out how to get access to this form, and I couldn’t. You cannot just download it. You have to order it because it has a serial number. You can’t just give your CPSO number. You have to have an official letterhead or whatever to order this form, so I think it is really complicated.”

The discussion also focused on the role of inpatient internists in addressing GOC conversations and completing the DNR-CFs. The participants discussed that for those patients requiring recurrent admissions with declining health status, inpatient general internists are in an excellent position to complete the DNR-CFs. Internists may offer fresh perspectives on patients with recurrent admissions, particularly when specialists may be too focused on disease treatment or cure. However, for patients with complex comorbidities, participants felt that relevant specialists best addressed GOC discussions. Similarly, some argued that PCPs are better positioned to have these conversations, given their longitudinal relationship with patients. Additionally, one physician did not believe that the inpatient environment was the right time to fill out the forms.

“The patient is not in a stable point of health where the discussion should be best held. If these discussions were to be held, it would be after they were stable.”

Additionally, they expressed concerns about the redundancy created by completing the DNR-CF, particularly if the code status is already documented on the electronic medical record (EMR). Additionally, nursing homes and hospices are already required to have them filled out.

Some questioned the validity of the DNR-CFs. Providers feared that patients may face harm if the forms did not reflect current wishes. Code statuses are fluid, and reassessment with the patient and their SDMs must be possible. Furthermore, initial GOC conversations may not be comprehensive enough (i.e., providers did not address all aspects of resuscitation in their code status discussions with patients), making it hard to trust the authenticity of these forms.

“When I have that form, I don’t trust the form to be comprehensive enough to provide the care that is required to the patient at the time.”

Finally, questions regarding the legally binding status of the forms were raised and considered a barrier to their completion. Table 2 provides a summary of the themes identified in the focus group in relation to the barriers.

Table 2

Focus group subthemes in relation to barriers in completing the DNR-CF

Focus group subthemes in relation to barriers in completing the DNR-CF

-> See the list of tables

Suggestions

The participants also suggested strategies to improve the completion rates of the forms as an inpatient. These included dedicating more resources and financial incentives. One physician suggested introducing these forms in palliative care workshops to raise awareness. Lastly, participants believe this form is not helpful for every admitted patient. They suggested the publication of a DNR-CF policy that identifies a specific demographic or set of patient criteria for whom these forms could be most beneficial (e.g., a patient suffering from the mid to end stages of their chronic illness rather than a patient admitted for a self-limited diagnosis).

Discussion

Overview of Key Findings

Our results demonstrate the low participation rate of inpatient HCPs in completing DNR-CFs at HHS and highlight specific barriers to their completion. These barriers span various domains, including personal and professional issues for HCPs, challenges with patients and SDMs, institutional policies, environmental factors, and the EMR limitations. Existing literature shows low awareness and completion rates of DNR-CFs among patients (16). Our study is unique in examining these trends from the perspective of inpatient HCPs. Combined with existing literature, our results confirm low awareness and completion rates of DNR-CFs among both providers and patients.

Clarifying Roles and Responsibilities

An important theme from the focus group was the uncertainty around who is responsible for completing the DNR-CF. While PCPs are generally assumed to complete the form due to their long-term relationships with patients, participants argued that subspecialists and inpatient providers also have important roles, especially when managing complex, chronically ill patients. Although GOC discussions often occur in outpatient settings, these conversations should also routinely happen during hospital admissions, especially as more patients face recurrent admissions due to chronic disease complications. In these situations, internists are uniquely positioned to revisit GOC discussions during hospitalization, as admissions could indicate a change in illness trajectory. Close collaboration between general internists, PCPs, and subspecialists is paramount to ensure these forms are completed at optimal times and with adequate context.

Misconceptions About Redundancy

Some focus group participants questioned the need to complete DNR-CFs in inpatient settings where code status is already documented in the EMR. However, it is important to reinforce that these forms are primarily intended to guide first responders during community emergencies. To avoid confusion, the DNR-CF should be limited to a binary directive for or against CPR and intubation, rather than including broader ACP preferences, such as ICU admission and non-invasive ventilation.

Timing and Validity of the DNR-CF

Concerns were raised about the timing of DNR-CFs and whether these forms accurately reflect current patient wishes. Participants emphasized that comprehensive serious illness conversations must proceed completion of the form, ensuring that patients and their SDM are fully informed. In an inpatient context, the optimal time to complete a DNR-CF may be after a serious illness conversation between the provider, the patient, and their family, ensuring they fully understand their prognosis and disease trajectory. Internists can play a role in reconfirming code status at discharge and facilitating the appropriate completion of the DNR-CF.

Focus group participants also questioned the legal standing of the DNR-CF. While it is binding for paramedics and firefighters under Ontario’s Basic Life Support Patient Care Standards and their standard operating procedure, other HCPs are bound by the Health Care Consent Act (HCCA) (32,33). Although section 5 of the HCCA describes that patient wishes can be expressed in any written or oral form, it also describes that “Later wishes expressed while capable prevail over earlier wishes” (34). The Ontario Ministry of Health and Long-Term Care states that HCPs who follow the DNR-CF “do so at their own risk” (32). Therefore, while the DNR-CF can provide guidance, HCPs should ideally verify current wishes with the SDM.

Addressing Barriers to GOC and DNR-CF Completion

Barriers to GOC discussions are well-documented in the literature, including patient’s difficulty accepting poor prognosis, misconception about life-sustaining treatments and family disagreements (35-37). Clinician-related barriers such as time constraints, lack of long-term relationship with patients, inability to reach the SDM, insufficient institutional support, and inadequate training are also common. An important systemic-related factor is absence of an appropriate confidential space for discussions (37-39). Addressing these barriers is crucial to enhancing DNR-CF completion rates.

Educational Opportunities

Our focus group also revealed a clear need for improved education around the DNR-CF and the processes related to its completion. Participants expressed uncertainty about when and how to complete the form, its legal implications, and where to access it. These insights suggest that an educational intervention could be a valuable next step. For example, a brief targeted education session offered shortly after focus group participation — when the topic is top of mind — may be a practical and timely approach to improving awareness and confidence. As part of future quality improvement (QI) efforts, we plan to integrate such an intervention into subsequent phases of this project. This may serve as a foundational step toward shifting the culture surrounding end-of-life planning and enhancing the appropriate use of DNR-CFs in inpatient settings.

Strategies to Improve Completion Rates

Low DNR-CF completion despite participation in GOC education suggests a need for multi-pronged strategies. These may include improving form accessibility, empowering allied HCPs to complete them, defining patient criteria for form eligibility, and embedding them in discharge checklists. Completing the form upon admission and providing it at discharge could reduce redundancy and increase uptake. Financial incentives may also help increase completion rates. Importantly, increasing the frequency and quality of ACP discussions is a foundational step toward improving DNR-CF completion. The form should represent the final stage of a broader, meaningful ACP process, in which capable patients are given the opportunity to reflect on and communicate their preferences. If ACP conversations are initiated early and routinely integrated into both inpatient and outpatient care, DNR-CF completion may follow naturally. A holistic, system-level approach that prioritizes proactive, and structured ACP discussions will help ensure that patient preferences are explored, clearly documented, and effectively guide future care.

Ethical Considerations

There are several ethical implications associated with DNR-CF completion. First, while these forms are crucial for respecting patient autonomy, it is essential to recognize that patients may not fully understand or be certain of their wishes regarding DNR status until they are confronted with the reality of their condition, such as impending death. Additionally, an existing DNR form that is not updated may no longer reflect the patient’s current wishes, undermining autonomy. This highlights the need for ongoing opportunities to revisit code status discussions and update DNR-CFs in inpatient and outpatient settings to reflect the patient’s current wishes accurately.

Second, CPR and intubation may carry significant risks, including physical injuries, like rib fractures, and potential psychological harm from prolonged suffering or diminished quality of life (15). Administering aggressive interventions to patients whose preferences have not been clearly established can lead to distress for both patients and their families, compromise dignity at the end of life, and contribute to moral distress among healthcare providers who are uncertain whether the care is wanted (40,41). Providing treatments that conflict with a patient’s values can result in unnecessary suffering and erode trust in the healthcare system (42). In addition, such interventions can lead to inefficient use of healthcare resources (43). It is thus essential that HCPs conduct frequent code status discussions to clarify these risks so that patients and caregivers develop a deeper understanding of the implications of their decisions. By completing DNR-CFs through comprehensive discussions, healthcare providers can help prevent unwanted harms, uphold patient-centred care and enhance informed consent.

Third, PCPs offer continuity of care, often have deeper rapport, and are more familiar with their patients, which provides an opportunity for better-informed consent regarding DNR status. However, the drawback is that PCPs may not be available when patients are acutely ill and facing potential mortality, a time when decisions may differ. On the other hand, hospital physicians have specialized knowledge and access to the immediate clinical context, supported by interdisciplinary teams and consulting specialists like palliative care teams. Yet, they face time constraints and often lack the long-term relationships that PCPs have with patients. Finally, providing care that does not align with a patient’s values and preferences is not only ethically concerning but also represents an inefficient allocation of healthcare resources (42,43).

Limitations

This study has several limitations. First, the small number of focus group participants limited thematic saturation. While individual instead of group interviews may have offered a broader range of perspectives, the focus group approach encouraged consensus-building through interactive discussion, which is valuable in quality improvement projects. Additionally, group settings can shape the discussions, as they may encourage consensus and limit expression of opposing and diverse views. Nonetheless, given that our study was a QI initiative aimed at identifying key barriers, achieving full thematic saturation was not a priority, and our findings still provide important insights. Future studies with larger sample size and individual interviews can improve the generalizability of our results.

Additionally, interpretation of survey responses may have varied across participants. For instance, some respondents who selected “did not think of it” as a barrier may still have considered or completed a DNR-CF in other contexts or may have selected additional barriers based on hypothetical scenarios rather than direct experience. There may also have been variation in how respondents interpreted “did not think of it” — for example, forgetting in the moment versus never having considered the form at all. The potential for varied interpretations of questions is a common limitation in self-reported survey data and reinforces the importance of careful interpretation of these responses.

Although the DNR-CF can be completed by various healthcare professionals, including registered nurses (RNs), nurse practitioners (RN-ECs), and registered practical nurses (RPNs). Our study focused predominantly on general internists as our primary interest was to explore the unique barriers faced in DNR-CF completion within inpatient settings where general internists often lead GOC discussions. However, this narrower approach to selecting HCPs may have limited the perspectives captured and overlooked barriers experienced by other providers who are also authorized to complete the forms. Future studies should include a larger range of HCPs to gain a broader understanding of the facilitators and barriers related to DNR-CF completion.

Conclusion and Future Directions

To address the issues identified in this study, it is essential to provide more opportunities for GOC discussions in both inpatient and outpatient settings, supported by reminders or strategies to prompt these conversations. Our findings lay the groundwork for future quality improvement studies aimed at implementing appropriate interventions to increase DNR-CF completion rates by inpatient HCPs. Future studies should explore DNR-CF completion across a broader range of inpatient settings, such as oncology and palliative care, and evaluate the role of allied healthcare professionals, in supporting form completion. Integrating structured ACP discussions into routine care — alongside improved provider education, clearer delineation of responsibilities, and institutional support — may help overcome identified barriers and ultimately improve the consistency and quality of end-of-life planning across the continuum of care.