Misconceptions and do-not-resuscitate preferences of healthcare professionals commonly involved in cardiopulmonary resuscitations: A national survey

Graphical abstract


Introduction
2][3][4][5] In the 2021 United States of America Heart Disease and Stroke Statistics assessing data from 2019, mortality rates until hospital discharge average 89.5% for out-of-hospital cardiac arrests (OHCA) and 73.3% for inhospital cardiac arrests (IHCA). 6][6][7][8][9][10] At hospital discharge, around 10-15% of OHCA-and 10-12% of IHCA survivors suffer from severe disability, usually requiring continuous care. 4,5,72][13][14][15] The question of whether a patient wishes to be resuscitated or not is commonly discussed on hospital admission in a shared decision-making conversation between the patient and the treating physician, as the hyperacute nature of cardiac arrest makes ad-hoc discussions of resuscitation preferences impossible. 16This discussion should be performed upon hospital admission including an assessment of the patient's health status, information about the individual prognosis in case of a cardiac arrest, potential consequences of a successful resuscitation and take into account patients' advance directives. 16Patients' CPR preferences are then usually documented in their medical records, enabling their implementation even when patients are unconscious or otherwise incapable of communicating.This is crucial for patients without readily available next of kin, where physicians often act as surrogate decision-makers. 179][20] This is especially important in the emergency and critical care setting where end-of-life decisions, and cardiopulmonary resuscitations are prevalent. 21,22Hence, it is the aim of this study to assess the DNR preferences of critical care-, anesthesia-and emergency medicine practitioners, to identify factors influencing decision-making, and to assess current conceptions regarding CPR outcomes.The results are then compared to a representative sample of the general population.

Study population
A multicenter web-based survey was conducted in Switzerland among healthcare professionals involved in the care of OHCA and IHCA patients and compared to a representative sample of the general population.In Switzerland, the following healthcare professionals are predominantly involved in advanced cardiac life support and post-resuscitation care: -Paramedics and prehospital emergency physicians -Emergency nurses and emergency physicians -Intensive care nurses and intensive care physicians -Nurse anesthetists and anesthesiologists Healthcare professionals or trainees from the aforementioned professions were eligible to be surveyed.

Survey administration
The national Societies of the respective subspecialties were contacted and asked to participate in the survey.All national societies in question consented to participate in the survey and to distribute the survey link using their email communication channel, except the Swiss Society for Anaesthesiology and Perioperative Medicine, which rejected participation in the survey.To compensate for this matter, anesthesia departments of four large Swiss tertiary care centers (University Hospital Zurich, University Hospital Basel, Cantonal Hospital St. Gallen, and Cantonal Hospital Aarau) were asked to participate instead.Also, as the Swiss paramedics are only incompletely represented in their national society, six large emergency services participated in the survey.A list of all participating societies and institutions can be obtained from the online supplement (Supplement 1).The emails were only sent once without a reminder, and the number of emails sent was registered to calculate the response rate (Supplement 1).

Questionnaire development
The questionnaire was developed in accordance with the Best Practices for Survey Research 23 and Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines. 24he questionnaire development has been previously reported in detail 25 ; in brief, the multi-stage process involved members of the public (including a hospital pastor) according to the Patient and Public Involvement strategy, 26 senior critical care physicians, a critical care nurse, and a member of the ethical counsel.The final version of the survey can be obtained from the online supplement (Supplement 2).

Outcomes
The primary outcome was the rate of DNR Code Status vs. CPR Code Status when taking the perspective from a clinical case vignette of a 70-year-old patient suffering an OHCA with a no-flow time (time from collapse to start of CPR) 27 of 10 minutes (Box 1).The key secondary outcome was the respondents' personal DNR preferences for themselves, independent of the case vignette.

Baseline characteristics and factors associated with preference for or against cardiopulmonary resuscitation measures
The questionnaire included the following baseline characteristics: age, self-reported gender, language, profession, nationality, specialty and subspecialty, region, highest educational degree, religion, grade of religiousness, comorbidities, years of work experience since degree, living conditions, number of children, and type of emergency service.
Additionally, the following factors were registered: -Estimated survival rates with independence in activities of daily living after an OHCA or IHCA according to a cerebral performance category scale (CPC) of 1 or 2. 28,29 In accordance with the original publication, a CPC of 1 indicates "Good recovery -(. ..)Resumption of normal life even though there may be minor neurological and psychological deficits."And CPC 2 indicates "Moderate disability -Disabled but independent" in activities of daily living, such as the use of public transportation or doing groceries". 28The survival with independence in activities of daily living estimates were then compared with IHCA and OHCA

Statistical analysis
Baseline characteristics and outcomes of healthcare professionals were stratified according to the primary and secondary endpoint.
We also compared the health care professionals' results with the results of a representative of the Swiss general population recently published by our group. 25For comparison, we used a two-tailed Student's t-test. 25Logistic and linear regression analyses were used to evaluate associations of these factors with endpoints.Multivariable models were adjusted for age and self-reported gender.Finally, to compare the different professions (i.e., physicians, nurses and paramedics), we used an analysis of variance (ANOVA).A p-value of <0.05 (two-tailed) was considered statistically significant.The statistics software STATA 15.0 (Stata Corp., College Station, TX, USA) was used for all analyses.

Baseline characteristics
Of Further predictors for a DNR Code Status in the case vignette included DNR Code Status in personal CPR preference, possession of an advance directive, a shorter no-flow time after which resuscitation should not be attempted anymore, not wanting to be mechanically ventilated in case of severe illness and respiratory failure, no belief in an afterlife, no symptoms of anxiety, lower perceived quality of life, and having written an own advance directive (Table 1, Supplement 3).Further predictors for a personal DNR Code Status preference included overestimation of OHCA/IHCA survival rates, a shorter no-flow time after which resuscitation should not be attempted anymore, not wanting to be mechanically ventilated in case of severe illness and respiratory failure, not believing in an afterlife, not having children, having an advance directive, and having more professional experience (Supplement 4).

Interprofessional differences
Regarding the primary outcome, the rate of DNR orders among physicians, nurses, and paramedics was comparable.However, regarding the secondary outcome (personal code status preference), physicians more often chose a DNR order than nurses and paramedics (adjusted OR 0.51, 95% CI 0.39-0.67;p < 0.001 and adjusted OR 0.6, 95% CI 0.48-0.75;p < 0.001, for nurses and paramedics respectively, Supplement 5).Physicians and paramedics had the highest proportion of correct estimations regarding OHCA outcomes (Fig. 3a, Table 2).When looking at IHCA outcomes, physicians expressed the highest proportion of correct answers (Fig. 2a, Table 2).Also, physicians were less likely to refuse mechanical ventilation than nurses and paramedics (Table 2).

Differences between the Swiss general population and healthcare professionals
The Swiss general population cohort included 1044 subjects.The mean age of healthcare professionals was slightly lower than the mean age of the Swiss general population with no difference in gender distribution between the two populations.

Discussion
In this multicenter survey of 1803 healthcare professionals commonly involved in cardiopulmonary resuscitations, a preference for DNR Code Status was found in 85% when taking the perspective of a 70-year-old patient with a substantial no-flow time of 10 minutes.Among the different professions (physicians, nurses, and paramedics) the rate of DNR orders in the personal perspective of the case vignette was comparable.When making a general personal decision, more than half of the healthcare professionals preferred a DNR Code Status for themselves.Notably, physicians more often chose a DNR order for themselves when compared to nurses and paramedics.The proportion of DNR Code Status was significantly higher among healthcare professionals when compared to the general population.This was true for both the case vignette and when making a personal decision for themselves.One could hypothesize that this difference might result from the frequent direct confrontation of healthcare professionals with poor outcomes after cardiac arrest.This is in line with comparable research in the field.Previous findings suggest that there might be substantial discrepancies between what healthcare professionals assume to be a reasonable treatment for themselves and what is considered reasonable for their patients when making clinical decisions involving invasive and burdensome treatments. 33- 35For example, in a recent Australian survey including 747 doctors and 233 nurses approximately 25% of ICU practitioners indicated continuing aggressive treatment for a hypothetical patient.Still, they would refuse the same treatment for themselves. 36This might lead to a substantial ethical dilemma between own beliefs, expectations and patients' preferences.In particular when providing perceived futile treatments such ethical dilemmas might cause moral distress for intensive care practitioners, potentially resulting in symptoms of burnout or a change of profession. 21,37,38Also, clinicians should keep in mind that many of the functional states (e.g., bowel/bladder incontinence or confinement in bed) commonly observed after critical illness are considered worse than death by a significant number of patients. 39n accordance with previously published results from a representative sample of the Swiss general population, healthcare professionals commonly involved in CPR over-and underestimated the survival rate with independence in activities of daily living after cardiac arrest. 25Although the surveyed healthcare professionals had a high exposure towards CPR, they substantially overestimated the no-flow Fig. 2 -Percentage of correctly estimated, overestimated, and underestimated in-hospital cardiac arrest survival rates.A. Percentage of correctly estimated, overestimated, and underestimated in-hospital cardiac arrest survival rates with independence in activities of daily living (CPC 1-2) among healthcare professionals.Abbreviation: CPC Cerebral Performance Category Scale.B. Percentage of correctly estimated, overestimated, and underestimated inhospital cardiac arrest survival rates with independence in activities of daily living (CPC 1-2) given by the general population and healthcare professionals.Abbreviation: CPC Cerebral Performance Category Scale.
time, after which resuscitation should not be attempted anymore.Approximately one third of healthcare professionals set the no-flow time cut-off for not attempting CPR anymore at >10 minutes, although, it is known that a no-flow time of around 10 minutes is associated with a <2% chance of survival without neurological sequelae, depending on the low-flow time. 27Still, compared to the general population, healthcare professionals gave lower cardiac arrest survival estimates and more often estimated survival chances correctly.
Interestingly, whilst over-and underestimation of survival rates and refusal of mechanical ventilation were predictive for a DNR Code Status the majority of healthcare professionals did not want to receive invasive mechanical ventilation in case of severe illness and respiratory failure.Healthcare professionals should be well aware of these issues when counseling patients regarding DNR preferences and end-of-life decisions, as poor prognostic estimation, lack of communication skills, and physicians' attitudes toward death have been shown to interfere with modern end-of-life care. 40terestingly, although intensivists and critical care societies advocate completing an advance directive, only 32.4% of healthcare professionals in our survey possessed an own advance directive. 41otably, possessing an advance directive was predictive for a DNR Code Status, which might indicate previous personal engagement with this topic.
The present study has several implications for clinical practice, personal reflections, and future research.First, healthcare professionals should be aware of their prejudices, choices, and ethical values when supporting patients and families in end-of-life discussions, such as code status preferences.This might potentially influence their counseling and shared decision-making.Standardized communication tools might be supportive in such situations.Currently, a multicenter trial assessing a checklist-guided shared decision-making process performed by our research group has just completed recruitment (https://classic.clinicaltrials.gov/ct2/show/NCT03872154).
Second, healthcare professionals should be aware that a reasonable number of professionals wrongly estimated survival with Fig. 3 -Percentage of correctly estimated, overestimated, and underestimated out-of-hospital cardiac arrest survival rates.A. Percentage of correctly estimated, overestimated, and underestimated out-of-hospital cardiac arrest survival rates with independence in activities of daily living (CPC 1-2) among healthcare professionals.Abbreviation: CPC Cerebral Performance Category Scale.B. Percentage of correctly estimated, overestimated, and underestimated out-of-hospital cardiac arrest survival rates with independence in activities of daily living (CPC 1-2) among members of the general population and healthcare professionals.Abbreviation: CPC Cerebral Performance Category Scale.
independence in activities of daily living and overestimated the duration of a reasonable no-flow interval.Thus, we advocate that healthcare professionals commonly counseling patients regarding code status and deciding about termination of CPR are aware of realistic outcome data and time intervals.Third, we suggest that healthcare professionals commonly involved in cardiopulmonary resuscitations engage personally and in-depth with advance directives, as only a minority of the surveyed healthcare professionals possess an advance directive.Additionally, as shown in the present study, previous mental engagement with the topic might influence personal decision-making.

Strengths and limitations
The present study has several strengths: First, to the best of our knowledge, it is the largest of its kind looking at healthcare profes-sionals' DNR preferences and comparing them to the preferences of a representative sample of the general population.Second, the present survey was developed in a multi-level iterative process applying the concepts of public and patient involvement and multiexpert input.Also, validated tools for the assessment of anxiety and quality of life were used. 31,32Third, the study integrates healthcare professionals from different professions and multiple centers and societies, thus resulting in a high external validity.
However, this study also has limitations: First, as the study was performed exclusively in Switzerland, the results might not be extrapolated to different countries or to populations with other cultural backgrounds.Second, the present study's design is observational, and the results are thus rather hypothesis generating.Third, the response rate of 26.5% might limit the generalizability of our results due to a selection bias.

Conclusions
Swiss healthcare professionals have a significantly higher preference for a DNR Code Status compared to the general Swiss population in both a hypothetical clinical case vignette and when making a personal decision for themselves.The estimation of outcomes following cardiac arrest and personal living conditions are pivotal factors influencing code status preferences in healthcare professionals.Healthcare professionals should be aware of cardiac arrest prognosis and potential implications of personal preferences when engaging in code status-and end-of-life discussions with patients and their relatives.

Fig. 1 -
Fig. 1 -Cardiac arrest survival estimates compared to the actual rate.A. In-hospital cardiac arrest (IHCA) survival estimates compared to the actual rate.*Survival with independence in activities of daily living (CPC 1 or 2) according to Virani et al. (2021) 6 Abbreviations: CPC, Cerebral Performance Category Scale; IHCA, In-hospital cardiac arrest.B. Out-of-hospital cardiac arrest (OHCA) survival estimates compared to the actual rate.*Survival with independence in activities of daily living (CPC 1 or 2), according to Virani et al. (2021) 6 Abbreviations: CPC, Cerebral Performance Category Scale; OHCA, Out-of-hospital cardiac arrest.
1822 healthcare professionals participating in the web-based survey (26.5% response rate, Supplement 1) 1803 were included in the final analysis.Of the 1722 healthcare professionals providing 8%) as heads of department.Average professional experience was (mean [±SD]) 14.2 years [± 10.4]) and 67.7% of participants reported a CPR experience of 21-50 cases Supplement 3.Primary endpoint: Code Status preference in the personal perspective of the case vignette within healthcare professionalsRegarding the personal perspective of the case vignette, 85% (n = 1532) of the 1803 subjects preferred DNR Code Status.The key predictor for a

Table 1 -
Predictors for Code Status preference regarding the case vignette within healthcare professionals.