The Complex Dynamics of Decision-Making at the End of Life in the Intensive Care Unit: A Systematic Review of Stakeholders' Views and Influential Factors

A lack of consensus resulting in severe conflicts is often observed between the stakeholders regarding their respective roles in end-of-life (EOL) decision-making in the ICU. Since the burden of these decisions lies upon the individuals, their opinions must be known by medical, judicial, legislative, and governmental authorities. Part of the solution to the issues that arise would be to examine and understand the views of the people in different societies. Hence, in this systematic review, we assessed the attitudes of the physicians, nurses, families, and the general public toward who should be involved in decision-making and influencing factors. Toward this, we searched three electronic databases, i.e., PubMed, CINAHL (Cumulative Index to Nursing & Allied Health), and Embase. A matrix was developed, discussed, accepted, and used for data extraction by two independent investigators. Study quality was evaluated using the Newcastle-Ottawa Scale. Data were extracted by one researcher and double-checked by a second one, and any discrepancies were discussed with a third researcher. The data were analyzed descriptively and synthesized according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Thirty-three studies met our inclusion criteria. Most involved healthcare professionals and reported geographic variations in different timeframes. While paternalistic features have been observed, physicians overall showed an inclination toward collaborative decision-making. Correspondingly, the nursing staff, families, and the public are aligned toward patient and relatives’ participation, with nurses expressing their own involvement as well. Six categories of influencing factors were identified, with high-impact factors, including demographics, fear of litigation, and regulation-related ones. Findings delineate three key points. Firstly, overall stakeholders’ perspectives toward EOL decision-making in the ICU seem to be leaning toward a more collaborative decision-making direction. Secondly, to reduce conflicts and reach a consensus, multifaceted efforts are needed by both healthcare professionals and governmental/regulatory authorities. Finally, due to the multifactorial complexity of the subject, directly related to demographic and regulatory factors, these efforts should be more extensively sought at a regional level.


Introduction And Background
Globally, a significant number (16-23%) of intensive care unit (ICU) patients die during their hospital stay, often following a decision to withhold or withdraw life-sustaining treatment [1,2].The aim of these end-oflife (EOL) practices should primarily be to relieve the patient's suffering when aggressive treatments are deemed futile and death comes as an unavoidable outcome of the disease [3].However, there is often a lack of agreement among physicians, patients, and families over their respective roles in EOL decision-making [4].The involvement of nurses has also been a point of interest in various studies [5][6][7][8][9][10][11][12][13][14][15][16][17], while the views of the general public have not been comprehensively examined [18].
EOL issues are one of the primary causes of conflicts originating mostly from the lack of knowledge of patients' wishes [19], comprehension of medical information [20], and discrepancies between the views, values, and beliefs of the stakeholders [21], leading to misperceptions and eventually communication breakdown [19].
Understanding stakeholders' attitudes regarding aspects of EOL and the resulting dynamics can contribute to a viable solution by promoting collaborative decision-making, family meetings, and tailored communication [22,23].
In turn, knowledge of social perspectives can be a useful tool to calibrate communicational and legislative strategies toward high-quality EOL care.This can be achieved by endorsing a cultural movement toward shared decision-making (SDM), especially in light of the recent pandemic events [24].Legislative and regulatory activity can also provide a solid basis by defining fundamental concepts [14], settling procedures

Search strategy and selection criteria
We systematically searched PubMed, CINAHL (Cumulative Index to Nursing & Allied Health), and Embase to identify studies exploring our research subject quantitatively.A broad search strategy was used by combining key terms and MeSH terms related to EOL decisions.No restrictions on demographics or research timeframe were applied.In addition, reference lists of any relevant research studies identified during the screening process were screened for additional relevant articles.The full search terms used for the literature search are noted in Appendix Table A1.

Study selection and eligibility criteria
Two investigators (SG and ED) independently searched titles/abstracts of retrieved references for eligibility, and when a consensus could not be reached, a third author (VK) was consulted.The same procedure was followed at each stage of the search, screening, and selection process.We included all types of studies addressing our population (P), exposure (E), outcome (O), and setting (s) framework (Appendix Table A2).
Studies were considered eligible only if they reflected the attitudes of physicians, nurses, family members, and the general public regarding who should be involved in EOL decision-making for critically ill adult patients (≥18 years of age) admitted in the ICU, reported original research data, and were available in English, full text.We considered EOL decisions as medical decisions to reduce suffering when treatments fail and the patient's quality of life cannot be maintained while recognizing that the possibility of death is an inevitable outcome of the disease's progression [3,36].
We excluded studies involving children or neonates and non-critical/intensive care settings.We also excluded studies that reported opinions about organ donation and transplantation decisions, euthanasia, or physician-assisted suicide.Organ donation decisions do not have a direct impact on the occurrence of death, while organ transplantation is a life-saving procedure considered a treatment [37].Euthanasia and physician-assisted suicide should not be confused with EOL practices, as they involve the doctor intentionally aiming for the patient's death, either actively or passively [3,38].
We also categorized the opinions expressed in the eligible studies based on the decision-making models they seemed to reflect.We adopted a broad interpretation of SDM comprising any form of collaborative decisionmaking where both experts and non-experts discuss or decide together [39].
Studies addressing practices, mere facts, or attitudes about who was actually involved in EOL decisions, e.g., ETHICUS studies [1,[26][27][28][29]40], were excluded.We also omitted studies examining factors that we considered not to be directly related to opinions about who should participate in these decisions.To limit assumptions and logical leaps, we excluded studies whose data did not clearly capture participants' opinions about who should be included, studies involving children or neonates and non-critical/intensive care settings, and those that reported opinions about organ donation and transplantation decisions or euthanasia or physician-assisted suicide.
Apart from the aforementioned, letters to the editors, opinions, qualitative studies, open-ended surveys, narrative analysis, focus groups, grounded theory, phenomenological and hermeneutic designs, and secondary research studies, such as systematic reviews and overviews of reviews, were excluded.

Study quality assessment
The quality of the included studies was assessed by two independent investigators (SG and ED) using the Newcastle-Ottawa Scale (NOS) [41] indicated in Appendix Table A3.

Data extraction
Data from all full-text papers reviewed were extracted into a predetermined Excel file (Microsoft

Methodological quality of the included studies
The overall summary of the assessment of the included studies according to NOS is described in Appendix Table A3.Overall, 19 studies were assessed as good/high quality and 14 studies were considered of fair quality.

Paternalistic Attitudes in National Studies
Findings from Portugal [52], Spain [7], Greece [6], France [19], USA [56], and Italy [54] highlight a paternalistic pattern on behalf of the physicians.Interestingly, Lomero-Martínez et al. [7] in Spain identified physicians supporting nurses' rather than relatives' involvement, while in Greece, Ntantana et al. [6] observed medical paternalism manifesting in a lack of information provided to the family.However, physicians advocated for nurse involvement.Notably, in the recent study by Giabicani et al. [19] in France, family presence was not thought to reduce potential conflicts, while in Portugal [52], physicians were found hesitant toward the involvement of any other party in decision-making.

Mixed Attitudes in National Studies
Our findings also highlight some mixed attitudes (Tables 1, 2).As reported by Weng et al. [46] in China, despite physicians being in favor of SDM, most of them opted to adjust medical information per occasion.They were also not keen on nurse involvement.Sjökvist et al. [18] revealed Swedish physicians' preference to collaborate in a competent patient scenario.Reversely, in the case of an incompetent patient, most of them thought themselves to be the sole decision-makers.In Turkey, Baykara et al. [43] reported physicians advocating for patient/legal representative's involvement, but not for families, while Giannini et al. [54] in Italy found that more than half (58.4%) of the respondents would often/always involve families, but not the patients' themselves and the nursing staff.Physicians in Croatia, as observed by Špoljar et al. [11], claimed that patient's autonomy should be respected and advocated for nurses' involvement.However, only 26.3% of them supported collaboration with the family.

Physicians' Attitudes in Multinational Studies
As shown in Table 3, multinational studies conducted in Australia and New Zealand by Tracy et al. [12] and in Europe and the USA by Metaxa and Lavrentieva [45], report physicians' inclination toward patients and families.Nurse involvement was supported by more than half of the physicians in the first study and by less than half (46%) in the second one.In the WELPICUS study [14], more than 80% (consensus) of the experts supported that SDM should be pursued, including patients/surrogates/families and nurses, while more than ¾ of the participants from every region agreed to this [13].The highest agreement rate was observed in Asia and North America and the lowest in Eastern Europe [13].
Both Vincent's studies [48,49] highlighted the fact that European physicians' perspectives differed by country.His first study [48] in the early 90s indicated that most physicians (66%) supported discussing donot-resuscitate (DNR) orders with the patient's family, but not with the patients themselves (32%).Although no uniformity could be observed, a follow-up study by the same researcher [49] revealed that physicians' attitudes evolved positively toward patient consultation.However, discussion with the patients was supported by less than half of the respondents in most countries.
In their multinational study in Asia, Park et al. [57] identified that most Japanese physicians supported SDM.Interestingly, Korean and Chinese respondents would involve families but not patients (5.9% and 5.1%, respectively).
Finally, almost half (46.5%) of the Middle Eastern participants in the relevant study [42] wanted the ultimate decisional authority regarding DNR orders.

Families' attitudes toward EOL decision-making involvement
Four of the included studies, three national [17,42,55] and one multinational [44], examined families' attitudes.In brief, findings in Brazil [42] and Canada [55] revealed that, overall, families prefer to have an active role in the decision-making process (Table 2).Similarly, Gerritsen et al. [44] found that most family members in Denmark and the Netherlands also favored SDM but not without physician input.Contrarily, less than half (47%) of the French participants supported their own involvement, as observed by Azoulay et al. [17] (Tables 2, 6).

General public's attitudes toward EOL decision-making involvement
Three national studies [18,50,51] assessed the attitudes of the general public (Table 1).Overall, participants from the French [50], Israeli [51], and Swedish [18] public supported family involvement in decision-making (Tables 2, 7).Notably, in a competent patient scenario, half of the respondents (50%) from the Swedish public wanted physicians to be excluded from the decision-making process [18].

Fear of Litigation
In Greece [6], France [5], and China [46], fear of litigation was found to influence healthcare professionals' attitudes toward EOL decision-making while also having a direct impact on the quality of the information provided by them to the patients and their relatives, especially in France and Greece.As reported by Westphal and Mckee [16] some physicians in the USA responded that family wishes should be followed even against patient's wishes due to fear of litigation.In addition, in Italy, Giannini et al. [54] reported that 52% of physicians' paternalistic attitudes were influenced by fear of legal consequences.
In their multinational study [57] in Asia, Park et al. identify that physicians from China, Korea, and Japan feel exposed to personal legal risk.This was observed mostly in China (49.7%) and Korea (44.1%)where most participants' attitudes endorsed paternalistic features.In the WELPICUS study [13], Sprung et al. stated that fear of litigation may be related to a lack of consensus about withholding or withdrawing agreement without consent in any case, especially in Asia and Eastern Europe.

Existence/Knowledge of the Existence or Absence of Relevant Legal Framework, Guidelines, and Protocols
In Sweden [18], physicians' paternalistic attitudes (incompetent patient scenario) appear to be correlated to relevant guidelines that were emphasizing on their own decisional authority.On the contrary, according to Baykara et al. [43], up until 2020, the absence of relevant legislation in Turkey placed physicians in difficult situations, obviously influencing their opinions on the subject.This is also supported by the findings of Ntantana et al. [6] in Greece.In China, Weng et al. [46] observed that physicians were generally unaware of the existence of relevant legislation, which was also the case in Spain [7], according to Lomero-Martínez M. et al., for both physicians and nurses, regarding their knowledge of the existence of relevant EOL protocols.Park et al. [57] in their multinational study in Asia observed that most Korean respondents preferred legislative measures to provide guidance regarding EOL issues, while in the WELPICUS study [13], Sprung et al. stated that legal reasons and lack of legal framework may be related to lack of consensus about withholding or withdrawing agreement without consent in any case.

Patients' State of Competence
In Sweden [18] and Brazil [42], differences were observed among the opinions of the physicians, nurses, patients' families, and the general public depending on the scenario presented to them (competent vs. incompetent patient).Most worth mentioning disparities were manifested in the attitudes of Swedish physicians who preferred to be the sole decision-makers in case of an incompetent patient but to collaboratively make such decisions in case of a competent one.At the same time, half of the Swedish public excluded physicians' participation when the patient is competent but advocated for physicians' involvement in a reverse case.

Perspectives by Respondents' Status
Lomero-Martínez et al. [7] identified that Spanish nurses were more supportive of collaborative decisionmaking than physicians.Four other studies (in France [5], Greece [6], Sweden [18], and Brazil [42]) also highlighted disparities regarding attitudes toward EOL decision-making between the examined population categories.A recent study by Špoljar et al. [11] in Croatia highlighted disparities among the perceptions of physicians and nurses toward family involvement with more than half of the latter being in favor of family involvement, a view supported by only 26.3% of the participant physicians.However, physicians were more inclined to respect patients' wishes than nurses with high school education.Interestingly, another study in France by Azoulay et al. [17] highlighted discrepancies between the opinions of the clinicians and the families.Hence, despite physicians and nurses being in favor of family members' involvement, this was supported by less than half of the latter.

Patient/Family-Related Factors
In five studies [5,6,19,45,46], patient and family-related attitudes, such as limited understanding of medical details, family involvement would not reduce the risk of conflict, family members do not always decide with the patient's wishes as an axis, concerns about adding to family's distress, patient's clinical condition and prognosis, the recipient's educational level, and the relatives' understanding and expectations, were cited as reasons for not disclosing full clinical information.As observed by Azoulay et al. [17], despite less than half of the French family members being in favor of their participation in EOL decision-making processes, those who did, claimed that this should be the case because they considered themselves as best knowledgeable toward patient's wishes, their opinion was important, and that this would help the ICU staff.Contrarily, those who did not express a desire for participation believed that their presence would not be useful.

Country of Practice
In four multinational studies [13,44,49,57], a country-related lack of uniformity was observed between the participants.

Country of Training
In the Middle East [59], western-trained physicians were more likely to override family and patient decisions about DNR orders.

Culture
In two multinational studies [13,57], culture was observed as an influencing factor.Sprung et al. [13] propose that lack of consensus toward treatment limitation without consent under certain circumstances may be influenced by cultural differences.In Asia, as observed by Park et al. [57], Japanese (29.5%),Chinese (93.8%), and Korean (74.2%) respondents highlighted that not discussing death with a critically ill patient was a part of their culture.

Medical Specialty
In Turkey, Baykara et al. [43] found that compared to physicians whose primary medical specialty was anesthesiology, more internal medicine physicians advocated for family involvement in EOL decisions.Additionally, Meltzer et al. [56] in the USA identified that more physicians (71.4%) in pulmonary/critical care supported that surrogate consent should not be required to discontinue venoarterial extracorporeal membrane oxygenation (VA-ECMO), compared to their colleagues with other medical specialties (general internists (52.8%), cardiologists (52.4%), and others (15.8%)).In the same direction, Forte et al. [53] noted that ICU physicians were more supportive of nurse involvement.In the WELPICUS study [13], it was observed that anesthesiologists were more likely to agree that withholding or withdrawing life-sustaining treatments without the consent of the patient or surrogate should be permissible under certain circumstances.

Sex and Age
Vincent [49] observed that female European physicians in the late 90s discussed DNR orders with patients less frequently than their male colleagues (17% vs. 28%).The same researcher, almost a decade earlier [48], also found that older physicians (>40 years) were more likely to be collaborative in EOL decision-making, which was also confirmed by the work of Baykara et al. [43] in Turkey.Younger age was also associated with attitudes advocating for nurse involvement among physicians in Brazil, as described by Forte et al. [53].In addition, Bodas et al. [51] reported that Israeli participants from the general public aged more than 70 years agreed to a greater extent (80.4%) toward family involvement than those aged 50 to 59 years (72.6%).Špoljar et al. [11] in Croatia also observed that female physicians and nurses were more inclined than men toward collaborative decision-making with the family.The reverse was observed by Cardoso et al. [52] in Portugal regarding physicians.Baykara et al. [43] found that experienced ICU physicians are more supportive of family involvement in EOL decision-making compared to less experienced physicians (less than two years of experience).In the Middle East [59], more experienced physicians were more inclined to override patient/family decisions about DNR orders, while in South Africa [15], nurses with more ICU experience supported family involvement to a greater extent than the rest of their nurse colleagues.In Portugal [52], physicians with >10 years of experience were more supportive of family and nurse involvement.

Religion/Religiosity
In his studies in Europe [48,49], Vincent found that Catholics were less likely to discuss DNR orders with the patients than protestants or agnostics and that almost all religious physicians were more inclined to involve family, nurses, and other ICU staff in EOL decisions.Additionally, Sprung et al. [13] propose that lack of consensus toward treatment limitation without consent under certain circumstances may be influenced by religious differences.Bodas et al. [51] identified that Jew participants from the general population in Israel agreed more than Arabs (77.6% vs. 68.6%,respectively) toward family involvement.Regarding religiosity, Sprung et al. [13] observed that less religious experts were more likely to agree toward life-sustaining treatments without consent under certain conditions.In Brazil, as noted by Forte et al. [53], more religious physicians were more supportive of nurses' involvement.In Portugal, Cardoso et al. [52] found that agnostic/atheist physicians would involve relatives in decision-making more frequently than Catholics.

Marital Status & Income
More divorced than single (78.4% and 66.7%, respectively) and more average income than below average (82.8% and 72.7%, respectively) respondents from the Israeli public would support family involvement in EOL decision-making, as addressed in the relevant study [51].

Discussion
In this systematic review, we aimed to exclusively synthesize and summarize stakeholder's perspectives regarding who should be involved in EOL decision-making and influential factors.
With one exception [17], the views of the nurses, families, and the public were found to be aligned toward SDM, with the nurses additionally supporting their own involvement.
Overall, a societal movement toward SDM was identified.Despite the latter being quite hesitant toward collaboration, to a different extent, nurses and physicians wish to have influence (though not necessarily absolute control) in EOL decisions.There may also be an underlying authority issue on the part of the physicians that nurses wish to balance.From their perspective, families are the most qualified to express their relatives' wishes and should be given the opportunity to be heard.
The COVID-19 period of hospital triage and autonomy limitation [24,62] stressed healthcare and legislative systems globally to the maximum, created vigorous ethical dilemmas, brought prognostic and communicational uncertainty, intensified conflicts, and brought barriers to SDM [24,63].A recent study in France [19] conducted in the aftermath of the pandemic revealed physicians' hesitancy toward family involvement in light of conflict.Contrarily, the Israeli public [51] supported relatives' involvement.Under these recent circumstances, stakeholders' attitudes toward aspects of EOL, especially decision-making, should be more than ever in focus.
Limitations of our review also deserve discussion.Firstly, due to discrepancies between facts or attitudes about what is being done and opinions about what should be done [11,12,52], as well as for needs of coherence and direct relevance with our specific objectives, important studies, like ETHICUS studies [1,20,26,27,34,40,[64][65][66] addressing practices, facts, attitudes toward other aspects of EOL decision-making and relevant factors, were excluded.To limit assumptions, we exclusively included studies that, in our view, illustrated opinions about who should be involved in EOL decision-making clearly and distinctly.However, this assessment was subjective.
Furthermore, the reviewed studies did not exclusively focus on the attitudes regarding decision-making involvement, and the attention given varied.We were also unable to identify any studies meeting our inclusion criteria that were conducted directly considering recent COVID-19 events.
Study categorization was subjective and influenced by the data provided, and response options given by the initial authors.Additionally, the reliance on closed-ended questionnaires in most of them restricts respondent input and may not fully capture the spectrum of their attitudes [5][6][7][8]18,42,48,49].Our findings derived exclusively from the provided data and future studies may introduce new input that could potentially modify them.
The uneven distribution of surveys and the focus on clinicians rather than a representative sample of the general population also restrict the broader applicability of the results.The timeframe of the studies is another important factor to consider, as the evolving legislative and societal landscape can influence the reviewed attitudes.Therefore, our results should be evaluated within the specific social, educational, legal, and scientific context of the time that the relevant research was conducted.

Conclusions
This systematic review highlighted the diversity and complexity of the stakeholders' perspectives as well as the dynamics that seem to arise.Overall, a societal movement toward SDM was identified, which aligns with international suggestions, and the findings of relevant literature toward practices.Globally, EOL scenarios occur in a complex context that cannot be uniformly harmonized, comprising personal values and beliefs, as well as diverse cultural and regulatory frameworks.Hence, applicable solutions against conflicts and toward overall high-quality EOL care should be more extensively pursued on a tailored regional level rather than a unified global scale.In turn, multifaceted efforts are needed both in clinical practice (regarding clinicians' handling of EOL situations) and across the governmental/regulatory spectrum to promote the practical implementation of the conceptualization of SDM, emphasize ethical training, and provide guidance.Finally, future studies should include various population categories within specific timeframes by using commonly accepted validated instruments for a more comprehensive illustration of EOL decision-making attitudes.

Additional Information
Author Contributions
of a competent patient, more than half of the physicians were willing to make decisions collaboratively (38% with the patient, 1% with the relatives, and 23% with the patient and the relatives) with a small percentage of them (8%) leaning toward patient's autonomy.In contrast, when dealing with an incompetent patient scenario, 61% believed they should make the decision alone.Only 36% expressed a preference for involving the family in the decision-making process.Nurses generally supported shared decision-making regarding the continuation of ventilator treatment in both competent and incompetent patient scenarios.70% of nurses believed that decisions about the continuation of ventilator treatment should be made jointly by the family and the physician in the incompetent patient scenario.In the competent patient scenario, 57% supported a collaborative decision between the physician, the patient, and/or their family while a smaller percentage (31%) supported patients being the decision-makers alone or together with their families.likely to discuss withdrawal of continued ventilation with the family when a patient is competent compared to when a patient is incompetent (74.8% vs. 60.7%).71.6% of the physicians believed that such decisions should be made collaboratively by the patient and/or their family together with the physician.In cases of an incompetent patient, 76.8% of the physicians expressed the opinion that decisions regarding ventilator withdrawal should be jointly made 75% of nurses believed that discussions about withdrawal of continued ventilation should be held with the family in the competent patient scenario, and 74% in the incompetent patient scenario.53.4% of nurses believed that decisions about continued ventilator treatment should be shared between the patient and/or family together with the physician in the competent patient scenario, and 78.4% in the incompetent patient scenario. of family members wanted the decision to be shared between 2024 Georgakis et al.Cureus 16(1): e52912.DOI 10.7759/cureus.5291210 of 29 by the family and themselves. of the physicians supported joint decision-making between physicians and nurses regarding the application of DFLSTs for the patient.Less than half of them (44.3%) agreed that relatives should actively participate in the decision-making process.88.9% of nurses believed that decisions to apply limitation of lifesustaining treatment (LLST) should be taken jointly by physicians and nurses.69.8% of nurses supported family involvement in the decisionmaking process.

TABLE 2 : Main findings of the studies toward EOL decisions by evaluated population in national studies.
attitudes Experts' attitudes (physicians, nurses, and non-clinical experts) Families' attitudes All of them supported 2024 Georgakis et al.Cureus 16(1): e52912.DOI 10.7759/cureus.52912withthe consent or agreement of the patient/surrogate/family.A consensus was not reached that it is permissible to withhold or withdraw treatment in situations where an agreement cannot be secured (incompetent 2024 Georgakis et al.Cureus 16(1): e52912.DOI 10.7759/cureus.52912Australiaand South Africa)patient and no family).However, most respondents from all regions (total >70%) agreed.Despite the fact that there was not an equal sample from all regions (ranging from 10 participants from Australia to 688 from Western Europe), this was less

TABLE 3 : Main findings of the studies toward EOL decisions by evaluated population in multinational studies.
N: number of participants per population group; EOL: end of life; SDM: shared decision-making; DNR: do not resuscitate.Physicians' Attitudes Highlighting a Collaborative Decision-Making Approach in National Studies2024 Georgakis et al.Cureus 16(1): e52912.DOI 10.7759/cureus.52912

Table A1 PubMed
(end of life decisions OR end-of-life decisions OR end of life decision making OR end-of-life decision making OR life-sustaining therapies OR withholding OR withdrawing OR DNR OR do-not-resuscitate OR de-escalation) AND (critical illness OR critically ill patients OR critical care OR intensive care unit OR ICU OR critical care unit) AND (adult intensive care OR adults) CINAHL TX (end of life decisions or end of life discussions or end of life communications) AND TX (intensive care unit or ICU or critical care or critical care unit) AND TX (adults or adult or aged or elderly) Embase ('treatment withdrawal' OR 'withdrawal, treatment' OR 'withholding treatment' OR 'end of life decision' OR 'end of life decision making' OR 'shared decision' OR 'de-escalation') AND 'intensive care unit' AND ([adult]/lim OR [young adult]/lim OR [middle aged]/lim OR [aged]/lim OR [very elderly]/lim)

TABLE 9 : Full search terms used for literature search.
CINAHL: Cumulative Index to Nursing & Allied Health.

Table A2 Population ICU physicians, nurses, family members of ICU patients, as well as the general public
ExposureThe perspectives and the influencing factors of the opinions of ICU physicians, nurses, family members of ICU patients, and the general public Outcome End-of-life decisions, decisions for withholding/withdrawing of life-sustaining measures, do-not-resuscitate orders Setting ICU for adult patients (≥18 years of age)