A
Nursing Caring for Terminal Patients in the ICU: A Systematic Literature Review on Humanistic Roles Beyond Artificial Intelligence
KoriLimbong1,2
FathiyaLuthfilYumni1
VinaPutriPatandung1,4
ErikaMartiningWardani1,5
MosesGlorino1
RumamboPandin3
Dr.
Ir.H.Soekarno1
Mulyorejo1
Kec.Mulyorejo1
Surabaya1
EastJava1
1
A
A
A
A
Doctoral Program of Nursing, Faculty of NursingUniversitas AirlanggaJl
2
A
Department of NursingHealth Polytechnic of KupangJl. Piet A. Tallo LilibaKupang
3Faculty of HumanitiesUniversitas AirlanggaJl. Dr. Ir. H. Soekarno, Mulyorejo, Kec. Mulyorejo60115Surabaya, East Java
4
A
Department of NursingSTIKes Gunung Maria, Jl. FlorenceTomohonNorth Sulawesi
5Departement of Nursing, Faculty of Nursing and MidwiferyUniversitas Nahdlatul Ulama Surabaya60237Surabaya, East JavaIndonesia
Kori Limbong1,2, Fathiya Luthfil Yumni1, Vina Putri Patandung1,4, Erika Martining Wardani1,5, Moses Glorino Rumambo Pandin3
1Doctoral Program of Nursing, Faculty of Nursing, Universitas Airlangga, Jl. Dr. Ir. H. Soekarno, Mulyorejo, Kec. Mulyorejo, Surabaya, East Java 60115.
2Department of Nursing, Health Polytechnic of Kupang, Jl. Piet A. Tallo Liliba, Kupang, East Nusa Tenggara 85111.
3Faculty of Humanities, Universitas Airlangga, Jl. Dr. Ir. H. Soekarno, Mulyorejo, Kec. Mulyorejo, Surabaya, East Java 60115
4Department of Nursing, STIKes Gunung Maria, Jl. Florence, Tomohon, North Sulawesi
5Departement of Nursing, Faculty of Nursing and Midwifery, Universitas Nahdlatul Ulama Surabaya, 60237 Surabaya, East Java, Indonesia
Corresponding author :
ABSTRACT
Background:
A
End of life care (EOLC) in intensive care units (ICUs) involves complex ethical, emotional, and systemic challenges that are increasingly influenced by the integration of artificial intelligence (AI) in healthcare. Nurses often experience moral distress when balancing life sustaining interventions and patient dignity while adapting to technological systems that affect empathy, accountability, and clinical judgment.
Objective:
This study aims to synthesize empirical evidence on nurses’ experiences, ethical dilemmas, and evolving roles in providing EOLC within the digital era, emphasizing the humanistic aspects that extend beyond AI applications.
Methods:
A systematic literature review was conducted following the PRISMA 2020 guidelines. Searches were performed in PubMed, Scopus, and CINAHL databases for peer reviewed articles published between 2019 and 2025. Nine eligible studies were analyzed thematically, covering ethical, emotional, cultural, systemic, and technological dimensions of EOLC nursing.
Results:
Nurses reported moral distress, limited palliative care competence, emotional fatigue, and inadequate institutional support. While AI enhanced efficiency and decision accuracy, concerns persisted about algorithmic bias, data transparency, and dehumanization of care. The findings highlight the irreplaceable human role of nurses in providing empathy, moral reasoning, and spiritual presence within AI supported environments.
Conclusion:
Effective EOLC in the AI era requires synergy between technological precision and compassionate nursing. Sustaining human centered values remains essential to ensure ethical integrity, emotional resilience, and dignified patient care.
Keywords:
End of life care
Critical care nursing
Artificial intelligence
Ethics
Empathy
Humanistic nursing
Digital transformation.
A
A
A
INTRODUCTION
End of life care (EOLC) in the context of intensive nursing practice presents increasing ethical, emotional, and systemic complexities, particularly with the advancement of digital technology and the growing integration of artificial intelligence (AI) in healthcare. International studies affirm that nurses working in intensive care units (ICUs) frequently face moral dilemmas in balancing life-prolonging interventions with the preservation of patient dignity (Palmryd et al., 2025).These challenges are often intensified by family pressure, interprofessional conflict, and the absence of clear ethical guidelines. Moreover, significant gaps remain in nurses’ understanding of palliative care philosophy and competence, as reported among ICU nurses in Oman and Indonesia ((Almahrizi et al., 2025); (Mediani et al., 2024).
Emotionally, ICU nurses experience moral fatigue, sadness, and guilt due to their recurrent exposure to death and dying (Aschale et al., 2025). Cultural factors also play a crucial role; in East and Southeast Asian settings, family values and religious beliefs profoundly influence medical decision making ((Xu et al., 2025). In addition, a study in France revealed that aggressive end of life interventions.
A
With the progression of digitalisation, a new dimension has emerged in nursing practice, AI integration. (Martinez-Ortigosa et al., 2023)demonstrated that AI applications in nursing can enhance efficiency and service quality; however, they also raise ethical concerns related to algorithmic bias and the potential dehumanisation of care. (Challen et al., 2019) emphasised that AI should function as an assistive tool within a human in the loop framework to preserve clinical safety and accountability. In nursing education, AI ethics awareness has become a vital competency for future practitioners. Yang (2024) found a strong correlation between digital literacy, moral sensitivity, and AI ethics awareness among nursing students.
Therefore, the challenges of end of life care now extend beyond traditional ethical and emotional issues to include how nursing professionals navigate the integration of intelligent technologies without compromising humanity, empathy, and professional responsibility.
METHODS
Study Design
This study employed a Systematic Literature Review (SLR) design to identify, evaluate, and synthesise empirical research related to nurses’ experiences, ethical challenges, and the role of artificial intelligence (AI) in end of life care (EOLC) within intensive care units (ICUs). The review followed the PRISMA 2020 guidelines to ensure methodological transparency, reproducibility, and rigour throughout the literature screening and synthesis process.
A
Fig. 1
The identification, screening, and inclusion procedures of the studies were available in Scopus, Web of Science, PubMed, and CINAHL databases.
Click here to Correct
Click here to Correct
Click here to Correct
Click here to Correct
Click here to Correct
Click here to Correct
Click here to Correct
Click here to Correct
Click here to Correct
Click here to Correct
Inclusion and Exclusion Criteria
Inclusion criteria included:
1.
Empirical studies (qualitative, quantitative, or mixed methods) published in peer reviewed journals between 2019 and 2025.
2.
Research focusing on the experiences of nurses or healthcare professionals in providing EOLC in ICU settings.
3.
Studies exploring ethical, emotional, competency-related, or technological aspects (including AI integration) within critical care nursing contexts.
4.
Full text articles published in English.
Exclusion criteria encompassed non empirical works (e.g., editorials or letters), single case reports, and studies unrelated to ICU populations or nursing practice.
Search Strategy
A comprehensive literature search was conducted across PubMed, Scopus, CINAHL, and Web of Science databases using combinations of the following keywords: “end of life care,” “critical care nurses,” “ethical challenges,” “palliative care,” “artificial intelligence,” “digital ethics,” and “nursing education.”
Additional searches of reference lists and grey literature were undertaken to ensure completeness and minimise publication bias. Articles meeting the relevance criteria were downloaded and screened by title, abstract, and full text. Screening and selection were performed independently by two reviewers, with discrepancies resolved through discussion and consensus.
Data Extraction Procedure
Data were extracted using a structured form that captured the following elements: author(s) and year of publication, country, study design, participants, main research focus, key findings, and identified ethical issues.
A thematic analysis approach was applied to categorise findings into five overarching themes:
Ethical and Professional Challenges, Knowledge and Skill Gaps,Emotional and Cultural Dimensions, and
Systemic and Technological Constraints and Human–Artificial Intelligence Synergy in End of Life Nursing Care
The synthesis process was conducted narratively by comparing contexts, designs, and outcomes across studies to identify common patterns and unique contributions. This analysis also integrated emerging perspectives on the role of AI and digital ethics in modern nursing practice.
RESULT
Study Characteristics
A
Table 1
Study Characteristics
No
Author(s) (Year)
Country
Design
Participants
Main Focus
1
(Almahrizi et al., 2025)
Oman
Quantitative (Cross-sectional)
131 ICU nurses
Knowledge and attitudes toward palliative care among critical care nurses
2
(Palmryd et al., 2025)
Sweden
Qualitative (Interpretive Descriptive)
20 Critical Care Nurses (CCNs)
Ethical challenges faced by nurses in end of life care (EOLC)
3
Agius et al. (2025)
France
Retrospective Observational
270 cancer patients
Intensity and economic burden of end of life chemotherapy
4
Aschale et al. (2025)
Ethiopia
Qualitative (Phenomenological)
12 ICU nurses
Emotional experiences and coping strategies in providing EOLC
5
Xu et al. (2025)
China
Qualitative (Descriptive Phenomenological)
13 ICU nurses
Cultural conflict and communication in end of life decision making
6
Mediani et al. (2024)
Indonesia
Qualitative (Phenomenological)
7 healthcare professionals (nurses and physicians)
Experiences in providing palliative care in the ICU context
7
Martínez-Ortigosa et al. (2023)
Spain
Systematic Review (PRISMA-based)
21 studies, total 10–230,936 participants (mean 14,948)
Applications of artificial intelligence in nursing practice (diagnosis, monitoring, education, and workflow efficiency)
8
Challen et al. (2019)
United Kingdom, USA
Narrative Review
Theoretical and clinical AI applications (no specific participants)
Analysis of algorithmic bias, safety, and clinical quality in AI based decision making systems
Main Findings
The synthesis of eight reviewed studies generated five overarching cross cutting themes that characterise the complexities of end of life care (EOLC) in intensive care settings and its intersection with technological advancement:
1.
Ethical and Professional Challenges
Nurses frequently encounter moral dilemmas when balancing the prolongation of life with the preservation of patient dignity. Ethical tensions often arise from pressure to continue futile treatments, the use of palliative sedation, and conflicts with physicians or family members regarding treatment withdrawal decisions (Palmryd et al., 2025; Xu et al., 2025). These dilemmas underscore the need for institutional ethical support systems and reflective decision-making frameworks in critical care environments.
2.
Knowledge and Skill Gaps in Palliative Care
A persistent deficit in knowledge concerning the philosophy and holistic principles of palliative care remains evident among critical care nurses. Gaps are particularly notable in addressing psychosocial and spiritual aspects of patient care (Almahrizi et al., 2025). Empirical evidence indicates that formal education and clinical exposure significantly enhance nurses’ competence and confidence in delivering EOLC.
3.
Emotional and Cultural Dimensions
Emotional distress manifested as sadness, guilt, and burnout was common across all settings (Aschale et al., 2025). In collectivist Eastern contexts such as China and Indonesia, cultural values related to family obligation and religious norms strongly influence end of life decision making (Xu et al., 2025; Mediani et al., 2024). These findings highlight the necessity of culturally responsive and emotionally supportive interventions for ICU nurses.
4.
Systemic and Economic Constraints
From a systemic perspective, aggressive medical interventions at the terminal stage, such as end of life chemotherapy, were found to elevate healthcare costs while diminishing patients’ quality of life (Agius et al., 2025). Additionally, organisational and policy limitations hinder the timely integration of palliative care practices in ICUs, particularly in low and middle income settings (Mediani et al., 2024).
5.
Human Artificial Intelligence Synergy in End of Life Nursing Care
The integration of artificial intelligence (AI) into end of life nursing practice highlights the critical need to balance technological analytical capability with human emotional intelligence. While AI improves diagnostic precision and workflow efficiency, nurses remain central in ensuring ethical judgment, empathy, and patient safety (Martínez-Ortigosa et al., 2023; Challen et al., 2019). This synergy reinforces the imperative that technological innovation should augment rather than replace the moral and compassionate dimensions of nursing care.
DISCUSSION
End of life care (EOLC) in intensive care units (ICUs) represents one of the most complex challenges in modern healthcare, requiring a delicate balance between life sustaining technologies and respect for human dignity. Palmryd et al. (2025) emphasised that ICU nurses often face ethical dilemmas when aggressive treatments continue despite a poor prognosis, creating significant moral distress. This finding aligns with the systematic review by Henrich et al. (2022) in Intensive and Critical Care Nursing, which revealed that ICU nurses frequently experience moral distress when performing interventions perceived as futile or inconsistent with their professional values. Institutional support, such as ethics consultations and team reflections, has been shown to reduce moral conflict and strengthen moral resilience among healthcare professionals.
Deficits in palliative care knowledge also emerged as a central issue across studies. Almahrizi et al. (2025) found that most ICU nurses in Oman demonstrated limited knowledge of palliative care principles, particularly regarding spiritual communication and the management of non physical symptoms. Similarly, Ma et al. (2023) in BMC Palliative Care reported that nurses’ competence in EOLC was strongly influenced by clinical experience, formal education, and exposure to terminal cases. A cross national study by Boucher et al. (2022) in palliative medicine confirmed that continuous education and interprofessional simulation training improve both clinical skills and empathy when caring for terminally ill patients. In Indonesia, Mediani et al. (2024) highlighted that the absence of hospital policies and structured training often leads healthcare workers to perceive palliative care merely as end stage management rather than a comprehensive approach that enhances quality of life throughout the disease trajectory.
A
Beyond knowledge gaps, effective communication between healthcare providers and patients’ families emerged as a key determinant of dignified EOLC. Aschale et al. (2025) and Xu et al. (2025) found that open and empathetic communication not only facilitates family understanding but also reduces decisional conflict. Evidence from a randomised controlled trial by Curtis et al. (2016) in the American Journal of Respiratory and Critical Care Medicine demonstrated that structured communication interventions in ICUs significantly improved family satisfaction by 27% and alleviated post loss anxiety. Similarly, Anderson et al. (2019) in Palliative Medicine confirmed that value based communication training centred on empathy and shared decision making is an effective strategy to improve the quality of EOLC.
Cultural context also plays a critical role in shaping end of life decisions. Xu et al. (2025) noted that in Chinese culture, the value of filial piety often compels families to continue all forms of medical treatment, even when the prognosis is poor. In contrast, Mediani et al. (2024) found that spiritual and religious beliefs strongly influence EOL decisions in Indonesia, highlighting the need for high levels of cultural and spiritual sensitivity among healthcare providers. Chan et al. (2021) in the Journal of Advanced Nursing further demonstrated that cultural competence significantly enhances patient and family satisfaction in multicultural ICU settings. A cultural humility approach defined as reflective awareness of patients’ beliefs and values proved to be more effective than rigid adherence to universal communication protocols.
Psychologically, repeated exposure to death and dying places ICU nurses at high risk of chronic stress, emotional exhaustion, and burnout. Aschale et al. (2025) and Palmryd et al. (2025) underscored the importance of internal support mechanisms such as debriefing sessions, ethical reflection, and peer support to help nurses manage grief, guilt, and emotional fatigue. Brooks et al. (2017) in Australian Critical Care showed that post death reflective sessions reduce burnout and improve psychological resilience among ICU nurses. Organisationally, Schwarzkopf et al. (2024) in Healthcare emphasised the necessity for hospital policies that provide structured emotional and psychosocial support as part of quality and patient safety strategies.
Roles Beyond Artificial Intelligence in EOLC Nursing
The emergence of artificial intelligence (AI) in healthcare has begun to transform clinical practice in ICUs, particularly in mortality prediction, early detection of organ failure, and clinical decision support systems. However, amid these technological advancements, fundamental questions arise concerning the nurse’s role “beyond AI.” Recent literature highlights that the unique strength of nurses lies in their capacity to integrate empathy, ethical reasoning, and contextual judgment dimensions that cannot be replicated by algorithmic systems. Topol (2019) in Nature Medicine asserted that although AI can automate diagnostic and predictive functions, decisions involving meaning, values, and human suffering remain inherently within the ethical and emotional domain of human caregivers. In the ICU context, roles beyond AI encompass emotional presence, therapeutic communication with families, and moral advocacy in end of life decision making.
Luo et al. (2024) in BMJ Health & Care Informatics reported that AI systems can assist in predicting terminal phases, yet without human interpretation, such recommendations risk dehumanised decision making. Here, the nurse’s role as a bridge between data and meaning becomes indispensable. Kim et al. (2023) in Nursing Philosophy further argued that AI integration in nursing practice must be guided by philosophical and ethical reflection to prevent the erosion of care’s humanistic dimension. Thus, roles beyond AI do not reject technological advancement; instead, they affirm the principle that AI should serve as an assistive rather than a substitutive tool supporting nurses’ moral, communicative, and spiritual competencies.
Consequently, nursing education and institutional policy must prepare nurses to become empathetic data interpreters professionals capable of using AI generated insights ethically while upholding patient values, spirituality, and humane interaction. This aligns with the WHO Human Centred Digital Health Framework (2023), which emphasises placing humans at the core of health innovation. Therefore, EOLC in the digital era should not focus on AI replacing human roles but rather on how technology can extend nurses’ reflective, ethical, and compassionate capacities. These roles empathy, moral advocacy, and spiritual presence will remain the enduring essence of critical care nursing in the modern ICU.
CONCLUSION
This systematic review demonstrates that end of life care (EOLC) in intensive care units (ICUs) represents a multidimensional challenge encompassing ethical, emotional, educational, systemic, and technological factors. Nurses frequently experience moral distress when balancing life sustaining interventions with the preservation of patient dignity, often in contexts lacking institutional guidance and ethical clarity. Deficiencies in palliative and spiritual care competence further constrain the delivery of holistic and compassionate services. Emotional exhaustion and cultural influences continue to shape decision making in critical care, underscoring the need for human centred and context sensitive approaches. The integration of artificial intelligence (AI) introduces both opportunities and ethical complexities: although AI enhances efficiency and clinical precision, it raises concerns regarding bias, accountability, and the erosion of empathy. Hence, nurses’ roles must transcend technical adaptation to reaffirm their humanistic capacities ethical reflection, emotional presence, and moral advocacy. Future nursing education and policy should strengthen digital literacy, ethical sensitivity, and interprofessional collaboration to ensure that AI remains an assistive, rather than substitutive, tool. Ultimately, sustaining a synergy between technological innovation and human compassion is essential to uphold the ethical integrity and dignity of end of life nursing care.
REFFERENCE:
Almahrizi, H. A., Alaloul, F., Al Mamari, O. K., Rani, E. K., Al Mahrizi, Z. A., Al Harthy, S. A., & Al-Naamani, Z. (2025). Empowering critical care nurses: Bridging knowledge gaps in palliative care. BMC Nursing, 24(1127). https://doi.org/10.1186/s12912-025-03699-1
Altaker, K. W., Howie-Esquivel, J., & Cataldo, J. K. (2018). Relationships among palliative care, ethical climate, empowerment, and moral distress in intensive care unit nurses. American Journal of Critical Care, 27(4), 295–302. https://doi.org/10.4037/ajcc2018252
Aschale, A., Gishu, T., Mengist, S., & Tsehay, M. (2025). Experiences of intensive care unit nurses in providing end-of-life care in public hospitals: A phenomenological study. BMC Nursing, 24(1185). https://doi.org/10.1186/s12912-025-03849-5
Brooks, L. A., Manias, E., & Nicholson, P. (2017). Barriers, enablers, and challenges to implementing end-of-life care in critical care settings. Australian Critical Care, 30(3), 161–165. https://doi.org/10.1016/j.aucc.2016.10.003
Chan, E. A., Wong, F., Cheung, K., & Lam, W. (2021). Cultural competence in end-of-life nursing care: A systematic review. Journal of Advanced Nursing, 77(4), 1873–1885. https://doi.org/10.1111/jan.14748
Henrich, N. J., et al. (2022). Moral distress and ethical climate in critical care nursing: A systematic review. Intensive and Critical Care Nursing, 68, 103136. https://doi.org/10.1016/j.iccn.2022.103136
Li, M., et al. (2023). Knowledge, attitudes, and practices of end-of-life care among ICU nurses: A cross-sectional survey. BMJ Supportive & Palliative Care, 13(2), 230–238. https://doi.org/10.1136/bmjspcare-2021-003012
Mediani, H. S., Sada, F. R., Nuraeni, A., & Subu, M. A. (2024). Healthcare professionals’ experiences in providing palliative care in an intensive care unit in Indonesia: A phenomenological study. Journal of Multidisciplinary Healthcare, 17, 4427–4439. https://doi.org/10.2147/JMDH.S486021
Palmryd, L., Rejnö, Å., Alvariza, A., & Godskesen, T. (2025). Critical care nurses’ experiences of ethical challenges in end-of-life care. Nursing Ethics, 32(2), 424–436. https://doi.org/10.1177/09697330241252975
Xu, D.-D., Li, J., Ding, X.-B., Ma, J., Hou, R.-T., Chen, N.-N., Cheng, X.-L., & Hu, F. (2025). Experiences of providing end-of-life care in adult intensive care units: A qualitative study. BMC Nursing, 24(768). https://doi.org/10.1186/s12912-025-03340-1
Topol, E. J. (2019). High-performance medicine: The convergence of human and artificial intelligence. Nature Medicine, 25(1), 44–56. https://doi.org/10.1038/s41591-018-0300-7
Challen, R., Denny, J., Pitt, M., Gompels, L., Edwards, T., & Tsaneva-Atanasova, K. (2019). Artificial intelligence, bias and clinical safety. BMJ Quality & Safety, 28(3), 231–237. https://doi.org/10.1136/bmjqs-2018-008370
Martinez-Ortigosa, A. (2023). Applications of artificial intelligence in nursing care: A review. [Journal], (note: full journal details to confirm) https://doi.org/10.1155/2023/3219127
Yang, Y. (2024). Influences of digital literacy and moral sensitivity on artificial intelligence ethics awareness among nursing students. Healthcare, 12(21), 2172. https://doi.org/10.3390/healthcare12212172
Total words in MS: 2931
Total words in Title: 18
Total words in Abstract: 220
Total Keyword count: 7
Total Images in MS: 1
Total Tables in MS: 1
Total Reference count: 0