Constrained moral agency,distress,and residue in Neonatal Intensive Care nurses

Aim

To explore how NICU nurses experience and enact moral agency under constraint during ethical decision-making at the threshold of viability, and how these experiences shape moral distress and moral residue.

Research question

How do NICU nurses describe organisational, relational, and clinical conditions that constrain or enable moral agency during threshold-of-viability decision-making, and how do these conditions shape experiences of moral distress and moral residue?

Objectives

(1) To describe nurses’ experiences of moral distress in threshold-of-viability care.

Design/approach

This study used an interpretive (hermeneutic) phenomenological approach,^35,36 to explore how NICU nurses experience and interpret moral agency under constraint during threshold-of-viability care. Hermeneutic phenomenology is appropriate when the objective is to develop an interpretive understanding of how individuals make meaning of ethically complex experiences embedded within specific relational and organisational contexts. Given the ambiguity, emotional intensity, and interprofessional power dynamics characteristic of viability-related care, this approach enabled in-depth examination of the meanings nurses attributed to ethical action, constrained voice, and enduring moral impact. ³⁷ The manuscript is reported in accordance with the COREQ guidelines. ³⁸

Theoretical framework

The study was informed by an integrated conceptual lens linking moral distress (as a constraint-based ethical phenomenon) with moral agency (nurses’ capacity to perceive ethically salient situations, deliberate, voice concerns, and act in accordance with professional values).^39,40 Figure 1 represents a dynamic pathway in which threshold-of-viability clinical conditions (e.g., prognostic ambiguity, rapid deterioration/improvement, time-critical turning points, heightened parental vulnerability) interact with multi-level constraints (organisational policies, interprofessional hierarchy, workflow compression, and communication norms). These conditions shape moral agency processes (moral perception, deliberation, perceived legitimacy to speak, advocacy, and participation), which, in turn, influence moral outcomes, including moral distress, moral residue, and ethical engagement. ²³ OPEN IN VIEWER

Consistent with interpretive phenomenology, the framework functioned as a sensitising concept rather than a deductive coding template. ⁴¹ It informed the research question and interview guide to ensure comprehensive coverage of key constructs while preserving openness to participants’ meanings. Analytically, the study adopted a primarily inductive stance: initial coding remained grounded in participants’ language; the framework was subsequently used during interpretation to examine how emergent themes mapped onto, refined, or challenged the proposed pathways. ⁴²

Data collection

Time frame and approach: Data were collected between June 2025 and October 2025 using purposive maximum-variation sampling, in-depth semi-structured interviews, and analytic field notes. Data collection and analysis proceeded iteratively to support ongoing assessment of conceptual sufficiency.

Setting and context: The study was conducted across four Level III neonatal intensive care units (NICUs) within one administrative region of the country. These facilities function as regional referral centres and provide care for infants at the threshold of viability. A multisite design was used to capture variation in institutional micro-cultures while maintaining comparability in level of care. The broader sociocultural and medico-legal context, including family roles in decision-making and policies related to resuscitation and treatment limitation, was treated as a contextual layer shaping moral agency and perceived clinical constraints.^43,44

Sample and recruitment: Purposive maximum-variation sampling targeted registered nurses with direct experience of threshold-of-viability care. Variation was sought across years of NICU experience (1–15 years), clinical role (bedside nurse/charge nurse), and hospital site (A–D). To minimise coercion, unit leadership served only as gatekeepers for distributing study information and did not know who participated. Interested nurses contacted the research team via a dedicated confidential email address. Of 45 nurses expressing interest, 32 met the inclusion criteria; 28 provided written informed consent (participation rate among eligible = 87.5%).

Sample size and conceptual sufficiency: Sample adequacy was guided by the principle of information power, considering the specificity of the sample, the clarity of the aim, the theoretical lens, the quality of interview dialogue, and the analytic strategy. ⁴⁵ Concurrent data collection and analysis enabled systematic assessment of conceptual sufficiency, whereby continued interviewing was evaluated against the extent to which additional data contributed novel meaning or only refined existing thematic patterns. ⁴⁶ The research team reviewed emerging concepts after each block of 4–5 interviews. By interview 24, the thematic structure was stable; four additional interviews were conducted across sites to test stability and to search for disconfirming cases. Recruitment concluded when no substantively new insights were generated regarding mechanisms of constraint or enactment of moral agency.

Eligibility criteria: Nurses were eligible if they ¹ were registered nurses employed in one of the four participating neonatal intensive care units; ² ; had ≥12 months NICU experience; and ³ had direct clinical involvement within the previous 12 months in the care of infants at the threshold of viability, defined as 22 weeks and 0 days to 24 weeks and 6 days’ gestation (22 + 0 to 24 + 6 weeks). Exclusion criteria were current internship/orientation status or exclusively administrative roles without direct patient care during the study period.

Data generation procedures: Semi-structured interviews were conducted to elicit rich accounts of ethically salient experiences, perceived constraints, and moral consequences. Participants selected Arabic (n = 22) or English (n = 6). Interviews lasted a mean of 65 min (range 45–95), were audio-recorded with explicit consent, and were accompanied by analytic field notes capturing contextual observations relevant to interpretation. The interview guide (Supplemental File S1) was developed from the sensitising framework and refined through expert review and piloting (n = 2). Pilot feedback informed minor refinements to improve the clarity and neutrality of prompts.

Translation and verification: Arabic interviews were transcribed verbatim in Arabic and translated into English by a certified professional.^47,48 A bilingual team member verified translations against the audio to ensure semantic and conceptual equivalence of ethically salient terms. A subset of transcripts (n = 5; 18%) underwent independent back-translation to further assess the fidelity of meaning.

COREQ interviewer and interview context: Interviews were conducted by a trained qualitative researcher from the study team with formal preparation in qualitative interviewing. ³⁸ The interviewer had no prior relationship with participants and held no supervisory or managerial authority within the participating units. Interviews were conducted remotely via Zoom, using password-protected meeting links and a waiting room to support confidentiality. Participants joined from a private location of their choice, and interviews were scheduled at times convenient to participants, outside clinical duties where possible, to minimise workplace influence and interruptions. No repeat interviews were conducted. Transcripts were not returned to participants. Analytic field notes were documented immediately following each interview. Remote interviewing was selected to enhance feasibility across four hospitals, support privacy for a sensitive topic, and reduce perceived workplace influence.

Data analysis

Data were analysed using hermeneutic phenomenological thematic analysis,^49,50 managed in NVivo 12. ⁵¹ Analysis proceeded iteratively alongside data collection to refine developing interpretations and to guide sampling toward conceptual depth. Consistent with hermeneutic phenomenology, analysis moved recursively between detailed textual segments and the developing whole (the hermeneutic circle) to interpret meanings situated within clinical, organisational, and sociocultural contexts. Themes were treated as patterns of meaning rather than frequency-based categories; theme development was used as an analytic strategy to articulate shared meaning structures across accounts while retaining attention to contextual nuance and individual variability. ⁵²

Transcripts were read and re-read alongside field notes (and audio where needed) to support interpretive immersion. ⁵³ Two researchers independently coded the first five transcripts to enhance analytic sensitivity and surface divergent interpretations; differences were discussed to refine code meanings while preserving openness to multiple readings. Coding remained grounded in participants’ language and concrete accounts. Codes were clustered into provisional themes capturing patterned meanings related to constraints, moral agency processes, and moral outcomes. ⁵⁴ The research team reviewed the developing thematic structure against coded extracts and the full dataset, refining boundaries, merging or splitting themes, and actively searching for disconfirming cases. ⁵⁵ Themes were then defined with analytic clarity and aligned with the study aim. The sensitising framework informed interpretation and theme refinement but was not used as a deductive coding template, ensuring themes remained grounded in the data. ⁵⁶ Analytic memos and an audit trail documented decisions throughout the coding and theme development process. Translated English transcripts were used for analysis; translation verification procedures are described in Section 4.3.

Rigour/trustworthiness

Trustworthiness was established using strategies aligned with Lincoln and Guba’s criteria for qualitative rigour (credibility, dependability, confirmability, and transferability), supported by ongoing critical reflexivity.^57,58 Procedures were selected to fit an interpretive (hermeneutic) phenomenological approach in which meanings are co-constructed, and interpretations are strengthened through transparency, reflexive scrutiny, and systematic documentation of analytic decisions. Table 1 summarises the criteria and strategies used.

OPEN IN VIEWER

Table 1.

Framework for ensuring trustworthiness.

Criterion	Definition	Strategies employed in this study
Credibility	Confidence in the plausibility and coherence of interpretations. 59	(1) In-depth interviewing (mean 65 minutes) to elicit rich experiential accounts; (2) maximum-variation sampling across experience (1–15 years), roles, and sites; (3) iterative analysis concurrent with data collection to refine interpretations and pursue conceptual depth; (4) investigator triangulation through early independent coding and ongoing team-based review; (5) search for disconfirming cases during late-stage interviewing and theme refinement; (6) participant reflections (n=4) to assess resonance of preliminary interpretations.
Dependability	Stability of the analytic process over time and across conditions 60	(1) Audit trail documenting recruitment decisions, interview guide refinements, analytic memos, coding iterations, and team decisions; (2) consistent procedures for interviewing, field notes, transcription/translation verification, and data management across all sites; (3) version-controlled NVivo project to track coding evolution.
Confirmability	The extent to which findings are grounded in the data rather than in the researcher’s preference 61	(1) Reflexive journaling by the primary interviewer/analyst to document assumptions, emotional responses, and interpretive shifts; (2) positionality statements developed prior to data collection; (3) peer debriefing with senior qualitative researchers to challenge interpretations; (4) verbatim transcription in the source language with documented translation verification to preserve meaning of ethically salient terms; (5) linkage of themes to auditably traceable extracts within NVivo.
Transferability	Applicability of findings to other contexts based on contextual similarity. 62	(1) Thick description of setting, sociocultural and medico-legal context, participant characteristics, and decision-making environment; (2) clear boundaries of the phenomenon (threshold-of-viability care) to support reader judgments of relevance to comparable NICU contexts.

Ethical considerations

Ethical approval was obtained from the relevant university institutional review board and the ethics committees of the participating clinical sites. Written informed consent was obtained from all participants before data collection. Participants were informed of their right to withdraw at any time without penalty. Confidentiality was protected through pseudonymization and removal of identifying details from transcripts and reports. Audio recordings, transcripts, and analytic materials were stored on encrypted, password-protected systems accessible only to the research team. Given the sensitive nature of discussions on ethically challenging care and moral distress, participants were reminded that they could pause or stop the interview at any point. A referral pathway for confidential counselling support was available if required, although no participant requested its use. Data management, retention, and disposal were conducted in accordance with ethics requirements and institutional policies.

Overview of the participants

Twenty-eight registered nurses from four Level III NICUs in the study setting participated. The sample was predominantly female (n = 25, 89%), included nurses across bedside and charge roles, and represented a broad range of ages and NICU experience (Table 2). Participants were distributed across all sites, supporting the interpretation of recurring patterns across different unit contexts. A senior nurse described the enduring emotional weight associated with viability-related decisions: ‘Whether you have been here two years or twenty, the weight of these decisions never really leaves you. It just changes how you carry it’ (Participant 19, 14 years’ experience).

Key themes identified

Four themes and ten subthemes were developed to address how nurses experienced moral agency under constraint during threshold-of-viability decision-making, how constraints shaped participation and advocacy, and how these experiences contributed to moral distress and moral residue. Themes and subthemes are summarised in Table 3.

Theoretical and Practical Implications

Theoretically, this study offers an empirically grounded account of moral agency under constraint in threshold-of-viability care. The thematic relationships illustrated in Figure 2 suggest that moral distress may be understood as part of a reinforcing process: prognostic ambiguity and ethically fraught care amplify moral tension, structural constraints narrow nurses’ perceived legitimacy to speak and influence decisions, and unresolved tensions accumulate as moral residue that shapes subsequent engagement, coping, and professional identity.

In practice, the findings indicate that exclusively individual-focused responses, such as resilience training alone, are unlikely to be sufficient when moral distress is experienced as structurally produced. Participants’ accounts point to the need for organisational strategies that formalise nursing inclusion in ethical deliberation and create protected spaces for processing ethically challenging cases. For example, the recommendation for interprofessional ethics rounds is grounded in participants’ descriptions of peripheral participation in formal deliberations, whereas the recommendation for structured post-case debriefing responds to their accounts of unresolved moral tensions that accumulate as moral residue. Potential strategies include interprofessional ethics rounds that actively solicit nursing perspectives, structured debriefing after viability-related turning points, and leadership development focused on psychologically safe communication and ethically attuned team practices. Such strategies require evaluation and local adaptation rather than assumptions of direct transferability across settings.

For health professions education, the findings suggest the need to prepare nurses, physicians, and allied health professionals not only in ethical principles but also in the practical enactment of moral agency within hierarchical clinical environments. This includes raising concerns, communicating across power gradients, participating in shared deliberation, and responding to ethical uncertainty in family-centred NICU care. Although this study was not designed to compare participants with and without prior training, most participants reported receiving no formal training in bioethics or neonatal palliative care. Those who had received such training described it as largely theoretical and insufficiently applicable to viability-specific clinical pressures. Educational interventions should therefore move beyond ethics theory alone toward interprofessional simulation and case-based learning that reflects prognostic uncertainty, family dynamics, hierarchical communication, and the relational complexity of NICU decision-making. Embedding this learning across health professions programmes, rather than only in nursing curricula, reflects the systemic and interprofessional nature of the constraints identified in this study.

Strengths and Limitations

A key strength of this study is the interpretive (hermeneutic) phenomenological design, which enabled in-depth examination of how nurses made meaning of constrained moral agency and its moral consequences. The multisite sampling across four hospitals and the use of trustworthiness strategies (audit trail, peer debriefing, reflexivity, and participant reflections) strengthen the credibility and transparency of the analysis.

Limitations should be considered. First, as a qualitative study conducted in a single region of Saudi Arabia, the findings are context-specific and not intended to be statistically generalisable. The cultural, medico-legal, and workforce-related specificities of the Saudi Arabian context, including familial decision-making norms, Islamic orientations toward life preservation, and the predominantly expatriate nursing workforce, should be considered when interpreting and transferring the findings to other settings. Second, self-selection may have favoured participation by nurses with salient experiences of moral distress, potentially shaping the range of accounts captured. Third, although confidentiality protections were emphasised, social desirability bias cannot be fully excluded. Finally, the study reports nurses’ perspectives; triangulation with physicians’ or parents’ accounts and with local policy documents could further enrich understanding of the decision ecology.

Recommendations and Future Research

Based on these findings, NICU leadership may consider piloting structured interprofessional debriefing models and formalised ethics deliberation processes that explicitly invite and operationalise nursing participation. Future research could test and refine the proposed interpretive model using mixed-methods or quantitative approaches, examine the prevalence of identified constraints in larger samples, and evaluate interventions designed to strengthen moral agency and reduce moral residue. Comparative qualitative studies across cultural and organisational contexts would further clarify which constraints reflect shared features of NICU work and which are context-specific.