Clinical Decision-Making Among Community Nurses: A Comparative Human and Generative AI-Assisted Qualitative Analysis

To explore how community nurses describe clinical decision-making across simulated cases and to examine how themes and quotations from human thematic analysis compare with those from generative artificial intelligence-assisted analysis.

Qualitative descriptive study using simulated clinical cases and semi-structured debriefing interviews.

Between May and July 2024, 68 academic nurses with at least 1 year of community nursing experience were recruited in Israel. Each nurse responded to four simulated cases and a debriefing interview. Three researchers analysed transcripts thematically. Separately, Claude Projects (Claude 3.5 Sonnet) analysed the same de-identified Hebrew corpus without access to the human themes. The researchers then compared corresponding themes and illustrative quotations.

Six themes described nurses' decision-making: perceived control over treatment management; challenges in responding to scenarios; consultation with professional colleagues; use of non-human support resources; training and experience and promotion of patient health and well-being. Both analyses identified corresponding themes and sometimes selected identical or related quotations. The comparison was not a test of agreement, reliability or validity.

Community nurses described decision-making as a context-dependent process shaped by experience, perceived control, consultation, protocols and a holistic orientation to health promotion. Generative artificial intelligence may assist in organizing qualitative material but requires transparent procedures and human oversight. The study does not establish that AI can reproduce or substitute for human qualitative interpretation.

Generative AI tools must remain subordinate to professional nursing judgement in community practice; AI-assisted qualitative analysis requires transparent, human-verified procedures.

The findings characterize nurses' accounts of decision-making in simulated community cases. The study did not evaluate clinical effectiveness, implementation or outcomes.

The study is reported in accordance with the Consolidated Criteria for Reporting Qualitative Research.

No patient or public contribution. Community nurses participated as research participants.