Artificial Intelligence for Medication Safety and Nursing Documentation in Emergency Departments: A Bibliometric Analysis
American Journal of Interdisciplinary Studies · 01.10.2024 · 14 (10) · 01-34
DOI: 10.63125/2krdb618 ↗Kurz gefasst
Fachartikel in der Zeitschrift American Journal of Interdisciplinary Studies, online erschienen am 01.10.2024. Verfasst von Md Mohsin Bhuyain.
Themen: Künstliche Intelligenz in der Pflege, Pflegeprozess und Dokumentation, Medikationssicherheit
Abstract (Originalsprache)
Emergency departments face medication-safety risks and incomplete nursing records, yet evidence concerning artificial intelligence (AI) remains fragmented across technologies, clinical settings, and implementation stages. This study examines whether AI supports medication-risk identification, nursing documentation quality, and information continuity, while identifying barriers to adoption. Guided by Sociotechnical Systems Theory, a qualitative, cross-sectional, literature-based multiple-case design integrates descriptive bibliometric indicators, thematic evidence synthesis, and comparative clinical application analysis. The intended corpus targets 96 peer-reviewed publications, although screening and extraction remain incomplete; four documentary cases address medication risk detection, medication reconciliation, AI-assisted documentation, and integrated clinical decision support. The underlying cases concern emergency and related hospital environments rather than cloud enterprises, and published estimates retain separate denominators, populations, and technologies. Key variables include prescription-screening discrimination, alert relevance, order modification, documentation coherence, clinical information continuity, professional acceptance, interoperability, and workflow compatibility. Analysis uses structured literature extraction, descriptive frequencies, source-specific performance measures, cross-case comparison, and qualitative assessment of six hypotheses; no original patient-level regression or pooled effectiveness estimates were produced. Previous medication-AI reviews found that 73 of 78 studies (93.6%) evaluated technical performance, whereas a nursing-AI review identified only nine implementation studies among 93 (9.7%) and one operational evaluation (1.1%). Hospital-based machine-learning prescription screening achieved AUROC 0.81 versus 0.65 for conventional decision support. In another investigation, 85% of alerts were clinically valid and 43% were followed by medication-order changes. A documentation prototype produced coherent, appropriately headed text in approximately 69% of generated paragraphs and reduced paragraph fragmentation by 23%. An emergency clinician evaluation involving 14 physicians reported perceived usefulness of 4.07/5, but task-speed benefit of 3.14/5. These results support technical potential but do not establish reduced emergency medication errors, improved nursing workload, or effectiveness of combined systems. Hospitals should prioritize nurse and pharmacist verification, interoperable records, targeted alerts, and prospective clinical validation before routine implementation.
Zitieren
APA
Bhuyain, M. M. (2024). Artificial Intelligence for Medication Safety and Nursing Documentation in Emergency Departments: A Bibliometric Analysis. American Journal of Interdisciplinary Studies, 14(10), 01-34. https://doi.org/10.63125/2krdb618
Vancouver
Bhuyain MM. Artificial Intelligence for Medication Safety and Nursing Documentation in Emergency Departments: A Bibliometric Analysis. American Journal of Interdisciplinary Studies. 2024;14(10):01-34. doi:10.63125/2krdb618
BibTeX
@article{Bhuyain2024,
author = {Bhuyain, Md Mohsin},
title = {Artificial Intelligence for Medication Safety and Nursing Documentation in Emergency Departments: A Bibliometric Analysis},
journal = {American Journal of Interdisciplinary Studies},
year = {2024},
volume = {14},
number = {10},
pages = {01--34},
doi = {10.63125/2krdb618}
}
EUVIMED