Scholarlink

Objectives <p>This study examined why artificial intelligence (AI)-based clinical decision support tools have had limited clinical translation in the emergency department (ED) and identified barriers and facilitators to their development and implementation.</p> Methods <p>We conducted a qualitative study involving semi-structured interviews with researchers who have expertise developing and implementing AI clinical decision support tools for use in the&#xa0;ED. We used purposive and snowball sampling to identify participants. We used platform-based AI transcription and anonymized transcripts manually. Using grounded theory framework, two coders iteratively analyzed transcripts in three stages (initial, focused, and theoretical) to identify barriers, facilitators, and themes. We adhered to SRQR and COREQ guidelines.</p> Results <p>We achieved data saturation after ten interviews conducted between October 15, 2024 and March 22, 2025. Participants ranged across a variety of medical and academic professions. We identified eight themes pertaining to developing and implementing AI clinical decision support in the ED, in descending frequency: team capacity; data infrastructure; defining the clinical problem and solution; research, ethics, and regulatory approval; legal and liability; model building and performance; time; and cost. We identified “engaging multiple healthcare end-users” and “sharing resources with other departments” as the highest yield facilitators.</p> Conclusion <p>Successful implementation of AI clinical decision support tools in the ED requires a clear clinician- and patient-defined problem, robust data infrastructure, and a diverse research team able to navigate challenges with regulatory, legal, and financial challenges over a long timeline. Anticipating barriers and leveraging facilitators early in the development process may increase the likelihood of successful implementation.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Barriers and facilitators to developing and implementing artificial intelligence-based clinical decision support in the emergency department: a qualitative study

  • 期刊论文

  • 发表日期：

    2026年6月22日

查看全文
  • Hashim Kareemi,
  • Alex Colak,
  • Krishan Yadav,
  • Wojtek Michalowski,
  • Christian Vaillancourt

摘要

Objectives

This study examined why artificial intelligence (AI)-based clinical decision support tools have had limited clinical translation in the emergency department (ED) and identified barriers and facilitators to their development and implementation.

Methods

We conducted a qualitative study involving semi-structured interviews with researchers who have expertise developing and implementing AI clinical decision support tools for use in the ED. We used purposive and snowball sampling to identify participants. We used platform-based AI transcription and anonymized transcripts manually. Using grounded theory framework, two coders iteratively analyzed transcripts in three stages (initial, focused, and theoretical) to identify barriers, facilitators, and themes. We adhered to SRQR and COREQ guidelines.

Results

We achieved data saturation after ten interviews conducted between October 15, 2024 and March 22, 2025. Participants ranged across a variety of medical and academic professions. We identified eight themes pertaining to developing and implementing AI clinical decision support in the ED, in descending frequency: team capacity; data infrastructure; defining the clinical problem and solution; research, ethics, and regulatory approval; legal and liability; model building and performance; time; and cost. We identified “engaging multiple healthcare end-users” and “sharing resources with other departments” as the highest yield facilitators.

Conclusion

Successful implementation of AI clinical decision support tools in the ED requires a clear clinician- and patient-defined problem, robust data infrastructure, and a diverse research team able to navigate challenges with regulatory, legal, and financial challenges over a long timeline. Anticipating barriers and leveraging facilitators early in the development process may increase the likelihood of successful implementation.