AI May Allow Physicians To Regain Their Humanity

Emergency medicine (EM) is a team-based specialty, where a diverse group works together to rapidly deliver acute, unscheduled patient care. In contrast to traditional teams that have the luxury of time for their members to build rapport, teams in the emergency department (ED) change every day.

Harvard professor Amy C. Edmondson coined the word “teaming” to refer to the dynamic coordination of individuals fluidly adapting to a constantly changing environment and working together without a fixed structure.1,2 As Edmondson described, ED teams disband almost as quickly as they are assembled. Her use of “teaming” as a verb aptly captures the fast-paced, flexible nature of teamwork in the ED. Crucial to the success of teaming is the ability for technology to stay as dynamic and fluid as the members involved.

Although technology like electronic health records (EHRs) were initially introduced as tools for improving team communication and streamlining information sharing, they have instead often isolated ED teams, siloing clinicians, who are hunched over keyboards rather than communicating at the bedside.3

Now, we find ourselves at the forefront of a new disruptive technology: artificial intelligence (AI).

AI stands to transform ED workflows, offering potential efficiencies that could enhance care delivery in this unpredictable environment. When designed with the pillars of teaming in mind (workload management, interdisciplinary problem solving, and psychological safety), the implementation of AI tools could significantly improve ED care delivery and may allow teams to function better.4    

Can AI Help with Teaming?

AI started to make its mark in the ED by maximizing the efficiency of certain tasks like patient triage, clinical presentation diagnosis, mortality prediction, clinical decision making, and operational workflow management (see Figure 1).5–10 These tools expedite administrative tasks such as recording patient encounters, facilitating quick retrieval of EHR data, and identifying incidental findings on imaging; they even play a role in shared decision making.11-14 As AI automates these routine tasks, physicians may be able to leave the computer screen and re-engage on the floor with the ED team.

Lessons Learned from EHRs

Although EHRs enabled streamlined administrative tasks, they also eroded aspects of team cohesion.15,16 Reduced face-to-face interaction led to decreased psychological safety and trust, and flaws in EHR design hindered usability, leading to frustration and errors like selecting the wrong patient.17,18 Rather than receiving the right information at the right time, emergency physicians now spend inordinate amounts of time sorting through often irrelevant information.

EHRs were meant to improve team dynamics by increasing transparency between clinicians and patients, reducing physician administrative burden, and improving health care quality with fewer medical errors and less paperwork.19 Instead, we have seen the rise of physician burnout, correlating with the widespread use of EHRs.20 EHR technology has consistently worsened physician professional satisfaction, with poor intuitive user design, time-consuming data entry, and decreased face-to-face patient care.21

As AI spreads across the health care sector, avoiding the errors made with EHRs with be critical. With strategic implementation, AI may overcome these challenges, enhancing team cohesion by reducing administrative burdens and creating more opportunities for clinicians to engage with one another and their patients.

How Can AI Benefit Teaming?

There is no question that AI is going to make its way into the ED; in many departments, AI is already present. For AI to transform how ED teams function, these tools must enhance Edmondson’s key pillars of teaming: workload management, interdisciplinary problem solving, and psychological safety.

AI can assist with clinician workload management by allowing all ED team members to direct more energy in direct patient care. As mentioned, AI has tremendous potential for automating non-teaming tasks such as documentation and data retrieval, and for managing ED workflow processes. These changes may free clinicians to engage in meaningful interactions with patients and among themselves, ultimately improving cross-functional teaming in the ED.

AI’s potential to analyze large amounts of patient data for clinician review will also help teams make high-stakes decisions for patient care using up-to-date information. For instance, mortality predictor tools can help direct resource allocation and monitoring of patients in EDs with limited resources. AI emergency radiology tools can lead to immediate results for high-acuity patients, reducing delay in initiating treatments for acute patients. The transformative effect from AI’s assistance will augment communication within ED teams as they anticipate and prepare for complex tasks and work together directly to address high-acuity cases. This unified, data-supported approach may enhance patient outcomes and strengthen team trust and communication for a smoother, more cohesive environment.

Psychological safety in a team describes an environment where team members feel free to express their thoughts openly and without fear of being penalized. As AI becomes more desirable in a high cognitive-load work environment, a teaming-focused implementation framework may bring the focus back to the patient, supporting a culture that prioritizes open discussion and shared decision making. With additional time and less cognitive load on all clinicians, ED team members may engage in more face-to-face communication, both with each other and with patients.

As we approach an AI-driven transformation, we encourage ED teams to adopt these tools thoughtfully, using them only if they genuinely support patient care in the right place, at the right time. Although the tools may evolve, our mission remains unchanged—the care of the patient.

Dr. Peabody is the director of the UCSF Acute Care Innovation Center.

Dr. Gailloud is a PGY3 resident at the George Washington University Hospital Emergency Medicine Residency program.

Mr. Obra is a third-year medical student working at the UCSF Acute Care Innovation Center.

References

1. 
   Edmondson AC. The importance of teaming. Harvard Business School Working Knowledge. http://hbswk.hbs.edu/item/the-importance-of-teaming. Published April 25, 2012. Accessed September 19, 2024.
    
2. Platts-Mills TF, Nagurney JM, Melnick ER. Tolerance of uncertainty and the practice of emergency medicine. Ann Emerg Med. 2020;75(6):715-720.
    
3. Tsai CH, Eghdam A, Davoody N, et al. Effects of electronic health record implementation and barriers to adoption and use: a scoping review and qualitative analysis of the content. Life (Basel). 2020;10(12):327.
    
4. Edmondson AC. Harvard Business Review. Teamwork on the fly. https://hbr.org/2012/04/teamwork-on-the-fly-2. Published April 2012. Accessed December 20, 2024.
    
5. Kirubarajan A, Taher A, Khan S, et al. Artificial intelligence in emergency medicine: a scoping review. J Am Coll Emerg Physicians Open. 2020;1(6):1691-1702.
    
6. Hinson JS, Taylor RA, Venkatesh A, et al. Accelerated chest pain treatment with artificial intelligence–informed, risk-driven triage. JAMA Intern Med. 2024;184(9):1125-1127.
    
7. McLouth J, Elstrott S, Chaibi Y, et al. Validation of a deep learning tool in the detection of intracranial hemorrhage and large vessel occlusion. Front Neurol. 2021;12:656112.
    
8. Chenais G, Lagarde E, Gil-Jardiné C. Artificial intelligence in emergency medicine: viewpoint of current applications and foreseeable opportunities and challenges. J Med Internet Res. 2023;25:e40031.
    
9. Gallo RJ, Shieh L, Smith M, et al. Effectiveness of an artificial intelligence–enabled intervention for detecting clinical deterioration. JAMA Intern Med. 2024;184(5):557-562.
    
10. Ehwerhemuepha L, Carlson K, Moog R, et al. Cerner real-world data (CRWD) – a de-identified multicenter electronic health records database. Data Brief. 2022;42:108120.
     
11. Abbasgholizadeh Rahimi S, Cwintal M, Huang Y, et al. Application of artificial intelligence in shared decision making: scoping review. JMIR Med Inform. 2022;10(8):e36199.
     
12. Fernandes M, Vieira SM, Leite F, et al. Clinical decision support systems for triage in the emergency department using intelligent systems: a review. Artif Intell Med. 2020;102:101762.
     
13. Cellina M, Cè M, Irmici G, et al. Artificial intelligence in emergency radiology: where are we going? Diagnostics (Basel). 2022;12(12):3223.
     
14. Leonard F, O’Sullivan D, Gilligan J, et al. Supporting clinical decision making in the emergency department for paediatric patients using machine learning: a scoping review protocol. PloS One. 2023;18(11):e0294231.
     
15. Amano A, Brown-Johnson CG, Winget M, et al. Perspectives on the intersection of electronic health records and health care team communication, function, and well-being. JAMA Netw Open. 2023;6(5):e2313178.
     
16. Ratwani R. Electronic health records and improved patient care: opportunities for applied psychology. Curr Dir Psychol Sci. 2017;26(4):359-365.
     
17. Ratwani RM, Benda NC, Hettinger AZ, et al. Electronic health record vendor adherence to usability certification requirements and testing standards. JAMA. 2015;314(10):1070-1071.
     
18. Adelman JS, Kalkut GE, Schechter CB, et al. Understanding and preventing wrong-patient electronic orders: a randomized controlled trial. J Am Med Inform Assoc. 2013;20(2):305-310.
     
19. Net Health. Electronic Health Records: A Comprehensive History of EHR Systems. https://www.nethealth.com/blog/the-history-of-electronic-health-records-ehrs/. Published September 16, 2021. Accessed November 10, 2024.
     
20. Budd J. Burnout related to electronic health record use in primary care. J Prim Care Community Health. 2023;14:21501319231166921.
     
21. Friedberg MW, Chen PG, Van Busum KRV, et al. Factors affecting physician professional satisfaction and their implications for patient care, health systems, and health policy. Rand Health Q. 2014;3(4):1.