--- title: "Artificial Intelligence in healthcare administration: A case study shows best practice results for implementation and application" language: "de" type: "post" original_url: "https://www.atlas-digitale-gesundheitswirtschaft.de/blog/2023/10/22/artificial-intelligence-in-healthcare-administration/" human_version: "../../../../../mensch/blog/2023/10/22/artificial-intelligence-in-healthcare-administration/" date: "2023-10-22" section: "Science Digest" categories: ["Automatisierung von Prozessen", "KI", "Krankenhaus", "Machine Learning"] author: "Jan-Patrik Novoa Lill" reading_time: "8 Min." description: "A significant potential of AI in healthcare lies in standardizing administrative tasks. How would a fully AI-supported hospital operate?" publisher: "Lehrstuhl für Management und Innovation im Gesundheitswesen, Universität Witten/Herdecke" --- # Artificial Intelligence in healthcare administration: A case study shows best practice results for implementation and application *22.10.2023 · Science Digest · Von Jan-Patrik Novoa Lill* AI’s impact on the workplace has been a topic of hot debate since its inception. Already, AI’s real-world applications in healthcare encompass various areas, ranging from predictive medicine to patient diagnostics and biomedical research [1, 2]. However, significant potential also lies in standardizing and automating administrative tasks. Considering this, how would a fully AI-supported hospital operate, and what implications would it have for managers and directors in healthcare? ### Humber River Hospital [Dicuonzo et al. (2023)](https://econpapers.repec.org/article/eeetechno/v_3a120_3ay_3a2023_3ai_3ac_3as0166497222000578.htm) explored this by studying the Humber River Hospital in Toronto, which opened in 2015 as North America’s first fully digitalized hospital [3]. The hospital serves a vast population and is fully equipped with integrated AI solutions, focusing on four main areas: Digitization, Communication, Patient Empowerment, and Overall System Automation. Semi-structured interviews with the organization’s leadership responsible for the transformation were held for the case study. To structure these interviews, a framework developed in 2002 for the assessment of health technologies [4] was used. The framework functions as a conceptual tool for decision-making about health technologies by considering population at risk, population impact, economic concerns, social context, and technology assessment information. Finally, newspaper articles documenting the digitalization process of the hospital and summary reports on current project results helped to confirm results from the case study and thereby increase its reliability and validity. ### Implementing AI-lead governance systems Based on the interviews, the researchers concluded that the implementation of a command center, using data and AI to support processes with 48h predictions on vital parameters such as bed availability, medical gas flows, care pathways, emergency room crowding, workload in operating rooms, etc., significantly enhanced care quality. They note a substantial reduction in medication errors down to 0.0001%, a more favorable nurse-to-patient ratio at 1-5, and a notable increase by a factor of 2 in nurses‘ available time to spend with patients [3]. Secondly, they found that to leverage AI’s benefits, healthcare institutions need a comprehensive business transformation. This involves updating administrative processes and the skills required from staff. Many healthcare professionals lack technological expertise, creating the need to outsource to third parties. The interview partners stated that this often results in solutions that are misaligned to internal strategic objectives or even fail to address any existing problems at all. Therefore, Humber River Hospital stands out, having in-house experts who guide external partners effectively [3]. ### Limitations It should be noted that the study has certain design limitations. Single case studies are generally considered less reliable with respect to replicability and can hardly serve as a basis for generalizable insights. In this case, this is further exacerbated by the fact that the interviews, containing 19 questions each, were only held with 5 people. Furthermore, all of these individuals were in leadership positions for the transformation, making biased results, due to the lack of interviewees affected by the transformation, likely. Finally, the often rather “free-form” approach to such studies can be prone to further bias due to the researchers‘ own expectations. ### Conclusion With all this in mind, the results should be seen as a first exploratory step for both research and professional applications towards understanding how AI might best be implemented to support effective and efficient administrative processes in healthcare. The case study suggests that key difficulties in this implementation process are the collaboration with external suppliers of AI applications due to a lack of in-house technological expertise, and ensuring that provided solutions fit the organization at hand. It is concluded that to achieve the business transformation required for the successful implementation of AI, healthcare leaders must prioritize technological proficiency. They will need tech-savvy professionals and should offer technology training and upskilling opportunities for staff across all levels of governance. Only with an appropriate blend of technical, strategic, and medical expertise can administrative and organizational problems be reliably identified, and appropriate solutions be designed. While outsourcing may provide a more cost-effective alternative, this effectiveness can quickly come at the cost of misaligned solutions. At Humber River Hospital, the case study shows how a staff-led implementation of technological innovation has allowed for significant increases in the quality of in-person care. With this in mind, it stands to reason that it can provide a best practice approach for appropriate, full-scale implementations of AI in healthcare. However, the digital upskilling process is a narrow tightrope to balance, given that some research suggests a decline in some health professionals‘ doctor-patient communications, examination skills, and clinical knowledge due to overreliance on technological solutions [5, 6]. Therefore, it is of utmost importance not to get carried away by the excitement over the potential of AI in clinical contexts and ensure strategic alignment of digital solutions. In other words, let machines do what machines do best, and let people do what people do best. References [1] Secinaro S, Calandra D, Secinaro A, Muthurangu V, Biancone P. The role of artificial intelligence in healthcare: a structured literature review. BMC Medical Informatics and Decision Making. 2021. DOI: 10.1186/s12911-021-01488-9 [2] Rong G, Mendez A, Assi E B. Zhao B, Sawan M. Artificial intelligence in healthcare: Review and Prediction case studies. Engineering. 2020. DOI: 10.1016/j.eng.2019.08.015 [3] Dicuonzo G, Francesca D, Fusco A, Shini M. Healthcare system: Moving forward with artificial intelligence. Technovation. 2023 [DOI: 10.1016/j.technovation.2022.102510](https://econpapers.repec.org/article/eeetechno/v_3a120_3ay_3a2023_3ai_3ac_3as0166497222000578.htm) [4] Kazanjian A, Green C. Beyond effectiveness: the evaluation of information systems using a comprehensive health technology assessment framework. 2002. Computers in Biology and Medicine. DOI: 10.1016/S0010-4825(02)00013-6 [5] Lu J, Will medical technology deskill doctors? 2016. International Education Studies. DOI: 10.5539/ies.v9n7p130 [6] Hoff T. Deskilling and adaption among primary care physicians using two work innovations. 2011. Health Care Manage Review. DOI: 10.1097/HMR.0b013e31821826a1 #### Verwandte Artikel - [KI-Unterstützung im Mammographie Screening](../../../2021/09/21/ki-unterstuetzung-im-mammographie-screening.md) - [KI zur Steigerung des In-Vitro-Fertilisation Erfolgs](../../../2022/07/18/ki-zur-steigerung-des-in-vitro-fertilisation-erfolgs.md) - [Prädiktion von Komplikationen nach der OP im Krankenhaus](../../../2021/06/07/praediktion-von-komplikationen-nach-der-op-im-krankenhaus.md) **Kategorien:** Automatisierung von Prozessen, KI, Krankenhaus, Machine Learning ## Bilder auf dieser Seite - shutterstock_620211089-min-1320x792.jpg: ../../../../../mensch/assets/bilder/wp-content/uploads/2023/10/shutterstock_620211089-min-1320x792.jpg --- Diese Seite ist für KI-Systeme optimiert. Version für Menschen: [Artificial Intelligence in healthcare administration: A case study shows best practice results for implementation and application](../../../../../mensch/blog/2023/10/22/artificial-intelligence-in-healthcare-administration/)