Organizing a Competence Center at a Public Hospital to Increase Artificial Intelligence Impact - Key Learnings from a Comparative Study Across Six Large Organizations
Examensarbete för masterexamen
Management and economics of innovation (MPMEI), MSc
With the threat of overpowering healthcare demands, a significant funding gap and systematic failure to attend to the increasing portion of chronically ill in the near future, Artificial Intelligence (AI) is showing great promise as a possible remedy. Despite the national interest in AI’s potential, Sahlgrenska University Hospital (SUH) are facing uncertainties regarding how to build AI into their organization, just like many others are. To help with this, SUH is pursuing an AI competence center (AICC) to help accelerate AI usage. This study aims to provide SUH with actionable guidance on how to organize the new AICC and to contribute to the nascent and sparse academic field of organizing the implementation of AI in healthcare. To achieve this aim, a qualitative multiple-case study was carried out, with semi-structured interviews as the main mode of data gathering. Data extracted from 23 managerial interviewees across six organizations, SUH included, was analyzed through thematic groupings. Complementing this is an extensive literature review, ultimately synthesized into a theoretical framework around AI innovation, diffusion of innovation and healthcare culture. The findings indicate that there are five major dimensions of approaching and organizing AI-enabled innovation in large organizations. To begin with, AI is argued to be highly complex, with unclear value, impact and boundaries. Organizations have subsequently constructed different degrees of direction for their AI work. Rather ubiquitous however, is the understanding of the large change and buy-in required. In addition, the value of external collaboration and knowledge is widely recognized. Finally, the organizational structure around AI plays a multi-faceted role in accelerating AI-enabled innovation. Key considerations and learnings from the data and theoretical framework underbuilt complementary recommendations for how SUH can organize around the AICC. Firstly, it is recommended that the AICC should focus on enabling AI to overcome the operative dominance currently challenging AI usage in healthcare. Secondly, the AICC is recommended to prioritize administrative AI higher than clinical AI, as it is easier to accelerate while yielding large value. Lastly, it is recommended that the center should be placed near top management, have regional sanction, contain different competences and act as a boundary-spanner to accelerate AI at SUH. Practically, the conclusions are valuable as guidelines for SUH moving forward with AI, possibly contributing to improving future healthcare delivery through AI impact. Theoretically, this study contributes to the field of healthcare AI implementation by exploring the ways a large hospital can organize around AI acceleration. The study’s results are believed to address important aspects of accelerating AI in large organizations, applicable with caution to other university hospitals in particular but possibly other large organizations as well.
Artificial intelligence, innovation, healthcare, organize, accelerate, center