Presentation
Proposing a change management model for Artificial Intelligence technology implementation in the operating room
SessionOptimize Your Telehealth
Event Type
Oral Presentations
TimeWednesday, June 8th3:30pm - 4:00pm EDT
Location
DescriptionTopic: This presentation proposes a new change management model for artificial intelligence technology implementation in the operating room (OR). As change management is a crucial part of the implementation process and change management efforts were sustained throughout all phases of the implementation, the presentation describes the implementation process and the change management strategies step by step.
Application: The change management model created for our quaternary academic hospital system can serve as a roadmap for navigating other similar processes and as a base model for customization to fit individual institution and project’s characteristics. The experience garnered and the change management model created have a high potential for transformative innovation and practice impact in patient morbidity and mortality improvement as well as surgical care improvement through enhanced and objective analysis of intra-operative performance.
Background: Artificial intelligence (AI) technology and applications are emerging in the operating room (OR) for surgical care improvement and surgical teamwork enhancement. An example of this is OR Blackbox, an AI-powered surgical analytical tool to visualize and aid in understanding and improving OR safety, performance, and efficiency (e.g., Jung et al., 2020; van Dalen et al., 2021). A few hospitals around the world have implemented the OR Blackbox (SST, Toronto). Some general feedback gained from the experiences of other hospitals who have implemented it, are that OR Blackbox is excellent in analyzing teamwork, and in turn in improving surgical staff’s non-technical skills, increasing OR efficiency and ultimately patient outcome (Boet et al., 2021). The OR Blackbox can synchronize audio and video inputs with other continuous or discrete intraoperative data such as room environment (ex. lighting), patient vitals and device data and to automate data analysis, which traditional surgical data recording technologies are not capable of (van Dalen et al., 2021). However, this new technology brings with it some implementation challenges.
Implementation is challenging as healthcare systems are complex adaptive systems (IOM, 2001) in general, which exemplify complexity, unpredictability, and varying characteristics especially during change processes (Braithwaite et al., 2018). In addition, hospital environments can be different in hospital organizational culture (Seren and Baykal, 2007), funding sources, and staff salary structure, making it challenging to clone the change management experience from elsewhere. The complexity of the healthcare systems, interacting with the novelty of the technology and patient privacy protection requirements, require significant change management strategies to acquire staff engagement and buy-in. These challenges are not adequately reflected and addressed in the several widely accepted change management models. Additionally, there is not a generalized enough change management model that is applicable for a variety of healthcare institutions yet specific enough to serve as a practical guide.
The authors share their experience leading the implementation of the OR Blackbox technology in three ORs at a large quaternary academic hospital system using a predictive and agile hybrid change management strategic plan, addressing the generic and institution-specific challenges faced during the technology implementation, especially during the change management, which permeated the implementation. The authors propose a customized change management model for reference for future AI technology implementations in the OR. The implementation steps and strategies applied in each step are specified below.
Step 1: Project Team Assembly. The core project team formed first with surgeon proponents from the Department of Surgery, administrative leaders, and research experts from the Kern Center for the Science of Health Care Delivery (especially ergonomists). Other team members were brought in when the ongoing project phases required their expertise.
Strategy 1: Assemble project team with necessary expertise. Non-core team members were dispersed as the phases that they contribute to were complete.
Step 2: Committee Approvals. Project team navigated a non-traditional hybrid practice and research approval process as the novel AI technology and its applications were relevant to both practice and research. Approvals from various committees and subcommittees were sought and granted.
Strategy 2: Make effort to understand and predict the challenges the project may face. The project team recognized the novelty of the technology and mapped a unique approval process – a hybrid of practice and research processes accordingly. The team predicted approval committees’ concerns regarding data security and confidentiality and therefore selected a core team member who is a clinical faculty from the project’s sponsoring department to deliver numerous presentations to committees to clarify the capability of the system (Strategy 5). Possible resistance to audio and video recording from staff was also accounted for (Guerlain et al., 2004) and minimized by deliberate socialization and consistent communication among project team, clinical leadership and OR staff (Strategy 6).
Strategy 3: Add agility to project planning and execution. Agile mindset was adopted; and ample time was allowed when the team was navigating the unconventional committee approval processes.
Strategy 4: Cater to institutional culture and regulations. The team aligned the project with the institution’s priorities and mission and presented the alignment properly in order to gain leadership approval and buy-in.
Step 3: Change Management. Change management efforts started at the beginning of the project and sustained throughout due to its importance in gaining early and consistent stakeholder engagement and buy-in.
Strategy 5: Early stakeholder buy-in and engagement. The early involvement, visibility, and commitment of senior leaders to the project sent a strong message across the enterprise that they would actively support this change with employees. This strengthened the leadership coalition’s visibility in conveying the vision and goals and in setting priorities for the change.
Strategy 6: Consistent communication. Information regarding the OR Blackbox project was controlled and only circulated among project team members during planning phase. It was then shared among approval committees before communicated to relevant staff leaders. Lastly, it was disseminated uniformly and consistently to OR staff members and other employees of the institution via the project FAQ and a consistent message. The FAQ contains knowledge regarding various aspects of the OR Blackbox technology and the implementation project on the Intranet. The consistent message was a one-page compelling elevator speech of the project, sent by managers to relevant OR staff
Step 4: System installation and go-live. Equipment was purchased and installed upon committee approvals. Data collection and analysis started and continued 24 hours a day and 7 days a week as soon as the OR Blackbox system went live.
This change management effort was deemed successful by project team members, stakeholders and OR staff users. The project stayed on budget, on time despite the ongoing pandemic, and the team has received mostly positive user feedback. This project is an example of leveraging the strengths of the Department of Surgery and the Kern Center for the Science of Health Care Delivery, relevant committees in both the research and practice endeavors of the institution, which can be applicable to other institutions. This type of partnership paves the way for future implementation and assessment of the impact of innovation and practice changes, especially those AI-enabled.
Overview of Presentation: After an overview of audio-visual recording in the OR and automated intraoperative activities and data analysis using AI, the benefits and challenges of implementing an AI-powered analytical tool in the OR will be presented. Implementation steps and strategies used throughout the implementation at our institution, and especially for the change management component of the implementation, will be discussed. Lastly, a new change management model for artificial intelligence technology implementation in the operating room will be presented.
Application: The change management model created for our quaternary academic hospital system can serve as a roadmap for navigating other similar processes and as a base model for customization to fit individual institution and project’s characteristics. The experience garnered and the change management model created have a high potential for transformative innovation and practice impact in patient morbidity and mortality improvement as well as surgical care improvement through enhanced and objective analysis of intra-operative performance.
Background: Artificial intelligence (AI) technology and applications are emerging in the operating room (OR) for surgical care improvement and surgical teamwork enhancement. An example of this is OR Blackbox, an AI-powered surgical analytical tool to visualize and aid in understanding and improving OR safety, performance, and efficiency (e.g., Jung et al., 2020; van Dalen et al., 2021). A few hospitals around the world have implemented the OR Blackbox (SST, Toronto). Some general feedback gained from the experiences of other hospitals who have implemented it, are that OR Blackbox is excellent in analyzing teamwork, and in turn in improving surgical staff’s non-technical skills, increasing OR efficiency and ultimately patient outcome (Boet et al., 2021). The OR Blackbox can synchronize audio and video inputs with other continuous or discrete intraoperative data such as room environment (ex. lighting), patient vitals and device data and to automate data analysis, which traditional surgical data recording technologies are not capable of (van Dalen et al., 2021). However, this new technology brings with it some implementation challenges.
Implementation is challenging as healthcare systems are complex adaptive systems (IOM, 2001) in general, which exemplify complexity, unpredictability, and varying characteristics especially during change processes (Braithwaite et al., 2018). In addition, hospital environments can be different in hospital organizational culture (Seren and Baykal, 2007), funding sources, and staff salary structure, making it challenging to clone the change management experience from elsewhere. The complexity of the healthcare systems, interacting with the novelty of the technology and patient privacy protection requirements, require significant change management strategies to acquire staff engagement and buy-in. These challenges are not adequately reflected and addressed in the several widely accepted change management models. Additionally, there is not a generalized enough change management model that is applicable for a variety of healthcare institutions yet specific enough to serve as a practical guide.
The authors share their experience leading the implementation of the OR Blackbox technology in three ORs at a large quaternary academic hospital system using a predictive and agile hybrid change management strategic plan, addressing the generic and institution-specific challenges faced during the technology implementation, especially during the change management, which permeated the implementation. The authors propose a customized change management model for reference for future AI technology implementations in the OR. The implementation steps and strategies applied in each step are specified below.
Step 1: Project Team Assembly. The core project team formed first with surgeon proponents from the Department of Surgery, administrative leaders, and research experts from the Kern Center for the Science of Health Care Delivery (especially ergonomists). Other team members were brought in when the ongoing project phases required their expertise.
Strategy 1: Assemble project team with necessary expertise. Non-core team members were dispersed as the phases that they contribute to were complete.
Step 2: Committee Approvals. Project team navigated a non-traditional hybrid practice and research approval process as the novel AI technology and its applications were relevant to both practice and research. Approvals from various committees and subcommittees were sought and granted.
Strategy 2: Make effort to understand and predict the challenges the project may face. The project team recognized the novelty of the technology and mapped a unique approval process – a hybrid of practice and research processes accordingly. The team predicted approval committees’ concerns regarding data security and confidentiality and therefore selected a core team member who is a clinical faculty from the project’s sponsoring department to deliver numerous presentations to committees to clarify the capability of the system (Strategy 5). Possible resistance to audio and video recording from staff was also accounted for (Guerlain et al., 2004) and minimized by deliberate socialization and consistent communication among project team, clinical leadership and OR staff (Strategy 6).
Strategy 3: Add agility to project planning and execution. Agile mindset was adopted; and ample time was allowed when the team was navigating the unconventional committee approval processes.
Strategy 4: Cater to institutional culture and regulations. The team aligned the project with the institution’s priorities and mission and presented the alignment properly in order to gain leadership approval and buy-in.
Step 3: Change Management. Change management efforts started at the beginning of the project and sustained throughout due to its importance in gaining early and consistent stakeholder engagement and buy-in.
Strategy 5: Early stakeholder buy-in and engagement. The early involvement, visibility, and commitment of senior leaders to the project sent a strong message across the enterprise that they would actively support this change with employees. This strengthened the leadership coalition’s visibility in conveying the vision and goals and in setting priorities for the change.
Strategy 6: Consistent communication. Information regarding the OR Blackbox project was controlled and only circulated among project team members during planning phase. It was then shared among approval committees before communicated to relevant staff leaders. Lastly, it was disseminated uniformly and consistently to OR staff members and other employees of the institution via the project FAQ and a consistent message. The FAQ contains knowledge regarding various aspects of the OR Blackbox technology and the implementation project on the Intranet. The consistent message was a one-page compelling elevator speech of the project, sent by managers to relevant OR staff
Step 4: System installation and go-live. Equipment was purchased and installed upon committee approvals. Data collection and analysis started and continued 24 hours a day and 7 days a week as soon as the OR Blackbox system went live.
This change management effort was deemed successful by project team members, stakeholders and OR staff users. The project stayed on budget, on time despite the ongoing pandemic, and the team has received mostly positive user feedback. This project is an example of leveraging the strengths of the Department of Surgery and the Kern Center for the Science of Health Care Delivery, relevant committees in both the research and practice endeavors of the institution, which can be applicable to other institutions. This type of partnership paves the way for future implementation and assessment of the impact of innovation and practice changes, especially those AI-enabled.
Overview of Presentation: After an overview of audio-visual recording in the OR and automated intraoperative activities and data analysis using AI, the benefits and challenges of implementing an AI-powered analytical tool in the OR will be presented. Implementation steps and strategies used throughout the implementation at our institution, and especially for the change management component of the implementation, will be discussed. Lastly, a new change management model for artificial intelligence technology implementation in the operating room will be presented.
Authors
Robert D. and Patricia E. Kern Director and Professor of Health Care Systems Engineering