CIC Charter

Michigan Medicine  

Clinical Intelligence Committee Charter 

Mission  

The Michigan Medicine Clinical Intelligence Committee (CIC)’s purpose is to provide governance and support for artificial intelligence (AI) tools associated with clinical systems and hospital operations. Artificial intelligence (AI) is a set of technologies that includes machine learning, deep learning, natural language processing and related technologies. AI can be used for various tasks including identifying patterns and abnormalities, categorization, personalization, recognition, human interactions, autonomous systems, predictive analytics and decision support. 

The mission of the CIC is to provide a framework and support for responsible use of health AI tools through governance, evaluation, consultation, and education, with goals of implementing safe AI practice, encouraging user trust, and improving quality of care. 

The CIC seeks to systematically analyze, compare, and provide recommendations on the adoption, implementation, and prioritization of artificial intelligence that address strategic priorities with respect to both the clinical and research missions in Michigan Medicine. For decisions requiring additional funding or resources, the committee collaborates with and reports to the Michigan Medicine Information Technology Operations and Management (ITO&M) Committee.  ITO&M and the CIC are chartered by the UMHS IT Steering & Oversight Committee (ITSOC). 

Vision 

It is the committee’s responsibility to identify whether an AI application applied to a clinical or administrative setting is appropriate and necessary to achieve the following: 

• Enhance the quality of care to improve health outcomes 

• Improve identification of at-risk patients relative to current standard practice 

• Enhance team workflow efficiency, decrease cost, or increase patient satisfaction 

Scope of Responsibilities 

• Review and provide recommendations around implementation of artificial intelligence technology within Michigan Medicine 

• Endorse or defer AI technology implementations within Michigan Medicine 

• Align AI Technology initiatives to Michigan Medicine strategic priorities  

• Assure that all relevant literature and best practices are assembled to support requests and their subsequent analyses 

• Curate, organize and publish resources supporting best practices, policies and guidelines for AI technology in Michigan Medicine 

• Assure that effective evaluation, education, and implementation of resources are outlined within the context of the development team, operational leaders, and HITS support 

• Assemble ad hoc experts to discuss analytic and implementation considerations 

• Provide actionable and accurate information to the ITOM Committee 

Endorsements 

Before an AI technology seeks Health Information Technology Services (HITS) resources, or is deployed into an operational or clinical workflow, the sponsors of that AI technology should seek endorsement from the Clinical Intelligence Committee (CIC).  

What does a model owner gain through endorsement? 

A CIC endorsement demonstrates an AI technology adheres to guidelines and policies of the institution. Michigan Medicine HITS will look to the CIC for guidance before prioritizing and allocating resources to support AI technology. 

What aspects of AI technology are evaluated in an endorsement? 

The CIC ensures AI Technology proposals address a demonstrated need, has executive sponsorship, has technical ownership, does not introduce bias, and is well documented.   

What is NOT endorsed by the CIC? 

General IT systems without AI fall under other Michigan Medicine governance processes and do not need to seek endorsement by the CIC. Only applications and services that leverage AI technology around patient care and hospital operations should be brought to the CIC for endorsement.  

How does an AI technology gain endorsement?  

Endorsements are gained through a formal vote at a CIC meeting. Endorsements require unanimous approval by the Voting Members in attendance. For a decision to go to vote, representation from a minimum of 6 of the 9 voting MM departments/units must be present. 

What comes after endorsement? 

An endorsed AI technology may request resources and funding through Michigan Medicine governing bodies. Requests for assistance may follow one of three paths: ITO&M Demand, HITS Demand, or HITS Incident. The CIC will assist AI technology owners in identifying the best path forward to meet their needs. 

Committee Structure and Membership 

Committee Meetings 

The CIC will meet monthly for 1 – ½ hours.  Meetings are facilitated by the chairs.  Additional meetings may be scheduled to address special topics.   

Member Responsibilities 

• Attend monthly meetings for 1.5 hours. In addition, offline work as needed. 

• Any member absent for four meetings in one year will be considered for replacement. 

• It is the CIC member’s responsibility to inform the Chair if they are no longer interested in serving so their vacant position may be filled. 

• Core Members represent voting departments within Michigan Medicine.  Core members should ensure that a representative from their department or unit is present at each meeting for endorsements. 

• Non-Voting Members do not vote in endorsements; however, they can provide input in their areas of expertise.  

• Consulting Guests are invited to attend CIC meetings as needed. 

Core Members 

Representative 

Name 

Chair 

Co-Chair 

Coordinator 

Data Science Lead 

Mike Burns  

Michael Sheppard 

Jessica Virzi 

Sean Meyer 

Executive Sponsors 

Ranjit Aiyagari (CMIO) 

Carleen Penoza (CNIO) 

Physician Informatics 

Paul Grant (ACMIO) 

Richard Medlin (ACMIO) 

Paresh Patel (ACMIO) 

Michael Lanham (ACMIO) 

Chief Medical Office 

Vikas Parekh (ACMO) 

Office of Innovation 

Dana Habers (Chief Innovation Officer) 

Nursing Informatics 

Annemarie Eichberger 

Pathology Informatics 

Ulysses Balis (ACMIO) 

Jerome Cheng 

Radiology Informatics 

Ken Buckwalter (ACMIO) 

Ben Mervak 

Jess Fried 

Ethics 

Jodyn Platt  

Health Information Technology and Services 

Beth Bertels 

Steve Schnobrich 

Quality and Learning Health Systems 

Sandeep Vijan 

Justin Pestrue 

Non-Voting Members 

Representative 

Name 

HITS Clinical Revenue Cycle (CRC) 

Andrew Gutting 

Emily Weatherup 

Kristen Shekut 

Tamir Baisa 

Tony Eccleston 

HITS Enterprise Data Information Systems (EDIS) 

Beth Plotts 

HITS Academic Information Technology (AIT) 

Jen Pantoja 

DOCTR 

Erin Kaleba 

The Michigan Medicine Capacity Operations and Real Time Engagement Center (M2C2) 

Max Garifullin 

Jennifer Pardo 

Department of Learning Health Sciences (DLHS) 

Andrew Krumm 

John Donnelly 

Research 

Michael Sjoding 

Cornelius James 

Maxwell Spadafore 

Caleb Smith 

Pharmacy 

Dana Habers 

Karl Renius 

Consulting Guests 

Representative 

Name 

Office of Patient Experience 

Devin Lippert* 

Office of General Council 

Kara Morgenstern 

Michigan Medicine Accreditation  

Natalie Plata 

Michigan Medicine Compliance 

Lauren Shellenberger 

Nursing Translation and Consulting 

Leah Shever-Doletzky * 

Leader Responsibilities 

Co-Chairs 

• Attend and facilitate meetings.  

• Ensure the committee member’s work is consistent with its purpose. 

• Be familiar with all aspects of the committee’s work, including the purpose, prior work, current budget, and activities. 

• Serve as a communication link with other committees and groups. 

• Adjudicate CIC requests. 

• Ensure that all members are engaged to ensure adequate representation of the constituent groups. 

• Chair and Co-Chair understand and advise the group about: 

• Its purpose and functions 

• Related policies, procedures, processes, rules, and organizational activities 

• Project planning process and related deadlines 

• Anticipate the committee members’ information needs. 

• Seek guidance when information beyond the chairs’ expertise is required. 

• Provide recommendations to the executive committee (ITOM) as appropriate. 

• Evaluate member performance, provide feedback, perform remediation, and elevate committee members based on expected responsibilities to the CIC. 

Data Science Lead 

• Acquisition and analysis of data from clinical systems will be needed to ensure the success of the committee’s mission. 

• Quality Analytics, HITS Reporting, and Analytics groups will provide support in obtaining data and carrying out necessary analyses. 

CIC Coordinator 

• Coordinate monthly committee meetings. 

• Author and distribute committee meeting materials. 

• Facilitate committee votes and record results. 

• Ensure the committee charter is accurate and up to date. 

• Coordinate committee communications to Michigan Medicine departments, committee members, and stakeholders. 

CIC Request Workflow 

All requests go to the co-chairs, and it is the responsibility of one or both to do initial analyses by following the steps below: 

• Meet with the requester for a detailed understanding of the request and ensure its need. 

• Communicate with the right teams to ensure the requested technology or something very similar does not exist to prevent duplication. 

• Review the needed data and its availability and ensure access to clean, reliable data to test and validate the technology. 

• Determine if one of our vendors already has an appropriate/similar technology. 

• Present findings to the CIC members to arrive at a final decision (endorse vs. defer). 

• For new requests, CIC will determine whether a technology should be built and run in the background for the purposes of prospective validation. After a 6 month period of the model running in the background, CIC will determine whether the model is appropriate for implementation into a clinical workflow. 

• The CIC may, at its discretion, approve multiple similar models to be built to run in the background for the purposes of comparison, although preference will be given to a single model for implementation into a given clinical workflow. 

• If the tool requires additional funding or resources, request time on the agenda with the ITOM to share our committee’s recommendation and rationale. 

Prioritization of Requests 

CIC uses the HITS prioritization guidelines to prioritize requests.  

Project Classifications 

Highest Priority 

Institutional Driver 

Strategic Objective  

Critical Operations  

Non-Critical Operations 

Lowest Priority 

Models Developed for Research Purposes 

• CIC will consult with the ACMIO for Research, Dr. Paresh Patel, to determine the prioritization category (per above) applies to a given request. 

• CIC will coordinate with MiChart Research where appropriate. 

Organizational Structure