Artificial intelligence (AI) has the potential to greatly impact our daily lives, especially in the field of health care. Because of the ethical and legal challenges of artificial intelligence-driven health care, there is a need for guidance and oversight as AI technology is created.
In December 2023, OSF HealthCare joined 38 other health care payers and providers to sign the Healthcare AI Commitments. This was done in collaboration with the White House to create a framework for using AI safely and responsibly in health care.
Interested in health care AI?
“We are pleased to join such a forward-thinking group of organizations focused on the advancement of health care through AI,” said Bob Sehring, chief executive officer for OSF HealthCare. “We have a long history of innovation and are excited to leverage our expertise to help shape AI use in health care on a national level.”
The commitments as outlined on the website are:
- We commit to vigorously developing AI solutions to optimize health care delivery and payment by advancing health equity, expanding access, making health care more affordable, improving outcomes through more coordinated care, improving patient experience and reducing clinician burnout.
- We will work with our peers and partners to ensure outcomes are aligned with fair, appropriate, valid, effective and safe (FAVES) AI principles.
- We will deploy trust mechanisms that inform users if content is largely AI-generated and not reviewed or edited by a human.
- We will adhere to a risk management framework that includes comprehensive tracking of applications powered by frontier models and accounting for potential harms and steps to mitigate them.
- We will research, investigate and develop swiftly but will do so responsibly.
Commitment 1 – AI Development
The Biden administration believes we have a “once-in-a-generation” opportunity to develop AI technology that improves health care outcomes. OSF HealthCare has a deep history of health care innovation, including the use of AI to improve patient outcomes, access to care and clinical workflows.
In 2023, Jonathan Handler, MD, a clinical informaticist for OSF, participated in a panel discussion that was used to advise President Biden on how technology can support cancer patient navigation. Dr. Handler’s invitation to participate follows his published research studies on breast cancer screening and the use of AI in health care to improve workload efficiencies for cancer patient navigators.
Commitment 2 – FAVES Principles
The FAVES Principles were created for the transparency of AI tools used in certified electronic health record systems. These same principles can be applied to AI models that assist clinical decision-making.
- Fair – Outcomes of the model do not exhibit prejudice or favoritism toward an individual or group based on their inherent or acquired characteristics.
- Appropriate – Model and process outputs are well matched to produce results appropriate for specific contexts and populations to which they are applied.
- Valid – Model and process outputs have been shown to estimate targeted values accurately and as expected in internal and external data.
- Effective – Outcomes of the model have demonstrated benefit in real-world conditions.
- Safe – Outcomes of the model are free from any known unacceptable risks and for which the probable benefits outweigh any probable risk.
Commitment 3 – Transparency in the use of AI technologies
As AI becomes more widely used in health care, it’s important to create guidelines for informing patients about its use in their treatment. This includes addressing legal and ethical issues in artificial intelligence and developing clear communication strategies that help patients understand their rights.
As laid out in the commitments, “We commit to informing users if content is largely or exclusively AI-generated unless such content is edited or closely reviewed by a human before being shared with end users. We also commit to sharing relevant best practices with industry and the public to accelerate responsible AI development in the health sector.”
Commitment 4 – Risk management
For AI in health care, effective risk management is necessary. The federal government is working on guidance, and the National Association of Insurance Commissioners (NAIC) is developing a risk-based framework to be used by insurers.
While each organization has its own approach to risk management, those that have signed the Healthcare AI Commitments agree to follow a common framework that includes three key practices:
- Risk analysis – identifying potential risks and adverse impacts of AI models.
- Risk mitigation – taking steps to reduce identified risks, such as testing models before deployment and monitoring outcomes over time.
- Governance – establishing policies and controls for AI applications, including how data is handled and used.
Commitment 5 – Responsible development and innovation
This commitment highlights that the benefits of artificial intelligence in health care are increased when the results are easily accessible and collaborative. To accomplish this goal, these organizations have agreed to find new ways to use innovative technology. They will focus on rapidly creating prototypes in secure environments and following HIPAA regulations for privacy.
Continuous monitoring of outcomes is important to ensure quality and accuracy. Embracing shared AI models responsibly while managing risks promotes long-term innovation. Training programs are also essential to help developers and users deploy AI safely and securely, making it more accessible and user-friendly.
Keep up to date on the newest AI developments from OSF here.
Last Updated: July 29, 2024