Weaving AI/ML into health.
Ensuring AI plays a helpful
role in healthcare
One of the largest global pharmaceutical companies was concerned that unregulated AI applications may be introducing unwanted biases into patient care. To remedy this situation, they asked us to convene a multi-stakeholder group to develop quality standards for clinical decision support systems.
- Identified unregulated areas within the FDA regulatory framework for AI on which the quality standards should focus.
- Led a cross-functional workshop with the client to confirm the objectives, scope and desired outcomes of the work, given that other areas of the business benefited from flexible regulation of AI applications.
- Recruited senior healthcare leaders including Robert Califf (while at Verily) along with senior leaders from the AMA, Mayo Clinic and Intermountain Healthcare, among others.
- Developed and refined a set of CDS Quality Principles that the group adopted during a CDS Quality Summit that we designed and chaired.
The CDS Quality Principles were widely shared, with John Halamka and Paul Cerrato of the Mayo Clinic noting that they exemplify the AI ecosystem’s gradual transition “from the ‘Wild West’ into a set of well-defined and repeatable processes that healthcare stakeholders can trust.
CDS Quality Principles