- Reports: Healthcare Artificial Intelligence Market CAGR of 47%-50%
“AI is now poised to disrupt health care, with the potential to improve patient outcomes, reduce costs, and enhance work-life balance for healthcare providers,” said Duke-Margolis Deputy Director for Policy Greg Daniel.
“Integrating AI into healthcare safely and effectively will need to be a careful process, requiring policymakers and stakeholders to strike a balance between the essential work of safeguarding patients while ensuring that innovators have access to the tools they need to succeed in making products that improve the public health,” Daniel noted.
AI-enabled CDS software, such as diagnostic support software (DxSS), could support clinicians’ decision making, assist them in making a correct diagnosis quicker, reduce testing and treatments resulting from misdiagnosis, and lower pain and suffering by starting treatments earlier, the paper found.
“AI-enabled clinical decision support software has the potential to help clinicians arrive at a correct diagnosis faster, while enhancing public health and improving clinical outcomes,” said Christina Silcox, managing associate at Duke-Margolis and co-author of the white paper.
“To realize AI’s potential in health care, the regulatory, legal, data, and adoption challenges that are slowing safe and effective innovation need to be addressed,” Silcox added.
The volume of data generated every day about disease, treatments, prevention, and wellness exceeds the ability of clinicians to absorb and process it all. AI-enabled CDS software can assist clinicians with knowledge management.
In addition, clinicians are feeling pressure to see more patients in a shorter amount of time. CDS software can help by distilling relevant information and delivering it to the right person at the right time in the right location in a format that supports clinical decision making.
The working group of experts identified several issues about the development, regulation, and adoption of AI-enabled CDS software that stakeholders will need to address:
- Evidentiary needs for increased adoption. This will include the impact of CDS software on patient outcomes, treatment quality, cost of care, and clinical workflow; the ability to use the software and its effectiveness in providing information in a format that clinicians find useful and trustworthy; and possible reimbursement by payers.
- Effective patient risk assessment. The degree to which software comes with details that explain how it works and the types of populations used to train the software could have significant effect on the assessments of regulators and clinicians about patient risk when the software is used. Regulators and stakeholders may need to reconsider product labeling, as well as the risks and benefits of continuous learning versus locked models.
- Assurance that AI systems are ethically trained and flexible. To ensure that software developed with data-driven AI does not exacerbate existing clinical biases, best practices should be developed to mitigate bias that may be introduced by the training data used to develop software. Developers will need to assess how the data inputs necessary for their software might affect scalability of their products to settings that are distinct from the original environment that generated the data used to train the algorithms. Finally, best practices and new paradigms are required about how to protect patient privacy.
For the potenial of AI-enabled CDS systems to be realized, the obstacles holding back effective innovation need to be addressed, and consensus standards need to be developed, the white paper concluded.
