Artificial Intelligence (AI) is often proclaimed as an approach to augment or substitute for the limited processing power of the human brain of primary health care professionals. However, there are concerns that AI mediated decisions may be hard to validate and challenge, or may result in automation biases, even rouge decisions.

While the use of AI in medicine should enhance healthcare delivery, we need to ensure meticulous design and evaluation of AI applications.

Primary Health Care is rapidly advancing and evolving technologically. The majority of PHC providers are now digitized and use health information systems as part of providing care. With the advances in technologies it is now possible to exploit these health information systems using Artificial intelligence (AI) concepts such as machine learning and deep learning. The primary care informatics community needs to be proactive and to guide the ethical and rigorous development of AI applications to ensure their safety and effectiveness.

Recently, CRE Chief Investigator Professor Teng Liaw was part of a team of researchers investigating the implementation of AI in primary care.  Using Delphi consensus building methodology, the researchers sought to form consensus about perceptions, issues and challenges of AI in primary care. They found that there was consensus in AI’s potential to improve managerial and clinical decisions and processes, but differing opinions on the need for AI applications to learn and adapt to clinician behaviours and the potential for harm to patients. They concluded that AI has potential to improve primary health care, but unsupervised machine learning is currently not sufficiently mature or robust to be confidently used without checks in place.

You can read the paper in the 2019 Yearbook of Medical Informatics.

Find out more about CRE partner University of New South Wales and the School of Public Health and Community Medicine here.

Oct 9, 2019