The Value of Artificial Intelligence for Healthcare Decision-Making — Lessons Learned

AI4HEALTH Article Reviews #07

The popularity of artificial intelligence in health and studies in this area are increasing day by day. Although the new technologies developed play an improving role in the delivery of health services, ethical and legal aspects still continue to be discussed.

How to evaluate the effects of these new technologies on the decision-making of healthcare providers is also an important point. In this article I reviewed, Whicher and Rapp offers an insight into on this issue with 3 important questions.

Can AI Be Evaluated Like Any Other Technology?

✅ Health technology assessment (HTA) bodies are facing new challenges in determining the value of AI devices. It is much more difficult to evaluate the efficiency of AI products compared to pharmaceutical products.

✅ If HTA is being used to decide how an AI-based tool is implemented, health economists should consider including as comparators all the potential ways that an AI could be used alongside clinicians.

✅ Profit and loss balance is essential. Technical evaluation alone is not enough. Many parameters should be considered, such as the competencies that healthcare providers will gain and lose.

What Is the Current State of Knowledge About the Efficiency of AI in Healthcare?

✅ Current state of scientific knowledge about the economic impact of AI in healthcare is low.

✅ Studies focusing on the effects of artificial intelligence devices in the literature mostly focused on their effects on health. Evaluations based on economic impacts are very limited and one of the reasons for this may be the inability to collect the data to make this evaluation.

✅ Evaluations on this subject may be biased according to the population to which they are applied.

✅ Studies were mostly conducted for less than one year. Although this period reflects short-term evaluation, it is insufficient for long-term evaluation and may have potential unforeseen effects.

What is the level of acceptance and adoption of healthcare AI tools?

✅ Although demonstrating the value and efficiency of healthcare AI tools is important, it is not sufficient for promoting the adoption and scaling of these tools in clinical practice.

✅ One of the challenges with development, deployment, and use that hinders the general acceptance and adoption of AI tools in healthcare is a lack of trust. Increased transparency, public communication, and stakeholder engagement in defining appropriate data governance will help increase healthcare confidence.

✅ Another obstacle in front of the studies is that the data is quantitatively increased, but qualitatively insufficient and rarely shared. To encourage appropriate data sharing, improved data governance structures are needed to ensure “high quality data are available [for relevant actors] and secure”.

The methods used in the evaluation of current clinical studies are so limited in evaluating health-AI devices. Producing assessment approaches specific to this field, educating users who will consider these assessments, and collecting quality data that will assist in these assessments are key parts of this issue and there should be specialists who will work on this issue.

🌺 Thanks for this valuable article:
Danielle Whicher, PhD, MHS, Thomas Rapp, PhD

📑 Click here for PDF of the article
Whicher D, Rapp T. The Value of Artificial Intelligence for Healthcare Decision Making-Lessons Learned. Value Health. 2022 Mar;25(3):328–330. doi: 10.1016/j.jval.2021.12.009. Epub 2022 Jan 31. PMID: 35227442.