This summer, AIMed announced the launch of AIMedConnect, a new strategic platform designed to connect innovators, healthcare practitioners and other stakeholders to facilitate collaborations and drive forward the development and deployment of artificial intelligence (AI). Jvion, a key clinical AI company, came on board as a founding member. The first AIMedConnect virtual event and roundtable discussions took place on 23 October.

Yesterday (17 November), Alexis May, AIMed Director of Content shared the very first AIMedConnect research findings with fellow audience who attended the latest AIMed webinar on Behavioral Health: Using AI to predict suicide, depression and opioid misuse before it happens. According to May, the survey-based research project looked at the extent which AI is being adopted in US healthcare organizations.

Specifically, the areas in which people hope for greater impact; the decision-making process, and differences in views between payers and providers. Guest speakers Karen Murphy, Executive Vice President and Chief Innovation Officer at Geisinger; Dr. Carlo Viamonte, Medical Officer at Anthem, and Dr. John Frownfelter, Chief Medical Information Officer at Jvion also took turn to express their views towards the research findings and general AI adoption.

The survey questions and findings

Many of the respondents involved in this survey-based research identified themselves as physicians. There were also representations from Chief Medical Officer (CMO); Chief Executive Officer (CEO); Chief Research and Information Officer (CRIO), and Chief Medical Information Officer (CMIO). The first survey question was “To what extent AI is part of your organization’s operational strategy?”. One-third of the respondents replied their organizations have not started with AI yet while 9% expressed their organizations had already deployed and are evaluating AI performance at the moment.

The second question is “Where you can see AI helping most within your organization?” Majority of the respondents regarded clinical performance as the area where AI could help most with. The next in line are revenue cycle management (RCM), patient experience, and staffing. Some of the respondents added quality assurance, patient safety, and operation forecasting are also the areas where AI could be useful. The third question focused on who is responsible for buy-in decisions made within the organization.

The received answers were relatively fragmented. Majority expressed AI adoption would probably be C-suite led in their organizations. Drivers such as cost and immediate return of investment (ROI) will affect the rate of adoption. In some cases, respondents said decisions could be made on a committee basis. Others believe AI procurement could be physician led or through research breakthroughs. Nonetheless, uncertainty was pretty obvious in all of the given answers. May quoted a respondent, who believed a lack of involvement within the organization had slowed down AI adoption.

The last question touched on the extent which COVID-19 had affected AI strategy. 50% of the respondents indicated absolutely not at all; 34% indicated perhaps AI development and adoption has been put on hold and the remaining 16% indicated COVID-19 had accelerated the process. AI providers and payers’ responses were also compared and indeed, they have different priorities and expectations of AI. A C-suite executive working in a provider organization proposed some form of a playbook, to guide interested parties.

Ample opportunities ahead

Murphy thought the findings is a clear demonstration that the healthcare industry is still early in the game of AI adoption and deployment even though some of us may feel like we have been talking about the subject for a long time. Dr. Frownfelter interpreted the results in a different way. On one hand, he thought the survey results will alter based on the demographics of the respondents. Particularly when it comes to regarding clinical performance as the area where AI could help most with.

On the other hand, he acknowledged healthcare is really slow in adopting AI, as compared to other industries. It is discouraging to know one-third of the respondents are not even contemplating AI in their organization. At the same time, he said we should be excited for the minority who are looking into AI deployment and evaluation and see the kind of change they can bring to healthcare. He agreed there should be a playbook acting as a form of governance to determine priorities, answer key questions, and making decisions that are impactful to the organizations.

Dr. Viamonte said the holy grail of healthcare – improve clinical outcomes and medical practice at a lower cost is far-reaching. There are many steps between where we are now and when AI can actually help us attain the ultimate goal. As such, healthcare providers should not consider the use of AI in operation as different from the use of AI in clinical setting. Perhaps, as Dr. Viamonte suggested, healthcare providers can start with the low-hanging fruits, using AI in the operational side of things, to generate excitement within the organization, before attempting the harder to achieve goals.

Murphy agreed. She said the breath of AI is so vast that we should not limit its potential within a specific domain. The three guest speakers continue their presentation on the use of AI to predict suicide, depression and opioid use which will be detailed in the next blog article. Please stay tune.

*

Author Bio

Hazel Tang A science writer with data background and an interest in the current affair, culture, and arts; a no-med from an (almost) all-med family. Follow on Twitter.