AI: A CEO's Perspective

 

New applications of AI have shown tremendous potential to transform healthcare, offering innovative solutions to long-standing challenges and wielding greater efficiency in how we approach patient care and treatment decisions.

Operationalizing this groundbreaking tool is a major challenge and an extraordinary opportunity. The global healthcare AI market is forecast to be worth almost $188 billion by 2030, according to a Statista report “AI in healthcare market size worldwide 2021-2030.” This growth comes at a critical crossroads in healthcare, with a projected shortage of nearly 10 million physicians, nurses and midwives globally by 2030 at the same time we are faced with the increased needs of an aging population. AI offers practical solutions that can help organizations overcome these challenges by improving outcomes, enhancing care and advancing health equity. 

Hackensack Meridian Health’s guiding principle in this journey has been that it must be highly strategic. The organization’s priority is to deploy AI safely, making sure the right governance mechanisms are in place. We are committed to ensuring AI is not acting without human intervention and oversight.

Here are a few ways teams were challenged to demystify the process.

They started by framing the discussion around key areas to guide them in selecting the best options for Hackensack Meridian Health’s network of 18 hospitals and more than 500 patient care locations. They also created a predictive health team that worked with stakeholders across the network to leverage AI technology and prioritize projects. 

The teams created a pipeline focused on three areas: clinical, operational and research, where AI would provide either diagnostic support or predictive analytics to enhance care delivery. They funneled all the projects their leaders suggested and ultimately selected five to pursue, with the purpose of improving clinical outcomes, enhancing operational efficiencies and advancing health equity.

The five pilot projects include:

1. Harnessing the power of AI to identify complex patterns in imaging data and to provide quantitative evaluation of radiographic traits. The goal is to better manage the radiology workflow queue, ensuring more complex and urgent images are prioritized. They may also detect modalities that can benefit from further inspection.
2. Testing the capabilities of AI to enable primary care providers to identify Stage 3 chronic kidney disease earlier to slow disease progression, potentially a major advance for health equity. Black Americans experience kidney failure at three times the rate of white Americans. This is vitally important, as America faces a tsunami of chronic disease, much of it undetected. In fact, an estimated 100 million people have an undiagnosed chronic disease, contributing to 90% of the country’s healthcare costs, or about $3.7 trillion, according to the Centers for Disease Control and Prevention.
   
   Despite our best efforts, there simply isn’t enough time for healthcare providers to support all the preventive, chronic and acute care needed. One study by Justin Porter, MD, with the Department of Medicine, University of Chicago, in 2022, “Revisiting the Time Needed to Provide Adult Primary Care,” factored that it would require physicians to work an impossible 26.7 hours per day to meet the need. AI has tremendous potential to improve the treatment of chronic illness.
3. Integrating AI for operating room time optimization. The goal is twofold. First, it is important to account for the “pairwise” familiarity—the number of past collaborations of pairs within the clinical team. According to the Harvard Business Review, researchers studying cardiac teams that conducted more than 6,000 surgeries over seven years found that the composition of the team made a significant impact on productivity. Second, they are developing recommendations of procedures for patients that can fill unused time.
4. Using Serious Illness Continuum Connect, which helps ensure patients get into the right care setting sooner, both to improve their care and decrease bed days and readmission rates. Too often, our industry does not provide palliative care at the most opportune time for patients because of administrative and cultural issues, among others.
5. Using AI concepts to generate a list of patients for case managers to consider for transferring to long-term acute care hospitals. These facilities treat patients with serious medical conditions who require care on an ongoing basis but no longer need intensive care or extensive diagnostic procedures.

Another promising area for AI is in drug discovery and development. Keep in mind that only about 12% of drugs entering clinical trials are ultimately approved to be introduced by the FDA, an expensive and challenging process. AI applications can analyze vast amounts of data to quickly identify high-potential drug candidates, predict their effectiveness and safety, and optimize their design. Hackensack Meridian Health’s Digital Technical Solutions team and leaders from its Center for Discovery and Innovation, along with Google, are working to develop a research data enclave and digital lab infrastructure to enable and accelerate the effectiveness of AI models.

While we should embrace AI as a groundbreaking tool, it is critically important to establish guardrails to ensure that it’s used effectively and ethically. 

For example, AI has the potential for bias. The effectiveness of AI models relies heavily on the quality, representativeness and diversity of the data used. Biased or incomplete datasets can lead to algorithmic biases and erroneous outcomes, potentially exacerbating healthcare disparities. There are also many ethical concerns related to privacy, security and informed consent. This is why Hackensack Meridian Health is joining with other organizations, including Google Cloud, to develop best practices. All the pilots they launched have a multidisciplinary team that oversees them. This includes clinicians, colleagues who are in the same field and administrative oversight from both their IT department and operations. A data governance committee of the board also provides oversight and makes sure safeguards are in place. 

Hackensack Meridian Health is also part of a nationwide commitment to harness AI’s potential while managing the risks posed by this breakthrough technology. The network joins organizations, including CVS Health, Boston Children’s Hospital and Houston Methodist, to define a set of voluntary commitments and guide their use of frontier models in healthcare delivery and payment. 

With more collaboration and a commitment to positive change, we can harness the power of AI to improve and transform healthcare in our communities, throughout our nation and ultimately around the globe. 

Robert C. Garrett, FACHE, is the CEO of Hackensack Meridian Health, New Jersey’s largest health network, which includes the Hackensack Meridian School of Medicine.