Fernando Schwartz, VP of Data Science, CitiusTech

home

authors

fernando-schwartz-vp-of-data-science-citiustech

Photo/Thumb have been modified. Courtesy of Artem - stock.adobe.com.

Artificial intelligence (AI) is taking over the world—or so it seems. According to a 

 by PricewaterhouseCoopers, healthcare industry advances in AI, including variants like machine learning (ML), will contribute significantly to an economic boost totaling 

. Productivity gains alone could contribute $6.6 trillion to that total.

It wasn’t long ago that electronic health records (EHRs) were hailed as 

 answer to improving medical care, given the ability to cut costs, making patient information more accessible. Instead of making information more accessible, 

 have created a patchwork system spread among hundreds of private vendors. Implementations mainly succeed in isolating patient data and making said data inaccessible to both physicians and patients.

The challenges are undeniable and create a need for optimism within the industry, considering the advantages AI plays in clinical data integration and interoperability. These are necessary to facilitate acquisition, accessibility and utilization of clinical data.

Eric Topol, M.D., founder and director of Scripps Research Translational Institute, put it 

: “If properly and humanely deployed, AI has the potential to restore efficiency to a wide array of burdensome healthcare processes, freeing up physicians to treat their patients in the way they deserve. The path won't be easy, and the end is a long way off. But with the right guard rails, medicine can get there."

The results of a recent survey by CHIME and CitiusTech, the 

, more than 50 IT decision makers from leading U.S. health systems reinforced this assessment. The leaders expect to gain measurable return on investment (ROI) for AI initiatives. While there are challenges to overcome, ultimately, health systems expect AI/ML initiatives to result in smarter and faster processes, enhanced clinical quality, lower costs and better care. Understanding these challenges can help organizations map out a strategic plan and start gaining the benefits of AI for their organization.

Look to Rules-Engine Success to Identify AI Use Cases 

For 58% of survey respondents, the lack of clarity around use cases caused IT leaders to stumble and diminish productivity. To alleviate challenges in the future, consider the precursor to AI: Rules engines. Rules-based technology is in wide use and performs well, so identifying rules-engine use cases where incremental improvement could provide a quick win is a recommended first step. By adding AI, those rules become more flexible, more dynamic and self-adaptable.

Consider how clinical data integration could contribute to a broader range of data, apply intelligence around it and then deliver it within a workflow.

For example, if a rules-based approach achieves 80% accuracy, applying AI might increase that to 90 or 95%, which translates to a significant value add.

Clinical performance holds the biggest promise for ROI, according to 80% of health system respondents, followed by operational performance (64%) and financial performance (44%). The key to delivering ROI, as seen in other industries, is integration within the day-to-day workflows. Consider how ubiquitous AI is—advertisers routinely present ads based on search history and interests. Even if we are not conscious of those decisions, we’re often taken in when items of interest appear before us. This same logic is within our reach in healthcare.

AI allows for vast amounts of patient data to be scanned and analyzed. Then, critical and relevant data can be quickly and efficiently presented at the right moment to the right care team member so the appropriate action can be taken to improve a patient’s health outcome. For example, testing a patient, performing a procedure and prescribing the right therapy sooner rather than later can make all the difference. AI can deliver insights that individuals, even physicians, cannot arrive at on their own. Common areas of focus include decision support, clinical documentation, population health, gaps-in-care, utilization management and denial avoidance.

As AI continues to demonstrate its ability to deliver ROI, organizations are committed to stepping up their investment in both infrastructure and staff—58% of those surveyed are expected to grow their team to five or more in the next three years. Mainstream adoption of AI technology will not only require stronger investment in data scientists with skills in AI and ML tools, statistical modelling, predictive analytics and big data processing, but also in the organizational structure needed to enable productivity and ongoing innovation. 

With AI’s potential to deliver savings from efficiency gains at more than 

 and make real-life improvements in patient health outcomes, both healthcare providers and payers are eager to move forward. Although critical decisions need to be made to implement an effective strategy for each organization, there are best practices for data integration and sharing that can help pave the way. By taking a pragmatic approach to identifying use cases that can provide quick wins, healthcare organizations can map a strategic plan to start gaining the benefits of AI for their organization, even earlier than planned.

AI Accurately Predicts Risk of A-Fib, Mortality from ECGs

AI: The Only Viable Solution to Healthcare's Data Problem

Articles

Survey results offer insights into AI challenges and solutions.

Fernando Schwartz, VP of Data Science, CitiusTech

Articles

Breaking Down the Barriers to AI Adoption in Healthcare