Technology solutions powered by artificial intelligence (AI) have become more prevalent in the healthcare sector in recent years. These solutions are used to automate burdensome administrative tasks traditionally performed by humans.
AI often performs these tasks faster and more efficiently. Yet, perhaps even more importantly, emerging AI-powered healthcare solutions are using technologies like image recognition, speech recognition and decision support to address challenges previously seen as unsolvable.
Becker's Hospital Review recently spoke with Detlef Koll, Vice President of Product Development at 3M, about AI in healthcare. He discussed how the number of AI solutions is growing and how AI is profoundly changing healthcare. AI applications are delivering significant value, improving the quality of care and even, perhaps counterintuitively, making healthcare more human. He said, “AI is transforming the experience of healthcare for all stakeholders – whether they be patients, doctors, nurses, CDI specialists, or coders.”
AI is already a reality in healthcare with meaningful innovations on the horizon
Clinical documentation systems are one of the most mature applications of AI in healthcare today. These solutions utilize speech recognition to automatically transcribe physicians' verbal comments into real-time documentation. Mr. Koll explained, "Users of clinical documentation systems expect these applications to accurately interpret the ASCII characters -- that is, the letters assigned to the words they speak. But there’s a lot more to it than that. . They also expect to be engaged intelligently during the process in ways that support clinical, billing and coding workflows."
Looking ahead, applications of AI in healthcare hold tremendous promise. Solutions like diagnostic support systems, knowledge discovery systems and advanced analytics will soon become the norm.
Mr. Koll is particularly excited about ambient documentation systems. 3M, now with M*Modal, has worked in this area for the past five years and industry interest is growing, as development of commercial solutions advances. "The promise of ambient intelligence systems is that they can create documentation automatically by observing the physician-patient interaction. Clinical documentation will not be a separate, burdensome task for physicians.. Documentation will become an automated byproduct of providing care. This has huge potential over the next three to five years," Mr. Koll said.
Ambient intelligence systems are different from passive documentation solutions. They simultaneously and seamlessly serve as advisor, agent and scribe. As a result, these systems can interpret conversations and provide contextually-relevant information to only capture the documentation and close care gaps, but also to support physicians' decision-making at the time of care. These applications can automatically enter orders, schedule follow-ups, look up patient information, create compliant documentation and code information.
Another emerging AI-based solution is clinical diagnostic support systems for radiology. The goal is not to replace radiologists with AI applications, but to integrate additional information and insights into the reporting workflow to further support radiologists. According to Mr. Koll, "AI capabilities are maturing in many areas. We are getting more intelligent about how AI should work within existing workflows to support physicians, rather than taking things away."
AI can drive outcomes for patients, physicians and healthcare organizations
Traditionally, creating clinical documentation in healthcare has been labor intensive and costly. AI applications are changing that. Speech recognition technologies save time for clinicians; ambient documentation systems may prove to be an even greater time saver.
Ambient documentation systems have the potential to save physicians several hours each day. This will create time to care, increase the quality of care and reduce physician burnout.
Mr. Koll noted, "Lookup and curation of contextually relevant information are incredibly valuable. If an application can identify one important data element in a sea of EMR information for a physician, that can improve the quality and consistency of care. The impact of ambient documentation systems on healthcare could be absolutely profound."
In addition, there are considerable inconsistencies in healthcare data collection today. As a result, existing healthcare metrics contain a great deal of "noise." Ambient documentation systems could increase the consistency and accuracy of documentation and coding, which would positively impact quality of care measures.
As AI matures, it will fade into the background
During his interview with Becker's, Mr. Koll referenced the book Deep Medicine by renowned cardiologist Eric Topol, MD. In the book, Dr. Topol observed that artificial intelligence has the potential to make healthcare human again. By this he means that instead of being consumed with technology and documentation, the physician can now focus on the patient as AI makes the technology seamless and invisible.
Mr. Koll said, "My perspective is that we'll see many breakthroughs achieved in healthcare with artificial intelligence technologies like deep learning and machine learning. If we are successful, in five years, no one will even talk about AI anymore because the supporting technologies will become completely invisible to end users. I look forward to the day when AI is no longer the hot buzzword because it is actually delivering value."