In 2017, an Accenture study revealed that artificial intelligence in healthcare could realize up to $150 billion in annual savings by 2026 - specifically, using AI in conjunction with clinical health IT applications. That’s nearly 4.3% savings on total healthcare spending in the United States per annum.
But does AI have the potential to generate actual savings in surgery? And if AI’s potential is best demonstrated as part of a larger, interoperable hospital technology ecosystem, how could AI’s actual impact - when viewed in isolation - be measured?
To better understand AI’s impact in surgery, a functional example would highlight AI’s value as an augmentation to established healthcare improvement technologies - data analytics, for instance.
At Medica 2019, caresyntax assembled a diverse panel of healthcare innovators and experts to discuss the relationship and combined value of AI in conjunction with healthcare data analytics. The panel included Pierre-Yves Antier, Head of Strategy, Innovation and Transformation at Relyens, Dr. Lorena Pingarrón, Oral and Maxillofacial Surgeon at University Hospital, Madrid, Dr. Heiko Ziervogel, Head of Medizin im Grünen and Björn von Siemens, MedTech entrepreneur and co-founder at caresyntax.
Von Siemens and his team at caresyntax are convinced that AI’s potential for healthcare is reflected in a number of possible and actual surgical use cases. Not least of all would be addressing surgical variability, where the benefits are manifold. Here is a brief excerpt from the discussion.
What exactly is surgical variability, and what role does AI and data analytics play in the surgical ecosystem?
Björn von Siemens: Technical skill varies from surgeon to surgeon, particularly when learning new or more complex procedures, and can have huge implications on patient outcomes and cost. AI can be a useful tool to reduce that variability, and help surgeons improve – even the best ones.
Our AI-enabled systems are designed to collect data and make it available for decision-making support, in the operating room and postoperatively. Based on learnings from the information collected, we are introducing “small helpers” - small decision-making assistance tools that forewarn clinicians about situations pertaining to the patient population or to a particular patient a surgeon is operating on.
Further, we are moving in the direction of using AI-enabled systems that can automate surgical decisions. We are hopeful that in some surgical disciplines, in five years or so, there will be more development in the field of surgical automation. But for complex surgical procedures, this might take many more years. In the future, we need surgical systems similar to the commercial aerospace industry where there is a lot of automation involved, but in the end, it's the pilot, or the surgeon in our case, running the show. The advantage of using AI-enabled systems is significant cost reduction, and the ability to treat more patients in a safer surgical environment.
Our primary goal or objective is to make surgery as safe as flying, and with automation in the workflow, frontline caregivers will have more time for patients. They can talk to the patients before and after surgery, educate them, and this will help to bring a more “human” side to the now very IT and systems-driven environment.
Who is responsible for gathering healthcare data? Is there enough relevant data out there?
Björn von Siemens: Data collection and aggregation are some of the key inhibitors to new development in the healthcare industry. Creating a data collection environment in the hospitals and having access to it is extremely challenging as most European countries have strict data privacy laws.
So, the need of the hour for the healthcare ecosystem - providers, hospitals, medical device companies, data companies, insurance companies - is to come together and form allies to get access to vast amounts of data, from hospitals, and combine this data from the population side to get a better understanding of the patient journey over a period of time. For example, an ideal scenario would be to gather data about a patient who underwent surgery a few years ago and again in ten years after the surgery. This kind of data that link to a patient’s post-operative and long-term health outcomes can help create powerful decision assistant solution for the caregivers and the hospitals. That’s an illustration of AI and data analytics working in conjunction to improve surgical care.
Is there enough awareness on the topic of AI-assisted surgery? Do you believe that AI and data analytics will be the major drivers to reduce surgical variability in the coming years?
Björn von Siemens: I believe there has been an accelerated growth and political awareness about AI has gone up as well over the last couple of years. Hospital owners and OR managers now want to harness data and profit from the developments as well. Many caregivers and doctors are already familiar with using different AI devices in their daily lives, and they would be ready to use them for surgical procedures if it helped them to improve care for their patients.
However, the healthcare industry is quite big and historically, has been slow to adopt new technologies. A lot of players that came in first created IT systems that involved more work for caregivers to maintain, rather than simplifying their lives. So, some stakeholders in the industry are skeptical about adopting new systems - whether they really save time, or are only being a burden, and not adding much value. That is why we work hard to educate the market - specifically, how the use of AI-enabled systems embedded in the workflow of surgical teams are meant to simplify their lives. This would help motivate caregivers to adopt new technologies, and it would become easier to access hospital data with greater precision, eventually improving quality and safety across the surgical care continuum.
Data analytics and machine learning techniques are now being used simultaneously to uncover critical insights from the millions of data points collected by endoscopic and laparoscopic surgical video. With the help of AI, this can also help reduce surgical variability, and over time, surgeons will be able to understand more clearly which techniques align with better outcomes.
Additionally, these insights can link to a patient’s post-operative and long-term health outcomes. Over the next few years, there will be rapid progress as health systems collect and integrate more data into their processes. Naturally, the roles of AI and data analytics will become increasingly important. With more and more data available to leverage, we can use AI and analytics to solve a number of problems facing the surgery department, including of course skills and outcomes variability.
