Why the real value of AI is hidden in plain sight

David Seda, GE Healthcare DigitalBy David Seda
Chief Marketing Officer
GE Healthcare Digital

In our world, attention is often directly correlated with impact. The more famous someone is, the more influence they have. If a story generates headlines, it must generate more change, right? While that’s often true, what about the thousands of little, behind-the-scenes stories? When it comes to smart technology, the most impactful work is hidden in plain sight.

Artificial Intelligence in healthcare is a topic that has garnered a great deal of attention in recent years. AI and machine learning are helping clinicians diagnose patients more accurately and determine the effectiveness of treatment, all while making headlines. This clinical AI has great value, there’s no doubt, but I would argue the more far-reaching value lies behind the scenes in the world of operational AI.

Meeting the challenge of cost and accessibility

Two of the bigger challenges facing the healthcare sector as a whole are cost and accessibility. Of course, outcomes in patient care are crucial, but we can’t get to the care part if patients don’t have access to it or if someone isn’t paying for it. To be sure, clinical AI can impact both of these.  For example, aiding in providing increased accuracy in detecting conditions can help reduce errors and therefore, costs. However, the game-changer in our industry will be in attacking inefficiencies — where operational AI and machine learning will be indispensable. According to OECD (2017), one-fifth of European health spending is wasteful and could be eliminated without undermining health system performance. In the U.S., published studies estimate that approximately 30% of annual health care spending—close to $1 trillion—may be considered waste.

A good degree of the waste comes in the form of unnecessary delays, poor coordination, mismatched resources and so on. To give a real-world example, a doctor will order a patient to be discharged at 7 a.m. but the patient doesn’t get discharged until 3 p.m. because someone forgot a scan or a prescription was missing, and now that patient is taking up a bed that could be used for a new patient. GE Healthcare’s Command Center Tiles address this and other challenges. The Tiles help support hospital and healthcare system staff to provide the best possible patient care by putting all available knowledge to work, near real time, for caregivers in a manner that’s easy, fast, smart, and personalized. Users can easily access Tiles through web browsers on PCs, phones, tablets, workstations on wheels, and shared screens using proper credentials. The Tile software pulls data such as a patient records, lab, radiology, pharmacy, cardiology, oncology from existing source systems like EMRs, orders for staffing and picture archiving and communication systems throughout a hospital or a network of hospitals into one place at the swipe of a finger. The “Patient Manager” Tile, a personalized Command Center for every caregiver, enables an entire care team to see the same critical information at the same time, which shifts conversations from collecting information to problem solving. The Tile helps clinicians to provide expedited care by identifying the most important discharges to break current bottlenecks in patient flow, orchestrating efficient care progression, reducing excess days in hospital?, and anticipating and resolving risks to protocol compliance.

Missing appointments or patient “no-shows” are another operational AI opportunity. One study found that no-shows cost the U.S. health care system more than $150 billion a year. In the UK, almost 8 million appointments were missed in 2018 (known as DNA – did not attend), costing the NHS £1 billion, equivalent to 257,000 hip replacements or 990,000 cataract operations.

One of the highest demand departments suffering no-shows is diagnostic radiology, where patients in some countries can wait months for MRI or CT appointments. Today, imaging departments use predictive AI-based smart scheduling software shown to reduce no-show rates by up to 70 percent, helping clinicians better manage the availability of their equipment by eliminating wasted scan slots. By not changing larger processes and instead making small tweaks to their daily operations, clinics like Chattanooga Imaging and many others become more efficient. With better scheduling, more people are afforded access to imaging, which is often a crucial step in a diagnosis.

Battling physician or clinician burnout

Even before the COVID-19 pandemic, healthcare systems were asked to do more with less. But “more” sometimes does not mean patient care. Clinicians in the US spend 49% of their time in the Electronic Health Record (EHR) system and as little as 27% delivering direct care to patients, while in the UK the figures are 60% in EHR and 17% with patients.

Operational AI and machine learning can help relieve clinicians of time spent searching in EHRs and other IT systems, trying to form a holistic view of a patient’s status. As shown in the Command Center example, intelligent software can sort through mountains of information in near real time, extract the clinical insights, and predict barriers to care in time for clinicians to circumvent them.  With operational AI creating day-to-day efficiencies, clinicians are offered a healthier margin to focus on patient care. That, after all, is what most of them entered the healthcare field to do. Operational AI doesn’t chop down the tree for us; it sharpens the axe.

This is where we really begin to see the breadth of the impact of operational AI. In these examples, more efficient processes lead to more optimal use of hospital assets, like scanners and beds. That leads to shorter wait times and lower cost of care. It also frees up clinicians to do more impactful and meaningful work, curbing burnout and decreasing turnover. The ripple effect goes further and further.

Seeing the same in the consumer world

This hypothesis extends beyond healthcare into the most common places in our daily lives. It’s no surprise to anyone reading this that AI has saturated our lives. But it holds true that the most impactful AI is what we use in nearly every moment of every day.

Recommendations on websites like Amazon and Netflix are powered by algorithms that learn what you like and influence your purchasing decisions and viewing habits. In fact, streaming platforms like Netflix create content based on these algorithms. They look at the data of what people watch, then produce shows and movies that fit what we want to see. It’s like making a TV show in a test tube.

GPS navigation is another great example. AI tells us where to go when we plug in an address, and factors in countless variables we rarely consider. It calculates how long it will take to travel, which routes have tolls, and considers other routes to avoid accidents—all in real time. Apps like Waze analyze user-submitted data to alert us if the police are nearby so we can slow down. We rarely consider the behind-the-scenes functionality of tools, but they are integral parts of our daily lives.

The little things make the biggest difference

Clinical and operational AI are vital to the AI revolution in healthcare and around the world. However, when you broaden the focus and see what’s causing the greatest changes in our world, it’s the little things that add up and cause other dominoes to fall.

AI can certainly knock down a few trees for us here and there with advances in specialized treatment, for example. But, what it does most often is continually sharpen our axes and strengthen our muscles so we can do our normal, daily work more efficiently—ultimately improving patient care.

David Seda is currently Chief Marketing Officer, GE Healthcare Digital. Prior to this, he was Vice President of Product Marketing at Persado, an artificial intelligence startup in Silicon Valley. David served 6 years as Vice President, Global Marketing at Calix, a telecom software company focused on network and subscriber analytics, and spent 13 years at Cisco in various leadership positions including General Manager based in Lagos, and head of Marketing for EMEA based in London. Also, in Europe, David served as Vice President, Marketing EMEA for Hewlett Packard.

Earlier in his career, David spent 11 years at Apple, first as a software engineer and later as head of Worldwide Communications, reporting to the Chairman and Chief Executive Officer. At different times during his Apple tenure, David was named both Business Development Executive of the Year by the CEO, as well as Systems Engineer of the Year.