2,314 exabytes
The estimated amount of health data that will be produced in 2020, according to the International Data Corporation. (one exabyte = one billion gigabytes)
CONTRIBUTOR
Michael Myser
Essential Insights contributor, healthcare writer
Featured expert
Jacob Haning
Senior Analyst and Software Engineer, Fuse by Cardinal Health
Artificial intelligence (AI), the simulation of intelligent behavior in computers, has previously been best known through science fiction movies and books. But now, artificial intelligence is showing real potential, especially in healthcare.
As the technology has matured over the past seven decades, there's mounting evidence that AI can help increase efficiency, streamline data-based operations and communications, eliminate errors and diagnose disease. But for all of the promise and futuristic ideas, AI is not yet widely used in healthcare – only about 35 percent of hospitals expect to implement it within two years, according to HIMSS Analytics, a subsidiary of the Healthcare Information and Management Systems Society that performs healthcare research and advisory work. And in reality, AI will have a very practical role.
“For health systems, and healthcare in general, AI will be an enabling layer implemented into already existing systems," said Jacob Haning, senior analyst and software engineer at Fuse by Cardinal Health. But this practical approach will have a massive impact on business, as Accenture recently found that AI could save the healthcare sector some $150 billion by 2026.
Are health systems ready for the AI revolution? Haning believes so and offered these five signs your health system should consider implementing AI solutions.
1. Does your health system have the data to support AI initiatives?
The answer to this question is nearly always a resounding yes. Not only is healthcare producing a massive amount of data – according to the International Data Corporation, an estimated 2,314 exabytes will be produced in 2020 – most of that is found in structured form in patient files, billing records, electronic medical records (EMRs) and a host of other sources. (Read "How big data is driving better patient support.") AI solutions are particularly suited for this type of standardized information, able to make quick sense of and learn from this massive amount of data.
But unstructured data like doctors' notes and high-res electronic images also presents a treasure trove of information that advanced AI can utilize. Companies like Arterys, VisualDx and CloudMedx, for example, are analyzing this information to aid in diagnoses and offering clinical insights, giving doctors and patients a digital “second opinion" in real-time.
2. Are your doctors, nurses or other staff performing repetitive tasks?
EMRs, patient intake, billing and a huge amount of other required paperwork often demands repetitive data entry into the health system. From simple auto-population of common fields, to auto-coding for billing and insurance paperwork, AI can handle a portion of those administrative tasks for staff, nurses and doctors.
“If you can remove some of this burden from doctors and nurses, they're freed up to spend more time with their patients," said Haning.
It's often hard to recognize the immediate opportunities AI presents in improving performance or saving time and money, but because there are measurable reported outcomes and commonalities from performing these basic, repetitive tasks, finding the easy AI wins here could prove a quick return on investment.
3. Can your health system be more efficient?
Get a free machine learning primer.
And, sign up for the newsletter that will help your organization thrive in the challenging, dynamic world of healthcare. Is patient intake slowing down emergency room performance? Do billing or insurance questions consistently block up payment? If so, AI solutions have a real role to play in reducing patient and ER wait times, increasing the speed of patient diagnosis and care, improving communication and resolving billing and payment questions. While AI won't remove these processes, you can add value and increase their efficiency with AI.
"At Fuse, for example, one of our jobs is to standardize file formats for our healthcare partners," said Haning. “While AI wouldn't replace the team of people performing the task, it could instead make them more efficient."
Fuse is also working on an R&D project that uses AI to predict if insurance payments will be made on certain claims, depending on a number of known factors like customer, insurance carrier and the medical equipment requested.
4. Can you get more from your EMR?
EMR files contain many free text fields, including notes from doctors and other care givers in the health system. That rich data provides an opportunity, if analyzed, to look at patient care and health trends in a certain population or area to offer new services across your health system, or prepare providers for treating the most common maladies.
At the individual patient level, AI could even assist doctors and nurses in gaining more insight by providing a summary of a patient's history without re-interviewing them each visit, or offering care options based on the massive amount of information found in EMRs. IBM's Watson EMR Assistant technology has done just that in a pilot program with the Cleveland Clinic, where its natural language capabilities can start to make sense of a “lifetime of clinical notes."
Additionally, human error, from simply writing the incorrect billing code for insurance claims and payment, to inputting notes in the wrong data field, can derail the flow of medicine and business. AI can not only guide these processes, but even auto-code diagnoses using natural language recognition to handle that responsibility.
5. Is your health system conducting research and clinical trials?
While very complex and time-consuming for health systems to implement, clinical trials can also be a boon for research and enhancing reputation. By examining records for patterns and health requirements, AI could help streamline the process of matching and enrolling existing patients into clinical trials. AI could also ensure those patients are adhering to trial protocols, with reminders through smartphones or devices like Alexa and Google Home, to help the research run smoothly and meet strict study requirements.
AI is ready for health systems
From smoothing over cumbersome and repetitive data entry processes to assisting with patient care and diagnoses, AI is now in a position to revolutionize healthcare operations. And interestingly, the biggest impact will likely be made by incorporating AI into the already-existing workflows and technologies that exist – EMRs, patient intake, billing, insurance claims and the like – helping to streamline health system operations.
More articles on AI
Why healthcare providers should embrace AI
AI is expected to disrupt many industries. Some providers may be concerned about the use of AI in healthcare, but here are three reasons to embrace it.