3 Ways Artificial Intelligence Is Transforming Healthcare (Part 2)

Welcome to the second part of our series focusing on the numerous ways that artificial intelligence (AI) is transforming healthcare. In our previous article, we delved into how AI is improving radiology, pathological imaging, and brain-computer interfacing. In case you missed it, you can read it here.

In this entry, we’ll take a look at AI’s impact on electronic health records (EHRs), how AI could help optimize EHRs as risk predictors, and last but not least, how AI can help assess the dangers of antibiotic resistance.

Ready to get started? Let’s dive in!

Better Electronic Health Record Systems

EHRs are a major tool for every provider across the board; they help keep records digital and easily accessible. But they come with their fair share of issues, too: difficult-to-learn interfaces, user burnout, complicated documentation, rote processes, and user overwhelm are only a few of the problems administrators have to deal with.

AI-powered health applications are helping to reduce these problems by introducing automation, alert systems, better interfaces, and easier sorting. One such AI tool that’s tackling these problems is natural language processing (NLP), although it’s still in its infancy. Still, that’s not stopping physicians and innovators from thinking about the next steps to take for integrating AI into EHRs even more.

Dr. Adam Landman, MD, is the Vice President and CIO at Brigham Health in Boston. He says, “I think we may need to be even bolder and consider changes like video recording a clinical encounter, almost like police wear body cams.” Landman thinks that AI and machine learning could then be applied to those videos again to retrieve more vital information.

Imagine a patient portal that has recorded videos of exactly what your doctor told you during your appointment. Physicians would also stand to gain back more time by not having to field calls with patients who forgot their treatment plans.

By integrating with EHRs, AI could also automate prioritizing tasks, like refills, appointment reminders, and to-do lists in the near future. And that’s just the beginning.

EHRs as Risk Predictors

EHRs contain a ton of information about each patient, from basic things, like height and weight, to more involved data, like lab results, pathology reports, and family histories. However, this information isn’t easy to extract, digest, and analyze quickly. Additionally, things get worse when you measure aspects like data quality, data formats, and incomplete records.

Dr. Ziad Obermeyer, MD, says that in his experience, “part of the hard work is integrating the data into one place. But another problem is understanding what it is you’re getting when you’re predicting a disease in an EHR.”

EHRs have enormous potential in calculating risk for diseases based upon past history and in displaying patient history as a timeline. But, according to Dr. Obermeyer, AI is still in its infancy with the wealth of information contained in EHRs. “You might hear that an algorithm can predict depression or stroke, but when you scratch the surface, you find what they’re actually predicting is a billing code for stroke. That’s very different from stroke itself.”

There might be some hope with the data from MRIs. “But now you have to think about who can afford the MRI, and who can’t? So what you end up predicting isn’t what you thought you were predicting. You might be predicting billing for a stroke in people who can pay for a diagnostic rather than some sort of cerebral ischemia.”

Perhaps most importantly, Dr. Obermeyer emphasizes that AI algorithms used in this regard must not contain any bias against a specific type of patient. In his opinion, this could represent the biggest obstacle in the way of employing AI for these predictive purposes.

Antibiotic Resistance Risks

Antibiotics are great when you’re experiencing a bacterial infection. But if you’ve had the infection before, and stronger doses of antibiotics aren’t working as well, you’re already experiencing a minor resistance to antibiotics.

In hospitals, however, antibiotic resistance is a more widespread and threatening problem. These superbugs don’t respond to any medicines, and they kill tens of thousands of patients globally every year.

The cost of maintaining cleanliness protocols and systems amounts to billions of dollars every year in the U.S. alone. Fortunately, a better solution could like in EHR data; patterns in patients and hospital locations could help predict the risk of infection before symptoms even appear. Unfortunately, this is easier said than done.

Dr. Erica Shenoy believes AI can still do a lot of legwork in using EHRs to predict antibiotic resistance, but that hospitals and clinics need to make major moves to consolidate and use their EHR data.

“AI tools can live up to the expectation for infection control and antibiotic resistance. If they don’t, then that’s really a failure on all of our parts. For the hospitals sitting on mountains of EHR data and not using them to the fullest potential, to industry that’s not creating smarter, faster clinical trial design, and for EHRs that are creating these data not to use them…that would be a failure,” she says.

Keep Your Eyes Peeled for Part 3!

Healthcare generates and stores an immense amount of data around each patient in EHRs. Within this information are invaluable insights just waiting to be unlocked. Hopefully, as AI development continues to grow and evolve, it will give us a much better understanding of how to deal with various maladies and streamline a better way for medical stakeholders to work together.

Stay tuned for the next article, where we’ll go over smart medical devices, wearables, and smartphones.