Why AI Won't Make Doctors Obsolete

In healthcare, there is constant change. Knowledge evolution and tech innovations make for a medical landscape that is always shifting. For practitioners, getting comfortable with one approach or system only means closer proximity to its next iteration. Over the past ten years, nothing embodies that ever-changing philosophy more than medical note taking solutions.

In just a short amount of time, we’ve seen numerous solution-based approaches. From in-person and virtual scribes to dictation devices and speech to text technology, the changes have been frequent and swift. Now, in the age of machine learning, we’re ushering in the newest solutions: AI-powered medical scribes.

Technology like this has the true potential to transform the lives of physicians and how they are able to provide care. But, it raises many difficult questions as well, in ethics and in practice. If we let AI sneak in the backdoor, what potential does it have to proliferate deep into medicine and disrupt the rest of the healthcare industry? More importantly, how will it affect the individual providers who allow the rest of us to flourish?

A few weeks ago, we outlined some of the ways advanced AI will change the future of reactive and proactive care. Its applications, while futuristic and advanced, are not totally unrealistic, and the ways in which they can infiltrate our daily lives are bound to change how care is imagined. 

But, if wearable AI devices come to replace regular check ups, what happens to primary care physicians? Or if systems of advanced sensors and machines develop to the point where they can detect, diagnose and treat, say, cancer, what happens to our radiologists and oncologists? Do they, just like scribes, become arbitraged out? Left for systems that maximize the bottom line and do a “better” job?

The answer, like all in this industry and far beyond, is layered. The biggest thing we must remember is that AI is here to help, not to take.

Recommended Reading: Ambient Intelligence in Healthcare — The Future is Near

AI for Humans, Not AI vs. Humans

There is a lot of commentary that suggests AI is being developed to take over and replace humans. While some developers and specialists might argue that computers will be better than humans at diagnosing conditions, monitoring patient info, and remaining unbiased, one element of that conversation that is often overlooked is how AI is being developed to help humans do a better job, not eliminate their jobs altogether.

In the case of EHR’s and medical documentation, this is especially true. Studies conducted over the last few years confirm what doctors and patients have been saying off-the-record for a long time: perception and satisfaction of care decreases with the integration of clunky EHR technology. A 2020 study conducted across six Mayo Clinic hospitals found that 100% of patients were affected negatively by the EHR switch (to differing degrees).

The primary reason behind these reactions is that a patients' perceived quality of care is just as important to them as the actual care they receive, and it’s all dependent on the engagement and presence they feel from their provider. As people, we need to build relationships and be heard to feel connected and maintain our well being — this is especially vital in the exam room. When those elements of a patient visit are muted, perceived quality of care plummets. Outright replacing physicians with advanced computers will only widen this gap. Instead, we must design technology that makes it easier for humans to do what they do best.

The widespread implementation of EHR systems, in a way, is a perfect example of how poorly designed technology, despite its intent, can have negative results. In certain respects, mountains of documentation tasks created by EHR’s have effectively turned physicians into computers, stripping them of the innate qualities that make them human and all that make healthcare beautiful and altruistic. In this new world, physicians are forced to spend more time as clerical administrators and less time providing engaged, personalized care.

When we dissect the data it becomes clear that, as humans, we prefer other real humans over computers. Not an earth shattering finding, perhaps, but one that is important to understand when discussing these topics and the future of healthcare. Humans possess the one power that AI can try to mimic, but will never be able to truly master: Empathy.

Empathy.

No combination of advanced language processing or machine learning formulas can replace the empathy that humans innately possess. In healthcare, in the exam room, anywhere. AI, at its core, is a system of binary questions asked in sequence that are used to eventually determine the “best” outcome or answer. The design exists to collect and filter as much data as possible in order to provide the most accurate solution. 

Humans, on the other hand, don’t operate on those same binary principles. Emotional intelligence and qualitative data all play a role in how we respond to questions and arrive at thoughtful solutions. That inherent ability to read and care for people is unique to humans and humans only and no amount of advancing technology could ever replace that in a real or meaningful way. And that’s a good thing for all of us, because we need other humans.

Let the last 18+ months of the pandemic be a glaring reminder that we do truly need human contact to survive; to thrive. Countless psychological experiments have proven how social isolation can alter the mind, and many more studies show a link between seclusion and increased risk of cognitive diseases like Alzheimer’s and dementia. The list is long, but it is clear that we are meant to be in the presence of others.

Understanding this reality is critically important. The practical, wholesome role of AI is to relieve us of our burdens so that we can get back to doing the human things that make us better, and wonderfully different, than any supercomputer. 

With DeepScribe, specifically, the goal for us is not to replace physicians and take away their jobs. In fact, it’s the total opposite. We are building an AI system that automates only the monotonous, computer-like tasks, not the human ones. We understand how valuable our care providers are, which is why we are doing whatever we possibly can to preserve their profession and maximize their skills so they can get back to providing passionate, engaged care.

We know that staring at a computer during a patient encounter or spending hours completing clinical documentation isn’t conducive to a healthy work-life balance. It’s time to stop asking healthcare providers to behave like computers, and time to let them just be human. And let them get back to doing what they love and do best: care for the rest of us.