The idea of artificial intelligence in medicine is not a concept for the distant future, but rather one that is gaining traction now.
It's already in hospitals, clinics, electronic health records, scheduling platforms, documentation tools, and even the conversations patients are having before they even step into an exam room.
So, what are the opinions of doctors regarding AI in Healthcare?
The answer is more complicated than either “doctors love AI” or “doctors are afraid of AI.”
I'm not sure there are many doctors out there who are sitting around debating whether AI will get smarter than humans. They pose more practical questions.
Will this help or simplify my day?
Can I trust it?
Will it enhance health care?
Will it generate another notification that I have to click through?
And, crucially: Who is at fault if the AI gets it wrong?
Physicians' attitudes to AI are a blend of excitement, skepticism, curiosity, and caution. And that might be just what healthcare requires.
Get enough time around doctors and one thing you will know: Doctors don't care about technology just because it is amazing.
They pay attention to its functionality.
If a time-constrained physician is forced to add three extra steps to the clinical workflow in order to use an advanced AI demonstration, they are unlikely to be impressed.
Having a predictive dashboard is not for a hospitalist who's treating several people. They want to see if it can be used to detect the downing patient earlier without setting up twenty meaningless alerts in the process.
That distinction matters.
The most significant application of AI in Healthcare might not be to replace complex medical decisions. It could be eliminating the myriad of trivial administrative tasks that take up doctors' time.
For example, documentation.
AI-powered Ambient documentation tools listen to clinical discussions and can generate notes. That can mean something for doctors who have to fill out charts for hours following a patient encounter.
AI can also streamline medical summaries, patient history, identify missing details, assist in discharge planning, automate administrative routines, and facilitate quicker access to key clinical information, all of which are areas of interest for physicians.
These apps might not seem as glamorous as an AI that diagnoses a rare disease.
However, they may be much more useful to the busy physician.
So, what do a lot of doctors wish for healthcare technology? The answer is quite simple.
Additional time for medical practice.
Health care has become incredibly complex over the years, and more than ever, it requires more administrative work. Physicians struggle through electronic health records, documentation requirements, insurance approvals, inbox messages, quality measures, compliance tasks and a host of alerts.
All of those tasks might have been added for a good reason.
But, collectively, they can leave physicians away from patients.
This is the realm where the power of AI in Healthcare truly shines.
With AI taking over repetitive tasks, doctors might find more time to spend on patient interaction and less on screens.
Suppose that there is an AI system that generates a helpful summary before an extended appointment.
Or one that finds relevant changes in a patient's medical record over dozens or hundreds of pages.
Or an assistant, who prepares a discharge summary as the doctor confirms his/her final medical decisions.
That's the type of AI many clinicians desire.
Not a robot taking over the doctor's job!
AI in parallel with the doctor.
While excitement abounds regarding AI, there is an underlying worry: health care is a space where errors are not taken lightly.
An ineffective marketing suggestion could be the loss of a few clicks for a company.
If a clinical recommendation is poorly generated, it can impact someone's health.
That is a change of the standard.
Physicians should be aware of where the AI is from, the level of confidence in the AI, and when the AI may be wrong.
Occasionally, large language models will produce very confident content that is false. That might be a nuisance to the non-geeky person looking up the internet.
It can be harmful for a doctor to make a clinical decision.
Hence the necessity of human intervention.
Rather than doctors taking orders from AI, the future of AI in Healthcare looks promising with the potential for collaboration and data-driven decision-making. It's more likely to include clinicians using AI for another source of information but still being responsible for interpretation, context and judgment.
Patients don't always fit into an algorithm, which is something that physicians know.
The 82-year-old who has heart failure, kidney disease, dementia, has limited family support and five hospitalizations, is not just a set of lab numbers.
There is uncertainty, values, social situations, communication, experience and judgment involved in medicine.
AI can be beneficial in those decisions.
It should not gloss over those complexities.
The healthcare industry has already seen the results of introducing technology without having paid attention to clinical workflow.
Simply ask physicians about alerts.
EHRs can produce alerts for prescriptions, orders, drug interactions, documentation, quality indicators, and clinical risk.
Some are helpful.
Many are ignored.
Why?
When all the time is an emergency, at some point, nothing is an emergency.
Doctors are concerned that the wrong kind of artificial intelligence could exacerbate this issue.
Can you imagine getting dozens of AI-driven risk predictions during the course of the shift? The natural feat of predicting something is not enough; what matters is that the clinicians are able to know what information is worth paying attention to.
To achieve successful AI in Healthcare, more than just the accuracy of a model is required.
It has to be carefully put into practice.
The right information needs to be delivered to the right person at the right time in a format that is useful to them in making a decision.
It is a workflow issue, rather than an AI problem.
The one thing that is becoming increasingly clear is that clinicians need to be engaged early in the healthcare technology space.
Technology is often created first and added to the doctors' repertoire afterward.
This can lead to products that might be great when presented as part of an academic program, but less effective in an actual clinical setting.
All health care providers (physicians, nurses, pharmacists, therapists, administrators, and patients) should have opportunities to give input during development and implementation.
This is where a robust Healthcare community is becoming crucial.
As AI enters the mainstream of healthcare, there is a need for spaces where professionals can share insights on the successes, challenges, and lessons learned that come with its integration.
In this conversation, platforms like medsocially can play a role by enabling healthcare professionals to connect with others, share experiences, discuss new technologies and gain insight from colleagues in various specialties and locations.
AI's future in medicine should not be decided within tech firms.
It should also be formed as a result of discussions taking place throughout the wider Healthcare community.
The concern about AI taking over the doctor is a good clickbait.
The reality is probably a lot more complicated.
With the increasing capability of AI technologies to access information, summarize records, identify patterns, and create documentation, there will be a shifting in medical work.The ability of AI to access information, summarize records, identify patterns, and create documentation will undoubtedly shift aspects of medical work.
However, medicine is more than information retrieval.
Physicians explain uncertainty.
They assist the patient in making the important decision when they are terrified.
They identify that an item does not fit into the expected pattern.
They consider evidence and take account of personal circumstances.
They interact with families.
They take responsibility.
Powerful technology makes those roles more relevant - not less relevant.
The doctor of tomorrow might not spend so much time looking for information and more time reading it.
The valuable skill could go from knowledge of information to contextualizing, challenging, and evaluating information as well.
As such, AI in Healthcare might even become even more significant in terms of human judgment.
AI is entering the healthcare sector in a multitude of ways, including document management, imaging, clinical decision support, patient communication, operations, research, and more.
Rather, it's whether doctors will see it.
They will.
The real question is, will the healthcare sector adopt AI in a thoughtful way?
Physicians desire tools that address real issues.
They are looking for proof that those tools are effective.
They desire to be aware of restrictions.
They wish to be protected from patient privacy issues.
They desire natural integration of systems in clinical workflows.
They want voice in decision making on how those systems are used that is meaningful.
It's not innovation resistance.
It's just good innovation.
The physician perspective on AI in Healthcare can probably be summarized in one sentence:
Do not give clinicians technology simply because it is possible. Give them technology that makes healthcare better.
The majority of physicians are aware that AI holds great promise.
They envision a future where AI not only minimizes documentation tasks, bolsters data accessibility, uncovers relevant clinical trends, aids in decision-making, but also frees up time for clinicians to spend with patients.
They can also be the risks.
The drawbacks of bias, incorrect suggestions, privacy concerns, subpar workflow integration, alert fatigue, and dependence on automation can't be overlooked.
This is why having doctors in the room is crucial when discussing healthcare AI.
It also requires nurses, patients, researchers, technologists, as well as Healthcare leaders and a connected Healthcare community willing to openly discuss what works and what doesn't.
Communities such as medsocially can help create those conversations as medicine enters this new era.
AI is likely to play a significant role in the future of healthcare.
However, medicine's human should never be removed.
The target should be to employ more effective technology so that health care providers can have more time, more information, and more abilities to engage in the truly human work of medicine.
From the very beginning of medicine, it has been a community enterprise. The formal learning of young doctors long before the advent of digital platforms was through rounds, con...
