Artificial intelligence (AI) is coming fast to the healthcare industry and while it does raise some logistical and ethical concerns, integrating AI into health care settings has the potential to boost productivity, minimize errors and improve patient outcomes.
Those are the conclusions of Curtiss Johnson, D.O., who received his medical degree from the Michigan State University College of Osteopathic Medicine (MSUCOM) in May of 2025. He is now a first-year anatomic and clinical pathology resident at Mayo Clinic in Rochester, Minn.
The Allen Park, Mich., native was inspired to write an article about AI use in the medical field while taking an online elective course in his last year at MSUCOM. The final assignment was to write a reflective piece about a medical ethics topic, and he naturally gravitated towards the subject of AI. Now he encourages everyone to take a deep breath – he is doing the same.
“I read and watch the news way too much, and there's a lot of fearmongering right now about AI, not just in healthcare, about AI replacing people's jobs in the future,” Dr. Johnson said. “That especially rings true in pathology and radiology, as well, because a lot of people believe that you can just train a computer algorithm to see what you see on the slide or on the screen. And that's true, but I don't think we'll ever actually replace a human being and the physician behind that.”
Dr. Johnson’s article, “Ethical considerations regarding AI use in healthcare: How much is too much?” was published earlier this year in the American Osteopathic Association’s (AOA) online publication, The DO. The article provides an overview of the pros and cons of AI and includes several comments from Casey Schukow, D.O., a resident pathologist at Corewell Health William Beaumont University Hospital, Royal Oak, Mich., and a MSUCOM graduate.
“As a physician who uses AI for academic writing and patient care purposes, I believe its emergence is an unfortunate necessity, given the current healthcare climate,” Dr. Schukow said. “The U.S. physician shortage has increased patient-to-physician ratios, overwhelming an already burdened workforce. This is due, in part, to declining reimbursement rates, which force physicians to see more patients, perform more procedures and work longer hours with fewer resources.”
According to Dr. Schukow, the problem is further exacerbated by the corporatization of healthcare, which often limits physicians to 15-minute appointments with patients, which is “barely enough time for proper examinations and documentation, excluding inbox and administrative tasks.”
Dr. Johnson points out that primary care physicians, especially, must dictate lots of notes and obtain prior authorizations with insurance. They deal with burdensome administrative requirements that take away from patient care, not to mention falling reimbursement rates, which force doctors to do a great deal more work to earn the same amount of money.
“The reality of where healthcare is going is unfortunate,” Dr. Johnson said, “but that’s the reality.”
It makes sense then, that more hospitals are introducing AI to help with administrative tasks, such as writing patient summaries or real-time dictation services.
“So, if you're sitting in the emergency room as a physician or a resident, and you have 10 patients to admit, the last thing you really want to worry about is writing a discharge summary for someone who has a broken ankle,” Dr. Johnson said. “It's easy to let an AI model do that, and it probably saves you, maybe 15-20 minutes that you could use to take care of other patients.”
Pros and cons of AI models in healthcare
Besides taking care of some of these administrative duties, AI models are very good at text recognition which can help ensure or improve patient safety, Dr. Johnson explained. For example, AI can look through a patient's chart to review their medical history to help a physician determine if a specific screening should be ordered. Or, if the physician is prescribing a new drug, the AI model can check to make sure no interactions would occur with the patient’s existing medications.
AI also has the potential to improve patient outcomes through predictive capabilities. For example, a 2024 JAMA study suggested that a properly trained and supervised AI model within an Electronic Medical Records (EMR) can help predict and detect sepsis, pressure ulcers and more.
AI models can track and analyze patient demographics and public health outcomes data, which can help predict or possibly even prevent the next disease outbreak, as well as be used for vaccine development. AI tools are also evolving in cancer detection, including an FDA-approved prostate cancer screening method for pathologists. But Dr. Johnson says the software is merely a tool that must be used responsibly by physicians with unique skillsets and intuition.
“Because AI can hallucinate, AI can make mistakes,” he said. “So, I think you're always going to have to have a physician on the end of that pipeline, so to speak, to verify what is actually being put out. At Mayo Clinic, they're actually trying to do this with gram stains, which is a stain for bacteria in spinal fluid or blood. They have an AI model that can point out areas of interest, and it'll tell you, ‘I think this is this.’ And then all you basically have to do is say yes or no. So that is kind of where it is now and where it can go, but I don't think you'll ever have a model that will just auto-verify and send it out on its merry way.”
Furthermore, AI models can play a valuable role in complex cases involving comorbidities. It’s not unusual for patients to suffer from multiple conditions, such as heart disease, hypertension, etc. As Dr. Johnson notes in his article: “As life expectancy increases, today’s patients are more complex; many have multiple chronic conditions and medication regimens that can rival the length of a CVS receipt.”
“There's a reason why doctors specialize, because there's no way you can know it all,” Dr. Johnson said. “There's no way you can do it all. So, AI can be a tool to help mitigate risk and navigate a complex database or knowledge base of information and make it readily accessible, to make sure that you're at least more informed about how to manage these complex patients with new medication regimens.”
Despite these potential benefits, AI raises privacy concerns for patient data. A 2018 study showed that a machine learning algorithm was able to successfully re-identify over 85% of adults and nearly 70% of children from a data breach in which patient data was anonymized and Protected Health Information (PHI) was removed. Dr. Johnson emphasizes that more research and legislation are needed to safeguard patients’ PHI entrusted to healthcare organizations and the AI models they use.
AI can’t replace what’s irreplaceable in medicine
While physicians need to use AI models to stay relevant, Dr. Johnson says physicians must not over rely on it because people will always want and need the human component of medical care.
As Dr. Johnson states at the conclusion of his article: “Most importantly, we must remember that no matter how advanced AI becomes, it will never replace the healing touch or warm words of a trusted healthcare professional; those truly human qualities which contribute so meaningfully to patient care. Healthcare is always evolving, and AI is evolving right alongside it, but in the end, the best decisions we can make are the ones that are best for our patients.”
by Lynn Waldsmith
Artificial intelligence (AI) is coming fast to the healthcare industry and while it does raise some logistical and ethical concerns, integrating AI into health care settings has the potential to boost productivity, minimize errors and improve patient outcomes.
Those are the conclusions of Curtiss Johnson, D.O., who received his medical degree from the Michigan State University College of Osteopathic Medicine (MSUCOM) in May of 2025. He is now a first-year anatomic and clinical pathology resident at Mayo Clinic in Rochester, Minn.
The Allen Park, Mich., native was inspired to write an article about AI use in the medical field while taking an online elective course in his last year at MSUCOM. The final assignment was to write a reflective piece about a medical ethics topic, and he naturally gravitated towards the subject of AI. Now he encourages everyone to take a deep breath – he is doing the same.
“I read and watch the news way too much, and there's a lot of fearmongering right now about AI, not just in healthcare, about AI replacing people's jobs in the future,” Dr. Johnson said. “That especially rings true in pathology and radiology, as well, because a lot of people believe that you can just train a computer algorithm to see what you see on the slide or on the screen. And that's true, but I don't think we'll ever actually replace a human being and the physician behind that.”
Dr. Johnson’s article, “Ethical considerations regarding AI use in healthcare: How much is too much?” was published earlier this year in the American Osteopathic Association’s (AOA) online publication, The DO. The article provides an overview of the pros and cons of AI and includes several comments from Casey Schukow, D.O., a resident pathologist at Corewell Health William Beaumont University Hospital, Royal Oak, Mich., and a MSUCOM graduate.
“As a physician who uses AI for academic writing and patient care purposes, I believe its emergence is an unfortunate necessity, given the current healthcare climate,” Dr. Schukow said. “The U.S. physician shortage has increased patient-to-physician ratios, overwhelming an already burdened workforce. This is due, in part, to declining reimbursement rates, which force physicians to see more patients, perform more procedures and work longer hours with fewer resources.”
According to Dr. Schukow, the problem is further exacerbated by the corporatization of healthcare, which often limits physicians to 15-minute appointments with patients, which is “barely enough time for proper examinations and documentation, excluding inbox and administrative tasks.”
Dr. Johnson points out that primary care physicians, especially, must dictate lots of notes and obtain prior authorizations with insurance. They deal with burdensome administrative requirements that take away from patient care, not to mention falling reimbursement rates, which force doctors to do a great deal more work to earn the same amount of money.
“The reality of where healthcare is going is unfortunate,” Dr. Johnson said, “but that’s the reality.”
It makes sense then, that more hospitals are introducing AI to help with administrative tasks, such as writing patient summaries or real-time dictation services.
“So, if you're sitting in the emergency room as a physician or a resident, and you have 10 patients to admit, the last thing you really want to worry about is writing a discharge summary for someone who has a broken ankle,” Dr. Johnson said. “It's easy to let an AI model do that, and it probably saves you, maybe 15-20 minutes that you could use to take care of other patients.”
Pros and cons of AI models in healthcare
Besides taking care of some of these administrative duties, AI models are very good at text recognition which can help ensure or improve patient safety, Dr. Johnson explained. For example, AI can look through a patient's chart to review their medical history to help a physician determine if a specific screening should be ordered. Or, if the physician is prescribing a new drug, the AI model can check to make sure no interactions would occur with the patient’s existing medications.
AI also has the potential to improve patient outcomes through predictive capabilities. For example, a 2024 JAMA study suggested that a properly trained and supervised AI model within an Electronic Medical Records (EMR) can help predict and detect sepsis, pressure ulcers and more.
AI models can track and analyze patient demographics and public health outcomes data, which can help predict or possibly even prevent the next disease outbreak, as well as be used for vaccine development. AI tools are also evolving in cancer detection, including an FDA-approved prostate cancer screening method for pathologists. But Dr. Johnson says the software is merely a tool that must be used responsibly by physicians with unique skillsets and intuition.
“Because AI can hallucinate, AI can make mistakes,” he said. “So, I think you're always going to have to have a physician on the end of that pipeline, so to speak, to verify what is actually being put out. At Mayo Clinic, they're actually trying to do this with gram stains, which is a stain for bacteria in spinal fluid or blood. They have an AI model that can point out areas of interest, and it'll tell you, ‘I think this is this.’ And then all you basically have to do is say yes or no. So that is kind of where it is now and where it can go, but I don't think you'll ever have a model that will just auto-verify and send it out on its merry way.”
Furthermore, AI models can play a valuable role in complex cases involving comorbidities. It’s not unusual for patients to suffer from multiple conditions, such as heart disease, hypertension, etc. As Dr. Johnson notes in his article: “As life expectancy increases, today’s patients are more complex; many have multiple chronic conditions and medication regimens that can rival the length of a CVS receipt.”
“There's a reason why doctors specialize, because there's no way you can know it all,” Dr. Johnson said. “There's no way you can do it all. So, AI can be a tool to help mitigate risk and navigate a complex database or knowledge base of information and make it readily accessible, to make sure that you're at least more informed about how to manage these complex patients with new medication regimens.”
Despite these potential benefits, AI raises privacy concerns for patient data. A 2018 study showed that a machine learning algorithm was able to successfully re-identify over 85% of adults and nearly 70% of children from a data breach in which patient data was anonymized and Protected Health Information (PHI) was removed. Dr. Johnson emphasizes that more research and legislation are needed to safeguard patients’ PHI entrusted to healthcare organizations and the AI models they use.
AI can’t replace what’s irreplaceable in medicine
While physicians need to use AI models to stay relevant, Dr. Johnson says physicians must not over rely on it because people will always want and need the human component of medical care.
As Dr. Johnson states at the conclusion of his article: “Most importantly, we must remember that no matter how advanced AI becomes, it will never replace the healing touch or warm words of a trusted healthcare professional; those truly human qualities which contribute so meaningfully to patient care. Healthcare is always evolving, and AI is evolving right alongside it, but in the end, the best decisions we can make are the ones that are best for our patients.”
by Lynn Waldsmith