Greensboro's African American Community Newspaper since 1967

Beware of using artificial intelligence to diagnose your ailments

Posted

When patients present for their visits, different dynamics may take place. Some patients arrive with their old problems ready to tackle them and get better. 

Other patients arrive fully aware that their condition has not improved yet they resist making lifestyle changes. Instead, they try to bargain with their health, often lacking the foresight to see that basic steps like eating well and exercising could substantially improve their condition.

One of the questions I try to understand is what are the patient’s goals? The goals they have for themselves, and the ideal goals they may have for their care may not be aligned. As a physician, I want to make sure their blood pressure, blood sugar and cholesterol are all at the recommended levels for healthy living. Maintaining such levels may require medication. Sometimes recommended changes in lifestyle are not what the patient is willing or able to give. Alignment is an important step in any relationship. In other words, when you see your healthcare provider, be honest with yourself and the provider. What effort are you really ready to put in to maintain a healthy lifestyle?

On the other side is the patient that comes in with an artificial intelligence (AI) generated diagnosis. They want me to write a prescription that AI says they need.  Some patients believe they have done all of the work and just have to deal with the nuisance of having to see a provider in order to obtain a prescription.

I appreciate a well-informed patient who has done some research into what the problem might be. However, I caution that the diagnosis AI may provide is often questionable. AI may not know to take certain information about that person into account. The information it provides may be out of context for a real situation. AI often does not have the full picture of a person’s life, their symptoms, their family history, or their lab results and the conclusions it creates may be erroneous. Therefore, nuanced medical expertise by a person who is board certified physician or a trained certified healthcare provider is the gold standard of appropriate and reliable healthcare. 

I have seen AI cause anxiety, worry or fright in many patients. I have seen self-diagnosis from AI lead to patients trying harmful self-treatment.

Yes, there is a place for AI in healthcare.  Let’s use it wisely and cooperatively. It is important to be an informed patient. It’s also important for patients to be open to ideas from your physician that AI did not generate. Understand that AI may not have enough information to provide you with a correct answer.

I suggest that when patients come for a check-up visit, it’s important to write down questions one may have for the healthcare provider. At the beginning of the visit, let the provider know that you have questions so the appointment time can be used wisely. AI, the patient, and the healthcare provider can all work together. It is important to remember that AI’s Claude, Chat GPT, Gemini, etc. did not go to medical school and do not replace the clinical judgment of a board-certified, licensed physician.


Dr. Veita Bland is a board-certified Greensboro physician and hypertension specialist. Dr. Bland’s radio show, “It’s a Matter of Your Health,” can be heard live on Wednesdays, 5:30 p.m. on N.C. A&T State University’s WNAA, 90.1 FM. Listeners may call in and ask questions. The show is replayed on Sirius 142 at 5 p.m. on Wed. Email Dr. Bland at ideas@blandclinicpa.com