Can Watson Really Practice Medicine

Artificial intelligence tools have the potential to help pharmacists.

At a recent conference, an IBM speaker talked about “IBM’s Watson, Supercomputing in Healthcare.” I was both mesmerized and skeptical at the same time. Mesmerized that artificial intelligence (AI) might be able to harness the drug therapy literature and apply it to drug therapy decision making at the patient level, but at the same time scared by the thought of a “brave new world.”

Watson’s original claim to fame was winning Jeopardy against 2 champions in 2011. IBM worked with Memorial Sloan-Kettering Cancer Center to develop products that will be able to make recommendations, in decreasing order of confidence, on what treatment is most likely to succeed. They have also been developing a product with WellPoint that will use Watson to consider what treatment should be authorized for payment.

The more I think about this type of development, the more excited I am. Health care seems to be the last industry to use the “Master Builder Concept.” The practitioner is expected to store information, recall it when needed, and then apply it at the point of care. But as pharmacists, we know that this model is broken. It is suggested that 1 in 5 diagnoses is wrong. Medical errors kill enough people daily to fill a 747. So anything that will improve the current system and do it cost-effectively is really valuable. Watson will not be practicing medicine, but will just help the medical practitioner do a better job. Such programs should change the speed of knowledge adaptation as well as improve the accuracy of the decision.

My generation has had a harder time with accepting these developments, but today’s youth are very comfortable with them. I turn to my grandkids regularly to help me solve a technology problem at home or work. This generation wants to be able to engage in their own health care using technology, so I assume we will see a much quicker adoption of AI in health care.

Of course, as I have been thinking about this topic, I realize I am surrounded with AI tools. The simple act of connecting with someone via a text message, e-mail, or cell phone uses intelligent algorithms to route the information.

So how will this apply to our practice of pharmacy? Getting over the fear of the unknown and the fear that because I might not be in total control, and also feeling that things are not as safe, is the first step. Whether it is telehealth, telemedicine, or telepharmacy, consumers are going to demand these programs, so we need to accept that these developments will come to pharmacy practice. You may be able to delay them for a short period by regulatory measures, but this is at best a short-term solution. These developments bring new players—currently outside of health care—into health care. Could efforts to stop telepharmacy lead to a push for national licensure or federal actions because these different players play by different rules? They do not care about the existing professional practices.

Our profession is very prescriptive. Pharmacists often see things as black and white, or at least they want their practice to fit that pattern. We need to be aware that disruptive technology based on AI is already here. Rather than fight it, maybe we should see how we can adapt it to help advance our practice and improve patient care.

I believe that these developments will lead to new care models. The entrepreneurial pharmacists or non-pharmacists will figure out how to do it and make these new care models accomplish better patient care—at a lower cost and with improved safety.

No, “Watson” will not be practicing medicine or pharmacy, but it will make physicians and pharmacists better practitioners. Are you going to embrace the changes or fight them? The choice is yours.

Mr. Eckel is a professor emeritus of the Eshelman School of Pharmacy, University of North Carolina at Chapel Hill. He is past executive director of the North Carolina Association of Pharmacists.