I believe it is quite clear to anyone who has been regularly reading my blog, that I believe that computers and robotics will replace the vast majority of the role of physicians. I imagine that physicians will morph into a very different type of professional. But the days of doctors running around in white coats, working 36 hours straight, handwriting charts, reading films while infusing coffee into their veins, will soon pass. It might take 10 years or even 50 years. But the days of this archaic system are numbered.

As I have mentioned many times in the past, computers will be masters of knowing and then implementing the best medical knowledge available. If medical science in 20 years from now, says to treat a patient with medication “X” for disease “Y”, then the computers will do just that. And if the patient has special considerations like renal failure or liver failure, the computers will take this into account as well.  But within this perfect scenario of perfect medical care delivered by perfect machines, there is still one very basic problem.

The algorithms that stem from the research that ultimately form the guidelines for managing the health care of the patient, will all be based on human work, for many decades to come. I can easily imagine human beings still running major research projects 100 years from now. Admittedly, computers will help the humans prepare the protocol for the study and will do the analysis of the data collected. But if there is a fundamental flaw in the way the human beings have designed the research studies, this will most likely be reflected in the actions of the computers.

Admittedly, in 100 years from now, I certainly hope that there are simulations of human populations that are so good and so accurate, that a great deal of actual human research will be unnecessary to uncover new treatments and medications. Nevertheless, somewhere along the line, there will be a human component. And the moment there is a human component, the potential for error arises.

When a computer will make a mistake, due to a human’s faulty input and/or programming, the computer may very well be able to identify that error and present it back to the human. The expectation will be that the human will correct the error so that it does not recur. It might even be that future computers will be so effective at self analysis and healing,  that they will correct human errors before they have had a chance to do any damage. Nevertheless, the possibility will continue to exist that a human component in the entire process of computerized healthcare, will lead to human patients being injured if not even killed.

I should point out that as time goes by and computers do more and more, the number of patients who will die due to human error will drop precipitously. But it is very likely that it will take a long time for that number to truly drop to zero.

When people ask me what the role of doctors will be when computers can “do it all”, I think I may have found it. Someone will need to walk into the room of the patient, speak to him or her, explain what happened in terms of the error, describe the consequences and then finally offer alternatives for repairing the damage. Especially in a time when people will have become accustomed to near zero error rates from computerized care, informing a patient of a mistake will lead to true culture shock. The patient will first need to internalize the concept that mistakes can still happen in such a computerized reality. But from that point on, the patient’s reaction will be equivalent to  a patient’s reaction in this day and age when he or she is informed of a medical error.

The patient will express anger and misunderstanding. He or she will demand an explanation and of course, a repair of the damage and even financial compensation that covers their pain and suffering. In this future time, the story will probably make front-page news. The patient will be interviewed over and over again, and each time will express his or her dismay at the possibility of such things still happening.

In this reality, it will be critical for a doctor, a human physician, to sit with the patient, to listen patiently and ultimately to apologize. The ability to apologize in a sincere and caring way, will not be a feature of computerized doctors for many decades, if not centuries to come. Any doctor who has erred in the care of a patient knows that pit in his/her stomach that never really goes away. I remember an Oncology patient that I cared for as a medical student. She needed an intravenous line but was known for having notoriously difficult veins. I still tried. It wasn’t just pride that pushed me to try and try again. I really did want to get the intravenous line into her vein so that we could give her the fluids she needed. But even after just the first try, it was evident to anyone but me that a far more skilled physician was necessary to treat this woman.

At the time, I apologized profusely to the patient. But clearly, that did not retroactively eliminate the unnecessary pain I caused. From that experience, I created a self-imposed rule that I would never try more than twice to place an IV line. If I failed twice, then it was time to pass materials over to someone else, to let him or her try instead of me. I wholly admit that  this kind of self-awareness and moralizing will be absent from computer systems for a very long time. But I think it is also clear that without a human soul, even the most powerful computer can cure a disease but kill the heart of the patient.

Will a doctor need to train a total of 12 to 15 years in order to play the role of the computer’s soul? I don’t know. I don’t know if a doctor will truly be able to empathize with an injured patient unless the doctor truly understands firsthand what it means to operate or to resuscitate a patient after a life-threatening deterioration. Only time will tell what role doctors will still play in the future when so much our healthcare is run and managed by computers.

I still want to be fair to the computers. We all know doctors who are distant and disinterested and who would never apologize for an error even if it was caught in real time on tape. These will not be the doctors of the future. The screening process for future doctors will emphasize heavily the empathetic side of care. Training will include anatomy and histology and physiology, but it will also include entire courses on how to speak to a patient, how to identify subtle clues of pain and concern, how to manage the family of the patient and the like. Perhaps in some way, the doctors of the future will recapture some of the nature of ancient physicians who were referred to as healers. People die. But for those left behind, they will need to heal. And it might very well become the primary role of the future doctor to help future patients/families to accept the tragedies that will still happen but to move on regardless

Fate definitely has its own sense of irony. It seems that we will have to surround ourselves with the metal of computer cases in order to rediscover our humanity.I also imagine that in the future, when I write letters to computerized healthcare providers, I will not end them in my classic way, specifically …

Thanks for listening.