Some weeks ago, a friend based in the United States called to discuss his decision to quit his job and start his own business venture. He was trying to work out a plan to develop algorithms, or computer programs, that might be useful in the medical field. His idea was to find some complex algorithm that would prognosticate the chances of somebody developing a chronic disorder such as diabetes or hypertension.
He wanted inputs on disease-determining factors that could be incorporated in the algorithm. I could understand him perfectly well till then, but lost track when he went further to suggest that he dreams of a world where doctors are not anymore the interface between patients and diseases! Since I am a clinician, the idea of totally replacing doctors with computer systems sounded blasphemous and instilled in me a train of thought that waged a war for and against each other. I started analysing how a computer would possibly outperform me in my daily practice. When thoughts cloud your consciousness too much, the only way to escape is to get some sleep, and I went to bed that night hoping that I would be blessed with at least an ounce of clarity and peace.
Case in the OPD
The very next day I saw a 35-year-old woman in the out patient department, who came with progressively increasing difficulty in using both her hands to hold objects. She changed her poor-quality flip-flops to sandals with straps, thinking that would help her better in holding the footwear to her feet. She was working as a live-in housemaid, doing all household chores. She was a single parent to her two daughters who were living with her 65-year-old mother in a far-off village. She has a brother who is 25 years old and has similar health problems. She remembers her paternal uncle as having a clawed hand but could not provide any details regarding his condition. She came to us to find out the nature of her problem, and she was worried whether it was a genetic disorder.
Using my limited intelligence and knowledge (compared to a computer, I knew and remembered fewer ‘algorithms’), I was able to figure out that she had peripheral neuropathy, which could be inherited. There are plenty of genes which could get corrupted and be held responsible for the health conditions in her family.
A test suggested
I finally suggested that she could get a gene panel test done, which would cover many of the genes implicated in the disease, and provided an estimate of the cost of the test. I also informed her about the yield potential of the test, that is, the chances of finding out the reason for her illness if she did the test. I could do this using the power of the limited data stored in my mind from experience and training in the field. I feel a computer would have probably used similar data or maybe an advanced algorithm to arrive at the same conclusion as mine.
Now comes the twist of events, which a computer can seldom recognise. She started sobbing uncontrollably, narrating her experiences in life as the wife of a man who was not concerned or bothered about her or their kids. She worked for a few years in a household in Delhi where the owner was not kind-hearted and abused her; she had to run away. She had to shell out all her earnings to educate her children and also look after her old mother.
After hearing all this from her, I decided to tell her that the benefit of spending so much money for the test is not significantly high for herself or her family. Of course it would be useful to know the exact reason for her symptoms and to predict whether her children could develop the disease later in life.
I suggested that she could seek monetary assistance from some organisation and promised her I would also try to help her find some financial assistance to get the test done. I listened to her intently and gave her time to offload her emotions, held her hands for a second and reassured her that the situation was not as hopeless as she felt. I thought she left the room feeling a bit more relieved than when she came in. This was the least I could do for somebody like her.
Question of empathy
I was thinking how artificial intelligence would help her or me in such a situation. If I had used a computer system, or if she had depended on a computer system which would predict what test needed to be done, I am not quite sure whether the computer would consider all the factors she spoke about, in making a decision for her. Would the computer, after predicting the tests which need to be done, listen to her sobs? I feel that the computer and its algorithm would have failed to empathise with her being unable to understand what it felt to be a single woman rearing two children, and doing household work with failing hands and feet.
A machine can only understand on the basis of what you feed into it; it cannot grasp the unknown and the unsaid. The computer might not perceive the agonies and apprehensions that she would have in her mind. I don’t think she came to our facility to get her tests done or to get the best practical single solution to all her problems. Often we don’t go to a doctor for that alone. Personally I do not want to go to a doctor who knows only algorithms and protocols, but I want a doctor to listen to me intently and comfort me. I have myself stayed away from doctors who are considered to be the epitome of knowledge and protocols but failed to see me as a human being and give me the deserving, rightful dignity of a human being. I feel this is what an artificial intelligence system would lack. It would fail to value human beings as they are. It would definitely lack the personal touch. It would dehumanise a patient.
What is important
The counter-argument to this could be that, if you put away your emotions like a machine, then you would be better equipped to make rational decisions. But I feel that arriving at a rational and right decision is less important compared to the support we extend to a patient in the process of arriving at a decision.
I am not against the use of technology in aiding medical diagnosis and treatment. If a machine can do certain things better than human minds, then I would not resist their use. But as long as human beings and diseases would exist, I would not dare to imagine a world free of doctors. I would not dare to dream of a world free of healers or counselors. The world does not need more of rationalists at present as much as it needs those who can empathise. And when the day comes when a machine or a computer system can empathise as effectively as or even better than humans, I will be open and ready to change my view and embrace such a system wholeheartedly.
(The author is a medical geneticist working as an Assistant Professor in the Department of Medical Genetics at the Nizam’s Institute of Medical Sciences in Hyderabad. dhanyalakshmi@gmail.com )
