While at King Edward Medical University (KEMU), one of my responsibilities was giving a series of lectures on cardiac surgery to the final year MBBS class every year. Cardiac surgery is not an examination subject so attendance in my lectures was rather thin and the atmosphere a bit relaxed. In my first lecture to the new class, I often started off with asking students three questions not related to cardiac surgery.
The first question was about Pakistan history. The question was, who was the second prime minister of Pakistan, and who was the second governor general of Pakistan. Sadly for six consecutive years not a single student could answer my question. The students in KEMU are of the highest merit among all students applying to medical colleges in the Punjab and they all must have done rather well on their ‘Pakistan studies’ subject before being admitted to KE.
The second question was about Urdu poetry where I asked the students to identify the poet of a particular verse and explain its meaning. Here again I was disappointed year after year when not a single student was able to identify the poet and explain the verse. Here again all these students must have done rather well in Urdu which happens to be a compulsory subject like Pakistan studies in their pre-med exams.
The third question was about the origin of the word ‘algebra’. One student one year did hazard a guess that algebra might be named after Jabir bin Hayyan, a Muslim alchemist known in the west as ‘Geber’. However algebra is named after the title of a book on mathematics by a Muslim mathematician, Al Khwarizmi. A short part of the title of the book in Arabic is ‘al-jabr wal-muqabila’. Al-jabr went on to become algebra.
Once I had explained this, I would then ask my students if they knew about the origin of the word ‘algorithm’. A general rule is that in English and European languages many words starting with the letters ‘al’ are of Arabic origin. So it is possible to think that algorithm is also of Arabic origin. Indeed it is. Algorithm was the Latinised version of Al Khwarizmi’s name and it led to the word algorithm. And that brings me to real topic of my discussion today.
Most of us use algorithms to arrive at different conclusions. In medicine almost all diagnosis is based on algorithms that physicians use even without putting a name on their thinking process. Today computer programmes or algorithms can often do as good a job as a physician or even better when it comes to figuring out what is wrong with a patient and what treatment is required. This has led to an increasing reliance on these computerised algorithms in how patients are diagnosed and treated in many hospitals.
Modern physicians in the United States for instance are facing competition from nurse practitioners and other ancillary medical staff that are taking over many of the simpler services provided by physicians. However, the ultimate competition is progressively a hospital administrator or a nurse employed by a hospital armed with a medical algorithm challenging a physician’s decisions about patient care.
If a patient does not do well then the physician providing care can be held responsible for the bad result but of course a computer algorithm cannot be sued for malpractice. Until there is a major change in laws governing medical practice, doctors will be needed to provide malpractice cover.
It is inevitable in my opinion that computer algorithms will eventually replace much of basic medical care. And it will not be doctors that will be replaced first but rather all the intermediaries that are providing increasing amount of basic medical care.
Instinctively I am a ‘Luddite’. That said I must also admit that computerised information has made life much easier for most physicians. Thirty years ago I had to remember all the dosages of medicines I prescribed and their possible interactions with other medicines a patient was taking. Mistakes with serious side effects were not uncommon. Today an algorithm/programme available on a smart phone decreases the possibility of mistakes.
A similar programme available to the pharmacy makes such mistakes rare even if the prescribing physician did not check for mistakes. The down side is that the pharmacist will question any prescription with the slightest problem even if the prescribing doctor has already considered that problem and decided that the medicine prescribed is necessary.
In a few years we might start seeing computer/robot medical outlets in commercial areas and even in pharmacies where people can walk in and interact with a computer to seek medical help. Much of the technology needed for such interactions is already available but it will require legal cover and initially some human medical backup before becoming widely available.
Perhaps the future medical computer will initially look a bit like an ATM where the interaction will take place. Eventually though it will become possible to access these medical algorithms from a cell phone or a home computer. In advanced countries these healthcare algorithms will of course undergo rigorous testing before they are made available for use by the public.
No, computers and robots are not going to replace humans just yet. Even the most basic medical algorithm has to be developed by a human physician and scientist and as the algorithms become more complicated, greater human input is needed. However, there will be a point when artificial intelligence (AI) will become advanced enough to develop self-learning computer systems that can actually replace human experts. Until then doctors will still be needed.
There is some concern about the loss of the human healing touch but already technological devices provide much of the social interaction between people. Also, even today most computer literate persons that have a medical problem consult computer data bases to learn more about what might be wrong with them. This is becoming a serious problem for doctors.
Patients often arrive for a clinic appointment after having read up on their problem and often they have looked up and researched a disease that they don’t really have. Much of a doctor’s time then is wasted in convincing a patient that they are not really dying from advanced cancer of some organ or the other.
A problem initially with medical algorithms is that the algorithm and patient interaction will be driven by the information a patient provides. The old saying about computers, ‘garbage in, garbage out’ (GIGO) will come into play. The ability of a computer programme to decide what information being provided is relevant and accurate will be the first major developmental hurdle that public health algorithms will have to face.
The ultimate hurdle will be the actual physical management of a sick person. Robotic technology and computer algorithms will have to develop much further along before a patient requiring intensive care or surgical intervention can be managed without human involvement.
Whatever will be, will be but one thing is for sure that algorithms run computers and computers run most things in the modern world. And that is only going to get worse or better depending on your point of view.