Well, it took a massively thought provoking article by Vinod Khosla on TechCrunch to jerk me out of my inertia of lethargy with blogging.
In his post “Do we need Doctors or Algorithms” VK embarks on a sci-fiesque prediction of a day when doctors will be replaced by automatons delivering care on the basis of pure evidence. As an enthusiast of Evidence Based Medicine, this is the EBM-Nirvana for me, and at the risk of making my profession irrelevant, I have to say this is too interesting to turn away from. I must say I find his vision very exciting and I totally believe that the future holds wonders for the dispensation of medical care. With personalized genetics coming into vogue, the day is not far off when every person is born with a genetic map. Just imagine the power of understanding why one particular drug works miracles for 99 patients but causes a fatal Steven Johnson Syndrome in the 100th! The possibilities are limitless and VK has dealt with them at depth in his post.
Which is why, for the sake of intellectual stimulation and stirring a debate, I am going to take a counter-position to his take, trying to find out what VK has glossed over in painting this immaculate picture of a perfect future! Oh well, this is gonna be a hard job… but here are the reasons why I think VK’s article makes for good, sensationalistic reading, but not realistic material, at least for now…
Dr. Greg House is a Bad Medical Model to emulate.
He states: “…we will be aiming to produce doctors like Gregory House who solve biomedical puzzles beyond our best input ability.” Bringing up the analogy of House he is clearly trying to curry favor with the media and the common man entranced by this paternalistic, misanthropic doctor. In my opinion, he dilutes the power of his observations by making this comment. If anything, the ideal world will have as few of Dr. Houses as possible. I maybe looking too much into an informal comment but in articles that take such a radical position such lapses are not worth the chuckle or media ripple they generate.
It is not that bad a situation.
The article somehow works on the principle that 80% of the doctors today are expendable. In my opinion that is a number picked out of a hat to satisfy the flow of logic in the article. In a piece where he touts the supremacy of evidence, he brings about a 20/80 dichotomy of doctors based on quality which seems to have been conjured out of thin air. The article moves on the assumption that medical professionals today are more or less replaceable with automatons because they are inept or below average. If that is the case, then why should we not extend the same logic to Judges in the Courts of Law?
It is not an either/or situation.
Rather than fighting over doctors OR algorithms, the right way to go would be Doctors WITH algorithms. In his article in the New Yorker about Peter Pronovost, the mercurial Atul Gawande waxes eloquent about the power of a simple Check List in improving the delivery of critical care. Extending this evidence, the power of algorithm-ized medical profession seems unbounded.
However, I am of the opinion, that even if VK’s dream world does come true, the need for doctors will not be obviated. They would still need doctors to test the technology, to keep quality assessments in place and to keep improving the system.
Adopting a very Engineering Look:
The other problem that I think holds the article back is that VK takes a very “engineering” look at a problem that has so many facets. Before I proceed any further, I would ask you to go through Dr. Abraham Verghese’s wonderful TED talk on the power of the Doctor’s Touch:
Of course it can be argued that the placebo effect that powers this issue is sometimes due to the perceived authority of the doctor (Steve Grimm of Facebook left an awesome comment in this regard), but as a clinical romantic (?rom-antique) I would like to believe that the physician’s healing touch has much to do with the process of cure. To be honest, most diseases are not amenable to cure; we had our chance with the infectious diseases, but now with the emergence of drug resistant superbugs, even that war is being lost faster than the speed of light (on another note, did you hear about the Total Drug Resistant TB cases? I know there is a lot of speculation on the matter, but still, this is a scary thought!). So, in most cases, as a doctor all we do is “keep the patient amused while nature affects her own cure”.
So unless a humanized form of the docomaton is created, with evidence based repertoire of empathy-words, I think the day when the physician’s human touch is dead is still far away. (but to be honest I think that humanized docomaton concept is rather creepy)
The Bug is in the System.
VK mentions a lot of the common plights that patients face when they go for a healthcare check up. Long waits. Regulations, etc. But truth be told these are not impediments stemming from the quality of the doctor, but from the institutional policies and regulations in place. So, the way to cure this is not to make algorithm regurgitating machines but to streamline a defensive, litigious system of healthcare delivery.
The 80% will be the 20%
Another erroneous conclusion VK draws is that but for the best 20% of doctors, the rest will die out. I think (like Rebecca, on the comments) that with the institution of an automated care system in place, the below average doctors will rise to the top. Basic rules of evolution dictates the survival of the fittest, and as it is, doctors are more often than not the best of the intellects out there. So, with this pressure to evolve, I believe doctors would fare better than most. Maybe not all of the “80%” average/below-average docs, but I would wager a healthy number would stay on!
Regulations, Liabilities, Opposition.
This technology has the potential to make the very profession extinct. So it would be nothing but naivety to not expect the medical associations to fight tooth and nail over it. As of now, it takes years, often a decade or so, for a new molecule to cross the various barriers of regulations put in place to become recognized as a drug. I wonder how much time it will take for a system that obviates the necessity of a doctor to pass through these treacherous waters of regulatory red tapism.
In conclusion:
However, two of the strongest points against VK’s vision are the destruction of empathy and the limitation of technology. I, for one, do not believe that any argument that bases its foundation on the limitations of technology available today will stand strong for too long. And I also believe come what may, people would rather get news of terminal disease or death or even good news from another fellow human than from a computer.
A time when doctors become second fiddle to machines may not be far off. As it is, today our over-reliance on investigations and testing and mechanical confirmation of our clinical conclusions has become an integral appendage of medical care. So, with the rapid strides genomic medicine, proteomics and artificial intelligence is making, the day may not be far off when doctors are just the face of Medicine while machines run the works.
