Sorry it's been a while -- but with the musculoskeletal final safely behind me I can turn to RMV again...
One question that occasionally passes through my mind is "What is medicine going to look like 30 years from now?" In my 50's, I'll probably be at about the highest arc of my career, but will that career be recognizable? After all, scientific progress proceeds exponentially -- in the 2040's, the present era will probably seem just as quaint as the 1890's do now. Many older physicians tell me that a great deal of what they learned in medical school is obsolete, a trend that has no sign of fading.
What brought this to mind was an article in Wired magazine (sorry, not online yet) about the Archimedes organization, whose goal is to produce a computer program that can quantitatively model the human body, completely. Quite an ambitious project, considering the untold intricacies of human physiology -- here is a chart that shows JUST the metabolic reactions! If successful, this would be a holy grail for medicine. Every drug's effect could be simulated before touching a human. Combined with genetic information, treatment algorithms could be personalized for each individual. Already, the designers have used their system to predict the outcomes of major randomized drug trials, successfully.
I personally think this was a great thing, but the article was not as positive. Will a computer take over the role of a physician? I doubt it. Physicians have always use tools, tests, and rubrics; this will just be a very powerful one. A human being is still the most adaptable and easily programmable computer we have, and physicians will always be needed to interpret for patients the results their science gives them. The doctor of the future will probably need to know much more about statistics then they are currently asked to -- the spheres of epidemiology, computer science, and individual medicine are rapidly colliding.
Interesting how our immediate response is to feel threatened by new technology, even when it could be extremely beneficial. In one of my classes recently, the professor (a clinical psychologist) mentioned that computers have been proven to be more accurate than humans at diagnosing mental illnesses. Humans tend to jump to conclusions and then only see the evidence that supports their idea. People also tend to ignore probability, diagnosing rare or unlikely disorders when there might be a more mundane explanation. In this sense, a computer could actually be the ideal clinician - weighing all the symptoms equally, and assigning the most probable cause objectively. However, despite the fact that machines could do a much better job of diagnosing patients than a human psychologist, nobody uses this method. It wouldn't put anyone out of a job - you'd still need to see the therapist - but it might just help in identifying the most effective treatment. Obviously there are some cases where a human can do a job better than a robot, but when we find a place where technology can save or improve lives, it seems unethical to refuse it!
ReplyDelete