I just watched a video about a software system called Watson (as in Sherlock Holmes's faithful and bright, but not nearly as brilliant, assistant). Developed by IBM in partnership with Memorial Sloan-Kettering Cancer Center, Watson is designed to assist doctors in their evaluation of patients, diagnosis of their conditions, and assessment of possible treatment plans. (To learn more, see the 8-minute video about Watson here.)
Even before the video started, I wondered if this was a good idea or a bad one. Either way, I knew Watson, or something like it, would be soon calling the shots in my healthcare (because of the relentless march of technological progress, not because of any governmental healthcare plan), so I'd better be prepared to take advantage of its benefits and shore up its weaknesses.
THE BRIGHT SIDE
Vast knowledge available quickly. At the top of my benefits list is the system's ability to search a vast array of medical books, journals and papers. Watson (or its imitators) could almost instantly seek out even the most obscure references to the symptoms presented and bring them the the physician's attention. Just being aware that all such references even exist would be impossible for a human being, let alone reading them all and recalling all their details months or years later.
This is a good thing. It means I don't have to worry so much about Dr. Smith's insomnia or hormone deficits or marriage problems that might degrade his/her ability to think clearly about my particular health issues.
And since I am one of those control-freak patients doctors hate, I too can use Watson's snapshots and drilldowns of data to see for myself what relevant evidence is out there. I may not understand it all, but my little bit of college pre-med and many years of medical research and writing will help me know what questions to ask.
Not (as) blinded by drug company propaganda. For decades, busy physicians have focused on seeing as many patients as possible in a given day, and had little time to read about advancements and revelations in their field. Many relied on the updates they got from their drug company reps, which naturally skewed the news in favor of the company's new and obscenely expensive drugs. With a tool like Watson, doctors can just as quickly, and more effectively, summon up not just the latest news and revelations, but older information that may have been obscured by the drumbeat of Big Pharma's sales pitch.
The caveat to this benefit is that even Watson's perspective will be skewed to some extent because Big Pharma sponsors the writing of (i.e., influences the content in) medical textbooks. However, I'd say that having easy access to ALL information (assuming the programmers feed everything available into Watson) still gives doctors and patients an advantage they don't have now.
THE DARK SIDE
Is there a Dr. House in the house? The death of the medical detective. With more patients pushed through the meat grinder healthcare system (because of the aging baby boomer hoards, Medicare payment cuts, fewer doctors going into family practice), tools like Watson will almost certainly replace the human medical Sherlocks whom Watson was designed to assist.
In this future, doctors like Gregory House will not be the ones who take on stubborn or mysterious cases, they will not awaken in the middle of the night with an a-ha! insight, or hear an unrelated comment about someone's lunch or pest-control problem and suddenly put together disparate puzzle pieces into a brilliant revelation that saves the day. In all likelihood this job will fall to Watson - emotionless, data warehousing, synthesizing and summarizing Watson. (Which may be preferable to being treated by a surly, rude, emotionally stunted Dr. House.)
I suspect that those stubborn or mysterious cases will be handled by everyday doctors who rely more and more on their Watsons for not just the legwork but for the skullwork as well.
And who knows, maybe we'll get better treatment and our everyday physicians may get better at being detectives when all the information is at their fingertips.
Doctors can "phone it in." This is not to say that with a tool like Watson all doctors will simply follow the computer's recommendations, but many no doubt will, especially on busy days.
I have already seen how cyber media (especially sites like Facebook and Twitter) tools have dramatically increased the shallowness of our conversations and our thoughts. And now it seems that some folks in politics want critical-thinking skills banned from our schools and our brains. We are rapidly become a society that doesn't think, doesn't probe, doesn't challenge or research or analyze. The only ones doing those things are the ones who want to manipulate us.
So it seems not only possible but likely that given a tool like Watson, doctors will quit carrying around the body of knowledge that today they are trained to store in their brains. Why force yourself to remember enormous volumes of data - cases and statistics, treatments and outcomes, odd little outliers that once resolved obscure sets of symptoms - when you can just summon up the relevant bits of information with a few keystrokes?
I predict that our doctors of the future will be far less able to think on their feet and less able to diagnose and treat without the aid of their computers. However, for those who don't need to fly blind, they may, in fact, have more successful treatment outcomes overall.
Dependency on the technology. Pilots who rely on instruments to land in poor visibility conditions may crash in those same conditions if their instruments are out. However, if they know their instruments are out, they will typically opt not to fly in the first place. And the reliability of modern instrumentation allows people to fly safely in a much wider range of conditions that ever before. Occasionally a pilot may be stuck with an unexpected instrumentation failure and have to get creative to land the plane. As long as he/she has the skills to handle emergencies, that's what matters.
The question is: Will our Watson-assisted doctors of the future be trained to handle situations when their computers fail? Will they recognize spurious data and recommendations that are evidence of viruses or hacking? Will they be able to treat people in emergencies when computer assistance is unavailable or power supplies are waning?
Voice dictation errors. The Watson video shows how physicians can dictate their notes directly into the computer. For anyone who has used the voice-to-text feature on your smartphone, you know the accuracy is laughable at best. Certainly a medical Watson will be trained to understand each specific user and will have in its databases thousands of words few normal humans could even pronounce.
But with similar sounding words and drug names, and new drugs and terms springing up daily, it will be a nightmare keeping Watson's database current. And the software will need to recognize not just the spelling and definitions of all those new terms and drug names - and their pronunciations, it will need to recognize the specific doctor's pronunciation of those terms/names.
Granted, Watson will allow doctors to read and make corrections - and will theoretically highlight terms/names that may be similar to others or that are not consistent with the context. Still, the ease of direct voice transcription (with all its flaws), coupled with the increasing pressure on doctors to go faster... may collide in an ongoing "perfect storm" where dictation errors never get corrected.
No outside-the-box, holistic information. It's possible that Watson's managers will see fit to upload holistic and complementary medicine's insights into its databases alongside its mainstream information. But in this respect, if the alternative data are left out, Watson will be no worse than the average (or even exceptional) physician who is locked in the mainstream mindset. And open-minded doctors may request alternative data modules for their applications, which I'm sure IBM or other developers will be happy to provide.
BOTTOM LINE
Watson and its kind are coming. I have no doubt of that. So what will I do about it? I will try to get access to the same information my doctors have when I feel I need it. (And any Watson worth its salt should make it easy for the doctor to print out key information for patients who request it.) I will continue to question every treatment recommended and explore alternative approaches. I will view drug hype with great caution. And I will monitor both my records and any prescription orders to ensure that no errors harm me.
Overall, I suppose I can either be a Luddite and resist change, or I can embrace the new technology and make the best of it. I choose the latter, following the rule: "Know what you're doing and do it on purpose."
Even before the video started, I wondered if this was a good idea or a bad one. Either way, I knew Watson, or something like it, would be soon calling the shots in my healthcare (because of the relentless march of technological progress, not because of any governmental healthcare plan), so I'd better be prepared to take advantage of its benefits and shore up its weaknesses.
THE BRIGHT SIDE
Vast knowledge available quickly. At the top of my benefits list is the system's ability to search a vast array of medical books, journals and papers. Watson (or its imitators) could almost instantly seek out even the most obscure references to the symptoms presented and bring them the the physician's attention. Just being aware that all such references even exist would be impossible for a human being, let alone reading them all and recalling all their details months or years later.
This is a good thing. It means I don't have to worry so much about Dr. Smith's insomnia or hormone deficits or marriage problems that might degrade his/her ability to think clearly about my particular health issues.
And since I am one of those control-freak patients doctors hate, I too can use Watson's snapshots and drilldowns of data to see for myself what relevant evidence is out there. I may not understand it all, but my little bit of college pre-med and many years of medical research and writing will help me know what questions to ask.
Not (as) blinded by drug company propaganda. For decades, busy physicians have focused on seeing as many patients as possible in a given day, and had little time to read about advancements and revelations in their field. Many relied on the updates they got from their drug company reps, which naturally skewed the news in favor of the company's new and obscenely expensive drugs. With a tool like Watson, doctors can just as quickly, and more effectively, summon up not just the latest news and revelations, but older information that may have been obscured by the drumbeat of Big Pharma's sales pitch.
The caveat to this benefit is that even Watson's perspective will be skewed to some extent because Big Pharma sponsors the writing of (i.e., influences the content in) medical textbooks. However, I'd say that having easy access to ALL information (assuming the programmers feed everything available into Watson) still gives doctors and patients an advantage they don't have now.
THE DARK SIDE
Is there a Dr. House in the house? The death of the medical detective. With more patients pushed through the meat grinder healthcare system (because of the aging baby boomer hoards, Medicare payment cuts, fewer doctors going into family practice), tools like Watson will almost certainly replace the human medical Sherlocks whom Watson was designed to assist.
In this future, doctors like Gregory House will not be the ones who take on stubborn or mysterious cases, they will not awaken in the middle of the night with an a-ha! insight, or hear an unrelated comment about someone's lunch or pest-control problem and suddenly put together disparate puzzle pieces into a brilliant revelation that saves the day. In all likelihood this job will fall to Watson - emotionless, data warehousing, synthesizing and summarizing Watson. (Which may be preferable to being treated by a surly, rude, emotionally stunted Dr. House.)
I suspect that those stubborn or mysterious cases will be handled by everyday doctors who rely more and more on their Watsons for not just the legwork but for the skullwork as well.
And who knows, maybe we'll get better treatment and our everyday physicians may get better at being detectives when all the information is at their fingertips.
Doctors can "phone it in." This is not to say that with a tool like Watson all doctors will simply follow the computer's recommendations, but many no doubt will, especially on busy days.
I have already seen how cyber media (especially sites like Facebook and Twitter) tools have dramatically increased the shallowness of our conversations and our thoughts. And now it seems that some folks in politics want critical-thinking skills banned from our schools and our brains. We are rapidly become a society that doesn't think, doesn't probe, doesn't challenge or research or analyze. The only ones doing those things are the ones who want to manipulate us.
So it seems not only possible but likely that given a tool like Watson, doctors will quit carrying around the body of knowledge that today they are trained to store in their brains. Why force yourself to remember enormous volumes of data - cases and statistics, treatments and outcomes, odd little outliers that once resolved obscure sets of symptoms - when you can just summon up the relevant bits of information with a few keystrokes?
I predict that our doctors of the future will be far less able to think on their feet and less able to diagnose and treat without the aid of their computers. However, for those who don't need to fly blind, they may, in fact, have more successful treatment outcomes overall.
Dependency on the technology. Pilots who rely on instruments to land in poor visibility conditions may crash in those same conditions if their instruments are out. However, if they know their instruments are out, they will typically opt not to fly in the first place. And the reliability of modern instrumentation allows people to fly safely in a much wider range of conditions that ever before. Occasionally a pilot may be stuck with an unexpected instrumentation failure and have to get creative to land the plane. As long as he/she has the skills to handle emergencies, that's what matters.
The question is: Will our Watson-assisted doctors of the future be trained to handle situations when their computers fail? Will they recognize spurious data and recommendations that are evidence of viruses or hacking? Will they be able to treat people in emergencies when computer assistance is unavailable or power supplies are waning?
Voice dictation errors. The Watson video shows how physicians can dictate their notes directly into the computer. For anyone who has used the voice-to-text feature on your smartphone, you know the accuracy is laughable at best. Certainly a medical Watson will be trained to understand each specific user and will have in its databases thousands of words few normal humans could even pronounce.
But with similar sounding words and drug names, and new drugs and terms springing up daily, it will be a nightmare keeping Watson's database current. And the software will need to recognize not just the spelling and definitions of all those new terms and drug names - and their pronunciations, it will need to recognize the specific doctor's pronunciation of those terms/names.
Granted, Watson will allow doctors to read and make corrections - and will theoretically highlight terms/names that may be similar to others or that are not consistent with the context. Still, the ease of direct voice transcription (with all its flaws), coupled with the increasing pressure on doctors to go faster... may collide in an ongoing "perfect storm" where dictation errors never get corrected.
No outside-the-box, holistic information. It's possible that Watson's managers will see fit to upload holistic and complementary medicine's insights into its databases alongside its mainstream information. But in this respect, if the alternative data are left out, Watson will be no worse than the average (or even exceptional) physician who is locked in the mainstream mindset. And open-minded doctors may request alternative data modules for their applications, which I'm sure IBM or other developers will be happy to provide.
BOTTOM LINE
Watson and its kind are coming. I have no doubt of that. So what will I do about it? I will try to get access to the same information my doctors have when I feel I need it. (And any Watson worth its salt should make it easy for the doctor to print out key information for patients who request it.) I will continue to question every treatment recommended and explore alternative approaches. I will view drug hype with great caution. And I will monitor both my records and any prescription orders to ensure that no errors harm me.
Overall, I suppose I can either be a Luddite and resist change, or I can embrace the new technology and make the best of it. I choose the latter, following the rule: "Know what you're doing and do it on purpose."
RSS Feed