Digitizing Medicine

15-24 Digitizing Medicine

Reed Pence: Just about every industry has been changed by computerization and the internet. Some fields have been revolutionized. And a lot of people expected that medicine would be one of them. After all, doctors have to know incredible amounts of information. They work with technology like MRIs and robotic surgery all the time. And health records can be extremely complicated. But computerization in medicine so far hasn’t lived up to what people expected.

Dr. Robert Wachter: No, and they were reasonable to think that, because I think we all got spoiled by our iPhones and we just came to believe that computerization in any industry is pretty easy. You take it out, and you download some apps, and off you go. We hope that would be true in medicine. I think, particularly in medicine, because we’re the most information intense industry that you can imagine, and what we do is high stakes and really important to people.

Pence: Dr. Wachter is Professor and Associate Chairman of the Department of Medicine at the University of California-San Francisco and author of the book, The Digital Doctor: Hope, Hype and Harm at the Dawn of Medicine’s Computer Age.

Wachter: We have been surprised, maybe even flabbergasted. There certainly have been wonderful things about it, and I don’t want to minimize that, nor do I believe that we should pull out the cords and bring back the paper. But there have been unintended consequences galore, and patients see that when they go to see their doctor and see their doctor looking more at the computer than at them. And doctors and nurses have seen that as well in that it’s changed their workflow and solved many problems but created a whole host of new ones and I think we just underestimated the complexity and challenges of this.

Pence: The problem, Wachter says, is that computerizing medicine isn’t simply a technical change. That would be a cookbook fix that’s quick and easy. Instead, taking healthcare into the digital realm turns out to be what management consultants call “adaptive change.”

Wachter: Adaptive change is complex change where the technology helps but it also changes everything about the way the people do their work and the way they talk to each other or don’t. This is the mother of all adaptive changes. I think that we underestimated how tricky it is, just the simple act. It’s not that simple of you coming in to see me in the office. Turns out to be an immensely complex act because I need to understand everything about your past history. I need to access the literature of medicine. I need to talk to you and find out a bunch of facts about you. I need to write a note about you, but at the same time, I’m writing about your clinical case. I’m also creating a bill, I’m trying to prevent myself from being sued, I’m also being measured as to whether I’m a good doctor. And when you go from paper and pen to digital you change every aspect of that.

Pence: And so if you simply bring in a computerized version of what’s still really a pen and paper process, you make it worse, not better.

Wachter: What you need to do, and it’s very hard, what you need to do is say, “Let’s rethink everything, let’s rethink the way patients get information, doctors get information, doctors speak to nurses.” And when you do that, you realize just digitizing the old process is not the right way. We have to rethink the entire ecosystem. That takes time and you have to work through the politics and the money and all that, and I think we’re just kind of in those early, clunky version 1.0 phases. And so from the standpoint of, I’d say doctors particularly, it’s sort of more irritating than helpful at this point. And in some ways, the downsides are more obvious than the upsides.

Pence: Picture, for example, what a visit to your doctor’s office is like these days. For a lot of your appointment, he may have his head down, tapping on the keys of a laptop, rather than looking at you.

Wachter: And it makes neither party happy. I mean, the patients get the sense that the doctor doesn’t care about me, the doctor cares more about the digital version of me than the real version of me. And the doctor obviously doesn’t want to do that. We went into medicine because we want to talk to patients. We want to learn about them, and yet the incessant need for… to feed the beast, to feed the computer, is just unrelenting. We’ve not gotten the technology and the workflow right, and part of what’s happened is that, now that there is a computer there, all of these third parties that want to peer into the exam room and be sure the doctors doing the right thing, and be sure the doctor is sending a bill to the insurance company, that’s backed up by what he or she did. Now say well, let’s just ask the doctor to document one more thing, but there are about a hundred and one more things.

Pence: Wachter says technology such as voice recognition in medical data entry programs would help smooth the process, but lacking that, the fastest-growing job position in medicine right now is “scribe,” someone to type into the computer for the doctor while he pays attention to the patient. Electronic medical records (EMR) have become one of the parts of medicine that many doctors hate.

Wachter: The EMR becomes a metaphor for a lot of the other changes that are really transforming the nature of what it’s like to be a doctor. And you may not like the software. You may not like the interfaces. You may not like all the checkboxes you have to check, but it’s also easier to be angry at your electronic medical record which is in your office and invaded your sacred space than it is to be angry at the government which is, you know, 2,000 or 3,000 miles away. So, there’s some of that, but I think we under-estimated. I tell the story in the book of an advertisement I saw last year for an emergency medicine job in Phoenix. And it went through the usual niceties. We got a lovely hospital, a beautiful setting, we’ve got fancy gizmos in the x-ray department. But there was only one part of the ad that was involved and it was clearly their biggest selling point, and it said, “No electronic medical record.” In 2014, a modern hospital thought that their best way of advertising for a new doctor was to say, “You can use pen and paper here.”

Pence: Still, even that hospital will have to go computerized someday. Wachter says the downsides will eventually be worked out, and the upsides are too great.

Wachter: There are huge virtues of not being electronic and one big example is patients have had a legal right for 20 years to see their medical record, but when it’s paper, try getting that. Try exercising that right. Go to a hospital and say “I want a copy of my record.” They’ll say, “well it’ll take us a month to photocopy it at fifty cents a page, and here’s the bill for six hundred bucks.” Once the record is electronic, all you need is a password. And there are about five hundred million patients in the United States under a program called Open Notes that now can not only see their laboratory tests and make appointments and e-mail their doctors, but they can actually read the doctor’s note. And I think that’s a healthy trend. I think you know the degree to which computers have democratized everything, from ordering a cab to making your train reservations or managing your finances. The same thing is beginning to happen in medicine. A lot of these are really wonderful trends. No one wants us to go back, but we have to do it more thoughtfully and with more attention to the nature of the work and what it feels like to be a user of these systems. I think that has not been emphasized in the early development of these machines

Pence: However, increasing patient involvement in their own care isn’t a smooth process, either. Doctors have to take care of more patients who’ve checked out their symptoms on the internet and are convinced that they’re now experts.

Wachter: My son last week called me (he’s 22 years old) called me and he was convinced that he was having diabetes and an attack of glaucoma because he made the mistake of going on WebMD and looking up his symptoms. And I had to talk him down and say, “no you don’t, you have a mild migraine.” This idea that patients can take over medical care, I think there are parts of medical care they probably can, and I think that’s good and healthy and will be cheaper. But there are parts of medical care that really do require, at least now require, an expert and then from the standpoint of the doctor-patient relationship well I’d rather that the patient come in and see me and say, “I looked this up on the web can you help me with it?” I’d rather that than them acting independently on it. But the web is a dark and mysterious and kinda crazy place and there’s a lot of stuff there that’s amazingly wonderful and accurate and there’s a lot of stuff there that’s crazy.

Pence: Computerization in medicine was also supposed to eliminate mistakes, and Wachter says it’s done that to a degree.

Wachter: I would not go to a hospital that didn’t have a computerized medication system, computerized ordering because the mischief in the old days of someone not being able to read my handwriting and giving the wrong dose is very real and we’ve killed a lot of people and so we have to do this.

Pence: However, digital systems have replaced some of those mistakes with new ones. For example, at his own hospital a few years ago, Wachter says a teenager received a 39-fold overdose of a common antibiotic. He lived but landed in intensive care for a week after a seizure.

Wachter: And it begins innocently enough. The doctor didn’t recognize the screen was set on milligrams per kilogram, meaning that the drug dose times his body weight as opposed to just pure milligrams. And part of this was a software problem. It should be unbelievably obvious which mode you’re in, milligrams versus milligrams per kilogram, in the same way, that it’s obvious that your caps lock key is on. But what happens subsequently was really breathtaking, and really illustrates the nature of the problem. Alerts fired to the doctor and ultimately the pharmacist and they both ignored them clicked out of them. And you might say “How could they be so careless?” Well they get tens of thousands of alerts a day, umm, in a month, and the vast majority of them is false alarms, so of course, they learn to click out of them the same way you don’t read the instructions when you download a piece of software.

Pence: Wachter says in the old days, common sense might have prevailed. The order would have gone to a pharmacy technician who would have to pour out 39 pills for a single dose. That might have prompted him to say, “something’s not right.” but now, Wachter says, the order goes to a robot.

Wachter: And the robot says, “You want 39 pills? I can do that.” And does that. Then it goes to a young nurse who sees an order for 39 pills, says, “This is a little bit strange.” But she says, I know to get to me it has to go through a doctor and a pharmacist and the robot and the alerts, it’s probably right. And then she barcodes it and the barcode, by that stage in the medication process, its job is to defend the order. In other words, if the pharmacist and the doctor said this is the right dose, the barcode machine signals to the nurse that that is the dose: 39 pills. And so she barcodes t39 pills. Think about that the next time you’re in the supermarket. Thirty-nine, one after another, she barcodes and gives them to the kid. That’s the nature of the kinds of errors we’re seeing. It’s not that the software is wrong, per se. But the complexity of the interaction between the people and the computers is breathtaking.

Pence: Wachter says in some other fields of business, the computerization process is much more mature. In aviation, for example, designers try out new software in flight simulators before it’s ever built into a plane.

Wachter: Other industries that are more mature about this know that you can be the smartest engineer in the world, but what you have to do is build the computer system that you think will work, and then watch the users use it for hundreds of hours because you are constantly shocked. You say, “Huh, I thought that was a good idea to fire the alert then.” But it turns out it’s a false alarm nine and a half times out of ten, and therefore it’s dangerous because every false alarm makes it less likely that people will pay attention to the real alarm. I better take that one out. In healthcare, we have not done that. The software manufacturers have not paid sufficient attention to the nature of the work, so you have systems that seem like a good idea if you’re an engineer at a software company. But in the real world of doctors and nurses and sick patients really don’t work very well and in some ways create safety hazards.

Pence: Wachter says it’s inevitable that computers will move into more and more portions of medicine. But they’ll change much more than just the work performed on the screen. They’ll change everything. The sociology, geography, and communications within hospitals and doctor’s offices will be altered as well. That transition will take longer. And it won’t be until then that computers in healthcare will reach their potential.

Wachter: Other industries that computerized and didn’t predict how long it would take for the computer to achieve their benefits, what’s known as the productivity paradox. They’ve all seen the same things, which is you bring the computers you think, “Wow it’s gonna have all these benefits within a year or two because we’ve seen that on paper or in PowerPoint slides.” And then the computers go in and for a year or two or five not much happens, and in some ways, you see some bad things. And then in year ten or fifteen you start seeing the magic, and the folks who have written about this have made clear that magic is the computers getting better technologically a little bit, but much more importantly it’s re-imagining the work. It’s recognizing that the workflow has changed, and we better think hard about what we want the workflow to be. How do we want doctors to communicate with nurses, how do we want frontline clinicians to talk to radiologists? And now that we have these digital tools, how do we re-imagine that in a digital environment. We’ve not yet done that hard thinking.

Pence: But Wachter believes that eventually, we will. You can find out more about Dr. Robert Wachter and his book, The Digital Doctor, through a link on our website, radiohealthjournal.org. You can also find archives of our shows there, as well as on iTunes and Stitcher. I’m Pence.