The following is an excerpt from the preface of my new book, which is tentatively titled: “Disrupted: Hope, Hype and Harm at the Dawn of Medicine’s New Age.” Author’s note and request to THCB readers.
If you’re a 24-year-old who does not plan on getting sick for the next couple of decades, this is probably not the book blog post for you.
By the time you need our healthcare system, it will be wired in ways we can’t imagine today. By then, computers will have transformed healthcare – as they already have retail, publishing, photography, and travel – leaving it better, safer, and maybe even cheaper. Most of the kinks, perhaps other than what our society will do with boatloads of unemployed dermatologists, radiologists, and hospital administrators, will have been ironed out. I hope to live to see this day myself. It’ll be, as my kids say, hecka cool.
But for the rest of us – both those who need our medical system today and those who currently work in it – the path to computerization will be strewn with landmines, large and small. The challenges are everywhere. Medicine, our most intimately human profession, is being dehumanized by the entry of the computer into the exam room. While computers are preventing many medical errors, they are also causing new kinds of mistakes, some of them whoppers. Sensors and monitors are throwing off mountains of data, often leading to more cacophony than clarity. Patients are now in the loop – many get to see their laboratory and pathology results before their doctor does; some are even reading their doctor’s notes – yet are woefully unprepared to handle their hard-fought empowerment.
In short, while someday the computerization of medicine will undoubtedly be that long-awaited “disruptive innovation,” today it’s often just plain disruptive: of the doctor-patient relationship, clinicians’ professional interactions and workflow, and the way we measure and try to improve things. I’d never heard the term “unanticipated consequences” in my professional world until a few years ago, and now we use it all the time, since we – yes, even the insiders – are constantly astonished by the speed with which things are changing and the unpredictability of the results.
Before we go any further, it’s important that you understand that I am all for the computerization of healthcare. I bought my first Mac in 1984, back when one inserted and ejected floppy disks so often (“Insert Excel Disk 2”) that the machine felt more like an infuriating toaster than a sparkling harbinger of a new era. Today, I can’t live without my MacBook Pro, iPad, iPhone, Facetime, Twitter, OpenTable, and Evernote. I even blog and tweet. In other words, I am a typical, electronically overendowed American.
And healthcare needs to be disrupted. Despite being staffed with (mostly) well trained and committed doctors and nurses, our system delivers evidence-based care about half the time, kills a jumbo jet’s worth of patients each day from medical mistakes, and is bankrupting the country. Patients and policymakers are no longer willing to tolerate the status quo, and they’re right.
For decades, healthcare’s immunity to computerization was remarkable; until recently, in many communities the local high school was more wired than the hospital. But over the past five years, tens of billions of dollars of federal incentive payments have raised the adoption rate of electronic health records from 10% to about 70% in both hospitals and doctors’ offices. When it comes to technology, we’ve been like a car stuck in a ditch whose spinning tires finally gain purchase: so accustomed to staying still that we were totally unprepared for lurching forward.
When I was a medical resident in the 1980s, my colleagues and I performed a daily ritual known as “checking the shoebox.” All of our patients’ blood test results came back on flimsy slips that were filed, roughly in alphabetical order, in a shoebox on a card table outside the laboratory. This system, like so many others in healthcare, was wildly error-prone. Moreover, all of the things you’d want your doctor to be able to do with laboratory results – trend them over time, communicate them to other doctors, patients or families, remind physicians to adjust doses of relevant medications – were pipe dreams. On our Maslow triangle of needs, just finding the right test result for the right patient was a sweet little triumph. We didn’t dare hope for more.
For those of us whose formative years were spent rummaging through shoeboxes, how could we not greet healthcare’s reluctant, subsidized entry into the computer age with unalloyed enthusiasm? Yet an amazing thing happened on the way to Clinical Nirvana. Once we clinicians started using computers to actually deliver care, it dawned on us that something was deeply wrong. Why were doctors and patients no longer making eye contact in the exam room? How did Kwashiorkor – the wasting, belly-bloating condition of children in famine-ravaged regions of Africa – start popping up as a common diagnosis in U.S. hospitals sporting marble lobbies and valet parking? How could one of America’s leading teaching hospitals (my own) give a teenager a 39-fold overdose of a common antibiotic, despite – check that, because of – a state-of-the-art computerized prescribing system? Logically, we pinned the problems on bug-ridden software, flawed implementations, muscle-bound regulations, and bad karma. It was all of those things, but it was also something far more complicated. And far more interesting.
As I struggled to answer these and other similar questions, I realized that I needed to write this book – first to explain all of this to myself, and then to others.
What I’ve come to understand is that computers and medicine are strange bedfellows. Not to diminish the miracles that are Amazon.com, Google Maps, or the cockpit of an Airbus, but computerizing the healthcare system turns out to be a problem of a wholly different magnitude. The simple narrative of our age – that computers improve every industry they touch – turns out to have been magical thinking when it comes to healthcare. In our sliver of the world, we’re learning, computers make some things better, some things worse, and they change everything.
Harvard psychiatrist and leadership guru Ronald Heifetz has described two types of problems: technical and adaptive. Technical problems can be solved with new tools, new practices, and conventional leadership. Baking a cake is a technical problem: follow the recipe and the results are likely to be fine. Heifetz contrasts technical problems with adaptive ones,
“…problems that require people themselves to change. In adaptive problems, the people are the problem and the people are the solution. And leadership then is about mobilizing and engaging the people with the problem rather than trying to anesthetize them so that you can just go off and solve it on your own.”
The wiring of healthcare is proving to be The Mother of All Adaptive Problems. Yet we mistakenly treated it as a technical one: simply buy the computer system, went the conventional wisdom, take off the shrink-wrap, and flip the switch. We were so oblivious to the need for adaptive change that we usually misdiagnosed the problem after failed installations, mangled workflows, and computer-generated mistakes; sometimes we even blamed the victims, both clinicians and patients. Of course our prescription was wrong – that’s what always happens when you start with the wrong diagnosis.
While this is a book about the challenges we’re facing at the dawn of healthcare’s digital age, if you’re looking for Dr. Luddite you came to the wrong place. Part of the reason we’re experiencing so much disappointment is that information technology in the rest of our lives is such magic. Even in medicine, I have no doubt that our awkward adolescence will ultimately mature into a productive adulthood. We just have to make it through this stage without too much carnage.
Of course, if you came looking for breathless digital hyperbole, you won’t find that here either. We are late to the digital carnival, but there are barkers everywhere telling us that this and that app will transform everything; that the answer to all of healthcare’s ills is being developed – even as we speak – by a soon-to-be billionaire twentysomething tinkering in a Cupertino garage. This narrative is seductive; some of it may even be real. But for now, despite some scattered rays of hope, the digital transformation of medicine remains more promise than reality. Having a few Millennials wearing Lycra bike shorts that can read their moods and count their steps is nifty, but it isn’t going to be the change that we need.
What you’ll find in these pages is an insider’s unvarnished view of the early days of the transformation of healthcare from analog to digital, with tales of modest wins, growing pains, and surprising bumps in the road, some the size of elephants. The answer to what ails healthcare is not going to be found in romanticizing how wonderful things were when your doctor was Marcus Welby. We can – we must – wire our world, but we need to do it with our eyes open, building on our successes, learning from our mistakes, and mitigating the harms that are emerging.
To do so effectively, we need to recognize that computers in medicine don’t simply replace my doctor’s scrawl with Helvetica 12. Instead, they transform the work, the people who do it, and their relationships with each other and with patients. Moving from “disruption” to “disruptive innovation” will take deep thought and hard work on the part of clinicians, healthcare leaders, policymakers, vendors, and patients. Sure, we should have thought of this sooner. But it’s not too late to get it right.
Excerpted Preface from “Disrupted: Hope, Hype and Harm at the Dawn of Medicine’s Digital Age” (posted with permission of McGraw-Hill)
Author’s Note
I have been in blog-silence mode of late, for which I am sorry. Rumors that I’ve taken my Elton John act on the road are, I’m pleased to assure you, incorrect.
Instead, I’ve been hard at work on my new book, tentatively titled “Disrupted: Hope, Hype and Harm at the Dawn of Medicine’s Digital Age.” I’m about one-third finished, and am on my way to Boston for a six-month sabbatical at the Harvard School of Public Health to keep working on it.
This is the most journalistic book I’ve ever attempted. I’ve already completed about 25 interviews for the book, and will do about 30 more by the time I’m done. And they have all been fascinating.
It seems a shame to leave so much great stuff on the cutting room floor. So for the next few months I’ll plan on posting some of the best, including interviews with Capt. Sully Sullenberger, Vinod Khosla, the head of Boeing’s flight deck engineering team Bob Myers, Abraham Verghese, Karen DeSalvo, and Gurpreet Dhaliwal.
Yet some of the best stories have come from chance encounters. At the Mayo Clinic, I met a physician who decided to leave his surgical training program for a career in informatics after a fateful internship night in which he found himself running four Code Blues simultaneously. He realized that medicine lacked the systems that we need to access information and communicate effectively. Another physician told me about trying to make sense of the clinical course of a wildly complex ICU patient. The notes in the EHR were such copied & pasted garbage that the only way he could tell what had changed from one day to the next was by printing out the notes and holding one sheet over the others against a window pane. And of course I’ve heard stories about scribes, Open Notes, Big Data, the death of radiology rounds, and much more.
I’m hoping you can help me. If you have an amazing story about how the computerization of medicine has transformed your life or your practice in any way, please do let me know, either by posting a comment (if you’d like to share it) or emailing me at bobw@medicine.ucsf.edu. These kinds of stories can help bring a subject to life.
To give you a sense of the range of topics I’ll be covering, as well as the book’s tone, I’ve pasted below a draft of the preface. If all goes well, “Disrupted” will be published in March, 2015. I’m doing my best to make it a fascinating and important book, and appreciate your help.
Categories: Uncategorized
I completely agree that today’s state of healthcare IT looks like some distant promise than a reality; however, history says that every major change in the human culture always started with some sarcasm, some motivation, some belief, but mostly pessimism from outside world, and same goes for Healthcare IT. Right now, everything is scattered and looks like a mess but with time, when all the pieces of the puzzle will fall in place, that’s the time, the real benefits of digitization of health IT will be clearly visible to all. It’s just a matter of time.
Good point Marva, and a problem that technology will not fix – may in fact paradoxically make worse.
The problem is effectiveness of communication. If I make a consult to someone, I call them directly – AND document the simplest point of the question in short and clear terms. IT does not fix defective communication. It allows the bulk transfer of massive amounts of irrelevant information to a consultant, which is daunting in its sheer volume.
It conditions people towards reduplication of data – if a CT was done at YOUR hospital, perhaps we should repeat it at OUR hospital so it can go into OUR EMR.
I am not sending out a blanket condemnation of using not-paper methods; rather, that poorly-planned EMR/EHR can make problems in communication worse.
PS and the dr. and patient will know promptly if a old or new problem is at hand and act accordingly pronto now not later.
Thank you.
Thank you.
Why should the patient “schlep” all of the paperwork x-rays MRIs ,CTs PET scans etc,that’s the referring doctors responsibility at least it should be but we the patient know we had better get our records together so that when we walk into the referred doctors office and he/she asks “well what are you here for? we can show them and get started or we can hear them say “I don’t have your records” and wait for phone calls and faxes.Machines are only as good as the one operating them.
The layout wasn’t the way I wanted:
“If we can just suppress the elements who are opposing change, then change will happen. Put on a brown shirt and join the fun,” say the bully-boys.
Home run with this – “The issue to me is that EHRs are not the problem or solution; people are not the problem or the solution – as we are, after all, error prone humans. The problem is our philosophy of science and value, and the how we use and conceptualize information. Numbers are meaningless without a model/philosophy of how to use them.”
There is a lovely old Paper publication called “To Err Is Human,” back from 1999 took a solid run at this problem. I thought it was well-written, and fairly human-centered. The press lifted the phrase “100,000 avoidable hospital deaths every year,” and went on to encourage those who make it worse, not better.
There is a terrible fascism of instrumental reasoning that has overtaken the information world regarding medicine. I am an MIT graduate, not a Luddite. But some of the arguments towards “progress” come with the tinkle of broken glass in the background. If we can just suppress the elements who are opposing change, then change will happen. Put on a brown shirt and join the fun.
The issue to me is that EHRs are not the problem or solution; people are not the problem or the solution – as we are, after all, error prone humans. The problem is our philosophy of science and value, and the how we use and conceptualize information. Numbers are meaningless without a model/philosophy of how to use them.
Thanks to everyone for their comments. I’d appreciate any stories — particularly vivid, remarkable, funny, sad, amazing ones — that illustrate some of the unintended consequences.
Today’s example: I was speaking to Barbara Drew, a nurse researcher at UCSF who studies alarm fatigue. In a one month study of every ICU bed at our institution, she found that there were 150 audible alerts PER BED PER DAY. Yes, that’s right, an alarm goes off, on average, every 10 minutes for every patient.
And that was just from the cardiac monitor (HR, BP, RR, O2 sat, ECG wave forms); it doesn’t include bed alarms, IV alarms, vent alarms, all of which are joining in the chorus.
Barbara asked one ICU nurse what the ideal alarm would be — one that would truly, unambiguously, let her know that something was wrong. “If all the alarms fell silent,” she said sadly. That’s what would signal that something was truly wrong.
You can’t make this up.
Thanks for your interest and help, and for the kind words about the preface. It’s a truly fascinating subject — my job is to do it justice.
Joe,
I would presume that if it was a separate billable event before it would be a separate billable event now – whether it was done digitally or not.
And I would presume that if listening to the chest was a separate billable event before (which I am fairly sure it was not) it should still be so.
Is there something I am missing?
Actually, since the iPhone EKG is just a rhythm strip, the doc still has to send the patient down the hall for a full 12 lead.
The real change in the physician’s workflow will be that, since the hospital is in debt up to its eyeballs paying for its EPIC EHR and they’ve laid off 75% of the techs, the MD has to wheel the patient down the hall, hook them up to the machine, do the EKG, and then wheel them back to the exam room, trying to do a little MU busy work at the same time.
Not for free, and not whether it is digital or not, but whether it is a separate billable transaction or just part of the patient visit with the cardiologist, who is of course skilled in interpreting EKGs. When the doctor puts a stethoscope on your heart and listens, is that a separate billable transaction or just part of the patient consultation?
“But a big reason would be that in a fee-for-service system the EKG done by the tech is billable, it’s a revenue source, and the one done by the doc with an iPhone probably is not.”
Why would interpretation of an EKG (which depends on training and is associated with with medico-legal risks) be done for free?
If I send my lawyer a contract to review digitally, will he/she review it for free?
If my Architect draws in CAD and doesn’t print out a hard copy is it free?
Etc.
Now we can argue about what the fee should be, but whether it is digital or not shouldn’t be the determining factor.
You’re not here for the hunting, are you?
Let’s take the EKG. Old way: Your cardiologist orders an EKG and sends you down the hall (or has you come back for a separate appointment) to a tech who performs the EKG, then you come back to the cardio (or have yet another appointment) to interpret the results.
New way: In the midst of the exam, the cardiologist snaps a cover on the back of an iPhone, shows you where to put your fingers on the back, and voila! Your EKG appears on the screen. Faster (nearly instantaneous), way cheaper, and an intimate part of the discussion with the doc. Many advantages.
All the advantages disappear if the doc sends you down the hall to have the tech do it. Yes, to gain the many advantages of medical gadgets, dongles, and apps, clinicians have to alter their workflows, usually in ways that make them simpler, quicker, and easier.
Why would the doc send you to the tech anyway if this new technology exists? In some circumstances, maybe the old EKG gives more information with the differences between its 12 leads, for instance. But a big reason would be that in a fee-for-service system the EKG done by the tech is billable, it’s a revenue source, and the one done by the doc with an iPhone probably is not.
So yes, if healthcare IT is to get us to better and cheaper healthcare, clinicians have to alter their workflows to take advantage of the simpler, faster, less expensive ways of working.
“if we shouldn’t start by
1. envisioning what we *really* want healthcare to look like”
Exactly.
Right now, we have health care that looks like ICD, CPT, P4P, MU, JCHO, and a broken med-mal system. And that is reflected in the HIT we have.
Until we fix the underlying broken structure of health care in this country, we’ll have broken HIT.
But then we come here and are told that, no, HIT isn’t working because doctors and nurses haven’t learned to change their workflows the way writers did when they spent 90 seconds learning to cut and paste . . .
Extremely interesting and well-written post, Bob. Your care with language and your serious thinking about the topic will make for a must-read book. The combination of your experience and the reporting you are doing have wonderful potential to produce a book that helps us figure out a better pathway forward. Cannot wait to read the finished product.
Well Bob, I certainly look forward to seeing what you find out.
I would highly recommend you interview Dr. Larry Weed, and also suggest you read his book Medicine in Denial.
It seems to me that it’s not just that medicine and computers are strange bedfellows. It’s also that medicine itself — the practice AND the business — are riddled with all kinds of terrible problems that make it hard to provide good care that is person centered, allows patients to participate as much as they can and want to, and gets decent outcomes for the money we spend.
Computers were never going to make these problems better on their own. Transparency and information liquidity — facilitated by the Internet & computers — has probably had the most impact on healthcare so far.
I personally wonder if we shouldn’t start by
1. envisioning what we *really* want healthcare to look like (has to be humanly feasible so no saying we want physicians to pay deep attention to panels of 5000, or to do shared decision-making in 10 min visits)
2. thinking about what kind of health IT infrastructure would support this, and
3. then thinking about how to get there from here.
Of course, what we want healthcare to look like partly depends on what tech allows us to do, but still.
I liked the vision Larry Weed outlined in his book, hence my hope that you’ll interview him.
“Do you know any writers who refuse to use cut, past and delete, who just write every word in sequence, write in edits by hand on a print-out, then re-type the whole document again?”
Oh, that’s right, I’d forgotten all those government mandates requiring writers to switch to word processors. And how they had to be paid 40K each to abandon pen and paper, because they couldn’t see any advantages in the new technology and just weren’t going to switch voluntarily.
Good analogy.
I am very familiar with reluctance to change. I was deeply involved in the adoption of PACS, Teleradiology and Voice Recognition in my specialty (Radiology). In many cases I was told how it was; impossible, undesirable and just plain bad to do things differently. And now, none of us would go back.
In my opinion, this phenomenon – not wanting to change – is not any more prevalent in medicine than in any other area. One key difference is that EMRs are (mostly) being forced on doctors by others – and without any concern about lost productivity. (To which Kaiser may be an exception)
I am a 60 year old Physician and during the course of my career have made many technological switches. But all of these switches have helped my productivity/quality. That does NOT appear to be the case with the EMR.
Having used EMR for the past several years, I feel the future is now for computers in medicine. I can look up information from anywhere in the hospital. I can read everyone’s handwriting and when I order a test or medication there is never a ward clerk error. I agree the systems of medical records could be more patient/doctor centered than billing, but that is going to take a change in our Fee for Service payment system. Now that would make a difference in the quality and quantity of care provided.
My blog talks at length about fee for service medicine and its pitfalls.
http://www.lakesidemedicalmusings.com
Absolutely. And the best systems have been redesigned that way. The Kaiser docs I know are relatively happy with their system, and I think that is because teams of clinicians from all across the system were intensely involved in the design at every stage.
But it needs to go both ways. When the available tools change, you redesign your workflow to take advantage of their vast new capabilities. Do you know any writers who refuse to use cut, past and delete, who just write every word in sequence, write in edits by hand on a print-out, then re-type the whole document again?
Joe,
I think Granpappy has a point.
You say: “medicine must be re-designed ……”
Why not: “computers, sensor and databases must be re-designed … ”
The current computerized systems were not designed to help physicians take care of patients, they were designed primarily to capture billing data and meet regulatory requirements. What about re-designing the current systems to primarily help physicians take care of patients with billing and regulatory requirements secondary?
Medicine of the Future:
1) Pay me $10,000 per family, per year, I will take care of your medical problems.
2) your kid needs stitches? I will sew her up on my kitchen table (in a completely sterile field) just as well (or with less MRSA) as anyone in the ER.
3) I will provide you with all your immunizations. I have injectable antibiotics and pain meds.
4) I will be available to you day and night. If I am on vacation, my one partner will see you in my stead. He will have all your records at his disposal.
5) I will keep ALL your medical records confidentially, in my safe in my office. NOTHING will be shared with the federal govt, or on the internet.
6) Unless you are deathly ill and MUST be hospitalized, I will provide your care. IF you have to go to the hospital, I will be at your side to guide you through the system.
7) you will not sue me.
8) I will not charge you for every Tylenol pill, or even for basic items of care like office visits.
– OK – now just substitute the “$10,000” figure of #1 above for $0.0, and you have medicine of the 1950’s and 1960s. Isnt the future great??
Yep, never loose sight of what’s important: Purposely misreading and misinterpreting other people’s posts for the sake of arguing, mocking and whining.
“medicine must be re-designed around the capabilities and problem sets of computers, sensors, and databases”
Yep, never loose sight of what’s important: don’t let those pesky patients and physicians get in the way.
Bob, you nailed it. Great post. You nailed it first by seeing the task as not just complaining about healthcare IT (which is pretty easy), but first understanding the nature of the problem, and then using that understanding to help lead us to fixing the problem.
Second, you nailed it in a key insight: We have treated healthcare IT as a technical problem, when it is first and foremost an adaptive problem, a problem of people. The industry has seen the problem as “digitizing healthcare,” as if we could take this extraordinarily complex human endeavor and just run it through a digitizing machine, doing everything the same way we did before, but now with computers, sensors, and databases. That manifestly does not work. Just as a photographer, a writer, or an oil pipeline engineer has a vastly different workflow than he or she did two decades ago, medicine must be re-designed around the capabilities and problem sets of computers, sensors, and databases. And it must be redesigned by the people doing it — by teams of doctors, nurses, and other clinicians.
It’s a problem we have barely begun to tackle.
There’s something quite millenarian about health IT brigade. Which is why it is so easy to mistake them for a religion.
Perfect analogy.
Or in other words, things are bad now, then along will come health IT and make things worse, then things will get better by becoming bad again.
Great. I can’t wait to return to status quo!
Excellent points and very well-written — will await the release of your book (even as the 20-something who hopes to have a healthy next 2 decades)!
With apologies to Monty Python
My theory is that a brontosaurus is very thin at one end, very thick in the middle and very thin at the other end.
That’s my theory
Interesting to read that Mayo is so understaffed that an intern is left alone to run four simultaneous codes. That’s one for the best practices guidelines.
In 25 years time there will emerge a technology that will rescue us from the disruption caused by widespread computerization.
I think it will be a disruptive technology.
That’s my prediction.
Until it actually is something the doc wants to use, it will bomb. Digital xray is a great thing. Computer prescription writing is not. Computer records are full of fluff and are not searchable. Useless.
When AI finally offers us acute intermittent porphyria and familial Mediterranean fever as differential diagnoses for appendicitis and acute migrating joint pain, respectively, we will finally have some worthy expenses for 17% of the GDP. Talk about lab testing, KRAS, will be one of thousands of new requirements for quality outcomes and value. Our finite real world budgets will finally force us to stop feeding 1/7 of the US as stakeholders of healthcare, and put the dough back from whispy progressive experiments in policy theory to helping trench workers diagnose and treat the correct diseases. But the hoorays for the computerization of medicine will await the solutions for its financing, I’m afraid.
And if you get rid of the radiologist, who are you gonna sue when that tumor gets missed. This is ‘Merica, there has to be someone to sue…
“By then, computers will have transformed healthcare …… leaving it better, safer, and maybe even cheaper. Most of the kinks, perhaps other than what our society will do with boatloads of unemployed dermatologists, radiologists, and hospital administrators, will have been ironed out.”
So I (a Radiologist) am soon to be rendered obsolete by a computer? We’ll see.
I remember talking to a graduate student in engineering in the early 80’s who predicted the demise of my profession in 5 years due to advances in computer technology. I should have made a bet with him – 30 years later I am still working as a Radiologist with no end in sight.
So I say to Dr. Wachter: “Would you like to make a bet about where my specialty will be 10 to 20 years from now?”. And how about your own?
My experience of the computerization of American medicine is to not trust it. I keep detailed paper records for our family, which are scanned and uploaded to a secure drive that I maintain. Here is in St. Louis there are four major health systems. At least two bought the same EHR. They then proceeded to modify the product so much that the two systems cannot communicate with one another. So, if you get referred from a provider in one to a provider in the other, you, the patient, must schlep all the paper because the tinkerers just could stop themselves.
I agree with Granpappy that this mess was completely predictable and our current pathway is not likely to clean it up.
Kind of like what we’re hearing from the Iraq War architects these days:
“Yes, I’ve been 100% wrong on everything so far, but my opinion is still very, very important.”
Sorry, but the disaster that is HIT was completely predictable by anyone who had their eyes open. And listening to Khosla and DeSalvo sure isn’t the way to clean things up . . .