MedEd in an AI Era

I’ve been thinking a lot about medical education lately, for two unrelated reasons.  The first is the kerfuffle between US News and World Report and some of the nation’s top – or, at least, best known – medical schools over the USN&WR medical school rankings.  The second is an announcement by the University of Texas at Austin that it is planning to offer an online Masters program in Artificial Intelligence. 

Credit: AAMC

As the old mathematician joke goes, the connection is obvious, right?  OK, it may need a little explaining. 

USN&WR has made an industry out of its rankings, including for colleges, hospitals, business schools, and, of course, medical schools. The rankings have never been without controversy, as the organizations being ranked don’t always agree with the methodology, and some worry that their competitors may fudge the data.   Last year it was law schools protesting; this year it is medical schools.

Harvard Medical School started the most recent push against the medical school rankings, based on:

…the principled belief that rankings cannot meaningfully reflect the high aspirations for educational excellence, graduate preparedness, and compassionate and equitable patient care that we strive to foster in our medical education programs…Ultimately, the suitability of any particular medical school for any given student is too complex, nuanced, and individualized to be served by a rigid ranked list, no matter the methodology.

Several other leading medical schools have now also announced their withdrawals, including Columbia, Mt. Sinai, Stanford, and the University of Pennsylvania. 

Now, I am no expert on the methodology and don’t have any particular love towards USN&WR, but I do find rankings to be informative.  As the USN&WR CEO said in response to the HMS withdrawal: “Our mission is to help prospective students make the best decisions for their educational future…we believe students deserve access to all the data and information necessary to make the right decision.”  I mean, who could argue that?

Evidently the medical schools.  I’ve seen lots of reasons cited for their withdrawals, but what I have not seen are suggestions for alternatives – how to make the rankings better, how to more accurately gauge “quality” of medical schools, how to fairly compare different medical schools.  I guess if you are Harvard or Stanford you believe your superiority is obvious.

I’ve brought this up on Twitter and gotten some interesting responses, especially from physicians – e.g., that medical school attended isn’t an indication of how good or bad a doctor will be, and that medical school actually doesn’t matter as much as where doctors do their residency. Those may be very valid arguments, but they leave me to conclude that we not only don’t know which medical schools are the “best,” we don’t even know if medical school has any real bearing on the quality/competence of the physicians it produces (not that we can measure that either). 

As with most things in healthcare, quality is too complex for the professionals to figure out, so they’ll punt to the patients to figure it out for themselves.

I’ve written before about how, in 2023, it makes no sense that we have parallel educational tracks for M.D.s and D.O.s, or, indeed, that our medical education system takes such a narrow and outdated view towards “health.”  Medical schools and graduate medical education programs have become an end unto themselves, and it’s no surprise that training physicians in the U.S. is a longer and more expensive process than anywhere in the world – not that we can show we have better physicians or those physicians achieve better outcomes as a result, of course.

We should be fundamentally rethinking how we train physicians, which brings me to the UT AI program. 

Credit: UT News

Online graduate school programs are no longer new.  There are a number of them now, for a number of degrees (and, in fact, USN&WR has rankings for them).  It’s not new for UT either; UT started offering an online masters program in computer science in 2019, and in data science in 2021.  But with the explosion of interest in AI, and taking advantage of $20 million in funding from the National Science Funding, UT is now adding this program.

The UT announcement brags that its Master of Science in Artificial Intelligence (MSAI) “will be the first large-scale degree program of its kind and the only master’s degree program in AI from a top-ranked institution to be priced close to $10,000.”  That is considerably cheaper than an in-person program.

The program will not require an undergraduate degree in computer science but candidates will need some technical expertise.  Professor Adam Klivans, director of the new program, told The New York Times the degree was “something working professionals can participate in to learn the expertise their companies need without leaving their jobs.”

He further says:

The fields of artificial intelligence and machine learning have seen unprecedented growth over the last 10 years. Our goal is to ensure that every qualified student can access a premier education in AI, one that is keeping pace with this rapidly evolving field. With the MSAI program, we have removed geographic barriers entirely and significantly lowered the cost barrier of graduate study. For our students, this a game changer.

Eric Busch, director of the Computer and Data Science Online program, added: “It’s not just an ‘online degree.’ It’s an immersive and connected community of learners and a credential from UT Austin that opens doors.”

Healthcare does have many online programs, but not, as far as I can tell, for medical school.  Medical schools are starting to use virtual reality, but only as a training tool, not as a replacement for in-person classes.  They’re tip-toeing when they should be taking great leaps.

Where are the medical schools that are seeking, to paraphrase Professor Klivans, to ensure that every qualified student can access a premier medical education, one that is keeping pace with that rapidly evolving field, to remove geographic barriers and to significantly reduce the cost barrier of medical education?

The future of MedEd? Credit: Medicine at Michigan

So to all the medical schools upset about the USN&WR rankings: yeah, keep worrying about that.  Keep raising your prices, while raising alarm bells about looming physician shortages (and associated need for funding increase).  Meanwhile, someone, somewhere, is going to take UT’s AI example and develop an online medical school program that is more geographically available, more open to a wider range of students, more immersive and interactive, and much cheaper. 

Welcome to MedEd in an AI Era. 

Ultrasound is Ultra-Cool

AI continues to amaze – ChatGPT is now passing Wharton Business School exams, Microsoft and Google are doubling down in their AI efforts – and I’m as big a fan as anyone, but I want to talk about a technology that has been more under the radar, so to speak: ultrasound. 

Ultrasound Direct Sound Printing. Credit: depositphotos/New Atlas

Yes, ultrasound.  Most of us have probably had an ultrasound at some point (especially if you’ve been pregnant) and Dr. Eric Topel continues his years-long quest to replace the ancient stethoscope technology with ultrasound, but if you think ultrasound is just another nifty tool in the imaging toolbox, you’ve missed a lot.

Let’s start with the coolest use I’ve seen: ultrasound can be used for 3D printing.  Inside the body.  

This news on this dates back to last April, when researchers from Concordia University published their findings in Nature (I found out about it last week).  Instead of the more common “Additive Manufacturing” (AM) approach to 3D printing, these researchers use Direct Sound Printing (DSP).  

The paper summarizes their results: “To show unique future potentials of DSP, applications such as RDP [Remote Distance Printing] for inside body bioprinting and direct nano particle synthesizing and pattering by DSP for integrating localized surface plasmon resonance with microfluidics chip are experimentally demonstrated.”

As lead author Mohsen Habibi explained it:

We found that if we use a certain type of ultrasound with a certain frequency and power, we can create very local, very focused chemically reactive regions.  Basically, the bubbles can be used as reactors to drive chemical reactions to transform liquid resin into solids or semi-solids.

The authors believe that DSP can have applications where AM cannot be used, particularly because sound can penetrate objects that light cannot (e.g., the human body).  Bioprinting inside the body is not, in itself, new, but has required open surgery, which DSP would not.  “DSP introduces the possibility of noninvasive deep inside the body printing,” they explicitly point out. 

"Also we can do the repairing of inside bio-organs. That's a future possibility,” corresponding author Muthukumaran Packirisamy said.  Here’s their video:

I’ve been fascinated with 3D printing for a long time, especially for its healthcare-related uses (e.g.., prescription drugs, blood vessels, prosthetic devices, even organs), but tell me we’ll be able to do those noninvasively, using sound waves – well, consider me entranced. 

-------------

All that is cool enough, but ultrasound is being used for many more healthcare applications, such as destruction of tumors.  In fact, that was one of the motivations for the Concordia efforts; Professor Packirisamy noted: “Ultrasonic frequencies are already being used in destructive procedures like laser ablation of tissues and tumours. We wanted to use them to create something.”

“Focused ultrasound” is the term commonly used; there is actually a Focused Ultrasound Foundation. “Focused ultrasound is a noninvasive therapeutic technology,” Dr. Neal Kassell, founder and chairman of the Focused Ultrasound Foundation, told CNN. “We’ve said that focused ultrasound is the most powerful sound you will never hear, but sound that someday could save your life.”

Credit: Focused Ultrasound Foundation

Some of the uses include:

• Earlier this month Canadian surgeons used ultrasound to deliver chemotherapy to an inoperable brain tumor, the first time this has been accomplished.  “Focused ultrasound is an innovative and non-invasive approach to more effectively deliver chemotherapy directly to the tumour,” one of the researchers said. “Our hope is that this continued research will bring us closer to enhancing treatments to help change the course of the disease.”
• Focused ultrasound has been shown to be equally effective in pain management and quality of life measures for painful bone metastases as external radiation beam therapy, with low adverse even rates. 
• Late last year the FDA approved focused ultrasound to treat the second side of patients with essential tumors; use for the first side was approved in 2016.  New research confirmed the long term effectiveness of its use for essential tremors.
• Researchers at UT Southwestern are using high-intensity focused ultrasound to treat medication refractory tremor in essential tremor and tremor-dominant Parkinson’s Disease, which the researchers believe “enables more precise targeting of the brain, decreases treatment times, reduces side effects, and improves treatment response.” 
• Researchers at West Virginia University Rockefeller Neuroscience Institute used Low-intensity focused ultrasound (LIFU) in the treatment of Alzheimer’s patients.  “This study is also a major step forward for the exciting possibility of combining focused ultrasound with targeted delivery of medications or antibodies that normally have limited capability to cross the blood brain barrier from the blood to the brain.”
• Similarly, researchers at Yonsei University College of Medicine (South Korea) found that focused ultrasound improved the delivery of Alzheimer’s drugs by over eight times.  “While there is no complete cure for dementia, we hope that open BBB [blood brain barrier] surgery using FUS surgery can help give hope to dementia patients,” the lead researcher said.
• LIFU is showing “promising results” for treatment of major depressive disorder, according to a paper from Delft University.  The paper describes LIFU as “an emerging neuromodulation method with disruptive potential since it allows for non-invasive stimulation across the whole brain with milimetre precision.”
• Focused ultrasound has been found safe and effective for intermediate risk prostate cancer. 
• A 2020 study suggested that focused ultrasound could be used for patients with depression or obsessive-compulsive disorder.  “We demonstrated that FUS is effective in significantly improving symptoms of patients with treatment-resistant OCD and depression,” the lead author said.

That is by no means a complete list. The Focused Ultrasound Foundation claims that focused ultrasound is currently (at this writing) being used by 65 device manufacturers, for 170 clinical indications, in 424 research sites and 895 treatment sites.  Impressive numbers, but still small in the scheme of healthcare. It warns:

Unfortunately, the evolution of a new therapeutic medical device from concept to standard of care can take decades. Complicated and inefficient, the process requires the interaction of many organizations with differing agendas and timelines. There are also numerous technology, economic, regulatory and reimbursement obstacles to overcome.

Too often, the mechanisms that healthcare has developed supposedly to protect us also work against us.  As the Foundation also warns: “Decades is too long for patients and their families to wait for medical breakthroughs.

-----------

Clifford Marks, MD, writes in The New Yorker about ultrasound replacing the stethoscope, citing miniaturization, lower costs, and application of AI as factors, but I think he’s not being ambitious enough.  As Diku Mandavia, MD, told him, “But ultrasound—it’s low-cost, no radiation, has so much value for patient care . . . it’s going to be ubiquitous.” 

Let’s hope so. 

Fighting the Wrong (Culture) Wars

News flash from the culture wars: they’re coming to take our gas stoves!

Well, actually, “they” are not, but the kind of people who got alarmed about it are a threat to our health, and to theirs.

Credit: Shutterstock

The gas stove furor started with a Bloomberg News interview that Richard Trumka, Jr, a Consumer Product Safety Commission commissioner.  “This is a hidden hazard,” he said. “Any option is on the table. Products that can’t be made safe can be banned.”

He was referring to the well known but little acknowledged fact that gas stoves emit various pollutants, especially nitrogen dioxide. Last year the AMA adopted resolutions about the risks of gas stoves, and urged migration efforts to electric stoves.  Shelly Miller, a University of Colorado, Boulder, environmental engineer has said:

Cooking is the No. 1 way you’re polluting your home. It is causing respiratory and cardiovascular health problems; it can exacerbate flu and asthma and chronic obstructive pulmonary disease in children...you’re basically living in this toxic soup.

So one can see why the CPSC might be concerned. But the outcry about Mr. Trumka’s comments were immediate and vociferous.  “I’ll NEVER give up my gas stove. If the maniacs in the White House come for my stove, they can pry it from my cold dead hands. COME AND TAKE IT!!” Rep. Ronny Jackson (R-TX) tweeted.  The Atlantic further reported:

Governor Ron DeSantis tweeted a cartoon of two autographed—yes autographed—gas stoves. Representative Jim Jordan of Ohio declared simply, “God. Guns. Gas stoves.” Naturally, Tucker Carlson got involved. “I would counsel mass disobedience in the face of tyranny in this case,” he told a guest on his Fox News show.

Ron DeSantis' stoves

Almost as immediately, Mr. Trumka clarified: “To be clear, CPSC isn’t coming for anyone’s gas stoves.  Regulations apply to new products.”  CPSC Chair Alexander Hoehn-Saric issued a statement making it clear that, while “emissions from gas stoves can be hazardous…I am not looking to ban gas stoves and the CPSC has no proceeding to do so.”  The White House issued its own denial.  Case closed, right?

Wrong.  House Republicans have already introduced a bill to block such a ban; 20 states, mostly Republican-controlled, have already passed such bans.  It has become, as Slate put it, “the culture war of the week.”  It joins, for example, masking, vaccines, abortion and climate change as issues that become political divides not on their merits but on the statement they make. 

As Brady Seals, a renewable energy expert at the Rocky Mountain Institute, told Jacob Stern of The Atlantic, “I don’t know if this discourse that we’re seeing now could have happened five years ago.” 

It doesn’t matter that gas stoves may be bad for the health of people in your house; it doesn’t matter that they’re bad for climate change either, with one study equating them to emissions of a half a million gas-powered cars.  Natural gas is good for the U.S. economy, proponents argue, and, in any event, if people want to use gas stoves, they have the right to do so.  Senator Joe Manchin, a Democrat but from a deeply Red state, tweeted: “The federal government has no business telling American families how to cook their dinner. I can tell you the last thing that would ever leave my house is the gas stove that we cook on.

Huh?

It’s similar to the arguments about regulating guns; demonstrably, they’re dangerous for the households they’re in and for the general public, but individuals’ supposed rights to them supersede rational discussion about the risks.  Or abortion; for all the impassioned talk about the sanctity of the life of the fetus, states with more restrictive abortion laws do worse for moms and young kids.

The culture wars about masking, shutdowns, social distancing, and vaccines had real consequences; COVID death rates were higher among Republicans, at the county, state, and national level.  It probably impacted the recent mid-term elections, blunting the expected Red Wave.  Similarly, no one should be surprised that childhood vaccination rates are falling.  We’re already seeing measles outbreaks and we can expect others.

Credit: Politico illustration; photos: Getty, AP

----------------

My “favorite” other recent example of health-related culture wars comes from climate change.  For all the denialism from the fossil fuel industry and the politicians who enable it, a new study found that not only did Exxon know about the risks of global warming for the past 50 years, its scientists had extremely precise predictions about exactly what that impact would be. 

Lead author Geoffrey Supran charged:

This is the nail-in-the-coffin of ExxonMobil’s claims that it has been falsely accused of climate malfeasance…Our analysis shows that ExxonMobil’s own data contradicted its public statements, which included exaggerating uncertainties, criticizing climate models, mythologizing global cooling, and feigning ignorance about when — or if — human-caused global warming would be measurable, all while staying silent on the threat of stranded fossil fuel assets.

The study’s authors concluded: “ExxonMobil understood as much about climate change as did academic and government scientists…Yet, whereas academic and government scientists worked to communicate what they knew to the public, ExxonMobil worked to deny it.”  One has to wonder how many other climate change deniers the same would be true of.  Fighting a culture war against climate change trumps the very real, and apparently widely known, risks of it. 

Exxon, of course, denies these latest findings too. 

Credit: Angela Weiss/AFP via Getty Images

---------------

America knows how to react when attacked by other countries (e.g., Pearl Harbor) or terrorists (e.g., 9/11), but we’re pretty terrible about more insidious risks.  The current pandemic would qualify as a national crisis, but aside from vaccine development and throwing lots of money at it, we’ve handled it pretty badly.  Our public health system is in a shambles a every level, our hospitals and healthcare workers are overwhelmed, and whatever warp speed our COVID vaccine development was at in 2020 is now more like impulse drive.   

And House Republicans and Republican Presidential candidates like Florida Governor Ron DeSantis appear more interested in fighting the culture war aspects of COVID than in, you know, fighting COVID. It’s not our health they’re focused on.

Add up the health risks from all the culture wars and it’d be a pretty scary number.  Culture wars may make great Twitter, but they make bad health policy. 

We're Disrupting Disruption

The Sunday Times featured an op-ed by Mark Britnell, a professor at the UCL Global Business School for Health, with the headline Our creaking NHS can’t beat its admin chaos without a tech revolution.  Substitute “U.S. healthcare system” for “NHS” and the headline still would work, as would most of the content.   

I wouldn’t hold my breath about that tech revolution.  In fact, if you’re waiting for disruptive innovation in healthcare, or more generally, you may be in for a long wait.

A new study in Nature argues that science is becoming less disruptive.  That seems counterintuitive; it often feels like we’re living in a golden age of scientific discoveries and technological innovations.  But the authors are firm in their finding: “we report a marked decline in disruptive science and technology over time.”

The authors looked at data from 45 million scientific papers and 3.9 million patents, going back six decades.  Their primary method of analysis is something called a CD Index, which looks at how papers influence subsequent citations.  Essentially, the more disruptive, the more the paper itself is cited, rather than previous work.       

The results are surprising, and disturbing.  “Across fields, we find that science and technology are becoming less disruptive,” the authors found, “…relative to earlier eras, recent papers and patents do less to push science and technology in new directions.”  The declines appeared in all the fields studied (life sciences and biomedicine, physical sciences, technology, and social sciences), although rates of decline varied slightly. 

Disruption is declining. Credit: Park, et. alia

The authors also looked at how language changed, such as introduction of new words and use of words that connote creation or discovery versus words like  “improve” or “enhance.” The results were consistent with the CD Index results.

“Overall,” they say, “our results suggest that slowing rates of disruption may reflect a fundamental shift in the nature of science and technology.”

“The data suggest something is changing,” co-author Russell Funk, a sociologist at the University of Minnesota in Minneapolis, told Nature. “You don’t have quite the same intensity of breakthrough discoveries you once had.”

It’s not all bad news, though.  The absolute number of highly disruptive works was surprisingly consistent over time, but they make up decreasing percentage of all papers.  Fortunately, “declining aggregate disruptiveness does not preclude individual highly disruptive works.”

Disruption isn't quite dead.  Credit: Park, et. alia

“A healthy scientific ecosystem is one where there's a mix of disruptive discoveries and consolidating improvements, but the nature of research is shifting,” Professor Funk said. “With incremental innovations being more common, it may take longer to make those key breakthroughs that push science forward more dramatically.”

The authors speculated about a couple possible factors for the decline.  One is that many researchers face “publish or persist” incentives that reward volume over innovation.  “A lot of innovation comes from trying new things or taking ideas from different fields and seeing what happens,” co-author Michael Park said. “But if you are worried about publishing paper after paper as quickly as you can, that leaves a lot less time to read deeply and to think about some of the big problems that might lead to these disruptive breakthroughs.”

A second contributor could be a narrowing of scope.  Papers were less likely to cite a wide diversity of sources, were more likely to cite the paper’s authors instead of papers by other authors, and tended to cite older works, “suggesting that scientists and inventors may be struggling to keep up with the pace of knowledge expansion and instead relying on older, familiar work.”  In any event: “All three indicators point to a consistent story: a narrower scope of existing knowledge is informing contemporary discovery and invention.”

Reliance on large research teams has also been cited as a culprit for the switch towards incremental change over disruption; a 2019 paper by Wang, et. alia warned “large teams develop and small teams disrupt.”   

“A healthy ecosystem of science and technology is likely to require a balance of different types of contributions,” Professor Funk told Physics World. “The dramatic declines that we observe in disruptive work suggests that this balance may be off, and that encouraging more disruptive work could help to push scientific understanding forward.”

The research makes me think about a recent article in MIT Technology Review by Shannon Vallor: We used to get excited about technology.  What happened?  “The goal of consumer tech development used to be pretty simple: design and build something of value to people, giving them a reason to buy it,” Professor Vallor laments, “whereas now it is about designing “a product that will extract a monetizable data stream from every buyer.” 

“The fact is, the visible focus of tech culture is no longer on expanding the frontiers of humane innovation—innovation that serves us all,” she worries. “Engineering and inventing were once professions primarily oriented toward creating more livable infrastructure, rather than disposable stuff.”  Now, she fears, it is all about “Take the money and run.” 

Healthcare needs a tech revolution, as Professor Britnell calls for, and I would argue that it needs disruption at every level, and instead we’re getting incremental change instead.  Professor Funk would not be surprised. 

Last year I wrote We Love Innovation. Don’t We?, highlighting research suggesting that for all we say we value innovation and creativity, in practice they can be a threat.  As one researcher noted: “Novel ideas have almost no upside for a middle manager — almost none, The goal of a middle manager is meeting metrics of an existing paradigm.”

Apparently, the same is true for most scientists as well. 

Professor Vallor believes: “When it stays true to its deepest roots, technology is still driven by a moral impulse: the impulse to construct places, tools, and techniques that can help humans not only survive but flourish together.”  There’s nowhere where that should be more true than in healthcare. 

If there has been a common theme in my writing over the years, it has been that there are good ideas outside of healthcare that should be considered for it/applied to it.  The new research from Professor Funk and colleagues confirms that disruptive ideas are out there; they just may not be in the expected field. “Relying on narrower slices of knowledge benefits individual careers,” the authors noted, “but not scientific progress more generally.”

Or progress generally. 

Disruption is hard.  It’s risky. It won’t come from committees or consensus.  But I’m more interested in ideas that jump us to a 22^nd century healthcare system than ones that incrementally just take us another year from a 20^th century one.     

Healthcare, Meet Southwest

Customers experiencing long, often inexplicable delays, their experiences turning from hopeful to angry to afraid they’ll never get back home.  Staff overworked and overwhelmed.  IT systems failing at the times they’re most needed.  To most people, that all probably sounds like Southwest Airline’s debacle last week, but, to me, it just sounds like every day in our healthcare system.

Credit: ABC 7 Chicago

Southwest had a bad week.  All the airlines were hit by a huge swath of bad weather the weekend before Christmas, but most airlines recovered relatively quickly.  Southwest passengers were not so lucky; the airline’s delays and cancellations numbered in the thousands and stretched into days.  Overnight, it seemed, Southwest went from being one of the most admired airlines – loyal customers, on-time service, happy employees, consistent profitability – to one with “its reputation in tatters.”

CEO Bob Jordan has had to do his mea culpas, and he’s not done. 

Some pointed to Southwest’s reliance on point-to-point flight schedules, in contrast to other airlines’ use of hub-and-spoke models, but most seem to agree that the root problem was that its IT infrastructure was not up to task – particularly its employee scheduling system, which told its employees who needed to be where when and how to accomplish that.   Casey A. Murray, president of the Southwest Airlines Pilots Association, told The New York Times: “Once one card falls, the whole house falls here at Southwest.  That’s our problem. We couldn’t keep up with the cascading events.”

Mr. Murray also told NPR: “We're still using not only IT from the '90s, but also processes when our airline was a tenth of the size. And it's really just not scaled for an operation that we have today.” Aviation analyst Scott Hamilton agrees, telling KERA News: “But the policies of being a skinflint on technology fall to Herb Kelleher and every succeeding CEO since then…I think this was benign neglect on the part of the leadership at Southwest across several administrations.”

There’s a term for this, as Zeynep Tufekci wrote in a NYT op-ed: “technical debt.”   Technical debt doesn’t mean that the underlying code is necessarily flawed or poorly written so much as it reflects decisions made at the time of development to defer getting the code “perfect” in lieu of simply getting it implemented.  It’s work that is still going to have to be done someday.

Product Plan has a number of definitions, such as Trey Huffine’s  “technical debt is any code added now that will take more work to fix at a later time—typically with the purpose of achieving rapid gains,” or from Holvitie, et. alia:

Technical debt describes the consequences of software development actions that intentionally or unintentionally prioritize client value and/or project constraints such as delivery deadlines, over more technical implementation, and design considerations

Conceptually, technical debt is an analog of financial debt, with associated concepts such as levels of debt, debt accrual over time and its likely consequences, and the pressure to pay back the debt at some point in time.

Credit: Melv1n

The point is that it is a debt that will come due, as Southwest painfully learned.  “It’s more than just the storm of the century,” said Andrew Inkpen, a professor of management at Arizona State University’s Thunderbird School of Global Management, told Washington Post.  “This exposed them, but they were going to get exposed at some point or another.” 

In a report for CISQ, Herb Krasner estimates that U.S. software technical debt is approximately $1.52 trillion (and the cost of poor software quality is $2.41 trillion).  Shocking but not surprising numbers, since a 2020 McKinsey CIO survey found that:

• 10-20% of the IT budget on new projects was diverted to technical debt issues;
• Tech debt amounts to 20-40% of the value of the entire technology estate (before depreciation).
• 60% felt their organization’s tech debt had risen perceptibly over the past 3 years.

No wonder that Professor Tufekci warns: “Without more government regulation and oversight, and greater accountability, we may see more fiascos like this one, which most likely stranded hundreds of thousands of Southwest passengers — perhaps more than a million — over Christmas week. And not just for a single company, as the problem is widespread across many industries.”

Including, and perhaps especially, healthcare.

Professor Tufekci goes on to blame short term profit incentives, like compensation based on quarterly earning and stock prices, for allowing technical debt to accumulate.  She notes:

…there are strong incentives to address any immediate problem by essentially adding a bit of duct tape and wire to what you already have, rather than spending a large amount of money — updating software is costly and difficult — to address the root problem. Then you can cross your fingers and hope that whatever catastrophe may be in the making, it erupts under someone else’s future tenure.

Ultimately, the problem is that we haven’t built a regulatory environment where companies have incentives to address technical debt, rather than passing the burden on to customers, employees or the next management.

Healthcare didn’t need a bad winter storm to expose its technical debt; it had a pandemic.  From front line caregivers to the healthcare supply chain to the public health agencies intended to lead the fight against such mass health events, healthcare’s technical debts became obvious.  We didn’t have the data, we didn’t have the communication, we didn’t have the right tools, we didn’t have the necessary flexibility.  Our healthcare system didn’t collapse, but it wasn’t far off, and it was only staved off by heroic efforts of healthcare workers and at an uncountable cost of lives. 

llustration: Bráulio Amado for Bloomberg Businessweek

But we didn’t need a pandemic to recognize healthcare’s technical debt.  Healthcare is heavily dependent on software, but find me someone who is happy about any of it -- from billing systems to scheduling systems to electronic records to claims systems to an explosion of digital health “solutions.”  Technical debts are obvious to anyone who has ever had to fill out the same information several times, get a CD to take test results from one healthcare organization to another, or send a fax.  Talk about 1990’s IT!

Healthcare has had its Southwest moment, and, unlike Southwest’s issues, people died.  Most of healthcare’s technical debts “only” result in extra work and/or inconvenience, but some of them have more serious consequences, including loss of life.  It cannot afford to have keep having technical debt issues that someone is pushing off until “someone else’s future tenure,” as Professor Tufecki put it.  

Healthcare CEOs: you do not want to be the next Bob Jordan.  Start paying off your technical debt.