Ben White

Regarding a study of remote work looking at 500,000 Americans with “remotable jobs,” Prof Emma Harrington, one of the ‘Home Alone’ study’s authors:

When one person is out of the office, it impacts people who remain in the office, so even an on-site worker in a remotable job may now be lonelier because their colleagues choose to work from home.

They conclude that remote work “substantially increases isolation and worsens mental health.” The reality is assuredly varied and more complicated, but the impact of remote work on onsite work is, as suggested, pretty clear.

// 08.31.26

Planning for the Radiology Oral Boards

08.23.26 // Radiology

I have literally no association with the ABR, but since they’ve been gradually publishing more info about the new Oral Boards (coming to a computer near you in 2028) and not everyone reads The Beam newsletter as religiously as I do, I thought I’d share the goods in one spot here as well. Here’s what’s available so far:

1. Written Sample Case content, which walks you through a couple cases in each subspecialty: what you’re supposed to get from the presentation images, differentials, further imaging requests, and key features. Despite being written, this is very helpful.

2. 46 min “simulation” video (where an experienced radiologist playacts being an examinee). Contains 3 cases/attempts (a strong pass, a clear fail, and a marginal fail). To be clear, you can fail individual cases and still pass the exam. Overconfidence is a cardinal sin, and the clear fail is a parade of whiffs. One example contains a partial scrollable CT with some PACS tools. After the three cases, they re-perform them with the opposite performance/outcome.

3. A logistics guide PDF (no, you won’t be able to have a stand-in take the exam for you).

4. An FAQ, of course. One interesting tidbit: all MRI sequences will be labelled. They say:

The ABR Image Quality Committee and the Board of Trustees think that having candidates spend exam time identifying sequences would not add value to the exam experience. The rare exception might be a case where recognizing the specific sequence is the concept being tested (e.g., in-phase vs out-of-phase imaging in the abdomen, or “silicone-only” image on a breast implant study).

Something tells me there will be “rare” exceptions? (They also happened to include unlabeled MRI sequences in the video above.) I would also be totally unsurprised if nucs studies are still often presented as unknowns, even though that would also never happen in real life, but we’ll see.

5. The usual, generally unhelpful “study guide” (which lists the exam content including such profound hints as “pancreas” and “normal variants”).

6. A Scoring explanation:

Performance on an individual case is scored as 68, 69, 70, 71, or 72. Although a candidate does not need to pass every case to pass the category, the average of the case scores must equal or exceed 70 to pass the category. Hence, a candidate who is shown five cases and receives scores of 69, 70, 70, 70, and 71 would pass because the average of those scores is 70.

Those numbers (68-72) correspond to absolutely unsatisfactory, marginally unsatisfactory, satisfactory, good, and outstanding, respectively. In response to the scoring guide, I did reach out to the ABR to ask why the oral exam would be graded so…distinctively, instead of, say, 1-5 or some other more typical scale.

The answer:

First, “70” is often (but not universally) used as a passing threshold for high school and college exams, so it has an intuitive feel for a pass-fail test that is intended to measure competency rather than relative performance such as “excellence” or “perfection.”

Second, a 3 is “200%” better than a 1 (i.e. proportionally “large”).  On the other end of the scale, a score of 5 on a case could be incorrectly understood to be “a lot better” than a 3.  This is less an issue for a 70 vs a 72 although, as you imply, it’s mathematically the same result.

In re-introducing the DR Oral Exam, we’re sticking with the past model to avoid making too many changes as once, and to parallel the RO, MP, and IR scoring that has been in continuous use.  All that said, I think we will consider going with the simpler version of 1-5 at some point in the future.

In other news, what’s old is new again.

Montaigne

08.17.26 // Reading

From Stefan Zweig’s short book on the famous sixteenth-century Frenchman, a passage where Montaigne rales against schooling as the mere transmission of knowledge:

“They never cease whining into our ears,” he gripes, “as if it were being poured into a funnel, and our task is merely to parrot back everything we have been told.” Instead of allowing his own thoughts to ripen in his mind, the masters fill it with a petrified knowledge and leave it “bereft of insight and consciousness”. And he angrily contests: “What use is it to have a bellyful of meat, if one cannot digest it, if it cannot transform us, if it cannot improve us and fortify us?”

…

“As when plants suffocate for too much moisture, and lamps from too much oil, so the action of the spirit is suffocated by an overload of study and matter.” Such knowledge is a burden on the mind, not an action of the soul: “To know by heart is not to know, it is to keep what they have given you and store it in your memory.”

…

Therefore, later, the mature man will hand out a bad mark and a good lesson to those masters who only seek to inculcate rules mechanically: “our masters”, he says in his last years, should judge the aptitude of a child not by his memory but by the testimony of his life… Let the young man read and sift and verify, not just accept the authority in good faith. A rich variety of opinions must be presented to him. He will make his choice, and if he cannot, then he will remain in doubt. He who sheep-like follows another follows nothing. He discovers nothing, because he seeks nothing.

As Montaigne may have hoped, modern pedagogy has moved more toward flipped classrooms and discussion sections. After all, there is a difference between information and education. Interesting, though, is how the cheapening of readily accessible knowledge seems to have stunted educational growth. Perhaps a driven, precocious polymath may not have needed those strict guardrails of a rigid system, but inadequate groundwork and the lack of constraint also don’t seem to, well, work.

Undeniably:

Books are, I find, the best provisions a man can take with him on life’s journey.

Zweig happily points out that Montaigne, like Marcus Aurelius, was learning in public, not issuing decrees from a self-help pulpit:

It is true: Montaigne achieved little else in his life aside from posing the question: “How should I live?” But what is so magnificent, so philanthropic in his case, is that when he posed this question: “How should I live?” he did not insist on following it up with: “This is how you should live!”

…

He had never sought to counsel others on anything he did not know intimately himself. “This is not my doctrine; it is my study, and no one else’s lesson but my own.”

Possible false humility aside, admirable restraint.

Hospital Myopia

08.13.26 // Medicine, Radiology

From independent interventional radiologist Dr. Kavi Devulapalli’s recent article, “Coming to Terms With Reality“:

I can build a program that did not exist before I got there, develop the referral relationships, create the clinic, generate the patients, perform the procedures and produce millions of dollars of new facility economics. The hospital can measure that value, present it to a board, use it to justify investment and congratulate itself on the growth of the service line. When it comes time to figure out what I should be paid, however, we are suddenly talking about how many wRVUs I produced, what fair market value is for call coverage, how many hours I spent doing administrative work, what other employed IRs make and what percentile my compensation falls into.

This is why I increasingly describe the game as rigged. I don’t mean there is some guy at CMS sitting around trying to figure out how to personally ruin my life. What I mean is that once you understand how the pieces fit together, the result becomes pretty obvious. CMS separates the physician economics from the facility economics. The hospital owns the facility and employs the physician. The physician can then create tremendous additional enterprise value for the hospital, while Stark and AKS make it difficult to simply reconnect physician compensation to the downstream economics that physician helped create. Compensation gets pushed back toward the defensible value of physician labor, which is then informed in part by surveys of what other physicians operating inside the same system are being paid. If that doesn’t strike you as at least somewhat circular, I don’t know what to tell you.

Value has always been hard to define, but even if RVUs were a perfect measure of physician work in direct patient care, they would never, ever tell the complete story. The hospitals know this, even if they pretend not to during negotiations or their usual budgetary gerrymandering.

Back in 2018, fresh out of training, I was asked, along with some physician finance bloggers, to contribute an answer to an article about early-career financial mistakes (my post about that post is here). That site is now defunct, but my answer (referring to my work as a student and pre-moonlighting resident) at the time:

“The ‘correct’ answer is usually to spend less on a car, not buy a house, put more away for retirement, and get disability insurance. But looking back, my biggest real regret is not being more purposeful with my discretionary time. Time outside of work obligations has both a monetary and nonmonetary value, and I was very free with spending too much of it to ‘maximize’ my income at rates that are trivial to me now. I was worth more on a per-hour basis back then, even if I didn’t realize it.”

// 08.10.26

Medicine is a Loser’s Game

08.06.26 // Medicine

In his 1971 book, Extraordinary Tennis for the Ordinary Player, Dr. Simon Ramo (also known as, of all things, the “father of the intercontinental ballistic missile”) popularized the concept of Loser’s Games and Winner’s Games:

The notion is that professional tennis is a Winner’s Game, where you win through high athletic achievement and hitting incredible shots, whereas amateur tennis is a Loser’s Game, where you win by avoiding unforced errors (i.e. you win by “not losing”).

Medicine, and especially radiology—which is especially prone to post-hoc criticism thanks to the permanent unchangeable nature of the images themselves—is much more the art of avoiding unforced errors than the art of hitting home runs. After all, common things are common. Horses, not zebras, as they say.

This is the purpose of search patterns and good habits: to build and achieve consistency.

It remains to be seen if and how and when AI might change what kind of game medicine is.

The Purpose of Perception

08.03.26 // Reading

From Douglas Hoffman’s “Interface Theory of Perception” as described in 2019’s The Case Against Reality:

The purpose of a desktop interface is not to show you the ‘truth’ of the computer — where ‘truth,’ in this metaphor, refers to circuits, voltages, and layers of software. Rather, the purpose of an interface is to hide the ‘truth’ and to show simple graphics that help you perform useful tasks such as crafting emails and editing photos. If you had to toggle voltages to craft an email, your friends would never hear from you. That is what evolution has done. It has endowed us with senses that hide the truth and display the simple icons we need to survive long enough to raise offspring. Space, as you perceive it when you look around, is just your desktop — a 3D desktop. Apples, snakes, and other physical objects are simply icons in your 3D desktop.

Interesting.

Steven Pinker essentially summed up the argument well back in 1997:

We are organisms, not angels, and our minds are organs, not pipelines to the truth. Our minds evolved by natural selection to solve problems that were life-and-death matters to our ancestors, not to commune with correctness.

As British statistician George E.P. Box once famously said, “All models are wrong, but some are useful.”