Ben White

Oregon passed a law attempting to ban the corporate practice of medicine. It’s probably not going to work? There are carveouts. It would still allow for private equity to use a puppet doctor to run the show equally poorly (already common practice). It doesn’t address hospitals/health systems taking over the world, which often use the same corporate playbook. And none of it addresses the fact that access to capital, regulation, and reimbursement make it really hard to be small in medicine.

But it’s a start.

// 06.13.25

Incidental Waste

05.26.25 // Medicine, Radiology

From the intro article for a new series on aging, by Joe Nocera in the Free Press:

Every six months for the last few years, I’ve been getting an MRI scan of my brain. There is a little dot in my left frontal lobe—”subcortical white matter,” my chart says, that “may represent a chronic microhemorrhage perennial.” Which, I admit, sounds pretty bad.

Not knowing anything else, this doesn’t matter.

The dot was discovered after I had an incident in which I blacked out while walking home from lunch one day. It was later diagnosed as a simple partial seizure, which also sounds pretty bad. My doctors—and at age 73, I have plenty of them—aren’t 100 percent sure that the dot is connected to the seizure, but they think it might be. Whether it is or not, they want to keep track of the thing, to see if it has moved or grown or done any other bad thing. So far, I’m happy to report, it hasn’t.

I won’t claim any insight into Mr. Nocera’s health, but a solitary chronic microhemorrhage in a 73-year-old shouldn’t be a cause for concern or be followed with serial scans, no matter what the folks at Prenuvo would like you to think.

Information is not understanding or knowledge. Here we have a patient (and his doctors!) who are distressed and utilizing resources for a non-entity: a practice that, more broadly, all radiologists see all of the time.

(I also know some reader is thinking: you know, an LLM could translate that ludicrously adorable phrasing of a microhemorrage “perennial” into English and explain that it is essentially an incidental age-related finding of little clinical significance. Indeed, I asked ChatGPT and, with a lot of other words, said it was “typically of low clinical concern in an elderly patient” and “most clinicians would note it as an incidental age-related change.” [boldface theirs]

While there are peddled software products marketed to translate jargon for non-physician readers, the reality is that imaging begets more imaging, and findings are psychologically incredibly hard to ignore. [I would also add that medical summaries are a fantastic example of a short-term business. AI is not a competitive moat, and this is a product feature, not a product itself.])

Of course we can (and do!) save lives with pictures. But screening narratives are complicated and full of salient, powerful anecdotes, and medical imaging—like this—is so often full of stressful waste.

 

Physician Survey Signup Bonuses

05.06.25 // Medicine

The current welcome bonus landscape:

  • ENOS is a new healthcare panel offering instant payments, (including $5 when screened out) and a $25 sign-up bonus.
  • M3 has three panels–M3 Global Research, M-Panels, and All Global Circle–and you can earn $25 for joining one panel, $40 for two, and $60 for joining all three (for the following specialties: Hematology-Oncology, Neurology, Gastroenterology, Nephrology, Cardiology, Urology, Surgery, Rheumatology, Obstetrics and Gynecology, Pulmonology, Allergy and Immunology, Family Medicine, Psychiatry, Dermatology, Ophthalmology, Endocrinology/Diabetes, and Pediatrics).
  • Sermo is again (finally, after years) offering readers of this site an exclusive $10 welcome bonus.
  • InCrowd is offering $10.

My complete list and more thorough descriptions can be found here.

In addition to being a way to earn extra money (and start a side business that enables you to take some business deductions or start a solo 401k), signing up through these links also helps support my writing. Thank you!

Who or What is the O-Ring? And Who or What Will It Become Over The Coming Years?

04.14.25 // Medicine, Miscellany

From Stripe’s 2024 Annual Letter:

Much as SaaS started horizontal and then went vertical (first Salesforce and then Toast), we’re seeing a similar dynamic playing out in Al: we started with ChatGPT, but are now seeing a proliferation of industry-specific tools. Some people have called these startups “LLM wrappers”; those people are missing the point.

The O-ring model in economics shows that in a process with interdependent tasks, the overall output or productivity is limited by the least effective component, not just in terms of cost but in the success of the entire system. In a similar vein, we see these new industry-specific Al tools as ensuring that individual industries can properly realize the economic impact of LLMs, and that the contextual, data, and workflow integration will prove enduringly valuable.

The “O-ring” of economist Michael Kremer’s original 1993 paper is a reference to the tiny failure point that caused the 1986 Challenger space shuttle disaster. Without diving into obscure references, we could summarize: a chain is only as strong as its weakest link, and, as a corollary: in workflows with multiple interdependent steps, errors can multiply.

The better we as humans perform and the better our systems become, the more each smaller failure begins to dominate the pain and inefficiency landscape.

Stripe is a massive payments processor and works with a large fraction of the world’s companies, and I suspect they are right.

Many have argued that AI models are likely to become the utilities that power new products and less likely to become the dominant products themselves. That remains to be seen, of course, but at least with current LLMs, some specialized wrapping is necessary to make these tools function in high-stakes, specific environments that have real work products outside of making a generic knowledge worker generically more productive or generate some fun art.

Healthcare, in particular, complicates this by adding in a long enterprise sales cycle and layers of slow human bureaucracy.

We’ll have to see what kind of durable moat—if any—any player from this nascent stage has. Capabilities are going up and training costs are going down, so even the simple question of how much wrapping you need to successfully deploy new tools is an impossible question given the unstable, constantly shifting sands at the frontiers of AI. With enough compute and time, perhaps dominant frontier players like OpenAI and Anthropic will simply make models that are robust enough to basically learn and do anything.

Stripe’s argument is that, like a plug adapter, we need those wrappers to solve the O-ring problems for implementing new products and improving processes. As a payments processor, it’s clear why Stripe would want to see a world where lots of companies make great businesses serving lots of other companies.

I’m not sure the current state of the art is useful for making longer-term predictions. Will we see a few novel foundational healthcare models as total farm-to-table solutions? Everyone perhaps choosing from a buffet of numerous efficient, cheaper, smaller narrower models a la carte from a handful of marketplace aggregators, like some of the nascent players have made possible? A few dominant models (e.g. OpenAI) but with numerous wrappers and customers mostly choosing between different implementations (largely ignorant of the machinery under the hood)? Or just perhaps one or two dominant models that are able to be or make everything for everyone? Understanding how long the wrapping stage of AI deployment lasts is likely both of function of how optimistic you are about hyperscaling and also a bunch of idiosyncratic industry, regulatory, and human factors for whatever use-case you care about.

Is wrapping a durable play or just a temporary necessity?

 

Open Loop Errors

03.10.25 // Medicine

From “The how we need now: A capacity agenda for 2025 and beyond,” published by the Niskanen Center think tank:

We need a new operating model for government if we are to restore our capacity to achieve our policy goals. This model must close the open loop we described in Part 3: a one-way system from law- and policy-making to implementation to real world outcomes that offers little space for learning and adjustment along the way. We can no longer rely on media coverage and elections, blunt tools that tend to be saved only for the most catastrophic errors, as the main corrective mechanisms.

Closing the loop means that we must apply test-and-learn approaches. This means conducting multiple small-scale experiments at the boundaries of policy and delivery — and doing this permanently, in pursuit of a policy intent or outcome. Incremental changes are scaled up once there is good evidence they work in reality. Test-and-learn does not mean simply running lots of pilots. A pilot implies starting with a phase for learning, which then ends as you move into “roll out.” Responsiveness is an embedded attribute, not a phase on a timeline. Closing the loop means the learning doesn’t stop at an arbitrary moment.

We briefly touched on this paper before and the concept of the Cascade of Rigidity “that occurs when well-intentioned laws and regulations become increasingly inflexible as they step down through bureaucratic hierarchies.” They discuss a healthcare-related Open Loop error with MACRA:

MACRA (Medicare Access and CHIP Reauthorization Act) was designed to pay doctors more for higher-quality care. But an implementation team at the Centers for Medicare and Medicaid Services (CMS) knew that doctors were already frustrated with the burdensome and confusing ways they had to report their data under the existing program, and many were so concerned that the new system would be just as bad that they were threatening to stop taking Medicare patients. Thus, a law designed to improve the quality of care threatened to degrade it, especially for patients in rural areas who relied on the small practices that were most affected.

Recognizing how challenging the administrative requirements could be for practices with fewer resources and limited Medicare revenue, one provision in the law exempted doctors who treated a minimal number of Medicare patients. But CMS’s initial interpretation of this provision would have required all providers to collect and submit a full year’s worth of data in order to demonstrate they fell below the exemption threshold. This meant exempt doctors would still have to comply with all the program’s requirements, including updating their systems and reporting data, only to be excused from all this at a later date. It’s not hard to see why this approach, while technically accurate, would have worked against the intent of lawmakers. Those doctors would have left the program, hurting the very patients the law meant to help.

Another provision allowed smaller practices to form “virtual groups” to gain advantages enjoyed by larger practices. Staff interpreted this provision as a mandate to create a “Facebook for Doctors,” a platform for doctors to find and connect with each other. A staffer on loan from the United States Digital Service, a part of the White House, doubted that Congress intended for CMS to create a social media platform, especially considering the limited time and resources available. She took the almost unheard of step of consulting the House Office of the Legislative Counsel, and confirmed that Congress simply wanted to make it easier for small practices to report together and had no intention of mandating a “Facebook for Doctors.”

Under more common circumstances, these and other overly literal interpretations of the law would have resulted in a burdensome, unwieldy, and ultimately unsuccessful implementation. Doctors would have simply opted out, leaving patients with fewer options, and some in rural areas with none.

Thanks to nimble actions by people at CMS and USDS to ensure that Congressional intent was realized rather than over-relying on literal interpretations, this outcome was avoided. But conflicts like these all too rarely resolve in favor of common sense. Agency staff are commonly taught to treat legal language as literal operating instructions, as if a programmer had written code and they were the computer executing that code. But as any programmer will tell you, code rarely works as intended on the first try. It works after trying one approach, testing it, adjusting, and continuing that cycle over and over again. That cycle of adjustment is very difficult to engineer within policy implementation today.

We run on an open loop, in which implementation teams neither test their programs in the real world nor loop back to the source for adjustments. We need to build the affordances for them to do both, thus closing the loop. Otherwise, the code will more often than not run exactly as Congress wrote it, even if that doesn’t result in what Congress wanted.

This is emblematic of the problem and also insane: a government-created Facebook alternative for doctors for the express purpose of dealing with a procedural nightmare created by a well-intentioned but completely untested, unproven, almost certainly unhelpful, and very gameable quality goals.

 

It is because doctors are understood to place patients’ interests above commercial ones that they have long enjoyed professional autonomy and public trust. The history of medicine is too littered with incompetence and immorality to believe that doctors have always been worthy of this status. Still, something profound is lost when we submit to the jaundiced view that medicine is a business like any other. There is value in striving for something higher.

From Dhruv Khullar’s “The Gilded Age of Medicine Is Here” in the New Yorker.

// 12.17.24

Screening Narratives

11.25.24 // Medicine

From Nassim Nicholas Taleb’s famous The Black Swan:

A life saved is a statistic; a person hurt is an anecdote. Statistics are invisible; anecdotes are salient.

A notable inversion: medical screening, such as the whole body MRI hyped by Prenuvo (previously discussed here) and Ezra as well as cancer blood tests like GRAIL.

The anecdotal lives saved are the salient stories (look, we found a renal cell carcinoma in Pete, we saved his life).

The harms—false positives, anxiety, wasteful and sometimes dangerous invasive testing and treatment, additional utilization of the overburdened health system, and extra costs of that healthcare distributed across the system—are mere statistics.

 

Paying Surveys for Doctors

11.07.24 // Medicine

Medical surveys are an easy way to make a few bucks at a good hourly rate (well, maybe at least for a resident), and there are multiple sites offering surveys to physicians. The caveat is that, of course, most survey sponsors are typically looking for board-certified physicians with multiple years of experience, particularly in sub-specialties. The less experience you have, the more you need to be prepared to get screened out of what seem like promising survey opportunities.

This article was originally posted way back on Feb 26, 2014 and last updated May 2025. This page contains referral/affiliate links (thank you for your support).

ENOS is a new healthcare panel with a novel premise:  members are paid instantly via Venmo, Paypal, or paper check—no delays or redemption thresholds. ENOS always pays for your time: Even if you’re ineligible for a survey after completing screener questions, they still send you $5. Readers get a $25 sign-up bonus.

One of the biggest survey sites is Sermo (also an online healthcare community), which is now offering my readers a $10 welcome bonus. The survey experience has been recently revamped, and once you maintain a balance of $100 in honoraria, you get preferentially invited to more surveys.

ZoomRx is also excellent and has a nice app and better/shorter-than-average surveys. No sign-up bonus for you, but there is a referral one for me if you’d like to support the site!

One of my very favorites is InCrowd, which has a slick mobile-friendly site and will send you survey opportunities by email or text message. These are always of the very short and painless variety (the fastest of all in my experience), so the payouts are small, but it’s good money for the time and basically effortless. You do have to respond quickly before surveys fill up, but you even get a buck when you get screened out. Being referred (like signing up through that link) will earn you a $10 bonus after you answer your first two microsurveys.

Curizon has been in the business a long time, but they recently completely revamped their website and platform. It’s a trusted site for well-paying healthcare surveys for physicians as well as other healthcare professionals. Every new registration is entered in a monthly drawing for $100.

M3 now has three separate very active research companies under its umbrella: M3 Global Research, M-Panels, and All Global Circle. You can earn $25 for joining one panel, $40 for two, and $60 for joining all three (for the following specialties:  Hematology-Oncology, Neurology, Gastroenterology, Nephrology, Cardiology, Urology, Surgery, Rheumatology, Obstetrics and Gynecology, Pulmonology, Allergy and Immunology, Family Medicine, Psychiatry, Dermatology, Ophthalmology, Endocrinology/Diabetes, and Pediatrics).

At the resident level, one of my old favorites has been Brand Institute, which almost exclusively sends out short surveys about potential drug brand names. Payouts are always on the smaller side ($15), but each one is quick (about $1 per minute or more) and screen-outs are rare. So if you get invited to a survey, then you can generally complete it and get the honoraria. No BS. The main style/format is nearly always the same, so you pick up speed as you do more of them. And that honoraria size is also significantly larger than what one can generally pull as a non-physician (e.g. SurveySavvy, the biggest most popular survey site around, usually pays a measly $2 per survey). The website, however, is clunky and terrible. You’ve been warned.

Additional legitimate additional survey sites, many of which are significantly less active, are below:

A Bias for Unfiltered Input

10.22.24 // Medicine

Henrik Karlsson, from “Scraping training data for your mind,” on the need for the right kind of information to learn new skills:

[You] want the input you observe to be as closely connected to a practice as possible. You want to see the process, not just the results. The results are often misleading. If you are a musician looking at successful bands, you might get the impression that you need to make music videos to succeed, whereas, in fact, a video only makes sense after you have achieved a certain level of success—so imitating that will lead you to misallocate your time.

You will also be led astray if you rely on advice and explanations unconnected to an ongoing process. Experts can rarely articulate the behaviors that allow them to perform at a high level. The knowledge is tacit. The explanations are post-hoc rationalizations; they do not produce the results.

But if you look at the performance, if you get close to it, it is all there.

Looking at the final product to reverse engineer the process can be an impossible task. This is why seeing doctors and surgeons practice medicine and operate is so important. Why it’s so critical to have good readouts when you first start radiology training.

You need to be exposed to the unfiltered expert approach. The “teaching” isn’t just the discrete teaching points and other pearls; it’s seeing the expert practicing natively that helps create the scaffolding to incorporate book learning, real world experience, and ultimately be able to learn the most from the emotional microtraumas of repeated failure that are an unavoidable part of medical training.

Coherent Professions

09.13.24 // Medicine, Miscellany

In The Happiness Hypothesis, Jonathan Haidt describes work by William Damon at Stanford that sought “to see why some professions seemed healthy while others were growing sick”:

Picking the fields of genetics and journalism as case studies, they conducted dozens of interviews with people in each field. Their conclusion is as profound as it is simple: It’s a matter of alignment. When doing good (doing high-quality work that produces something of use to others) matches up with doing well (achieving wealth and professional advancement), a field is healthy.

In their study, modern journalists were suffering in the era of market consolidation and the growing attention economy:

Many journalists who worked for these empires confessed to having a sense of being forced to sell out and violate their own moral standards. Their world was unaligned, and they could not become vitally engaged in the larger but ignoble mission of gaining market share at any cost.

Contrast that with genetics, where doing good and doing well are usually the same thing.

Decide if you think being a healthcare worker in the now-typical corporatized environment reflects a coherent or incoherent profession. This is essentially the premise behind the “moral injury” reframing of physician burnout.

Unfortunately, recognition doesn’t really help because reclaiming coherence is hard:

A coherent profession, such as genetics, can get on with the business of genetics, while an incoherent profession, like journalism, spends a lot of time on self-analysis and self-criticism. Most people know there’s a problem, but they can’t agree on what to do about it.