National Board of Physicians and Surgeons

There is a line often attributed to Lenin: “The capitalists will sell us the rope with which to hang them.” A recent email from the ABIM suggests they may be building the physician version. It promotes a study finding that physicians with higher LKA scores have patients who are slightly less likely to receive “low-value” services. The payer-friendly message is obvious: higher-scoring, board-certified doctors may cost less. But the doctors are the ones paying the annual fee, spending the time, and supplying the performance data used to make the case What did they claim the study showed? From the blog post - “Patients whose physicians scored higher in the first year of the American Board of Internal Medicine’s (ABIM) Internal Medicine Longitudinal Knowledge Assessment (LKA®) were significantly less likely to receive low-value healthcare services, which cumulatively cost the U.S. healthcare system an estimated $100 billion each year. The authors say the findings, published in JAMA Internal Medicine, highlight an association between greater physician knowledge and judgment and lower use of tests and interventions that provide little to no benefit to patients.” That $100B is doing a lot of heavy lifting. I will come back to that. LKA is a form of maintenance of board certification, where the ABIM sends about 120 questions per year to participating physicians. Think of it like a timed, open-book “pop quiz” for doctors. The actual publication was titled: “Low-Value Services and Longitudinal Knowledge Assessment Performance” jamanetwork.com/journals/jam… What did that study show? I only have access to the abstract and a few press releases, but I could glean a few details The study was not a trial of LKA. It was a cross-sectional analysis of about 900K Medicare beneficiaries cared for by ~7,000 outpatient general internists who had enrolled in LKA. Its primary endpoint was whether a patient received at least one of 25 claims-defined “low-value” services. Low value services were largely the usual suspects, such as you might see in the “Choosing Wisely” campaign, including: Certain cancer screening in elderly patients Certain labs such as Vit D testing, homocysteine levels and T3 Certain imaging such as CT for headache and carotid screening Certain meds and cardiovascular testing (specifics not available) The headline difference: 28.6% of patients of physicians in the highest LKA-score quartile versus 31.0% in the lowest quartile received at least one flagged service. That is a 2.4-percentage-point difference. That could represent better care. It could also reflect patient mix, local practice culture, health-system rules, or other unmeasured factors. The study cannot tell us whether taking LKA, studying for it, or scoring better on it caused the difference. There was no comparison to non-participating doctors. This is an association from a claims-based observational analysis, not evidence that LKA participation—or improving an LKA score—causes physicians to provide better care, reduces total spending, or improves patient outcomes. At what cost? To participate in LKA, ABIM charges $220 per physician per year But physicians also spend *time* on this. For 120 questions at 5’ a question, that is about 10 hours a year Assuming a doctor’s time is worth at least $250/hour, that is $2500 / year Net, this is ~$2700 per year, per physician. Multiplied by roughly 1M physicians in the US (assuming they all participated) that cost the system of ~$2.7B in direct and opportunity cost What does investing this $2.7B get us (based on this study)? ABIM’s frequently cited $100 billion figure is doing substantial rhetorical work. It is a broad national estimate of low-value care across payers, settings, specialties, and services—not a calculation of savings attributable to LKA or even to these 25 measures. If one makes every favorable assumption—that the observed association is causal, that it generalizes nationally, that each flagged episode produces proportional savings, and that no offsetting harms or costs occur—then the approximately 8% relative difference might be portrayed as an extreme ceiling of roughly $8 billion against a $100 billion national baseline. But that is not an estimate; it is a deliberately overbroad thought experiment. A more scope-concordant benchmark is Medicare spending on similar claims-defined low-value-service measures: approximately $2.0 billion to $5.9 billion annually, depending on how strictly the services are defined (medpac If the study’s 7.9% relative difference were causal and translated proportionally into spending—a series of generous assumptions—that implies gross annual savings of roughly $160 million to $470 million. That is before the cost of LKA, physician time, implementation, downstream utilization, or any missed or delayed necessary care. Compared to the ~$2.7B in physician fees and opportunity cost, it fails to demonstrate real value to the system as a whole. Why did I write all this? ABIM wants people to think that LKA and maintenance of certification have value. This analysis does not show that LKA improves patient-important outcomes. Instead, it offers a payer-relevant association between higher LKA scores and lower use of a selected claims-defined utilization metric The format of this study makes it impossible to show causation, but that does not matter. It is enough that it can be spun that way. Because even though I was the recipient of the email, it is clear who ABIM’s real audience is. The payors. It signals to payors that if they enforce board certification for participation in their panels, it may save them money And that’s why the Lenin quote - because we are basically asking doctors to pay for the privilege of ABIM to do this to us A credentialing system should not ask physicians to finance its own payer-facing value proposition without first showing that it improves outcomes patients actually experience.