The point of a bone-density scan is the fracture that never happens. The test is quick, it is cheap, it is the one every guideline names — and the population it is built for, Medicare-age adults, has grown every single year since 2005.
So the direction of this number should have been up.
D X A scans per 100,000 Medicare beneficiaries, 2005 to 2023 — down 35.4%
Harvey L. Neiman Health Policy Institute, from a 5% national sample of Medicare fee-for-service claims: nineteen years, more than 5.2 million unique beneficiaries. Published this week in Osteoporosis International.
Meanwhile the odds of getting a D X A rose among patients carrying the heaviest comorbidity burden — up 58% over the period. Follow the two directions at once: the test is drifting away from the healthy-but-at-risk and toward the already-sick. That is a screening test turning into a confirmation test.
Declines were disproportionately larger among racial/ethnic minorities and socioeconomically deprived groups, while use increased among highest-acuity patients. These findings highlight widening inequities in DXA testing and underscore the need for targeted policy and public health interventions.
Between 2006 and 2010, Medicare reimbursement for office-based D X A fell by more than half. The study window opens in 2005. The decline and the pay cut share a starting line, and the institute draws that connection itself. Association, not proof — but it is the obvious suspect standing at the scene.
Buried in the same claims data is the lever. A preventive-care visit in the prior year multiplied the odds of getting scanned nearly tenfold. Not a new machine. Not a new guideline. An annual wellness visit.
If you run imaging, the audience for this is your referrers, and the message is almost embarrassingly concrete: the patients not being screened are not refusing the test — they are never in the room where it gets ordered. Every referring practice that runs Medicare annual wellness visits is already sitting on the fix.
My position: this is the rare study that converts to a one-paragraph note to referring physicians this week. Screening D X A is collapsing nationally; it is collapsing fastest in your underserved panels; and the single strongest predictor that it happens at all is the wellness visit your practice already bills. Send that note. It costs nothing and it is the whole intervention.
Kept honest. I read the institute’s own release and independent trade coverage, not the journal manuscript — Osteoporosis International holds it behind a paywall. Every figure here appears in the Neiman Institute’s announcement of its own study — the authors’ institution speaking — and matches the independent trade report; per this letter’s rules, that is said out loud rather than papered over. The +58% is odds, not volume — an increase in the odds of the highest-comorbidity group being scanned, not “58% more scans.” Utilization is not appropriateness — this data says who got scanned, not who should have; some mid-2000s scanning may have been excess, and the reimbursement-cut explanation is association, not causation. And the sample is Medicare fee-for-service — Medicare Advantage patients, now most of the program, are not in this data, so the national picture could differ.
Sources, fetched today, August 7, 2026. (1) Harvey L. Neiman Health Policy Institute, New Study Finds DXA Testing Osteoporosis Screening Declined 35% Among Medicare Beneficiaries, the institute’s own release on its own study, August 6, 2026 — neimanhpi.org. Source of the 35% decline, the 7,255 → 4,690 per-100,000 figures, the −67% / −47% / −43% subgroup declines, the −42% among beneficiaries without prior diagnosis, the +58% odds increase among highest-comorbidity patients, the near-tenfold preventive-visit odds, and the >50% office-based reimbursement decline of 2006–2010. (2) Pelzl CE, et al., Harvey L. Neiman Health Policy Institute, in Osteoporosis International — the underlying journal article, which is the origin artifact; it is paywalled and was not fetched, which is why the institute’s release and an independent report are used together and disclosed. (3) Radiology Business, Osteoporosis DXA scans have declined by more than 35%, August 6, 2026 — independent cross-check; figures agree with the institute’s release, including the study-quote passage reproduced verbatim above.
The gold-standard test for catching osteoporosis before the fracture just posted nineteen years of numbers, and they run the wrong direction. DXA screening among Medicare beneficiaries, 2005 to 2023: down 35%. From 7,255 scans per 100,000 beneficiaries to 4,690 — across a population that aged INTO the test the entire time. The distribution is worse than the total: Hispanic beneficiaries: −67% Black beneficiaries: −47% Asian beneficiaries: −43% People with no osteoporosis diagnosis yet — the population screening exists for: −42% And one group's odds of being scanned went UP: patients carrying the heaviest comorbidity burden, +58%. Read those two directions together. The test is drifting away from the healthy-but-at-risk and toward the already-sick. That is a screening test quietly turning into a confirmation test — finding the disease after it announces itself, which is precisely the job screening exists to make unnecessary. When did the slide start? The study window opens in 2005. Between 2006 and 2010, Medicare reimbursement for office-based DXA fell by more than half. Association, not proof — but it's the obvious suspect standing at the scene. Now the part worth acting on. In the same claims data, one variable multiplied the odds of getting scanned nearly TENFOLD: a preventive-care visit in the prior year. Not a new machine. Not a new guideline. The annual wellness visit that referring practices already run. The patients not being screened aren't refusing the test. They're never in the room where it gets ordered. If you run an imaging operation, this study is a one-paragraph note to your referrers, and I'd send it this week: screening DXA is collapsing nationally, it's collapsing fastest in underserved panels, and the strongest predictor it happens at all is the wellness visit you already bill. Kept honest: figures are from the Neiman Institute's own release of its own study (Osteoporosis International has the manuscript paywalled), cross-checked against independent trade coverage. The +58% is odds, not scan volume. And the data is fee-for-service Medicare — Medicare Advantage isn't in it. Pelzl et al., Osteoporosis International, via the Harvey L. Neiman Health Policy Institute, August 6, 2026.
Use the MRP Personal Design System (Signature). Create a 1080x1350 animated piece titled "It Stopped Screening." Bone field (#F4F1EC), Ink type (#141414), Coral Deep (#D9401F) for kickers, labels and small rules; Coral Bright (#FF6A4D) spent EXACTLY TWICE — once on the falling figure in beat 3, once on the final line of the end card. Cormorant for numerals and end card; Manrope Light for labels, captions and footer; Italiana for the SignatureMark. Tall, airy, left-aligned. Max 3 type moves. The piece is TWO NUMBERS MOVING IN OPPOSITE DIRECTIONS, AND WHAT THAT INVERSION MEANS. No skeletons, no bones drawn, no medical iconography, no crosses. Typography only. Sequence (about 41 seconds): 1. (0-5s) Kicker top-left, letter-spaced uppercase Manrope in Coral Deep, types on: "NEIMAN INSTITUTE · OSTEOPOROSIS INTERNATIONAL · MEDICARE CLAIMS 2005–2023". Below, one Manrope Light Ink line fades up: "the test that exists to prevent the fracture". Hold. 2. (5-14s) Two Cormorant Ink figures set wide apart: "7,255" left, then a thin Ink rule draws slowly rightward, then "4,690" sets right. Manrope Light caption beneath: "scans per hundred thousand, nineteen years apart". Then a Coral Deep small-caps label fades in under the caption: "DOWN 35 PERCENT". Hold. 3. (14-23s) Clear to Bone except the kicker. Two lines set stacked, Cormorant, with generous space between: first "no diagnosis yet −42%" — and the −42% warms to Coral Bright (moment 1 of 2). Beneath it: "heaviest comorbidity +58%" in Ink. Manrope Light caption below both: "a screening test drifting toward the already-sick". HOLD, unvoiced, at least two seconds. The inversion is the piece. 4. (23-32s) The two lines dim to pale Ink and hold. Beneath, three short Cormorant Ink figures stack, each preceded by a small Coral Deep rule: "−67 Hispanic" / "−47 Black" / "−43 Asian". Manrope Light caption: "steepest where the barriers were already highest". 5. (32-41s) Clear. End card in Cormorant, three lines, the third warming to Coral Bright (moment 2 of 2): "The fix is not a machine." / "It is the wellness visit —" / "ten times the odds of being scanned." Footer in Manrope Light Ink: "PELZL ET AL., OSTEOPOROSIS INTERNATIONAL · FIGURES FROM THE INSTITUTE'S OWN RELEASE · FEE-FOR-SERVICE CLAIMS — MEDICARE ADVANTAGE NOT INCLUDED". SignatureMark bottom-left: Italiana "Meschelle Peterson" + code63labs. Motion language: fades, one rule drawing between two fixed figures, a dim-and-hold before the subgroup stack. No bounce, no scale-pop, no counters ticking, no glitch. The design rests on beat 3 — two numbers, two directions, held in silence.
FIRST SOURCE · one verified original-source finding, composed for one reader · this edition: radiology and imaging — written for the people whose referrers decide which fractures never happen