before the five findings
Hospital billing looks like a hospital-wide policy, not a market fact
Charges vary enormously for the same care, barely track what Medicare pays, and the hospitals that bill the most relative to their payments are largely the same ones in both settings. That pattern is the signature of a hospital-wide pricing choice, not a service-level or regional quirk.
- Charges vary enormously for the same care — and barely track what Medicare actually paysscenes 1–2
- The highest-billing hospitals are largely the same ones in both settings — far more than chance would producescenes 3–4
- Outpatient high-charge behavior concentrates in a handful of services, so the leverage points are identifiablescene 5
- Contracting — providers with persistently high charge-to-payment ratios in both settings are the priority negotiation targets: network discounts create the most value there.
- Steerage — outpatient high-charge flags concentrate in a few services; steering members to lower-charge facilities for those services moves the most dollars.
- Screening — extreme charge-to-payment persistence in both settings is a cheap first-pass screen for claims review: a flag for follow-up, not a finding on its own.
descriptive FY2023/CY2023 patterns; non-causal — method and limits on the Approach page
Maximum billing variation in the 2023 exploratory universe
Among the 315 FY2023 inpatient DRGs with at least 30 valid providers, the maximum CV_b — the spread in what hospitals ask to be paid for a service across providers — is 1.5306 for Revision of Hip or Knee Replacement with MCC. The outpatient extreme is Level 2 Urology and Related Services (CV_b 1.1156).
One hospital can bill far more than another for the same operation — the widest gap is in hip/knee replacement revisions.
We followed up on this result beyond this project's data. A peer-reviewed study of chargemasters at the top-20 US orthopaedic hospitals (Pollock et al., JAAOS Global 2023) finds DRG 466's listed price spans $107,582–$472,517 — the widest range of any arthroplasty DRG — with no correlation to Medicare payment, local cost of living, or income. Implant costs alone vary roughly 7× between hospitals (PubMed 22878562), and revision-with-MCC cases are inherently heterogeneous and low-volume. External evidence independently confirms this service as an extreme charge deviator.
descriptive FY2023 maximum; historical value is non-equivalent context
coefficient of billing variation · inpatient max 1.53, outpatient max 1.12
States that charge more don't get paid more
For each service we average hospital charges and payments by state, then ask: do the two move together? Inpatient, the match is only modest (median correlation 0.44 of a possible 1.0); outpatient it is essentially zero (0.11). What a hospital charges is largely its own choice — Medicare's payment barely follows it. (State cells need at least five hospitals; Maryland is excluded.)
Charging more does not mean getting paid more — especially for outpatient care.
FY2023/CY2023 state-service ecological summaries · Maryland excluded, sensitivity documented
state-service ecological correlation distribution · n = 299 current summaries
The same hospitals have the biggest bill-to-payment gaps in both settings
For each hospital, we compare what it bills to what Medicare pays. Among the 2,812 hospitals measurable in both settings, 496 (17.6%) sit in the top quarter of that gap in inpatient and outpatient care alike — far more than chance would produce. This is a billing-vs-payment comparison, not a cost or overcharging measure.
The hospitals with the biggest gap between bills and payments are largely the same ones in both settings.
cross-dataset proxy comparison · FY2023/CY2023 · join on provider_ccn
dual-setting providers · proxy top quartile in both highlighted
High-charge hospitals repeat — 1,134 appear in both settings
Within each service, a hospital is flagged when its bill is in the top quarter for that service. Of the 1,304 hospitals flagged inpatient, 1,134 (87%) were also flagged outpatient (66.63% of the 1,702 outpatient-flagged hospitals). Unlike the gap measure in scene 3, this compares each hospital with its peers on the same service — a list overlap, not a cost or intent measure.
Almost every hospital on the inpatient high-charge list also appears on the outpatient one — the pattern repeats.
project-defined high-charge flags · selection and opportunity effects apply
directional overlap of project-defined high-charge cohorts · orange = both settings
Outpatient high-charging is a concentrated problem; inpatient is diffuse
Half of all outpatient high-charge flags come from just 16 of ~70 services; inpatient flags spread across hundreds of DRGs — the top 16 hold only 23%. The most-flagged service is Comprehensive Observation Services outpatient (689 flags) and Septicemia or Severe Sepsis with MCC inpatient (666 flags). Counts are provider-service flag pairs, not payment dollars, so they are not comparable with OIG's historical 41% payment-dollar benchmark.
In outpatient care, high charging is a few services' problem — in inpatient, it's everywhere. One setting tells you where to look; the other doesn't.
APC 8011 — Comprehensive Observation Services — is the bundled payment for patients held in hospital (usually from the emergency room) while staff decide whether to admit them. It is one of the most common outpatient services, billed by the hour with no standard charge structure, so it tops the flag list largely through volume and billing heterogeneity. It is also one of Medicare's most contested billing areas: the "two-midnight rule" and readmission-penalty pressure give hospitals incentives to classify short stays as observation, and patients can be left with large out-of-pocket bills and no skilled-nursing coverage (AMA Journal of Ethics 2023; OIG OEI-02-15-00020). High flag counts here are not evidence of overcharging.
project-defined flag-pair concentration · not an OIG payment-dollar comparison
share of high-charge flag pairs by service · concentrated outpatient, diffuse inpatient