HBOT Suite

Machine learning tools for hyperbaric oxygen therapy programs, built on national Medicare data.

Fraud RiskWhere a billing profile sits on the national fraud-exclusion risk scale.Open →Denial RiskChance a group of HBOT claims is heavily denied, by contractor and setting.Open →Practice TrajectoryNext year's HBOT payments, and the chance of a surge or an exit.Open →

About the HBOT Suite

The HBOT Suite turns a national study of Medicare hyperbaric oxygen therapy (HBOT) claims into three calculators. Fraud Risk places a billing profile on the national scale of providers later excluded for fraud. Denial Risk estimates how likely a group of HBOT claims is to be heavily denied. Practice Trajectory forecasts next year's HBOT payments and the chance of a surge or an exit.

All three are trained on public data: Medicare Physician and Other Practitioners files, the Physician/Supplier Procedure Summary (all Part B claims) and the Office of Inspector General exclusion list. Each is validated on years it never saw.

Fraud Risk: EasyEnsemble of gradient-boosted models on 1.66 million Medicare providers; AUROC 0.84 on 2020 to 2022, years it never saw.
Denial Risk: Gradient boosting on Part B HBOT claim groups, validated year by year (rolling origin) from 2015 to 2025.
Practice Trajectory: Gradient boosting on HBOT clinician billing histories, 2013 to 2024, temporal train and test split.
Data: CMS public use files and the OIG LEIE. No patient-level or diagnosis data.

Research demonstration only. Not legal, billing or compliance advice. Fraud Risk predicts exclusion, which reflects detected and prosecuted fraud, not fraud itself.