Specialty & HCPCS Concentration
Specialties and HCPCS codes with unusual year-over-year Medicare payment growth from the public CMS Physician PUF. These are concentration / outlier signals — investigative leads for further review, not findings of wrongdoing. No PHI.
Payment construct: Line-level Medicare payment is average Medicare payment × total services (Avg_Mdcr_Pymt_Amt × Tot_Srvcs). Specialty and HCPCS views sum those amounts. Large % changes on thin volume (few NPIs or services) need extra caution.
Attribution: CMS Medicare Physician & Other Practitioners (by Provider and Service).
Specialty — largest YoY $ growth (latest year)
Filter requires a prior-year payment base of at least $1M so tiny specialties do not dominate %.
Specialties with YoY (latest)
Growing ($)
Shrinking ($)
Specialty — highest YoY % (with volume floor)
HCPCS — largest YoY $ growth (latest year)
Procedure/code concentration can flag coding or product uptake shifts. Treat as leads only.
HCPCS codes with YoY (latest)
Growing ($)
Soft methodology
| Item | Detail |
|---|---|
| Sources | mart_medicare_spend_yoy_by_specialty, mart_medicare_spend_yoy_by_hcpcs |
| Grain | Specialty × year; HCPCS × year |
| Payment | Sum of provider×HCPCS Medicare payment (Avg_Mdcr_Pymt_Amt × Tot_Srvcs) |
| Framing | Concentration / outlier leads only |
| Caveat | High % on low volume is noisy — volume floors applied in headline charts |
| YoY accounting | Dense: codes/specialties new in a year count with prior $0 (yoy_status = new_in_year); codes no longer billed get a $0 row (absent_in_year). Per-year sums of YoY $ reconcile with the state mart (dbt test assert_national_yoy_totals_reconcile). KPI counts above cover codes/specialties billed in both years. |