Glossary
Every term and label used on this site, in plain language. Deeper explanations live in the methodology.
Prices
List price
The hospital's full chargemaster amount (published as "gross charge"). Almost no one pays it directly.
Cash price
The amount for patients paying without insurance (published as "discounted cash").
Negotiated rate
The dollar amount a specific insurer agreed to pay the hospital for a service (published as "negotiated dollar"). One hospital can have many negotiated rates for one service — one per insurer contract.
Price row / observation
One published amount in a hospital's file: one source row and one price type. A single charge item usually produces several price rows (its list price, cash price, and each insurer's rate).
Derived dollar
A dollar figure calculated from a published percentage or formula rather than directly quoted. Shown as context, never ranked against quoted prices.
Services and contexts
Service context
The exact combination prices are compared within: billing code + care setting
- billing type + modifier set (+ drug units) + price type + payment methodology (for negotiated rates). If any part differs, prices are never mixed. Every comparable context has its own stable page address. See how comparison works.
Payment methodology
How a negotiated rate is paid: fee schedule (per item/service), case rate (one amount per episode or bundle), or per diem (per DAY of inpatient care — a daily amount, not the price of a full stay). Rates with different methodologies price different things and are never compared with each other, and a per-diem is never labeled above or below a cash price.
Insurer contract
One insurer + plan + methodology combination inside one hospital's file. A contract's repeated rows collapse to a single amount before any statistic; a contract carrying several different amounts for one exact context is excluded from statistics as ambiguous (visible in data quality).
Hospital-weighted
How every market statistic here is computed: one representative price per hospital ("one hospital, one vote"), never raw row counts — so a rate repeated hundreds of times in a file gains no weight.
Billing code
A standardized identifier for a service. Code systems used here include MS-DRG (inpatient stays), CPT (procedures), HCPCS (procedures/supplies), and APC (outpatient payment groups). Internal chargemaster codes ("CDM") are not comparable across hospitals.
Care setting
Whether the price covers inpatient or outpatient care. "Not specified" means the hospital's file did not say.
Billing type
Whether the price covers the facility fee or the professional (clinician) fee. "Not labeled by hospital" means the file omitted it; such rows are only ever compared with other unlabeled rows.
Modifiers
Billing codes' add-on flags that change what is being priced — for example professional-only (26) or technical-only (TC) components. Rows with different modifier sets are never merged.
Description not available
The service's code has no displayable description. For CPT/CDT codes this is a licensing constraint (AMA-licensed descriptions cannot be republished); for other systems the public-domain reference data is not yet loaded. The code itself still compares correctly.
Comparison vocabulary
Hospitals (n)
The number of hospitals with a safely representable price for the exact service context — the denominator behind every market statistic, each contributing exactly one representative price. Shown next to every comparison on this site. Hospitals that published rows but could not be represented safely are counted separately as excluded.
The 3-hospital floor
No market statistic is computed from fewer than 3 comparable hospitals in the exact same context. Below the floor, individual prices stay visible but the context is labeled "Too few hospitals." Why.
Comparison status
Whether a service context qualifies for cross-hospital comparison: Comparable (described) — comparable with a service description; Comparable (code only) — comparable by billing code, description unavailable; Too few hospitals — below the 3-hospital floor.
Comparison confidence
How solid a context's comparison is: High (10+ hospitals, described), Moderate (5–9), Limited (3–4), Low (below the floor). Details.
Typical (median)
The middle value: half of hospitals price above it, half below. More robust than an average against extreme values.
Lower / Upper (10th / 90th percentile)
The price below which 10% (or 90%) of the context's prices fall. At small n these are close to the minimum and maximum — that is why they only appear above the floor.
Price spread (×)
The 90th-percentile price divided by the 10th-percentile price. A spread of 4× means the high end is four times the low end.
Variation band
How spread out a context's prices are: very high variation (spread ≥ 3×), high variation (≥ 2×), moderate variation, or not ranked (below the floor).
Market position
Where one hospital's price sits vs. the corpus median for one exact context: well below / below / near / above / well above market, using ±10% and ±25% thresholds. Details.
Blocker
A named, stable reason a row or context is excluded from stricter comparison — never an undocumented filter. All 11 are explained on the data quality page.
Hospitals and files
Data usability score
A hospital's 0–100% score for how usable its published file is for price comparison: the average of freshness, code coverage, dollar-amount coverage, insurer-name matching, and comparison readiness. Not a legal-compliance finding. Details.
Data confidence
The usability score as a band: High (≥85%), Moderate (≥70%), Limited (≥50%), Low. Describes the published file, not the hospital's care.
File recency
How recently the hospital's file says it was updated: within 90 days / 180 days / 1 year / over a year old.
Snapshot
One captured version of a hospital's published file, identified by a snapshot id. Only each hospital's current snapshot feeds comparisons.
Source file
The hospital's own machine-readable standard-charges file. Every hospital page links to it so any number can be checked against the primary source.
Insurers
Matched insurer
A published payer name that the pipeline confidently matched to a canonical insurer identity (e.g. "BCBS TN" → BlueCross BlueShield of Tennessee). Only matched rates appear in insurer views.
Insurer market median
For one insurer and one service context: the median of that insurer's negotiated rates across hospitals (needs the 3-hospital floor).
Rates above cash price
Contexts where an insurer's negotiated rate exceeds the hospital's own cash price. A signal to investigate, not proof of overpayment. How to read it.
Site-wide
Corpus
The set of hospitals included in an export. The site names the corpus and limits every claim to the hospitals it contains.
Comparability funnel
The count of price rows surviving each comparison rule, from "published" down to "meets the 3-hospital floor." Shown for the whole corpus and per hospital. See it.
Build identifier
The exact version of the open-source pipeline that produced the published data, shown on the downloads page for citability.