Methodology
How ClearCost standardizes hospital pricing files for fairer comparisons.
Download each hospital's machine-readable CSV/JSON file; store the source URL and retrieval timestamp.
Map codes to CPT/HCPCS, standardize descriptions, hospital identities, settings, price types and payer/plan names.
Detect missing amounts, malformed rows, duplicate entries and implausible values.
Compute min, max, median, IQR, spread and coefficient of variation; flag outliers.
Serve comparable procedure-level prices through versioned API endpoints.
Statistics
For each procedure, price type, setting and (where relevant) payer, we compute the minimum, maximum, median, interquartile range, spread ratio (max ÷ min) and coefficient of variation (standard deviation ÷ mean). Outliers are values more than 1.5× the IQR beyond the first or third quartile (Tukey fences).
Data quality and comparability scores
Completeness is the share of expected records (procedure × price type × payer) that contain a usable amount. Freshness reflects the age of the hospital's published file. The preliminary comparability score (0–100) weights completeness 60%, freshness 30% and file format/structure 10%. Weights will be revised once real files are analyzed.
Comparison limitations
Hospitals bundle services differently: one MRI price may include the radiologist's reading, another may not. Inpatient prices may be per-case or per-diem. Payer names and plan tiers are inconsistent across files. Until normalization accounts for these differences, comparisons should be treated as directional, not exact.
Published prices are not patient bills, and nothing on this site is a guarantee of cost.