Nursing Home Inspection Ratings Dataset Data: Dataset Structure and Field Coverage

Within that family, two records matter most for inspections. The Nursing Home Provider Information file carries one row per active Medicare-certified home — about 14,700 of them per the Care Compare coverage counts, running to roughly 15,000 rows by the Provider Data Catalog's own estimate — across 99 columns and about 9 MB of CSV. Its Health Inspection Rating field distills three cycles of state inspection deficiencies into the same 1-5 scale used by the Overall Rating, Staffing Rating and QM Rating beside it.

Because every row also carries geocoded latitude and longitude, certified bed counts, ownership type and chain affiliation, an inspection ratings extract becomes a mappable quality layer rather than a flat scorecard. The companion Hospital General Information file holds 5,419 hospitals with their own overall star ratings, so post-acute and acute quality screens can run off one catalog.

CMS ships a PDF data dictionary with each release (NH_Data_Dictionary.pdf under the catalog's data_dictionaries/ path), so field meanings are documented rather than inferred. Verified fields relevant to an inspection-ratings workflow include:

• Identity - CMS Certification Number (CCN), Facility Name, Address, City/Town, State, ZIP Code, Latitude, Longitude.
• Inspection rating - Health Inspection Rating, the 1-5 star score derived from three cycles of health inspection deficiencies.
• Composite scores - Total Weighted Health Survey Score, the weighted composite of those rating cycles.
• Ratings context - Overall Rating plus the Staffing Rating and QM Rating components, each on the 1-5 scale.
• Enforcement - Number of Fines and Total Amount of Fines in Dollars, cumulative federal penalties over the last three years.
• Staffing inputs - Adjusted RN Staffing Hours per Resident per Day, case-mix adjusted.
• Structure - Ownership Type (For profit, Non profit, Government) and Number of Certified Beds.

That combination lets one table answer questions enforcement data usually spreads across sources: whether a low Health Inspection Rating coincides with thin RN hours or a large fine history, filtered to any state, county or chain. Sample records show how granular it gets — an Alabama row carries CCN, city, ownership category and certified bed count in a single line, which makes facility-level joins trivial once you key on CCN.

Three workflows fall straight out of the single CSV. First, a national quality screen: sort the roughly 14,700 homes by Health Inspection Rating and Total Weighted Health Survey Score, then cut by Ownership Type to see whether for-profit, non-profit and government homes distribute differently across the five stars. Second, a chain rollup: group rows by chain affiliation to produce per-operator counts of facilities, certified beds, average RN staffing hours and the share of each portfolio rated 1 or 2 on inspections — the shape of diligence work no directory feed supports.

Third, enforcement triangulation: join the fine fields against the inspection component to flag homes where citations and penalties diverge from their published stars. Because the geocodes ship in-file, all three outputs plot directly without a separate geocoding pass.

The same quarterly-refresh logic scales across settings. Benchmarking facility quality across the US means comparing Care Compare star ratings and measures for ~5,400 hospitals, ~14,700 nursing homes and ~12,500 home health agencies in one quarterly CSV cycle — the nursing home file is simply the largest and most inspection-heavy member of that set.

Inspection stars explain more joined to supply, demand and money. Three companions in the same industry slice cover each: