Data Sources

Fields used

• Facility name, address, county, city, state, ZIP, phone
• CMS Certification Number (CCN) — the federal primary key
• Number of certified beds; number currently occupied
• Ownership type (for-profit / non-profit / government)
• Ownership change date (where reported)
• Overall Five-Star rating + component star ratings
• Special Focus Facility status (SFF / SFF Candidate / none)
• Abuse icon flag (last 24 months)

Known caveats

• CCN is reused if a facility closes and the bed inventory is re-certified under a new operator — historical records under that CCN may not be the same business.
• Ownership change date often lags reality; brokered transactions can take months to surface in CMS data.

Fields used

• Daily hours per resident day (HPRD) by role
• Registered Nurse turnover (12-month rolling)
• Administrator turnover
• Weekend staffing levels
• Reporting compliance flag

Known caveats

• PBJ records hours paid, not hours physically present — a nurse on call but not in the building still appears as staffed time.
• Agency-staff hours appear under contract worker categories; some facilities under-report agency hours.
• Quarterly publication means the freshest data is up to four months stale.

Fields used

• F-tag (federal regulation cited)
• Severity letter (A–L, where I/J/K/L = immediate jeopardy)
• Scope (isolated, pattern, widespread)
• Deficiency narrative (free text)
• Date of survey, date of correction

Known caveats

• Severity is assigned by surveyors and varies meaningfully by state survey agency. A 'D' in one state may be cited as an 'F' in another for the same conduct.
• Some facilities aggressively dispute citations and have them reduced or removed on appeal — disputed citations remain in the public file until resolved.

Fields used

• Complaint-driven citations (with severity and scope)
• Substantiated complaint count
• Date investigated

Known caveats

• Complaint volume reflects who is willing to report, not just what is happening — affluent areas with engaged families surface more complaints than under-resourced ones.

Fields used

• Penalty type (CMP per-day, CMP per-instance, DPNA)
• Amount (USD)
• Date imposed
• Underlying deficiency severity

Known caveats

• Penalty amounts are pre-appeal; many are reduced on appeal but the original amount stays in the public dataset.

Fields used

• Owner name (individual or organization)
• Owner role (operational / managerial / financial / property)
• Percentage ownership where reported
• Owner address

Known caveats

• Real-estate-only owners (REITs, private equity property arms) often do not appear here unless they also hold an operating role.
• Layered LLC structures resolve to disclosed names but the parent ultimate beneficial owner can be obscured.
• Disclosure is self-reported and audited inconsistently; HHS-OIG has documented systemic underreporting.

Fields used

• Long-stay QMs (8 measures)
• Short-stay QMs (5 measures)
• National and state averages for benchmarking

Known caveats

• QMs are risk-adjusted but the model is imperfect — facilities serving sicker or higher-acuity populations may look worse than they are.
• Reporting is from MDS (Minimum Data Set) assessments which are completed by facility staff; assessment quality varies.

Fields used

• State ombudsman office contact info
• Regional/local ombudsman contacts where published
• Complaint data (where state publishes)

Known caveats

• Complaint data publication varies — some states publish summary statistics, most do not publish facility-level complaint counts.

Fields used

• Facility license number, license type, license status
• Inspection findings (where published)
• Complaint records (where published)

Known caveats

• Coverage is highly uneven — California publishes RCFE inspection reports online; many southern states publish nothing. See the per-state availability table at /methodology/data-availability.
• Some states require inspection reports be requested in writing and are not available in bulk format.