These tools use public CMS datasets to help users explore Connecticut nursing home staffing patterns and facility-level survey and enforcement context. They are designed for screening, comparison, and question-building.
CMS Payroll-Based Journal Daily Nurse Staffing dataset files are used for staffing hours, census and resident-day denominators, and quarter-by-quarter staffing trends. PBJ is quarterly public staffing data, not real-time staffing information.
CMS Nursing Home Provider Information dataset supplies facility metadata such as name, address, certified beds, ownership type, CMS case-mix staffing comparison fields, and CMS Care Compare rating context when available. These values are contextual and may not align exactly with every historical PBJ quarter.
For the Statewide Comparison county filter, the tool uses a separate CMS Certification Number (CCN)-based geography crosswalk built from the April 2026 Provider Information county fields. County is current Provider Information context, not historical quarter-specific geography. Facilities without a Provider Information county match are labeled as Unknown / needs review.
CMS nursing home quality-measure data supplies facility-level claims-based quality-measure rows, including measure descriptions, scores, footnotes, rating-use flags, and measure periods. These measures are contextual Care Compare quality-measure data and are not calculated by this staffing tool.
CMS Skilled Nursing Facility Enrollments dataset records provide legal organization, doing-business-as name, NPI, proprietary/nonprofit status, organization structure, and affiliation entity context. Shared affiliation entity data does not by itself prove common day-to-day control or legal responsibility.
CMS nursing home survey and enforcement data supplies citation-level Health, Fire Safety, and Emergency Preparedness deficiency records plus fine and payment-denial enforcement events. These records are shown as historical context and remain separate from staffing classifications.
HPRD means hours per resident day. It divides reported staffing hours by resident days for the same time period.
For example, if a facility reported 300 nursing hours across 100 resident days, the result would be 3.00 HPRD. Higher or lower HPRD can be meaningful, but it should always be interpreted with context such as resident acuity, facility size, missing data, survey findings, complaints, and resident experience.
Historical CMS Payroll-Based Journal staffing is available in the public tools from Q4 2017 through Q4 2025. These rows are PBJ-derived staffing measures for each facility-quarter, including total nurse HPRD, RN HPRD, LPN/LVN HPRD, nurse aide HPRD, contract staff percentage, CT direct-care HPRD estimate, and CT licensed HPRD estimate.
Current contextual CMS snapshots remain separate. Provider Information, Care Compare ratings, CMS case-mix comparison points, and Quality Measures Claims are April 2026 context. SNF Enrollment affiliation context is from May 2026. Those current values are not historical quarter-specific values for each PBJ staffing quarter.
Connecticut nursing home staffing regulations, Title 19, Sec. 19-13-D8t, set nursing-staff requirements in subsection (m). The current 3.00 total nursing and nurse's aide direct-care hours per resident day and 0.84 licensed nursing hours per resident day comparison points used here are described in Connecticut Department of Public Health's amended 3.0 staffing implementation notice.
The Connecticut materials also explain that director of nursing and assistant director of nursing hours are not counted toward those minimum requirements. The staffing tools create PBJ-derived screening estimates to compare against those points.
For historical PBJ display, the tools use a conservative public-applicability rule: Q4 2017 through Q4 2021 are shown without applicable CT threshold status labels; 2022 quarters are not assigned public CT status under the current method; Q1 2023 is transitional/partial-period context; Q2 2023 through Q4 2025 may show PBJ-derived CT screening-status labels with caveats.
These estimates exclude PBJ DON and administrative categories: Hrs_RNDON, Hrs_RNadmin, and Hrs_LPNadmin.
The case-mix comparison point value is imported from CMS Nursing Home Provider Information. The field used is Case-Mix Total Nurse Staffing Hours per Resident per Day. The CMS Nursing Home Data Dictionary describes that value as case-mix total nurse staffing HPRD combining aide, LPN, and RN staffing.
The app compares PBJ-reported actual total nurse HPRD to that CMS-published point. The actual-minus-benchmark and percent-of-benchmark text are calculated by the app.
These current Provider Information comparison points are not historical quarter-specific values. Historical case-mix persistent-pattern analysis is not available unless historically aligned CMS benchmark snapshots are later acquired.
CMS Care Compare ratings shown in the facility explorer are imported from CMS Nursing Home Provider Information. The tool does not calculate these star ratings and does not use them to replace PBJ-reported staffing metrics, Connecticut direct-care screening estimates, or case-mix comparison points.
When available, the export may include overall, health inspection, staffing, RN staffing, quality-measure, long-stay quality-measure, and short-stay quality-measure ratings. These ratings provide contextual CMS summary information and should be interpreted alongside the staffing metrics, source timing, survey findings, complaints, resident experience, and other review materials.
Individual claims-based quality measures come from the CMS Nursing Home Quality Measures Claims file. The facility explorer imports measure-level rows by CMS Certification Number (CCN), including measure code, description, resident type, adjusted score, observed score, expected score, footnote, whether CMS marks the measure as used in the Quality Measure Five-Star Rating, measure period, and processing date.
CMS rating summaries and individual quality measures are different. Rating fields are summary values imported from Provider Information; individual quality measures are measure-level rows imported from the Quality Measures Claims file.
The Facility Explorer uses nursing home survey and enforcement files published by the Centers for Medicare & Medicaid Services (CMS). The records are joined to current Facility Explorer facilities by the six-character CMS Certification Number (CCN).
The current build uses the May 2026 CMS files below. Their available date windows differ, so totals across sources do not represent one uniform observation period.
NH_HealthCitations_May2026.csv; survey dates through .NH_FireSafetyCitations_May2026.csv; survey dates through .NH_CitationDescriptions_May2026.csv; a reference lookup used with citation records.NH_Penalties_May2026.csv; penalty dates through .The compact snapshot defines the recent window as through , inclusive. The end date is the maximum date across the three detailed runtime datasets, and the same three-year period is applied across sources for a consistent summary window.
“Recent” does not mean that every record occurred during a facility's latest survey cycle. All current enforcement rows fall inside this window because the penalties source begins after the recent-window start date.
For Health F-tag deficiencies only, the tool translates CMS Scope Severity Codes into the following plain-language groupings:
This grouping is not applied to Fire Safety K-tags or Emergency Preparedness E-tags.
The Fire Safety source contains citation rows for K-0211 and K-0133, but the current CMS Citation Descriptions file does not contain matching lookup entries for those codes. The application preserves the deficiency descriptions supplied in the Fire Safety source and labels the mismatch as an official-source lookup gap. It does not invent or manually substitute a description.
Fine and Payment Denial rows are distinct enforcement types. Fine Amount applies to Fine rows; Payment Denial Start Date and Payment Denial Length apply to Payment Denial rows. Penalty Date, payment-denial start date, and processing date are separate fields and should not be treated as interchangeable.
Three duplicated full-row signatures exist in the current Connecticut penalties source. The build preserves those published source rows and flags duplicates rather than silently removing them. Aggregate totals therefore reflect the published source rows unless a display is explicitly labeled otherwise.
The normal Facility Explorer page load requests a compact one-record-per-facility summary. Detailed citations and enforcement events load only after a user chooses View detailed findings, and only the selected facility's CCN-specific file is requested. The full statewide detailed files are not downloaded during the normal page load. This improves performance and avoids unnecessary data transfer.
The current Facility Explorer includes 196 facilities. Health records cover 191 current CCNs, Fire Safety or Emergency Preparedness records cover 178 current CCNs, and penalties cover 102 current CCNs. Five current facility records have no records in the included survey or enforcement files; they remain in the Explorer with zero counts and unavailable latest dates.
Affiliation entities come from the CMS Skilled Nursing Facility Enrollments dataset. They are useful for grouping facilities and reviewing patterns across connected records.
Affiliation entity data does not by itself prove shared management decisions, day-to-day operational control, or legal responsibility. Small affiliation groups should be interpreted carefully because one facility can strongly affect the group summary.
The project includes an independent calculation audit script. The audit re-read raw PBJ files, recomputed the staffing metrics, and compared the results to the current generated export.
The current calculation audit found zero unexplained discrepancies. The production historical PBJ staffing file was also independently audited against raw PBJ source files. Technical users can review docs/nursing_home_staffing_calculation_audit.md, docs/nursing_home_staffing_history_calculation_audit.md, scripts/audit_nursing_home_staffing_ct.py, and scripts/audit_nursing_home_staffing_history_ct.py.
Review one facility over time.
Compare latest-quarter facility rows.
Review endpoint-to-endpoint changes.
Find repeated screening patterns.
Review CMS affiliation entity groupings.