A complete comparison with HCAI’s October 1 roster found 55 consolidated campuses, 15 distinct hospital units and five parent facilities missing beyond the earlier Kaiser Modesto correction. All 517 open hospital records in the defined roster scope are now listed. The original 445 records retain their values; the directory has 520 entries including three older records outside the open roster.
Corrections & change log.
This page records material corrections, evidence updates, and changes to how the Observatory presents its work. A website update does not mean the underlying hospital data was refreshed; every figure keeps its own reporting year and source.
Latest presentation review: October 12, 2026.Hospital campuses now have their own directory entries.
Search matches names, aliases and cities by separate words, including in Compare. Added campuses display their own identity, licensed beds, source date and parent relationship; unmatched RN hours, travelers and staffed beds remain UNKNOWN. A build check prevents roster omissions or duplicate facility IDs.
Central-line records on 43 hospital profiles.
The count pilot now includes 172 calendar-year records for 2022–2025. Each hospital matches a dated September 16, 2026 CDPH parent-facility identity and an exact facility ID and name in all four annual sources. Source zeros, missing values and reporting coverage remain distinct; rehabilitation-unit records stay separate.
Unresolved identity and name changes remain under review. Each published profile names the CDPH reporting facility and links to preserved source rows and checksums.
Four years of central-line infection counts.
Doctors Medical Center Modesto, Memorial Medical Center Modesto, Alameda, Alta Bates–Alta Bates Campus and Children’s Oakland now show calendar-year 2022–2025 reported CLABSIs, central-line days and months reported. Each record matches its CDPH facility ID to the September 16, 2026 state identity snapshot. Primary files, row numbers and source notes are inspectable.
State records retain their own reporting periods alongside existing CMS outcomes. This count pilot adds no infection-ratio trend or hospital safety grade.
Pending losses now have their own category.
Maternity, NICU, pediatrics, psychiatric beds, behavioral-health clinic and emergency services. Proposals, announced closures and suspensions keep separate statuses. Date labels distinguish proposed dates from report dates.
Less digging. More useful answers.
Three findings up front. Three job questions. Detailed evidence folds away.
Separate CMS SNF turnover for Alameda and Oakdale. Acute-hospital turnover stays UNKNOWN. Vacancies, overtime, pay and orientation keep their own status.
Ten RSS feeds carry profile changes and service events. Email delivery remains unconnected.
Shorter headings and plainer explanations.
Repeated slogans and abstract wording were shortened across About, Contact, Methodology, source dates, reporting gaps, regional records, and the Observations introduction. Essay bodies, the public data transparency thesis, hospital figures, source links, and evidence findings were preserved.
Newer inspection findings and recorded resolutions are now visible.
Alameda, Alta Bates–Alta Bates Campus and Children’s Oakland now show 20 survey events, including 13 with citations. All 37 citation rows and 20 initial-comment rows are preserved with original narrative text and source links.
Children’s Oakland’s July 2025 and April 2026 patient-protection findings include the reported removal of Immediate Jeopardy on August 5, 2025 and May 14, 2026. Actual correction plans and later practice remain to be assessed.
Alameda’s December 2025 acute-provider finding is separate from its SNF record. Alta Bates’ June 2025 amended nursing-supervision report retains differing causal accounts. Overlapping tags and state monetary penalties are not added into an incident count or safety score.
Alameda, Alta Bates and Children’s Oakland now have deep evidence reviews.
The first three cards now preview the evidence findings. Each profile connects staffing, finances, emergency access, licensing, enforcement and institutional changes.
Acute and skilled-nursing identifiers, locations and capacity are separated. CMS staffing, turnover, ratings and 16 citation rows are shown with dates and corrections. All 90 Q1 2026 PBJ days separate RN work from administration, with monthly and weekend comparisons.
Alta Bates’ proposed Emeryville move and Children’s Oakland’s rebuilding and employment transition appear alongside the actual financial periods and current licensed services.
Fifteen institutional documents, three annual periods and five quarterly periods per hospital, and 18 additional parent-facility penalty records are inspectable. Alameda’s separate historic SNF citation is also assessed.
Deeper records for Doctors Modesto and the two Memorial hospitals.
Three annual periods, five newer quarters through March 2026, and collected Modesto cost-center labor extracts now appear with actual filing periods, source locators and review status.
CDPH records were matched through official facility identifiers. Exact duplicate API rows were removed. Penalty dates, categories, amounts and historical appeal status remain inspectable.
Bakersfield’s county and CDPH letters, application, notice, webpage capture, EMSA directory and de-identified follow-up request are assessed. The Lost Log now distinguishes proposed destinations from actual transfers and the closed burn center from separately listed wound care and hyperbarics.
Original deficiency narratives, newer enforcement outcomes, unit staffing and post-closure access records remain unresolved. No unreceived request response is treated as a finding.
Clearer staffing ratios and service-review status.
RN ratios above the care-group upper quartile plus three interquartile ranges carry an “Outlier · review pending” label, with RN hours and patient days beside the figure.
Hospitals without a documented service review now display “Service review not documented.” A “No changes found in reviewed sources” result requires a completed review with sources and dates.
Each metric keeps its reporting dates; the broad “Latest data” row is removed. The directory displays 24 cards per page and searches all 444 hospitals.
Hospital acronyms and apostrophes are corrected. Navigation icons are consistent. Personal bylines and career biographies are removed.
Readmissions, mortality, and a cleaner traffic baseline.
Condition-specific 30-day readmissions and mortality now appear in hospital outcomes, with source windows, cohort sizes, CMS comparisons and exact records.
Healdsburg’s October 2024–March 2025 ambulance-offload record is matched through its historical HCAI facility name and retained facility ID.
A browser opt-out now applies to both traffic counters. TikTok referrals and tagged TikTok links have separate attribution categories.
A place to follow CNO.
The shared footer now collects requests for CNO updates. Addresses are held privately, duplicate requests are combined, and readers can remove their address through the update-list page.
Patient experience and infection measures.
Hospital profiles now include patient-reported nurse communication, legacy staff responsiveness, and six infection measures below the nurse dashboard.
243 nurse-communication values, 244 legacy responsiveness values, and 230 reporting units with published infection ratios. Each result retains its reporting period, CMS status, CCN, source row and preserved file. No composite score or new reviewed-profile designation was added.
More hospital histories, fewer unknown trends.
Three annual reports now show RN hours per patient day, historical charts, and trend arrows in hospital profiles, search cards, and comparisons.
Contract RN share of nursing hours now has a 2022–2024 series alongside staffing. Existing pre-pandemic baselines remain available.
The complete 2022 and 2023 HCAI API responses are downloadable. Hospital source tables show reporting dates, RN hours, patient days, contract hours, record identifiers, and checksums.
A simpler way into the hospital records.
The NurseSignal label was retired. “Find a hospital” now opens the same statewide directory, and old links lead to the new address.
The homepage now leads with what public records can show nurses and where important questions remain unanswered.
The selected hospital now opens with four key measures and practical questions. Detailed source checks, peer context, and additional metrics remain available in expandable sections.
Hospital figures, source periods, review status, and underlying evidence were not changed in this update.
Lost Log pilot published.
The Lost Log now documents the June 30 closure of Bakersfield Memorial Hospital’s Grossman Burn Center acute-care beds and outpatient program, with the stated reason, source links, and unresolved questions kept together in one public record.
The Log records documented service changes. It does not rank hospitals, infer causes the record cannot establish, or convert missing information into a conclusion.
This is a new public record and presentation feature. It does not change the Observatory’s hospital workforce dataset, reporting periods, or hospital-condition classifications.
Data freshness became inspectable.
Six source families now record the period currently loaded into CNO, their coverage, refresh cadence, dependent signals, external-check status, and next maintenance action.
The ledger explicitly distinguishes the newest period inside the Evidence Engine from the newest release that may exist outside it.
The public can inspect freshness and maintenance boundaries without exposing future master workbooks, automation, or internal research queues.
One definition. Every nurse-facing tool.
The former NurseSignal interface and Compare began using the same definitions for patient demand, RN intensity, contract reliance, staffed capacity, evidence status, and known public-data gaps.
A signal definition is maintained once and reused across interfaces, reducing the risk that two CNO pages describe the same hospital differently.
This was an interpretive-architecture and presentation update. Source years, hospital records, and evidence boundaries were not changed.
The governing rules remain unchanged: incompatible reporting periods stay separate, unknown does not become zero, and no composite hospital safety score is produced.
Clearer identity, pathways, and freshness.
The Observatory moved to californianursingobservatory.org, and its search metadata, sitemap, crawler instructions, and structured identity were updated to match.
The project name was standardized as California Nursing Observatory. The hospital interface initially used the name NurseSignal; that name was retired in September 2026.
At that release, the homepage linked to the former NurseSignal interface, hospital comparison, investigations, and methodology. The current interface is named “Find your hospital.”
Every page now distinguishes the date the website was reviewed from the reporting periods attached to individual hospital figures.
These are website and presentation changes. They do not represent a new release of the underlying hospital data.
Open evidence. Protected engine.
The frozen Pilot V1.0 workbook and methodology remain available for public inspection. The redundant full-package ZIP was withdrawn, and future master-engine versions and internal research machinery are no longer presumed public.
Noncommercial research, journalism, education, public-interest use, and independent verification are permitted with attribution. Rebranding, wholesale republication, and commercial distribution of substantial CNO work require permission.
The reconstructed development estimate now exceeds 100 hours. It remains an estimate rather than a time-clocked record.
Public-source facts remain public. This update protects CNO’s original analysis, workbook architecture, documentation, writing, and presentation without concealing the published methodology.
The reconstruction became visible.
The downloadable model was checked directly: 27 worksheets and 739 formulas. Development time is identified as a reconstructed estimate, not a tracked timesheet.
Central Valley records now explain consolidated and mixed-facility reporting, including the combined Kaiser Modesto–Manteca filing and bed totals that include separately listed behavioral-health or nursing facilities.
The site now distinguishes source fields—including HAFD inputs—from the matching, reconciliation, missingness rules, denominators, and peer-comparison work performed in the public review model.
To report a possible error, include the hospital, disputed figure, reporting year, and any public source in an email to californianursingobservatory@gmail.com.