StarlynnCare · Methodology
How we rate facilities
Every number on StarlynnCare comes directly from California state inspection records. We do not invent data, accept paid placements, or produce a single letter grade. Instead, we show four independent signals and let families weigh them.
Primary data sources
Facility licensing, inspection, and complaint data are drawn from the California Department of Social Services (CDSS) Community Care Licensing Division via two public endpoints:
- CDSS CKAN Open Data — statewide RCFE roster, including license status, bed count, and operator.
- CDSS Transparency API — per-facility LIC 9099 inspection reports, including visit dates, Type A/B deficiency citations, inspector narratives, and complaint outcomes.
Data is refreshed via automated scrapers. Each facility record shows a "last updated" date reflecting the most recent ingest run.
Current coverage
StarlynnCare currently covers Alameda County, California — specifically RCFEs (Residential Care Facilities for the Elderly) that serve residents with memory care needs. This is a deliberate geographic beachhead; we add new counties as data quality and coverage allow.
Only facilities with an active California license and at least one published inspection record are shown. Facilities missing key data fields are held back until the next ingest run resolves them.
Sample size for current benchmarks: 14 publishable Alameda County RCFEs. Percentile thresholds will shift as coverage expands.
The four metrics
Each facility profile shows an "At a glance" panel with four rows. Here is exactly how each is computed.
Compliance record
Formula: Total deficiencies ÷ number of routine (non-complaint) inspections on file
Tier logic: Compared against the distribution across all 14 Alameda County facilities. At or below the 33rd percentile = Strong; 34th–66th = Mixed; above 66th = Concerns.
Why it matters: Routine inspections are the most comparable measure of regulatory compliance across facilities. Complaint-triggered visits are excluded here because they are initiated by external reports, not by the state's regular cycle.
Severity record
Formula: Count of Type A deficiency citations on file (all time)
Tier logic: Same percentile cutoffs (33rd / 66th) across all county facilities.
Why it matters: California distinguishes Type A (actual harm or immediate risk of death/serious bodily injury) from Type B (potential harm). A Type A citation is the most serious regulatory outcome short of license revocation; even one warrants a closer look.
Dementia-care specificity
Formula: Boolean: any §87705 or §87706 citation in the past 5 years
Tier logic: No tier comparison — shown as 'For reference' with the date of the most recent qualifying citation.
Why it matters: California Code of Regulations Title 22, §§ 87705–87706 govern RCFE dementia-care programs specifically. A citation under these sections means a facility was found out of compliance with requirements designed to protect residents with cognitive impairment. This is surfaced separately because it is uniquely relevant to memory care placement decisions.
Complaint pattern
Formula: Substantiated complaints ÷ total complaints that received a CDSS outcome determination
Tier logic: Same percentile cutoffs, restricted to facilities with at least one complaint that has a recorded outcome. Facilities with no complaints-with-outcome show 'For reference'.
Why it matters: CDSS investigates complaints and issues a formal outcome (Substantiated, Unsubstantiated, etc.). A high substantiation rate relative to peers is a meaningful signal, but many complaints are routine and do not result in findings — context matters.
Tier color key
All numeric metrics are compared against the current county distribution using 33rd and 66th percentile cutoffs. "Strong" means better than most peers; "Concerns" means worse than most peers. These are relative assessments within the current sample, not absolute benchmarks.
At or below 33rd percentile — better than most peers
34th–66th percentile — near the county middle
Above 66th percentile — worse than most peers
Boolean metric with no peer comparison
Why we don't publish a composite score
A composite score implies that one weighting of metrics is universally correct. It isn't. A family with a parent who has advanced dementia may weight §87705 citations more heavily than a family whose parent is largely independent. A family that has already toured three facilities and spoken to staff may weight a slightly elevated complaint rate differently than the aggregate numbers suggest.
Composite scores also obscure: a facility with one catastrophic Type A event and otherwise clean records will look identical to a facility with a pattern of minor Type B deficiencies if both are averaged into the same number.
We believe four honest signals, clearly labeled, serve families better than one opaque number.
Limitations and disclaimers
- Inspection records reflect conditions at the time of a specific visit. They are not a real-time assessment of today's care quality.
- Facilities that have improved since a citation was issued will still show that citation. Corrective action dates are shown where available.
- A small sample (14 facilities) means percentile thresholds are sensitive to individual outliers. Tiers may shift as coverage expands.
- StarlynnCare is an information resource, not a licensed placement agency or healthcare provider. Nothing on this site constitutes medical or legal advice. Always tour facilities, speak with staff, and consult a licensed care advisor before making placement decisions.
Questions about methodology or data errors? hello@starlynncare.com