census-analysislisted

# Census Analysis Turn a facility's census data into a decision-ready brief. The audience is SNF administrators, admissions directors, and regional ops — often non-technical. Do the math for them, explain what the numbers mean, and always end with concrete actions, not just statistics. ## Privacy first Census data can contain PHI (resident names, MRNs, room-level detail tied to identity). Before analyzing: - If a file appears to contain resident names or identifiers, remind the user they can strip names/IDs before sharing — aggregate counts are all this analysis needs. - Treat any data provided as confidential. Never echo resident names or identifiers back in output. Report only aggregates (counts, percentages, averages). ## Step 1 — Gather inputs Accept either: 1. **A census export** (CSV/Excel — a daily or midnight census report from the EHR, e.g. PointClickCare, MatrixCare, Netsmart). Read it with available file/spreadsheet tools. Typical columns: date, unit/room/bed, payer, admission date, and sometimes anticipated discharge date. Identify the date range and whether rows are resident-days or one row per resident. Aggregate exports also work: one row per day with payer-count columns (date, Medicare, MA, Medicaid, … , total) — no resident-level detail needed. 2. **Typed numbers** — current census, licensed beds, counts by payer, admissions/discharges for the period. Always establish, asking if missing: - **Licensed beds** — the state-li