BRANMOOR
THURSDAY · 14 MAY 2026
ED Boarding & Throughput Watch

Methodology

Source data

The register is built from the CMS Timely and Effective Care - Hospital dataset, refreshed quarterly. It covers every Medicare-participating U.S. hospital that reports ED throughput measures — 4660 hospitals at last refresh.

What v1 surfaces

  • OP-18b — median time (minutes) from ED arrival to discharge for discharged patients. The canonical ED-throughput measure. National median 148 minutes
  • OP-18c — same measure restricted to psychiatric / mental health patients (often dramatically higher due to capacity-of-care issues for behavioral health)
  • OP-22 — left-before-being-seen rate. The direct patient signal of ED overcrowding
  • EDV — CMS volume bucket (very low / low / medium / high / very high)
  • SEP-1 — severe sepsis & septic shock 3- and 6-hour bundle compliance. ED throughput is the upstream gate
  • STK-02 / 03 / 05 — stroke care timeliness measures, also throughput-dependent
  • SAFE_USE_OF_OPIOIDS — concurrent prescribing safety, an acute-care risk-management measure

Branmoor's signal aggregation

For each hospital we compute a per-measure comparison against the national P25 / median / P75 distribution and aggregate into a single signal flag:

  • Elevated ED-throughput stress — 3 or more measures in the worst quartile (above P75 for lower-is-better measures, below P25 for higher-is-better), or composite capacity-stress score > 0.5
  • Mixed signal — 1+ measure in the worst quartile
  • Strong ED throughput — 2 or more measures in the best quartile and none in the worst
  • Neutral — otherwise

The composite capacity-stress score combines OP-18b (arrival-to-discharge time) and OP-22 (left-before-seen rate) into a single number that approximates relative ED overcrowding. Methodology: each measure's relative deviation from the national median is summed.

What v1 is not

v1 surfaces the CMS-published acute-care substrate. It is not the real-time capacity surveillance product the full ED Boarding & Ambulance Diversion Surveillance scope (per the original spec) delivers. v1 does not include:

  • State EMS diversion data — some states require ambulance diversion reporting (varies wildly by state). Per-state ingest. Effort: XL
  • EMTALA transfer patterns from claims — Medicare and Medicaid claims expose the transfer signal at the regional level. DUA-gated for full claims
  • FAA flight data + helipad mapping — flight data is clean public; tail-number-to-helipad mapping is manual curation but finite work
  • Real-time hospital staffing signal — CMS staffing data is lagged. Real-time requires partnerships or scraping hospital workforce dashboards (where public)

See the product page for the paid v2 layer and the EMS-agency / payer / health-system sales motion.

Update cadence

CMS refreshes the Timely and Effective Care dataset quarterly. Branmoor re-ingests on each site deploy.

Limits

ED throughput measures lag 6–12 months from the measurement period. The register reflects sustained operating patterns, not real-time capacity. For real-time signal, the paid layer is the right channel.

Hospitals below CMS minimum-volume thresholds for a measure are flagged as Not Available rather than scored.

‹ All hospital EDs