Infrastructure›Hospitals

Hospital cohort

Loading…

Projected cumulative loss
2025–2030 (unsubsidized)[?]
—
 
Weighted operating margin
—
 
Cohort NPR · FY —
—
 
Cohort OpEx · FY —
—
 
Geographic distribution

· colored by ownership type, active cohort highlighted. Click a dot for details. Geocoding: IL Hospital Report Card[?].

Cohort NPR vs. Operating Expenses

Solid lines: actuals from NASHP HCT[?]. Dashed: COVID-controlled OLS projection median. Shaded bands: 80% confidence interval via 10,000-draw parametric bootstrap.

Cohort weighted operating margin

Cohort-weighted margin. Reference lines at 0% (break-even) and −25% (SNWT critical threshold[?]). Grey-shaded years are COVID disruption (2020-2021) controlled out in the projection.

Cohort payer mix FY—

NPR-weighted average across the active cohort. Source: NASHP Hospital Cost Tool[?]. Medicaid share ≥ 30-40% is the strongest signal for safety-net status[?].

— hospitals reporting

Methodology lens how per-hospital trends borrow strength

Each hospital has only ~8 years of margin data, so its raw trend is statistically noisy. Empirical-Bayes shrinkage[?] pulls each hospital's estimate toward a peer average — trading a little bias for much less noise. The lens is the choice of peer group; it re-keys every per-hospital slope, trajectory and projection in the profiles and comparison views.

Methodology & caveats

Cohort taxonomy
24 Chicago-area safety-net hospitals organized into 4 ownership buckets reflecting different financial dynamics: Independent Non-Profit (12; mission-driven, no system backstop), Government-Sponsored (4; tax-base backstopped), Health System Non-Profit (4; system-level cross-subsidies), Privately Owned For-Profit (4; PE/equity-firm owned). Geographic sub-buckets (South / West / North) are available within Independent Non-Profit; La Rabida, a children's specialty hospital serving the whole city, is not assigned a single geographic slice.
Subsidy isolation[?][?]
The non-patient revenue that offsets operating losses (government grants/appropriations, investment income, contributions, other revenue) is measured per-hospital, COVID years excluded, as a median of recent normal years. Two sources by entity type: independent non-profits via IRS Form 990 Part VIII; government, health-system, and for-profit hospitals via CMS HCRIS Worksheet G-3 (the only per-hospital source for entities that file no 990 or a consolidated system-level 990). Both exclude Medicaid (incl. the Illinois Hospital Assessment Program), which flows through net patient revenue, so the measure is HAP-clean. Cook County's Stroger and Provident carry a system-wide property-tax appropriation context note. Where no usable measured figure exists, the fallback is the prior estimate (8.52% of NPR).
Source data[?][?]
NASHP Hospital Cost Tool (FY2011–2022) for 22 of 24 hospitals — a cross-validated aggregation of CMS HCRIS, AHA Annual Survey, and Medicare Cost Reports. CMS HCRIS direct for UI Health and Chicago Reed. Quality, ownership and geocoding from the Illinois Hospital Report Card[?].
Forecast methodology — v0.2[?][?]
Per hospital: OLS regression of operating margin against fiscal year with COVID indicator — margin = α + β·(year − 2022) + γ·is_covid_year + ε. The β captures underlying trend net of COVID disruption; γ captures the COVID-era shift. Projections forward use only (α + β·t). Per-hospital estimates are shrunk toward a cohort mean via empirical Bayes (see the methodology lens). Uncertainty is propagated via 10,000 joint MVN bootstrap draws.
Limitations
La Rabida and Chicago Reed lack sufficient years of margin data for a credible fit and appear as "insufficient-data." NASHP HCT FY2023 not yet released. No balance-sheet (Worksheet G) data — we report "probability of breach −25%" rather than "years to insolvency." Small buckets (N = 4) and small geographic slices carry wide bootstrap intervals and should be read with that uncertainty in mind.