Lesson 02 — The Pareto is sharper than you think

Hospital material spend follows 80/10, not 80/20

The Pareto principle is so familiar it’s lost its edge. Most people quote “80/20” and stop there. Hospital material spend is sharper.

Rank a hospital’s catalogue by total billed CHF and the distribution lands consistently at:

• Top 1% of materials → ~50% of total CHF
• Top 10% → ~85%
• Bottom 50% → ~2%

That is closer to 80/10 or even 80/5. A typical 10–15 K-material catalogue has its margin and inventory exposure concentrated in roughly 100–200 items.

What’s driving it

The catalogue is two populations stitched into one table:

• The expensive few. Orthopedic hardware, oncology drugs, cardiac devices. Individually CHF 1–5 K. Consumed in dozens to thousands per year. A single item easily exceeds 100 K CHF/year of total spend.
• The cheap many. Gauze, syringes, gloves, saline. Cheap, high-volume, but collectively a few percent of CHF.

Add a third quiet contributor: one-shot specialty items (rare drugs for orphan diseases, single-use trauma implants). They bloat the catalogue count without moving CHF at all.

The three populations do not blend into a smooth distribution. They are structurally different, and the high-value population dominates outcomes by orders of magnitude.

What changed in the analysis

Three habits that work for normal-distribution data fail on this curve:

1. Outlier detection over-fires in the top decile. signal_billing_amount_outlier will concentrate findings in the top 1% population by definition — that’s where the CHF live. The fix: tier the outlier logic by Pareto band (top 1% / next 9% / bottom 90%), each band scaled to its own population.
2. Cycle-counting every catalogue item is wasted effort. A/B/C inventory tiering on the Pareto curve is the right discipline. “Where do we focus governance?” has a quantitative answer.
3. Catalogue rationalisation at the bottom doesn’t move CHF. Removing low-decile items saves master-data effort but doesn’t reduce spend. The high-leverage rationalisation is at the top — consolidating duplicate premium items, negotiating volume on the few that matter.

What to watch for

Two divergences from the Wisdom worth investigating:

• Pareto flattening (top 1% share decreasing) — could mean catalogue diversification (good) or invoice under-billing of premium items (bad). The signal is the same; the cause matters.
• Pareto sharpening (top 1% share increasing) — could mean treatment-mix shift (e.g., more oncology) or premium price drift. Investigate before celebrating.

Why this is in a Lessons journal

Because the Pareto’s sharpness is so consistent that not knowing it produces a category error. A new analyst who reads the catalogue as a uniform population — applying the same review cadence, the same alert thresholds, the same tracking discipline to every item — wastes effort across the long tail and under-weights the few items that actually drive cost outcomes. The Wisdom says catalogue items are two structurally different populations, treat them differently. That sentence is what stops the category error.

Sources

• Wisdom: smebit_material_spend_pareto.yaml
• Provider: a hospital CFO with multi-site finance responsibility
• Anchor signal: signal_billing_amount_outlier
• Date Wisdom captured: 2026-05-08