Hypersphere HealthPatient 3-D™ See it on your data

Evidence

Measured, reproducible, cited.

An academic health systemRegion one
Lives
2,299,052
Scored
2,088,115
Care-management eligible
911,181
Compounding-activePatients with at least one active interaction between their conditions.
31.2%
Re-rankedPatients who move higher on the priority list once those interactions are counted.
17.4%
An academic health systemRegion two
Lives
190,387
Scored
172,318
Care-management eligible
77,544
Compounding-activePatients with at least one active interaction between their conditions.
31.9%
Re-rankedPatients who move higher on the priority list once those interactions are counted.
18.2%
A Midwest countyAll-payer
Lives
158,867
Scored
151,181
Care-management eligible
78,630
Compounding-activePatients with at least one active interaction between their conditions.
35.7%
Re-rankedPatients who move higher on the priority list once those interactions are counted.
21.3%
Median multiplier
1.62×
Multiplier at the 95th percentile
4.5×
Installation · clean machineInstalls and runs in one step

One step installs it and runs it from start to finish, with nothing else added to the machine.

Method and reproducibility.

  • Every figure traces to a run record.
  • Every software change must pass 43 automated checks before it is accepted.4
  • Same results across five software releases.
  • Cohorts are synthetic, built from public data. No PHI.
  • The same engine runs on your extract.

Watch it run.

Eight steps, one illustrative patient, about a minute and a half. Every step appears with its stamp.

—
Run it on your data

Sources.

  • CDC. About Chronic Disease. cdc.gov/chronic-disease/about
  • America’s Health Rankings (United Health Foundation), analysis of BRFSS.
  • U.S. Joint Economic Committee. Chronic Conditions.
  • Boyd CM et al., JAMA 2005;294:716-724.
  • Barnett K et al., Lancet 2012;380:37-43.
  • Tinetti ME et al., JAMA 2012;307:2493.
  • Uhlig K et al., JGIM 2014;29:670.
  • CMS. 2024 Rate Announcement.
  • White paper: The Architecture Problem (PDF)

Sources

  1. Academic health system cohort, two regions: 2,299,052 lives, 2,088,115 scored, 911,181 care-management eligible; and 190,387 lives, 172,318 scored, 77,544 care-management eligible. Compounding-active 31.2% and 31.9%; re-ranked 17.4% and 18.2%. Source: Patient 3-D population analysis of the cohort. Synthetic cohort; no PHI.
  2. Midwest county cohort, all-payer: 158,867 covered lives; 151,181 scored; 78,630 clinically eligible; 35.7% compounding-active; 21.3% re-ranked; median multiplier 1.62×; 4.5× at the 95th percentile. Rerun on five software releases; all 492 checked values came out identical every time. Synthetic cohort built from public census data; no PHI.
  3. Installation test: a release package installed, checked and run on a clean machine with nothing else added, in one unattended step that finished without errors. Source: Hypersphere installation test record.
  4. Automated checks: 50 run on every software change, and 43 of them must pass before the change is accepted. Source: Hypersphere software build records, release of Oct 12 2026.