Healthcare and hospitalsWhat we do for hospitals, and one case
Hospital care can be improved massively, inside Epic, without clinicians noticing.
We map the care pathway as it really runs from the hospital’s own records, set it against clinical standards, and land the fixes where clinicians already work.
What a hospital can gainPublished results, with their sources
The gains are documented.
1.6 days
shorter length of stay with an endocrine consult within 48 hours
Sheahan et al., Endocrinology, Diabetes & Metabolism, 2021
39%
fewer high-glucose patients with a virtual glucose service in Epic
Rushakoff et al., Annals of Internal Medicine, 2017
7.7 h
in Epic per 8 h of clinic for endocrinologists, the 2nd-highest specialty
Holmgren et al., Journal of General Internal Medicine, 2024
4–11%
of hospital costs: the AI savings potential ($60–120B a year)
Sahni, Stein, Zemmel & Cutler, NBER Working Paper 30857, 2023
How we work in a hospitalWhy the head of endocrinology says yes
How we win the clinical champion.
Epic is the process map
Mapped from Clarity, Caboodle and audit logs. No interviews.
Invisible in the workflow
Fixes land where clinicians already work: BPAs, order sets, In Basket.
Zero human in the loop
Runs on its own. No added staff, no training, no new screen or login.
Hyper-tailored
Tuned to your endocrine protocols and CMS glycemic-harm measures, mandatory from the 2026 reporting period.
Case in pointTeaching hospital, inpatient diabetes care
$4.2–4.8M saved a year.
Results at month 12 · 0 clinician training hours.
58% → 92%
insulin orders to ADA standard
Teaching hospital, results at month 12
4.1 → 2.3
hypoglycemia per 100 patient-days
Teaching hospital, results at month 12
19% → 15%
30-day readmissions
Teaching hospital, results at month 12
How we got thereThe six steps of the hospital case
From EHR orders to live checks on every patient.
Step 01: Gather the data
EHR orders, glucose meters, pharmacy, lab, nurse notes
Step 02: Map the process
As-built glucose pathway: 14 variants across 9 wards
Step 03: Map the data
Meter to order to dose; 3 feeds with no owner
Step 04: Assess the gap
vs ADA Standards of Care: sliding-scale only, late basal insulin
Step 05: Design best practice
Tailored order-set defaults, signed off by the head of endocrinology
Step 06: Implementation monitoring
Live QC on every patient, silent to clinicians