BRENNEN

Industries / Health / Healthcare

HealthTech

HealthTech makes and sells the software that hospitals, clinics and patients use to keep medical records, advise clinicians, watch over patients, book and bill care, and pass health data from one system to another.

1965

Lockheed Missiles and Space Company and El Camino Hospital in Mountain View, California, began building one computer system for a whole hospital; an IEEE Milestone plaque dedicated in May 2026 calls it "the first hospital-wide computerized medical information system".

ethw.org

42

Milestones achieved across 6 domains.

17

To watch for: announced, not yet confirmed outside.

16

Open deliverables. Nobody has shipped them.

In one minute

Where it started

Hospital software began in the 1960s as large projects at single hospitals: Lockheed's system at El Camino Hospital, where doctors placed orders by pointing a light pen at a screen, and MUMPS, a programming language for medical databases written at Massachusetts General Hospital and taken up by sellers such as Meditech and InterSystems. From 1969 small hospitals shared billing mainframes, and in the 1970s F. T. de Dombal's team in Leeds and Clement McDonald in Indianapolis showed that computers could suggest diagnoses and remind doctors of steps they had missed. In 1987 a new standards group, HL7, began writing the message format that hospital computers still use to talk to each other.

Where it stands

Software now holds the chart, checks orders and sends claims: HL7 says its version 2 messages are used by 95% of US healthcare organizations, Epic says more than 325 million patients have a current record in its software, and by the US health IT office's count 464 million documents had crossed the national exchange network TEFCA by the end of 2025. Many of the industry's lessons come from failures in use: deaths rose at a Pittsburgh children's hospital after order-entry software was switched on in six days, a new VA record system sent more than 11,000 orders to a queue no one could see, and a 2024 ransomware attack on one clearinghouse left practices across the US unable to send claims. AI tools are arriving in every domain, but outside tests are few: in a randomized trial at Kenyan clinics, an AI adviser improved diagnoses and treatment plans as judged by reviewers but did not significantly reduce treatment failures.

What is open

Most products are judged by their makers' figures: order checkers, AI coding and prior-authorization tools, record converters and the AI answer tools clinicians consult have few public, outside tests. The open jobs below build public answer keys and test benches from made-up patients (from the open-source generator Synthea), open-source record systems, simulators and volunteers, so that anyone can score a product without touching a real patient. AI coding agents and web agents now make it cheap to write the converters and run the thousands of test cases these need.

Domains

AchievedWatch forOpen

Advances to build with

2 research notes

Open in healthtech

3 of 16 open deliverables

Open · Records

What a chart move loses

A public test that moves the same synthetic patients between open-source record systems and counts, item by item, what is lost, changed or duplicated on the way. Generate 1,000 patients with Synthea from a fixed seed, load them into OpenEMR, move them to a second system such as OpenMRS or Medplum and back, using each system's own export and import tools and any community migration module. Compare every allergy, problem, medication, immunization, lab result, vital sign, procedure and encounter with the Synthea original, and log each difference as missing, altered (time, unit or code) or duplicated. Only made-up patients are used.

Built with what exists

Synthea, an open-source generator of made-up patient records; OpenEMR (GPL), OpenMRS (MPL 2.0) and Medplum (Apache 2.0) from GitHub; the oe-module-migration importer; Python for the item-by-item comparison; an AI coding agent such as Claude Code to write converters where no import path exists, with its losses reported beside those of the hand-built tools. Coding agents make it cheap to write a converter for each pair of systems, which is most of the work.

Credit · unclaimed

Open · Decision support for clinicians

An open order-check test

A public set of at least 200 test cases in the style of the Leapfrog test, using made-up patients only. Each case is a synthetic patient (allergies, kidney function, age, current drugs) plus one draft order, labeled "should alert" or "should not alert" by two pharmacists. Cases cover the test's 10 published categories: drug-allergy, drug-drug, wrong route, single dose, duplicate therapy, drug-lab result, daily dose, drug-age, drug-diagnosis and a missing companion order, and at least 40% are safe orders, so that needless alerts are counted. Each case is packaged as a FHIR record bundle sent through CDS Hooks, the HL7 standard by which a record system calls an outside advice service, at the "order-sign" moment when a clinician signs orders. The same cases then run against any order-checking software or AI model, and the scores are published.

Built with what exists

Synthea, MITRE's open-source generator of made-up patient records in FHIR R4; the CDS Hooks order-sign hook; the 15 high-priority interacting drug pairs that an expert panel said every record system should flag and FDA drug labels from openFDA as the basis for the expected answers; OpenEMR and AHRQ's open-source CQL Services, which answers CDS Hooks calls with rules written in CQL, as conventional engines; AI models through their APIs, including an open-weight model such as Mistral Large 4 (see the Note "A trillion open parameters").

Credit · unclaimed

Open · Monitoring and alarms

A NEWS2 answer key

A public set of at least 2,000 synthetic observation sets, each with the correct NEWS2 sub-scores, total and response level worked out from the Royal College of Physicians' charts. It covers every boundary value of the six measurements, both oxygen saturation scales (the second is for patients at risk of type II respiratory failure), extra oxygen, new confusion, a single measurement scoring 3, and missing, zero or impossible entries. Run it through every NEWS2 calculator that can be reached: open-source code, free web and phone calculators, test versions of hospital e-observation systems with their makers' permission, and general AI chat models. Publish each one's error rate after giving makers time to fix what is found.

Built with what exists

the RCP's NEWS2 charts and its 2021 additional implementation guidance; a short reference scorer in Python, checked by hand against the charts; Browser Use 1.0, a web agent, to type each case into web calculators and read back the result (see the Note "Deleting 80% to start over"); Gemini Decision 1, which scores up to 1,000 fixed-answer questions in one request, to put the cases to a chat model in two requests (see the Note "A thousand questions, one pass"). Synthetic cases only; no patient data.

Credit · unclaimed

Where it sits

Broader

Healthcare