Healthcare IT project management is the job of taking a hospital system from a charter to a go-live that clinicians will actually use: scope, stakeholders, risk, a blended clinical/IT team, and hypercare after the switch. It is not a radiology worklist. It is not how you write the code. It is not a PYCAD project-management product.
If you meant the imaging worklist / interruptions / PACS–RIS routing → radiology workflow optimization. If you meant scanner idle / report TAT → operational efficiency in medical imaging. If you meant the engineering discipline (HIPAA as code, DICOM / HL7 / FHIR) → healthcare software engineering. If you meant how you run the SDLC → medical software development process. If you meant what “going digital” means → digital transformation in healthcare.
An imaging-IT project — a PACS cutover, a web DICOM viewer inside the EHR, a teleradiology hop — is one of the projects this page is about. PYCAD builds the viewer or the model. It does not run your PMO.
What the project is
Retail software can ship a feature and iterate in public. A hospital cannot. The new EHR module, the new archive, the new viewer has to talk to what is already there, keep PHI inside the fence, and not stop the floor on Tuesday morning. Time and budget still matter. They are not the success metric. Adoption is.
Write the charter before anyone opens Jira. What problem, for whom, what is out of scope, and which number will prove it later. “Install a viewer” is not a charter. “ED trauma CTs open inside the chart in under N seconds, from any authorized browser, without a second login” is.
People before tickets
More of these projects die on people than on a failed interface. Map the humans before you map the interfaces.
| Camp | What they care about | What they will kill the project over |
|---|---|---|
| Clinical champions | A faster or safer read. Time back at the bedside | A tool that adds clicks to a shift that is already full |
| Administrators | Budget, downtime, a go-live that does not make the news | An open-ended vendor, a surprise interface fee, a second archive |
| End users who have seen “the next big thing” | Whether Friday still works | Training that assumes they have an afternoon |
| IT / security / infrastructure | Auth, audit, the bus that already runs, capacity | A box that cannot speak HL7 / FHIR / DICOMweb, or that stores PHI without a BAA |
You need two translators, not a weekly status meeting. A clinical liaison who lives the workflow and will tell you the spec is wrong. A technical translator who asks why the request exists before they build it. When those two people sit in the same stand-up, the surgeon’s “images load too slow” becomes a hanging-protocol and a prefetch job, not a ticket that says “make it faster.”
Risks that are not “late and over budget”
- Security. A breach is not a defect. It is a reportable event and a loss of the only currency the hospital has with the patient. Encryption, role-based access, and an audit trail are in the charter, not a phase-two slide. The transfer how-to is HIPAA-compliant data transfer.
- Interop. A brilliant system that cannot talk to the EHR or the PACS is a paperweight. The products that move the data are healthcare interoperability solutions. This page is the project risk: you test the live feed, not a vendor demo.
- Adoption. A technically correct tool that the floor works around is a failed project. Pilot with the people who will hate it. Change the spec. Do not “train harder” a bad hanging protocol.
Data migration, a sudden rule, and a stubborn legacy EMR are the usual landmines. Break the work into a WBS you can actually staff: discovery, interface map, build, validation, training, hypercare. Do not pretend the landmines are not there.
How you run it
Waterfall fits a large, already-specified cutover — an EHR go-live, a PACS swap with a kill date on the old store. Requirements are known; the regulator wants a paper trail. Agile fits a new patient-facing or viewer feature you will learn from the first ten users. Most imaging-IT work is a hybrid: a dated cutover on the outside, sprints on the inside.
The daily stand-up is three questions and fifteen minutes: done yesterday, today, blocked. Executives get milestones, burn, and red risks — not a 100-row board. The team gets the board.
| Tool | What it is for | When it is the wrong pick |
|---|---|---|
| Jira | Sprints, bugs, a validation workflow you can show | A hospital-wide training calendar |
| Asana | Cross-team dates: training rooms, departmental readiness | A custom clinical build with a lot of defects |
| Trello | A small task force or the go-live help-desk lane | A multi-year EHR |
| Smartsheet | Budget, headcount, a HIPAA-eligible grid an exec will read | Developer bug tracking |
Pick the tool the team will open. A pilot on a two-week slice beats a six-month “platform decision.”
Go-live is the start
Launch day is not the finish. It is the first morning the floor has to work in the new system. Write the playbook before the date: last data check, who is on call, how you tell the building if it is slow, what you roll back to. Then hypercare — experts in the room or one call away for the first one to two weeks. A ticket queue no one watches is how you lose the champions you spent six months making.
Measure what the charter promised, not uptime alone. Report TAT, a second login that disappeared, a re-key that stopped. Those numbers are how the next project gets funded. Imaging-department TAT as an ops job is 7571 — a different URL.
What this page is not
- Not 694. Worklist design, scanner idle, hospital beds, and cost playbooks stay on those URLs.
- Not 702. How you write HIPAA-aware code, and how you run an SDLC, already have homes.
- Not a market slide. Dropped $360B / $730B, remote-PM percentages, “61% on time,” and the Deloitte-as-a-reason-to-buy link.
- Not a PYCAD PMO. We do not staff your stand-up. If the missing piece is a viewer or a model on a project that already has a charter, that is the imaging piece. Case studies.