Every PACS go-live has a moment — usually around 5 AM on the morning of cutover — where you realise something was not on your checklist.
After 15+ hospital PACS deployments, those moments have been catalogued. This article presents the full 47-point verification framework, organised by domain, that forms the backbone of every go-live we run.
This is not a generic IT checklist. Every item maps to a specific failure mode observed in production.
Why Go-Live Checklists Fail
Most PACS administrators approach go-live with a checklist. Most of those checklists are inadequate.
The common failure patterns:
- Checklists built from vendor documentation rather than production experience
- Items that are too high-level to be actionable ("verify DICOM connectivity" vs "send a test CT study from each modality AE and verify it appears in the PACS worklist within 60 seconds")
- No assigned owner per item — everybody's responsibility is nobody's responsibility
- No sign-off mechanism — items get verbally confirmed but never formally closed
- No sequencing — critical path items aren't distinguished from nice-to-have verifications
The framework below addresses all of these.
The 47-Point Framework
Section 1: Infrastructure (9 items)
1. Confirm all PACS application servers are running at target specification. Verify CPU, RAM, and available disk space against design requirements. No server should be above 70% capacity baseline.
2. Confirm database server connectivity from all application tier servers. Run a test query against the PACS study table. Record response time — anything over 100ms for a simple indexed query warrants investigation.
3. Verify SAN/NAS storage is accessible from all application servers. Confirm read/write permissions. Test with a file write of >100 MB to fast-tier and near-line storage.
4. Confirm network VLAN configuration is complete. Imaging traffic VLAN separated from general hospital network. Verify with a packet capture that DICOM traffic is routing correctly.
5. Test all firewall rules for DICOM traffic: PACS to each modality subnet, modality to PACS, PACS to any external DICOM destinations (teleradiology, referring hospitals).
6. Verify backup jobs have run successfully within the last 24 hours. Confirm backup logs. If backup has not run, do not proceed to go-live until it has.
7. Confirm DR/failover configuration. For active-passive setups, verify the secondary node is synchronised and failover has been tested within the last 7 days.
8. Verify UPS coverage for all PACS infrastructure. Confirm runtime under load. Any server in scope without UPS coverage is a go-live risk.
9. Confirm monitoring and alerting is active for all PACS infrastructure components. Verify alert routing to on-call team. Test with a simulated alert.
Section 2: DICOM Connectivity (10 items)
10. Send a test DICOM study from every modality AE Title configured in the PACS. Verify each study appears in the PACS worklist with correct patient demographics and study metadata. Document AE Title, IP, port, and test result for each modality.
11. Verify Modality Worklist retrieval from every scheduled modality. Create a test scheduled procedure in the RIS for a test patient. Verify the worklist entry appears on the modality within 60 seconds. Check that patient name, MRN, accession number, and procedure description are correctly populated.
12. Test DICOM C-STORE association from each modality. Verify the following DICOM header fields are correctly populated in received studies: (0010,0010) Patient Name, (0010,0020) Patient ID, (0020,000D) Study Instance UID, (0008,0060) Modality, (0008,0050) Accession Number.
13. Verify DICOM transfer syntax negotiation is successful for each modality. Check PACS logs for any "Presentation Context Rejected" entries from modality connections.
14. Test DICOM C-FIND (study query) from radiologist workstations. Verify studies can be located by patient name, MRN, accession number, and date range.
15. Verify DICOM C-MOVE (study retrieval) from radiologist workstations. Retrieve a test study and confirm images display correctly at full resolution.
16. For any modalities using DICOM TLS: verify certificate validity, verify TLS handshake succeeds, confirm cipher suite compatibility.
17. Test DICOM print SCP if in scope. Send a test print job and verify output.
18. Verify DICOM storage commitment is functioning. Confirm the PACS is sending N-EVENT-REPORT responses to modalities that use storage commitment.
19. Confirm DICOM audit logging is enabled and audit records are being written. Verify audit log storage capacity.
Section 3: HL7 Interfaces (8 items)
20. Verify every HL7 interface is live in production and in a running state. Check interface engine monitoring for each connection. Any interface in an error or stopped state must be resolved before go-live.
21. Send a test ADT A01 (patient admit) message and verify patient record is created in PACS with correct demographics.
22. Send a test ADT A08 (patient update) message and verify patient demographics are updated in PACS.
23. Send a test ORM O01 (new order) message and verify a worklist entry is created in PACS. Confirm scheduled procedure maps correctly to PACS procedure dictionary.
24. Send a test ORM O01 (cancel order, ORC-1=CA) and verify the worklist entry is cancelled in PACS.
25. Verify ORU R01 (result/status) messages are being generated by PACS and received by the RIS when a study is completed and a report is signed.
26. Test interface engine failover: stop the receiving PACS HL7 listener, send 10 test messages, verify messages queue in interface engine, restart listener, verify all 10 messages are delivered in order.
27. Confirm HL7 message archive is enabled in interface engine. Verify archived messages are searchable and retrievable for audit/troubleshooting purposes.
Section 4: Workflow Validation (8 items)
28. Complete an end-to-end workflow test for each major study type: outpatient radiology, inpatient portable, emergency/STAT, and scheduled procedure. Track from order creation in RIS to signed report in PACS.
29. Verify radiologist worklist configuration: worklist filters, sort order, prior study display, hanging protocols for CT, MRI, CR/DX. Have at least one radiologist from each specialty sign off on their worklist configuration.
30. Test report distribution: verify signed reports route correctly to RIS, HIS, and any configured result distribution endpoints. Confirm fax routing if in scope.
31. Verify STAT/emergency workflow: create an emergency order, verify it appears with correct priority on the radiologist worklist, verify expedited routing.
32. Test cross-modality comparison: retrieve a prior CT study while reading a current MRI. Verify prior loads without manual intervention.
33. Verify downtime procedures are documented and accessible offline. Confirm staff know where to find them. Test that the downtime worklist is accessible.
34. Test patient demographic correction workflow: correct a patient name in the HIS, verify the correction propagates to PACS via ADT A08.
35. Verify image quality: spot-check display of CT, MRI, and CR studies on radiologist workstations. Confirm window/level presets, zoom, and measurement tools are functioning.
Section 5: Training and Readiness (6 items)
36. Confirm all radiologists have completed go-live training. Minimum: 4-hour hands-on session with sign-off. Identify any radiologists who have not trained and arrange catch-up before go-live.
37. Confirm all imaging technologists have completed modality-specific training. Each modality type (CT, MRI, CR, etc.) requires separate training confirmation.
38. Confirm PACS helpdesk staff have been briefed on the top 10 expected go-live issues and escalation procedures. Verify on-call engineering contact list is current and distributed.
39. Distribute go-live day contact list to all stakeholders: PACS vendor support, infrastructure on-call, interface engine support, radiology department lead, IT management escalation.
40. Confirm clinical super-users (1–2 per department) are identified and available on go-live day. These are the first line of support for workflow questions.
41. Verify go-live communication has been sent to all affected departments: radiology, ED, ICU, inpatient wards, outpatient clinics. Include helpdesk contact number prominently.
Section 6: Go-Live Day (6 items)
42. T-1 hour: run a complete DICOM connectivity test across all modalities. Document any failures and confirm resolution or mitigation before proceeding.
43. T-1 hour: verify HL7 interface queue depths are at zero (no backlog). Any queued messages indicate an issue that must be resolved before cutover.
44. T-0: confirm system switchover sequence has been followed per the cutover plan. Verify legacy PACS is in read-only mode (if parallel migration) or offline (if big bang).
45. T+30 minutes: verify first real study has been received and is accessible from radiologist workstations. Confirm MWL is serving real patient orders correctly.
46. T+2 hours: review PACS error logs and interface engine logs for any unexpected errors. Address any issues that have emerged.
47. T+4 hours: confirm with radiology department that workflow is stable. Confirm with clinical leads in ED and ICU that urgent image access is functioning.
Using This Framework
Each of the 47 items above maps to a test procedure, an owner, a pass/fail criterion, and a resolution path.
The Excel-based PACS Go-Live Checklist in the Majware product catalogue formats all 47 items into a structured sign-off workbook with:
- Owner assignment per item
- Pre-go-live / day-of / post-live phase classification
- Pass/Fail/N-A status tracking
- Notes and issue log per item
- Automated completion summary dashboard
It has been used across hospital systems in [region] and updated after each deployment. If something goes wrong on your go-live that isn't covered, we want to know about it.