Modality Worklist (MWL) is the bridge between the RIS order and the DICOM image. When it works, the technologist selects a patient from the console and all demographics flow correctly into every image header. When it fails, images arrive at PACS with wrong names, missing accession numbers, or no demographics at all — requiring manual correction and delaying the radiologist.
This is a field-tested list of the 15 failures you will encounter in any MWL deployment.
How MWL Works (The 30-Second Version)
The modality sends a DICOM C-FIND request to the MWL SCP (usually hosted by the RIS or a gateway). The C-FIND query includes search keys — typically AE Title, date, and sometimes patient ID. The MWL SCP queries the RIS order database and returns matching worklist items. The modality displays them for the technologist to select.
The MWL item carries the demographic and order data that the modality embeds in the DICOM image header. The two most critical fields carried from MWL to image are (0008,0050) Accession Number and (0010,0020) Patient ID.
The 15 Failures
1. Empty Worklist Despite Valid Orders
Symptom: Modality shows "No Results" even when orders exist in the RIS.
Most common cause: AE Title mismatch. The modality queries with its own AE Title as a search key ((0040,0001) Scheduled Station AE Title). If the RIS has a different AE Title entered for this modality, the query returns nothing.
Fix: Verify the AE Title entered on the modality console matches exactly what is configured in the RIS for this modality. This is case-sensitive and must be ≤16 characters.
2. Worklist Shows Other Departments' Orders
Symptom: Modality shows orders for every patient in the hospital, not just its own.
Cause: The modality sends a wildcard or blank Scheduled Station AE Title query. The MWL SCP returns all unfiltered orders.
Fix: Ensure the modality is configured to send its own AE Title in the C-FIND query. Check the modality's DICOM configuration — there is typically a field labelled "Worklist Query AE Title" or "Station Name for Filter."
3. Today's Orders Not Appearing
Symptom: Yesterday's orders still visible; today's orders absent.
Cause: Date range query not configured. The modality is querying a fixed date or a relative date that doesn't match the RIS timezone.
Fix: Configure the modality to query a date range of today ± 1 day (to catch overnight orders and allow for timezone drift). Verify system clock and timezone match between modality and RIS.
4. Orders Visible but Demographics Wrong
Symptom: Worklist shows correct patient name but wrong date of birth or no patient ID.
Cause: RIS has demographic data quality issues. Patient was manually entered on the RIS without an ADT feed, or the ADT-to-RIS demographic sync is broken.
Fix: Investigate the ADT feed from the EMR to the RIS. Verify the A01/A08 messages are flowing and that the RIS is updating demographics on update events. See HL7 ADT Messages Explained.
5. Missing Accession Number in Images
Symptom: Images arrive at PACS with blank or auto-generated Accession Number (0008,0050).
Cause: Modality not using the MWL item — technologist entered demographics manually instead of selecting from the worklist.
Fix: Enforce worklist use through workflow policy. Some modalities allow "mandatory worklist" mode that prevents manual entry. Consider adding a DICOM gateway that validates Accession Number is present and non-empty.
6. Duplicate Patient Records in PACS After MWL Change
Symptom: After a RIS upgrade or Patient ID format change, patients appear twice in PACS.
Cause: Patient ID format changed in the RIS (e.g., leading zeros removed). PACS creates a new patient record because the ID no longer matches.
Fix: Implement a patient identity reconciliation step before any RIS Patient ID format changes. PACS Patient ID matching is almost always an exact string match — 00123456 ≠ 123456.
7. Modality Can't Reach MWL SCP
Symptom: "Cannot connect to Worklist Server" error on modality console.
Cause: Network connectivity, firewall, or wrong IP/port configuration.
Fix checklist:
- Ping the MWL SCP from the modality's network segment
- Verify port 104 (or configured DICOM port) is open in firewall
- Confirm the IP address and port on the modality configuration page
- Verify the MWL SCP service is running on the RIS/gateway
8. C-FIND Returns Results but Modality Doesn't Display Them
Symptom: Network capture shows C-FIND responses with matching items, but modality console shows empty.
Cause: Character set mismatch. The MWL SCP returns UTF-8 encoded names but the modality only supports Latin-1.
Fix: Configure the MWL SCP to send a compatible character set, or ensure the modality has the correct (0008,0005) Specific Character Set configured. For sites with non-Latin patient names, this is critical.
9. Worklist Items Appear After Exam Is Already Done
Symptom: Technologist starts scanning manually, then the worklist item appears 10 minutes later.
Cause: High latency between order entry in EMR, HL7 ORM transmission, RIS order creation, and MWL availability.
Fix: Measure the end-to-end latency: order entry → ORM sent → ORM received by RIS → RIS creates order → MWL SCP cache updated → modality C-FIND query. Any leg > 60 seconds needs investigation. Common culprit: RIS batches MWL cache refresh every 5–10 minutes.
10. Wrong Body Part on Worklist Item
Symptom: Technologist selects correct patient but the protocol shows the wrong procedure.
Cause: RIS procedure catalogue has wrong body part or modality mapping for this procedure code. Often a configuration error introduced during a catalogue update.
Fix: Audit the RIS procedure catalogue for the affected procedure. Verify (0018,0015) Body Part Examined mapping in the MWL response matches the expected procedure. See the Body Part Examined tag reference.
11. MWL Returns Multiple Items for Same Patient
Symptom: Technologist sees the same patient listed twice or three times on the worklist.
Cause: Duplicate orders in the RIS (often caused by duplicate ORM messages from the EMR), or the RIS is not correctly de-duplicating when an order is modified.
Fix: Investigate ORM deduplication in the HL7 interface engine. Check ORC-2 (Placer Order Number) — duplicates here always cause duplicate worklist items.
12. Cancelled Orders Still Showing on Worklist
Symptom: Technologist selects an order that was cancelled in the EMR; results in wrong study being performed.
Cause: ORM^O01 with ORC-1 = CA (Cancel) not processed by the RIS, or MWL SCP not removing cancelled items.
Fix: Verify the HL7 cancellation flow: EMR sends ORM with ORC-1=CA → RIS marks order cancelled → MWL SCP removes or flags the item. Test this explicitly — a missed cancellation is a patient safety issue.
13. Scheduled Procedure Step Not Closed After Exam
Symptom: RIS shows exams as "In Progress" indefinitely. Reporting queue never moves.
Cause: MPPS (Modality Performed Procedure Step) N-SET with status COMPLETED not sent by modality, or not received by RIS.
Fix: Verify the modality has MPPS configured and pointing to the correct SCP. Check MPPS SCP logs for incoming N-CREATE and N-SET messages. This is related to the Performed Procedure Step Status tag (0040,0252).
14. Worklist Works in Test, Fails in Production
Symptom: MWL worked perfectly during UAT, fails after go-live.
Cause: Production RIS has different AE Title, IP, or procedure catalogue configuration than the test system.
Fix: Maintain a configuration checklist that documents every MWL parameter (AE Title, IP, port, query fields, character set) and verify it against both test and production environments before sign-off.
15. Worklist Stops Working After RIS Upgrade
Symptom: Everything worked before the RIS upgrade; now the modality gets empty results or connection failures.
Cause: RIS vendor changed MWL SCP port, AE Title, or C-FIND response structure without communicating it.
Fix: Treat every RIS upgrade as a potential MWL interface change. Capture a baseline DICOM network trace before the upgrade. Compare with a trace after the upgrade to identify any changes in the C-FIND response structure.
Prevention: MWL Monitoring in Production
The best approach is not reactive troubleshooting but proactive monitoring:
- Set up a synthetic monitor — a small script that sends a test C-FIND query every 5 minutes and alerts if the response is empty or takes > 5 seconds
- Track "manual entry" rate — PACS can report what percentage of arriving studies have a blank Accession Number; this is a direct proxy for MWL bypass
- Monitor MPPS completion rate — RIS should show near-zero "In Progress" studies after shift end
- Alert on worklist queue depth — if more than 20 orders are on the MWL unselected at shift start, something is wrong with order flow
A functioning MWL is invisible to the clinical team. The only time they notice it is when it fails. Keep it healthy and it stays invisible.
Related: DICOM Tags Reference for PACS Engineers · HL7 Integration Guide · Accession Number tag