An imaging order is the most common place a healthcare interface has to cross
standards. The order is written in HL7 v2, the exam is performed and stored in
DICOM, and the report comes back in HL7 v2 again. Nothing translates between
them automatically. This page follows one order across both seams.
Two standards, one order. The first two hops and the last are HL7 v2; everything between them is DICOM. Most imaging interface work happens at those two seams, not inside either standard.
The handoff is not a conversion step. The RIS receives an HL7 order and, some
time later, a modality asks a DICOM service for a worklist. Those are two
separate transactions against the same underlying order, and the only thing
tying them together is the data the RIS copied from one into the other.
That copy is where imaging interfaces break. If it never happens, the modality
queries an empty worklist. If it happens with the wrong values, the study
arrives at the PACS carrying identifiers the
order system does not recognise.
The ordering system sends an ORM message naming the patient, the procedure and
the accession number. The ORM message reference
covers the order control codes and the full lifecycle; what matters here is
which fields the worklist will be built from.
An ED physician orders a CT of the abdomen and pelvis with IV contrast to evaluate acute right lower quadrant pain.
What the worklist is built from. Each ORM field lands in a specific DICOM attribute, and the accession number is the one both sides have to agree on: it is what lets the finished study find its order again.
The accession number is the join. It is assigned by the order system, carried
in OBR-18, written into the worklist entry as DICOM
Accession Number, and returned in the completed
study. Every later step that has to associate images with an order resolves
through it.
The modality asks a worklist provider what it is scheduled to do, using the
DICOM query services rather than anything HL7. The
DICOM service classes page covers how those
queries are structured; the operationally interesting part is what happens when
one comes back empty.
The failure everyone meets first. A query that returns nothing does not stop the exam, it just moves the data entry to the console, which is how a study ends up with an accession number the RIS has never seen.
A technologist who gets nothing from the worklist will key the demographics in
by hand and scan the patient anyway, because the alternative is sending the
patient away. The exam is now correct and the metadata is not, and the study
will not reconcile against its order without manual repair.
While the exam runs, the modality reports progress with MPPS, and on completion
sends the images to the archive. The archive now holds a study; the order
system still has an open order. The loop closes when the reading radiologist
signs a report and the RIS sends it back as an ORU message.
A radiologist finalizes a 2-view chest X-ray report, returning the narrative findings and impression to the ordering EHR.
Radiology information system (RIS) originating the report
MSH-4
Sending Facility
MAIN_HOSP
MSH-5
Receiving Application
EPIC
Ordering EHR that receives the result
MSH-7
Date/Time of Message
20260514093000
YYYYMMDDHHMMSS, when the report was transmitted
MSH-9
Message Type
ORU^R01^ORU_R01
Unsolicited Observation Result
MSH-10
Message Control ID
RAD00781
Unique message identifier for ACK matching
MSH-11
Processing ID
P
P=Production, T=Training, D=Debug
MSH-12
Version ID
2.5.1
MSH-15
Accept Acknowledgment Type
AL
AL=Always send an accept ACK
MSH-16
Application Acknowledgment Type
NE
NE=Never send an application ACK
Field
Name
Value
Note
PID-3
Patient Identifier List
MRN78214^^^MAIN_HOSP^MR
Medical Record Number: primary matching key
PID-5
Patient Name
TANAKA^MARCUS^LEE
Family^Given^Middle
PID-7
Date of Birth
19711203
YYYYMMDD format
PID-8
Administrative Sex
M
PID-11
Patient Address
812 CEDAR LN^^FAIRVIEW^OH^44016^USA
PID-13
Phone (Home)
^PRN^PH^^^555^3014477
PID-18
Patient Account Number
ACCT55310^^^MAIN_HOSP^AN
Visit/encounter account
Field
Name
Value
Note
PV1-2
Patient Class
O
O=Outpatient: exam performed in the imaging department
PV1-3
Assigned Patient Location
RAD^XR2^^MAIN_HOSP
Department^Room^Bed^Facility
PV1-7
Attending Doctor
REF2244^PATEL^ANJALI^^^^MD
Ordering/attending physician of record
PV1-10
Hospital Service
DIAG
Diagnostic Imaging
PV1-18
Patient Type
OP
Outpatient
Field
Name
Value
Note
ORC-1
Order Control
RE
RE=Observations/Performed (results being returned)
ORC-2
Placer Order Number
ORD44120^EPIC
Ordering EHR reference
ORC-3
Filler Order Number
FIL77820^RADIANT
RIS reference
ORC-5
Order Status
CM
CM=Completed
ORC-12
Ordering Provider
REF2244^PATEL^ANJALI^^^^MD
Field
Name
Value
Note
OBR-1
Set ID
1
OBR-2
Placer Order Number
ORD44120^EPIC
OBR-3
Filler Order Number
FIL77820^RADIANT
OBR-4
Universal Service ID
36643-5^XR Chest 2 Views^LN
LOINC code for chest radiograph, 2 views (PA and lateral)
OBR-7
Observation Date/Time
20260514084500
When the imaging study was acquired
OBR-16
Ordering Provider
REF2244^PATEL^ANJALI^^^^MD
OBR-22
Results Report/Status Change Date
20260514092600
When the report was finalized
OBR-24
Diagnostic Service Section ID
RAD
HL7 Table 0074: RAD=Radiology
OBR-25
Result Status
F
F=Final result (HL7 Table 0123)
OBR-32
Principal Result Interpreter
PRI8890&OKAFOR&DAVID&&&&MD&&RADIANT
Radiologist who interpreted the study and signed the report
Field
Name
Value
Note
OBX-1
Set ID
1
OBX-2
Value Type
TX
TX=Text data: free-text narrative
OBX-3
Observation Identifier
18782-3^Radiology Study observation^LN
LOINC code for a radiology study observation
OBX-4
Observation Sub-ID
1
Groups this line as the FINDINGS section
OBX-5
Observation Value
FINDINGS: The lungs are clear and well expanded without focal consolidation, effusion, or pneumothorax. Cardiomediastinal silhouette is within normal limits. No acute osseous abnormality. Visualized upper abdomen is unremarkable.
Radiologist narrative description of the study
OBX-11
Observation Result Status
F
F=Final
OBX-14
Date/Time of the Observation
20260514092600
OBX-16
Responsible Observer
PRI8890^OKAFOR^DAVID^^^^MD
OBX|2|TX|19005-8^Radiology Imaging study^LN|2|IMPRESSION: No acute cardiopulmonary process.||||||F|||20260514092600||PRI8890^OKAFOR^DAVID^^^^MD
That ORU carries the same accession number the order started with, which is how
the report lands against the right order in the chart rather than becoming an
orphaned document.
The accession number does not match. The most common cause, and the
easiest to prove: compare OBR-18 on the order against
Accession Number in the stored study.
The patient identifier does not match. The worklist entry was built from
a stale PID, or the modality overwrote it during
manual entry.
The order never reached the worklist provider. Check the ORM was
delivered and acknowledged before looking anywhere downstream.
The study was performed outside the scheduled window. Worklist queries
are usually date-scoped, so a rescheduled exam can be invisible to the
modality without anything having failed.
Allow us to enhance your experience. We use cookies to understand
what's useful and keep making the site better for visitors like you.
Privacy Policy
"Accept all" allows analytics and advertising cookies, including
HubSpot (CRM and visitor analytics) and Google Ads (advertising and
remarketing tags), to help us understand how visitors use the site
and to deliver relevant advertising. "Essential only" enables
functional features such as live chat, without analytics or
advertising tracking. "Deny" refuses every optional cookie. See
our Privacy Policy for details.