PaLM F2F Minutes 2018-Nov-12

Back to IHE Pathology and Laboratory Medicine (PaLM) Domain

Back to IHE Pathology and Laboratory Medicine (PaLM)Technical Committee Page

Attendees

	Nov 12 AM	Nov 12 PM	Nov 13 AM	Nov 13 PM	Nov 14 AM
Francois Macary	x	x	x	x	x
Riki Merrick	x	x	x	x	x
Raj Dash	x	x	x	x	x
Kenichi Takahashi	x	x	x	x	x
Naomi Ishii	x	x
Filip Migom	x	x	x	x	x
Jim Harrison	x	x	x	x	x
Mary Kennedy	x	x	x	x	x
Gunter Haroke			x	x
Alesandro Sulis	x	x	x	x	x
Francesca Frexia	x	x	x	x	x
Andreas Vendel			x
Dan Rutz		x	x
Ed Heierman		x
Francesca Vanzo	x	x	x	x	x
Megumi Kondo	x	x	x	x	x
Mario Villace	x	x	x	x	x
Ed Heierman		x
John Hobson			x	x	x
Nick Haarselhorst			x	x
Mikael Wintell			x	x

Minutes

Presentations are available on Google Drive https://drive.google.com/open?id=1EjH9C0s0i_RvsN2qQXLZ6P22kC56ealn

November 12, 2018

Introductions were given and the agenda and goals for the meeting was reviewed by Francois.

PaLM Deployed Around the World

Japan update slides available here

Kenichi reviewed the IHE activities in Japan. The October 2018 Connectathon in Tokyo had 250-300 attendees from 10 companies with 45 instances of established connections. The 2019 Connectathon will be October 7-11 in Yokohama, Japan. Several dates were presented for the 2019 IHE PaLM F2F meeting in Tokyo. The most likely dates will be May 27-29. Other dates include: June 3-5 (no Riki); June 19-21; June 24-26. A final decision will be made in the next month of

Europe update

(https://www.ihe-europe.net/news/events)

The 2019 Connectathon will be April 8-12 in Rennes, France. Roche might bring their middleware; MIPS is planning to test TMA (bloodbank) (hopefully EPIC will be able to join MIPS for TMA);

Implementations: France big hospitals are implementing LTW and LCSD profiles. In Sardinia, Arsenal is working to get labs of the entire Veneto region to adopt LOINC; they have software that offers LOINC mapping; LOINC is used for results. Orders use national codes.

North America update

(https://www.iheusa.org/ihe-na-Connectathon-2019)

The 2019 NA Connectathon will be held January 21-25 in Cleveland, Ohio.

IHE PaLM Board Report Board Report Available Here

Riki reviewed the draft PaLM Board report. The following issues were discussed

• Technical Committee Co-Chair in Japan – Kenichi may be the co-chair pending a decision by IHE Japan and JAHIS
• Mary will have Sabrina update the 2018 meeting roster
• Inpeco – LBL deployment – Alessandro to find out in the next few days
• Japan deployment - Megumi provides during the meeting, the webpage links also for IHE Japan and JAHIS
• Belgium- KHMER based on XD-Lab – content is XD-Lab with an XML wrapper. So KHMER is replacing the IHE XDS wrapper, but XD-LAB is used plainly within this KHMER envelope. – in trial proof of concept (end of 2019 should have final specification including LOINCs) – this will be removed from the draft report for now
• The Netherlands - LAB-1 and LAB-3 for LTW operational with EPIC and plan implementation of TMA next quarter University of GRONINGEN (UMCG). Filip is the contact.
• IHE presentation – introduction to IHE PaLM activity and technical framework at JACLaS October 11-13, 2018 (Kobe, Japan)
• Presentation of IHE basics and DICOM Digital Pathology Visions meeting November 5 – 6, 2018 (San Diego, CA)

- There was support of a joint IHE and DICOM Connectathon – now looking for date for the Connectathon in 2019: first IHE option April 8 – 12, 2019 in France; this may be too early, a better plan is for North America Jan 21-25, 2020. An alternative is to host it at the next Digital Pathology Visions meeting in October 6-8, 2019 in Orlando, FL. DPA would support that.

- PaLM is invited to the European Pathology Congress for a joint meeting with DICOM WG 26 on April 11 – 13, 2019 Warwick, GB

- Will LAW and LIVD be CLIA recommendation? There was a CLIAC meeting was last week; Mary to find out if it was on the agenda.

- OO has a project to provide mapping v2 to FHIR elements; possibly after that and when FHIR has workflow worked out we could target work on LOI and LRI

- Remove the harmonization bullet but add a bullet about discussion to create new LCSD profile based on FHIR catalog resources

- Need to still get feedback on the LIVD and LAW piece from Ed

- Include future plans for TMA

• There was a comment ballot in May 2018, have to publish as Trial Implementation – need to check with Dan to get that done, then we have the technical team to build the profile in Gazelle
• Also has possibility of implementation at university hospital in Belgium (Leuven + Antwerpen)
• In December Filip to meet with Epic Netherlands about implementing TMA (December 19th : MIPS offices in Ghent Belgium)

Specimen Event Tracking (SET) Vol 2 Review (SET) Powerpoint Available Here

Alessandro reviewed SET to date:

• Will work in parallel on HL7 CR and Vol 2 editing using z messages until Hl7 approves the new message structures
• Looking at the message structure document

- Z01

• More than 1 PRT segment following the EVN?
  • Idea there is to use it when a participation is valid for ALL the following segments
  • We may have to remove PRT after EVN and then create a specimen group that has PRT and OBX
  • And a group for container prep an have PRT after container, too
  • For each specimen, would there be more than one collector?
    • PRT can track person, organization, location and device, so if there was interest in tracking the person and the device used for the collection = assume ONLY 1 PRT, since matrix ONLY lists the person at this point
      • Example US guided biopsy (the needle, the laparoscope and the US used?) – do we need to expand the use case described?
• For all PRT segments following the EVN (if EVN = act, then following v3 logic it could be there, but how to then declare when more than one specimen is collected or more than one device was involved in preparing the container) – since we are creating messages for tracking an event
• Currently we lean towards having PRT following the EVN, since 1 event is being described
• => create different options for the HL7 CR
• Need to also expand the HL70912 codes to support this
• Adding in the ORC and OBR – just to have the order numbers (ORC-2/OBR-2, ORC-3/OBR-3, ORC-4) and the test ordered (OBR-4) – make the order group optional
• No need for the TQ1 segment, unless we need the priority for the order
• For container preparation – do we ALWAYS know the specimen already?
  • Yes for SET
• Specimen collection missed event:
  • Where to place the missed reason – OBX/SPM, or add NTE somewhere – NTE-9 supports coded, NTE-4 would be coded to express reason
  • As alternative could make different message trigger (in MSG.2) and use the EVN-4 and not have NTE
• SAC-27 is container material – as well as any additives – there is a CR at OO that may rename that field
• EVN-4 is the event reason code – we are using that to describe the event type, since EVN-1 is withdrawn – we may have to create trigger events for each, so that the MSH-9.2 would have a different trigger event

- Z02

• If physician updates a specimen to move the patient from one location to a different one – may miss the shipment,
• Missing the shipment status? Have
• What does the shipment ID represent?
  • Each message is ONLY about 1 shipment
  • Use the PRT segment to declare the from and to location following the EVN or SHP?
  • MUST have at least 1 PRT to cover at least the from location (PRT-9), but could have more like person shipping, or organization shipping as well as intended recipient
  • If the message is about just 1 shipment, then no { for the shipment group)
  • Need to define new codes in HL70912 – depending on how we define the codes, may be able to document shipping location and organization and person in one PRT? Also, could add code for intended recipient of specimen or something like that.

IHE LAW and CLSI AUTO16 update

Ed updated the group on CLSI’s incorporation of LAW into AUTO16 as well as LIVD and FDA:

• For inclusion into the IHE PaLM Board report:
  • AUTO16 – has been sent out for public review – due by 12/10
  • After that editing and approval assume Q1 of 2019
  • LIVD on FHIR is getting close to getting stable definition, will have Connectathon track Jan 2019 and ballot May 2019
  • FDA is reviewing a grant proposal to have funding for proof of concepts to implement both LIVD (FHIR) and LAW (vendors, LIS and middleware participants)
  • CLSI automation and interoperability expert panel is working with FDA to create an interoperability and guidance document focusing on LAW, LIVD and potentially some of the other IHE PaLM profiles (LTW) – maybe eDOS (FHIR) potentially LOI/LRI, too = this plays into potential updates to LAB-1 and LAB-3 due to digital pathology needs.
  • Spell out FDA, MDIC the first time in the report

Transfusion Medicine A (TMA) update

Filip and Dan presented an update on TMA:

• TMA plans
  • Have completed public comment
  • Needed to add sample messages before publishing Trial for Implementation
  • Want to have Connectathon soon, but would need it for Gazelle team as soon as possible
  • Dan will try to finish any edits by end of this week
• 2 additional use cases for new profiles:
  • Ordering of blood products (create a new profile – how to adjust LAB-1 to accommodate the existing message structure in the base (OMB)
    • University of Leuven has joint venture to deliver software services for 26 hospital services in Belgium
    • MIPS working on that with EPIC (12/19 meeting with Dutch colleagues including standardizing the dispense information)
  • Dispense = once order is placed in lab – follow up of information in the lab status of blood bag after crossmatch, but before TMA is triggered, no underlying standard exits

2019 Call for Proposals

We will open the call for proposals on November 16 and close it on January 3, 2019. Decisions regarding new proposals will be discussed on the January 9th 2019 conference call.

Laboratory Clinical Communication (LCC) update

Riki and Jim gave an update on the LCC:

• The document was reviewed; Riki will update the document per the notes and will share it with the group
• We will review again on Wednesday; the plan is to send to Mary Jungers by end of this meeting; expect it will be published within a week or so (could be longer due to holidays)
• Jim and Mary will write an announcement to send to CAP committees asking for their input
• Have open for comment till Feb 8th, so we can review any on call Feb 13th

The meeting adjourned for the day at 5:00PM local time.

November 13, 2018

Francois reviewed the agenda for the day. Introductions of new attendees were made.

Digital Pathology Update (IHE white paper link and PPT from Raj)

Raj lead the discussion for the day on the work of the digital pathology white paper (https://wiki.ihe.net/index.php/APW-EDM_White_Paper)

• Image Acquisition profile was the focus of the day with the following discussion/decisions:
  • The Actor transaction diagram from Francois was displayed. It is ambiguous how the roles for image archive and image manager are connected – they are grouped in one system.
  • At the DICOM WG26 Connectathon and F2F meeting in San Diego there was a summary of the 2 year progress. DICOM now wants to expand the workflow and is looking for reasonable steps to implement rather than trying to build it all. There is a WG with biweekly meetings to move these profiles forward in one year. At the joint IHE PaLM/DICOM26 meeting in Helsinki, it was decided to focus on image acquisition and to design a joint Connectathon around that.
  • Agree tp split the image manager and image archive into 2 separate actors for AP (example: Duke will use a radiology archive, but have different image manager for AP); the image archive can hold DICOM and non-DICOM images
  • Should we include transactions between those 2 actors? It is the same as the image archive manager in slide #2 in the ppt; decided to keep those boxes together – Raj was thinking of the acquisition manager
  • Image display is missing from the transaction diagram on slide 3 – because it was not covered in top priority
  • Rename “Procedure request” to “work order request” (to be in-line with PaLM TF vocabulary) – this comes from the AP Framework:
  • Add image from AP TF – Figure 1.13
  • Procedure request can be split into multiple workflow steps, so keep procedure request
  • Do we keep image display in scope – will be helpful for Connectathon success as well as imperative for the work, so it was decided to keep it in scope
  • Need to add 1 more actor – with 2 transactions QIDO-RS (Query Images for DICOM Object) or RAD-14? = WADO-RS (Web Access to DICOM Object = RAD-55) RS = restful – name of these transactions query for digital asset / retrieve digital asset
  • The Image manager is better split per Mikael, so need transactions between them
  • Acquisition modality is also a part of the image manager – this needs to be reconciled.
  • Image manager to image archive C-MOVE, C-FIND, C-GET, C-STORE (is part of radiology workflows)
  • In future supplements we can describe interactions between the image manager and the image archive – for this profile we define those actors as grouped; even for the use case of exchanging images between 2 image archives, both those systems have both actors

• • For the Connectathon we would have to have vendors for all these actors – we have interactions between PACs viewers from the DICOM Connectathon, which covers the image display, so we can keep that in
  • Acquisition Modality has the biggest burden in the Connectathon – there are three actors to connect
  • Agreed that both broadcast and query mode are needed.
  • Check if we need to use RAD-14 (is not using QUIDO-RS; the elements that are the same in both transactions) or QIDO-RS with DICOM WG27
  • RAD-16 was not ideal for WSI, so we wanted to adjust that transaction – this is outdated, as we want to use the web access to the digital image to support retrieving of specific parts of the WSI
  • Need to reverse the arrow on RAD-55 (arrow comes from where the transaction starts)
  • Add final image = SLIDE 2 FROM ppt
  • In the other slides rename LAB-2 to LAB-3, (but that is out of scope for this profile (no LAB-1, LAB-3, LAB-35, LAB-36)?)
  • Do RAD-8, RAD-10, RAD-49 and RAD-11 fit the DICOM supplements for WSI (144) and specimen (122)? Need to check.
  • Also need to keep in mind the ad hoc use case of scanning a slide without a procedure request- what if the technologists loads the same wrong slide to the acquisition modality – would that be caught here?
  • For the Actor table:
    • Need to keep order filler
    • LAB-x = Acquisition Procedure Request
    • LAB-y = Acquisition Procedure Status Change (slide arrived as option/ slide scan complete = must support that – includes the IDs that help you retrieve the image, could also include a link to the location of the image in some systems)
  • Assumption for this profile is that ALL slides have a unique ID
  • What is the difference between the procedure request and the IWOS?
    • Request to perform imaging procedure = what magnification, what kind of images to make from each slide – 1 request per slide – do more than one workflow off that physical slide resulting in multiple digital assets – so each one of those are IWOS
  • Consider the digital slide a derived specimen
  • For gross images, we need to figure out how to have unique IDs – can be pictures of the same gross sample. When we get a vendor that supports gross imaging we should include this use case – it is a different acquisition modality type = there may be different data elements for gross / macroscopic imaging management that for slide scanners
    • We don’t have a participant yet. We have reached out, but so far no firm commitments from vendors. Possibly Spot Imaging is interested.
    • The software does generate unique IDs for each image that are derived from the case number. They allow annotations on the images, currently storing as jpeg; in the future we may need to add a DICOM wrapper. DICOM already has a definition for metadata that they could use; also, that would require the unique ID for the DICOM object, so this is a benefit for vendors that don’t otherwise do that yet.
  • LAB-U vs LAB-28 = Imaging Work order Step (IWOS) = MUST have a unique physical asset ID the IWOS must contain additional information about the IWOS.
  • Option to have acquisition modality report the status of their workload – keep for future workload
  • LAB-V vs LAB-27 = ad hoc slide in the scanner
  • For the communications between the acquisition modality and the acquisition modality manager we have either just LAB-U/LAB-28 or you may have LAB-V/LAB-27 AND LAB-U/LAB-28
  • LAB-W vs LAB-29 = scanner can provide as part of the status change a thumbnail
  • Image available message – who is the receiver of this information? The Acquisition manager or order filler?
  • When is the “steps performed” message sent that the scan is complete?
  • Do we need to send the scan complete message - before we know where the image has been stored?
  • The Image available must go to acquisition manager (DICOM) and from there the order filler can get the procedure request complete (which is HL7 and can include thumbnail / link to the stored location / all information needed to retrieve the image)
  • Does this image available message have to be DICOM? It would be the only DICOM format that the acquisition manager has to receive – we need to find out what data elements we need for the availability notification to determine or if we can use HL7 message (preferred) or if we must use DICOM.
  • All transactions are mandatory
  • Need to update the names to match he transaction table
  • For the Actor descriptions, add the AP viewpoint, so we can later update the Actor descriptions in the Appendix.
  • Is the staining done before the order filler initiates the acquisition procedure request? That is what is happening now; what about manipulations of images (unstained slides can be made to look like stained)
  • This profile deals ONLY with the scanner process – not the staining workflow
  • In the lab workflow now, the staining occurs for every tissue that is produced – all without human input. Bu for some others there needs a user input by identifying which slide (deepness of slide) needs to be stained with what stain
  • For some stains you need to have a specific scanner; we need to clearly list the exclusion of creation of the glass slide as out of scope – also include a statement that a slide will need unique ID
  • Can we reuse the ?? transactions to cover some of the transactions mentioned – RAD-49, also some procedure update messages – Francesca will check
  • In the Image archive, for sending annotations, there needs to be a description for the actor role, even though the creation of annotation is NOT part of this profile. DICOM covers that part; manipulation of original image is captured in DICOM today – we should forward that as well
  • For Actor grouping, in this profile it is required for image manager and image archive because these actors are usually tightly coupled as part of a common system
  • The use cases supported are whole slide image (WSI) AND gross imaging – or just WSI? The pathologists want to have grossing included, but we need vendor participation (Gunter mentioned he has a contact with a small German company that converts image into DICOM images).
  • List as an Open issue that we have removed the gross image UNTIL we can have input from vendors
  • Security Section:
    • CT should be required
    • ATNA (Audit Trail and Node Authentication)– doing the audit trail is hard in this profile; to do this will require the bar to participation is raised for a Connectathon. However, it should be covered in the end product. ATNA requires a central audit trail repository – the LIS currently don’t do that. They don’t have to have the external repository – overall, the system needs something similar. It should be made a desired, but not a required element = SHOULD
  • How are we going to handle Patient Admin? The actor that is closest to the patient shall be grouped with PAM actors; this applies to the order filler. How do we deal with patient reconciliation (when demographic data is updated)? Do we need to add the image archive? Are you allowed to update an authorized and stored image? For example, the change of name due to marriage is handled by linking. What about the scenario, where the case is created, in an emergency of an unconscious patient – how would that be updated? If we need to, it can be checked using the patient reconciliation profile (PIR) from radiology.
  • Completed Vol 1

Vol2:

• We will create 1 section for each of the previously described transactions
• LAW is for analytical workflow – does this suit our need? We can copy from LAW and remove a lot of the guidance in LAW that is related to micro etc., then we will renumber it. We will change LAB-27 – LAB-V, LAB-28 to LAB-U, LAB-29 to LAB-W and AWOS to IWOS
• In the message structure we will use OML^O33 since we are limiting to 1 IWOS for these messages
• The ORDER comes in from the placer to the Filler, from Filler on. We use Acquisition Procedure Request and Image Acquisition Workflow.
• Inside the LIS tasks are assigned on a particular physical asset that goes to that lab worklist that is being executed
• Order placer to order filler - 1 order on multiple specimen = OML^O21
• LAB-X = For the procedure request: 1 slide with multiple steps done on it = OML^O33
• LAB-U = IWOS = 1 slide and 1 step = OML^O33
• Can the glass slide contain only 1 specimen? 1:1 relationship to a block from 1 patient, or it may include a control tissue (but in the context of the image that would still be 1 image)
• O33 = specimen centric that has information about the container, the order is on the specimen; it is less likely to accidently break the 1:1 between the specimen and the container
• DICOM Supplement 122 are container-centric, so we should use OML^O35. The order is on the container, but the container is still underneath the specimen (one or more)??
• We will EXCLUDE more than 1 specimen from different patients
• Should allow more than 1 specimen from the same patient on a slide that will be imaged?
• Allowing double stains on the same specimen is common when they are 2 AB with different colors, but if applying 2 different stains (HE and PAS) on the same specimen, would those be considered two specimens?
• The HL7 message currently does not have a message structure that has orders on container or that tracks the specimen under the same container. We would need to do this.
• The data model is in DICOM Annex NN
  • NN4.2 is the only model supported with this profile
  • Do not support NN4.3 (different patient samples on the same slide)
  • NN4.5 is more often for TMA – excluded right now
• We are not currently assigning specimen IDs to pieces on a slide, so we consider the slide the specimen. It is not clear if we could describe the different stains on one slide
• Use O33 – allow multiple orders on Procedure request, single order on IWOS
• Procedure request status change:
  • Should handle thumb nail or pointer in OBX – OUL^R22
  • LAB-U = QBP^Q11, limit parameter to slide ID = SAC-3
  • LAB-V = RSP^K11
  • LAB-W = OUL^R22
  • Also added in the respective ACKS for LAB-X and LAB-Y
• LAB-X:
  • How do we get parameters for scanning across in the order message?
  • We reviewed the TCD segment (only 1 and currently tailored to Clin Lab automation)
  • Option 1: create a list of orders that create combination the parameters – codes in OBR-4
  • Option 2: create generic order code and use OBX under OBR to send the parameters for the specific order to be executed. Some of the attributes may have other spots in the HL7 message, like specimen type = SPM-4, staining = Additive (SPM-6)
  • Send request to vendors to collect ALL applicable configuration attributes (configurable parameters)
  • Then we will decide where in the HL7 message those will be represented and with what codes. Does DICOM already have a value set for this?

The meeting adjourned for the day at 5:00PM local time.

November 14, 2018

Francois reviewed the agenda for the day.

Specimen Event Tracking (SET) Vol 2Review Powerpoint Available Here

Alessandro reviewed the change to SET based on the Monday meeting.

Review message structures

• Specimen movement tracking – package centric
• Also add 2 PRT segments after the EVN segment to cover from location and to location
• Remove PRT below PAC and Specimen
• Why would the specimen come before the container?
  • Can have 1:1 Specimen:SAC, or you can have 1 :many (example 24 hour urine in more than one bottle)
  • Can have pooled samples, but that would still be a single specimen
• PAC segment is used for all the boxes within the box it is shipped in (example: box of 24 containers that hole DNA samples)
• After arrival the shipment is examined and a decision made to accept or reject. Should this be covered with the same message structure?
• An example: if a shipment is received with 4 packages and only 2 of the specimen in the entire shipment of 10 are not usable, do not reject the entire shipment then. Instead send this at the SPM level
• We need to be able to indicate that every container has been received and contained a specimen
• If we reject only some of the specimen in a shipment, use SPM-21 to indicate that. You also have SPM-24 specimen condition to make a statement about the specimen
• Can we reject 1 container of 2 containers that contain the specimen? Use SAC-8 container status - missing or unavailable.
• When a container has no specimen, use SPM-21 to indicate no specimen in container. But, H70490 does not have missing sample (SNR code from HL70493 was deprecated in 2.8; this should be added to 490 in next harmonization cycle)
• Specimen de-identified / re-identified use case
  • Replacing specimen IDs
  • What about the parent specimenID and parent container IDs – would these also need to be removed / reassigned?
  • A de-identified sample would be assigned a different ID
• No need for the PRT after the SAC; just use the PRT after EVN
• Specimen archived, and specimen retrieved
• SPM-12 = Specimen Collection amount and SPM-25 – Specimen Current Quantity
• Could you use these fields for AP since containers are relatively standardized? This should work. If it doesn’t then we will submit a CR when we have the details
• Specimen Processing Tracking:
  • How do you state what processing procedure was performed? We need something that represents an act. Maybe a completed OBR? Not really a good solution though.
  • If we use OBX, then the OBX-3 would be the question “Type of processing” and OBX-5 would be a code representing the processing type that was done. Maybe we should be using SCT from the procedure hierarchy. We will also check DICOM for a set of processing activities (or BRIDG may have a value set, too)

Next Steps for SET, LSH and Digital Pathology

Francois and Riki reviewed the next steps for SET and LSH

1. SET will be updated and reviewed on the next conference call
   • Write up the HL7 CR for the new message structures and submit to OO for review – including the list of the needed triggers to cover SET events (MSH-9.2)
2. For LSH, create a brief summary of LSH and link it to the wiki page
   • Diagram is missing communication arrow between automation manager and the transportation manager. How quickly can this communication happen to reduce the turnaround time? (Roche is working on that interface -workload balancing on the fly)
   • What is the difference between a pre-post processor manager and the transportation manager?
   • Communication between the actors inside the automation manager is out of scope for the LSH. (may be a new profile in the future)
   • The example of an implementation diagram would be useful to add into the supplement template (at explanatory bottom sections)
   • There are cases where the processing or analyzer does not read the barcode (so no sample ID) how would the transporter fulfill that requirement – needs example and consider how to figure that out
   • For the Connectathon, we will have to figure out how to deal with the elements that are physical since these are control interfaces. We should consider building simulators for testing this part (currently have some that simulate instrument performance in the lab – different scenarios can be “played). Mario will check with his simulator team at Roche.
   • We could potentially set up an on-site test with remote connectivity to Gazelle to check the conformance of the messages exchanged between the participants

   - This is similar to the LCC where the replacement suggestion times out. This is not at the socket level; the transport manager would know. If a time out was defined the turn-around time would know if something was wrong

   • The system sends empty packages to verify that the connection is still present.
   • Is there precedence to detect loss of connection in other IHE profiles?
   • Specimen track MUST stop if the specimen acquisition complete message was not received. We still would need to define the timeout for that or set up a time window by which the instruments is queried for its status to test both connections (in and out). That would be the preferred route. We should consider the time needed for testing to set query intervals and combine that with unsolicited OOS status messages
   • We could require use of MLLP/TCPIP protocol for connection status (in IHE we make the assumption that we are using internet – not serial connections)
   • Why not use the status query as a requirement – or ACKs and NACKs. If ACK is not received, then the systems know something is wrong
   • If the Analyzer manager sends AWOS and no ACK/NACK received, then the automation manager knows something is wrong
   • Device is communication with analyzer manager, so analyzer manager is letting the transport manager know that there are issues with the analyzer manager
   • The consensus is to use the status request queries and allow unsolicited status reports to be included.
3. Digital Pathology white paper/profile
   • For the digital pathology white paper/profile, after counting the number of transactions, we decided to reduce the scope of the profile and focus on the IWOS.
   • We will create different profiles for LAB-X and LAB-Y (workorders)
   • There will not be an actor for the Connectathon for this profile, but we will have 2 profiles for testing.
   • There was a question of taking out the image display out of this profile, too.
   • In the white paper we will explain how we are splitting the workflow into several profiles and planning to bring them to the Connectathon by ? (need date)

Wrap Up (Francois)

Riki and Francois gave the wrap up.

1. SET:
   • Review HL7 proposal on the December call (2nd hour)
   • Review Vol 2 on call (2nd hour) in February 2019
2. The revised LSH draft will be shared before the December conference call.
3. The Board report is being updated. The report will be shared on PaLM Google Groups before the DCC December call.
4. Digital Pathology / 1st profile Image Acquisition
   • We will set up a vendor survey to flesh out the parameters for the imaging work order steps (survey monkey)
   • Raj will share the updated digital pathology profile before the December call
   • We will review this in 1st hour and try to finalize the open issues for the vendors
5. Riki will update the LCC and send it to the Google Groups before the December call.
   • It will then be sent to Mary Jungers for comment publication
6. TMA: Dan will finalize the TMA supplement for Trial Implementation and send it to Mary Jungers by November 24 Action Dan: keep Filip informed too.
7. For the next calls:
   • Dec 12th: Wrap up decisions review, Dig Path white paper and smaller updated profile1st hour, SET 2nd hour
   • Jan 9th : Dig Path 1st hour, LSH 2nd hour
   • Feb 13th : Dig Path 1st hour, SET 2nd hour

The meeting adjourned for the day at 12:00PM local time.