IPM Interface Project

Update on October 29th:

Cerner quote for 3 database connections (enabling 3 more users to connect) and 3 PICS Plus Licenses (enabling IPM Report PDF upload into CoPath) received.

Possible need for 2 more interfaces identified:

1. Billing interface for technical codes of IPM testing out of CoPath to Millennium (?) [It is possible to add a second outbound billing interface to CoPath in addition to the currenttly functional Professional one; see page 105 of attached "(Billing)-30TRGEXT022267 Cerner CoPathPlus 2014.01 Interface Specifications.PDF" below.)

2. Order+Results interface from Millennium to Standard Molecular (treating Standard Molecular's solution as an "outside laboratory") or from Standard Molecular's solution to Millennium (unifying all IPM testing within Standard Molecular's solution) discussed. Concensus appears to have been reached to define this interface in "Stage 4" of implementation (after Stage 1: CoPath Procedure Order Interface, Stage 2: "In Lab" Order Message out of CoPath, Stage 3: Epic to CoPath order interface) during the October 26th meeting.

Update from Val October 26th:

Please check below my answers for ##1, 2, 3, 4, 7.

Messages examples (##5,6) I hope to get by the end of this week.

1. Can the array of variables vary from message to message (so that CoPath just treats it as rows of Column A (variable names) and column B (variable values) or will IPM have to send the exact same number of variables ("rows") with exactly the same variable names every time in every separate result (with some values populated in some results)?

No fixed record template is needed.  The number of fields received can vary per message.  Source system have to supply the specific results and optional values that are needed for each test. The inbound message can have any number of discrete results segments.  Each segment is an entry in the results dictionary.  With an optional value or comment supplied.  Any number, no limit.

2. If a single detected mutation has a set of multiple associated discrete variables (and a single result message can contain a report with multiple detected mutations, how would you prefer to receive it in the HL7 message?

The additional value is received and stored separately with each test.  Alpha or numeric values can be used.

3. Is there a limit to the number of variables in the array above that CoPath could accept as a procedure result in order to file the message to the database, display the discrete variable names and values, and send them via the outbound message?

Unlimited array. No limit to the number of discrete results.  All are stored in the results table for the procedure and available to be sent out via an outbound interface.

4. Is there a limit to the possible number of values in a value set associated with a single variable name ("Gene name: ____") that CoPath can accept (or is the expectation to accept the values as they are if the value set is too large?

Only limit is value has to be 255 characters or less.  All values stored as string characters.

7. While IPM Requisition Forms are available for review, there is a need to transmit answers to order questions along with an outbound procedure order from CoPath. These values may be contained in discrete fields associated with a Part in CoPath (manually entered) that are enclosed in the associated HL7 Procedure order message, or asked as "Order Questions" that are asked by CoPath at the time of placing the Procedure order in CoPath. The specific questions to be answered (and answers to be enclosed with the Procedure Order) are:

Diagnosis [Possibly contents of the CoPath Final Diagnosis Field]

Percent Tumor: [percent from 1% to 100%]

Neoplastic Content: [free text]

Sending additional data enclosed with a procedure order will be possible via a single text or "comment" field.

Minutes for the Call with Val on Friday September 25th at 2pm: 

Present: Suman, Victor, Val

Val Provided the overview document (See attachment named "CoPath-Procedure discrete results.docx" below on this page) of the CoPath 2014 discrete data capabilities that involves appropriate configuration of the following CoPath dictionaries:

 A. Discrete results dictionary: lists items with each item having a: Name, descr, abbreviation, abnormal/normal and interface code

 B. Quick text dictionary: enables linking of value to text

 C. Result normality dictionary: defining abnormal flags for the interfaces

 D. Procedure dictionary: select appropriate filed: procedure results or procedure intereptation

Val has received the draft list of discrete variables in the "Updated-by-VB-Report information for exome test_v1_20150312.docx" file attached at the bottom of this page.

Val is still working on obtaining the following information from the Cerner Team in order to generate the updated Procedure Interface quote:

1. Can the array of variables vary from message to message (so that CoPath just treats it as rows of Column A (variable names) and column B (variable values) or will IPM have to send the exact same number of variables ("rows") with exactly the same variable names every time in every separate result (with some values populated in some results)?

2. If a single detected mutation has a set of multiple associated discrete variables (and a single result message can contain a report with multiple detected mutations, how would you prefer to receive it in the HL7 message?

3. Is there a limit to the number of variables in the array above that CoPath could accept as a procedure result in order to file the message to the database, display the discrete variable names and values, and send them via the outbound message?

4. Is there a limit to the possible number of values in a value set associated with a single variable name ("Gene name: ____") that CoPath can accept (or is the expectation to accept the values as they are if the value set is too large?

5. Please send us example HL7 messages containing discrete data that CoPath would expect to receive as inbound procedure results.

6. Please send us example HL7 messages containing discrete data that CoPath would expect to send as outbound specimen/accession-level [procedure] results.

7. While IPM Requisition Forms are available for review ( IPM Requisition Form.pdf and Precision Medicine Patient Requisition[3][10].pdf attached to this page at the bottom ) there is a need to transmit answers to order questions along with an outbound procedure order from CoPath. These values may be contained in discrete fields associated with a Part in CoPath (manually entered) that are enclosed in the associated HL7 Procedure order message, or asked as "Order Questions" that are asked by CoPath at the time of placing the Procedure order in CoPath. The specific questions to be answered (and answers to be enclosed with the Procedure Order) are:

Diagnosis [Possibly contents of the CoPath Final Diagnosis Field]

Percent Tumor: [percent from 1% to 100%]

Neoplastic Content: [free text]

Val agreed to review the aforementioned documents including the sample Genomic reports ("PM347_Z2_1_Case_HALO_Report_20150309.pdf" and "PM332_Z3_1_Case_HALO_Report_20150218.pdf" attached at the bottom of this page), the list of discrete variable names using in construction of these reports (listed in "Updated-by-VB-Report information for exome test_v1_20150312.docx" attached at the bottom of this page), and the three order quotesions above and to let us know if he has any questions.

If there are no questions, we will await the answers to the questions for Val above as well as the updated quote for the Procedure Order interface.

Val also stated that he is not aware of any methods to make a path to a file clickable from within CoPath to enable opening of PDFs "linked" from a given accession number. Outside of this capability, Val will look into the feasibility of the "mini-project" to add a path to a PDF report resting on a network share to a given accession number in CoPath and to output that path to a PDF discretely in the outbound result message, enabling the receiving system to receive that link and make it clickable to the end use in that system.

Victor 09/25/2015

"IPM/Precision Medicine Interface call" Minutes for Thursday, September 17th at 10am:

We were able to touch on all agenda items below.

If you have any questions or corrections to the meeting summary, please send them to me to avoid long email threads. At the same time, don't hesitate to reply to all if you have an announcement.

Meeting summary:

1. Val from Cerner (CCed) deemed the target state of the interfaces described by the diagrams feasible, but Val stated that he needs to modify Cerner's  "CoPath outbound Procedure Order" interface quote to include the work necessary to "pass-through" the Epic or Eclipsys EMR order number associated with an accession number as a discrete value in the corresponding outbound HL7 procedure orders.

To Do: Val to adjust the quote and Cerner to issue an updated quote document (updating one of the two quotes issued by Cerner for this project). Thus only one of the two quotes we have on hand (Epic->CoPath orders) is currently actionable.

2. Val from Cerner (CCed) needs to investigate whether it is at all possible to use the "discrete data" features available in CoPath version 2014 to receive, file, display, and "pass-through" genetic discrete data formatted as a non-complex array of key/value pairs sent in the body of the HL7 message by IPM/Precision medicine in full compliance with the HL7 standard (To help visualize the key value pairs: {"A":"5"; "B":"blue"; "See":"This Note"} etc.)

To Do: Val to discuss with Victor the discrete data capabilities of the latest CoPath version 2014 and finalize the answer regarding what is possible (if indeed possible, what is the maximum number of key/value pairs, do all "keys" have to be added to a CoPath dictionary for the message to file, do all keys have to be always sent in the same order and quantity, etc).

3. Val from Cerner (CCed) introduced another capability of CoPath: to receive digital results for a procedure different from the one that was ordered, thus enabling IPM/Precision Medicine (or Millennium) to perform a different laboratory test from the one ordered (if necessary) and still send result to CoPath digitally while including the proper order identifiers. Let me know if this capability needs to be incorporated into the diagrams. This detail may be important for you David and for Standard Molecular to make sure this is implemented correctly (please let us know if you need additional details from Val).

4. Aside from the items above, Val from Cerner (CCed) stated that the two quotes from Cerner indeed do cover the work necessary to complete the project (form Cerner's perspective). Thus the Procedure Order quote includes the receipt of the procedure results digitally.

5. Val from Cerner (CCed) stated that to enable CoPath to accept the link (to an IPM PDF report) into a discrete field associated with a specific accession number, and for CoPath to output this link via the outbound report interface discretely as an interim solution would require a separate quote from Cerner and this work would be handled as a separate project. The alternative in the short term is to import the IPM/Precision Medicine PDF reports into CoPath (as opposed to linking them) since this feature is already available and requires only the purchase of the PICSPlus image import module for every PC where import is to occur.

6. After the meeting, Rob Kim asked the question: "What will happen when the lab QC's the tissue and determines that there is not enough DNA or the quality is too poor to run sequencing assays?  In the current (non-clinical) IPM workflow, feedback is given to the Pathologist, who will identify another specimen or determine there is no other material available.  Will the already complex diagram need to have a version for troubleshooting scenario’s?" It appears to me that this scenario does not impact the envisioned interfaces or accessioning steps shown in the diagrams, but please do let me know if you feel otherwise.

Additionally, after the meeting, Richard Davis ( the Director of NYH Corporate Systems Project Management Office ) has now received the diagrams and had the following comments:

A. "eGate" middleware has been replaced by "Viaduct", but the functions remain the same.

B. "Corepoint" middleware necessity: to be determined by POIS

C. "Please be advised that project priority scheduling and resource forecasts are indicating that all interface work required for the Precision Medicine project presently under discussion will have to be outsourced.  Please make note of this when you are budgeting and scheduling for the project."

To Do: Richard Davis to provide actionable quotes for the work proposed to be outsourced.

7. Rob Kim is working on initiating the transfer of funds from WCMC to NYH for the Epic->CoPath order interface quote.

To Do: Harlan (CCed) to issue a PO for the Epic->CoPath order interface quote prior to the quote's expiration date.

8. Michael - we should meet with Bulent to familiarize him with the project, get his input, and make sure he is prepared for the interface work to be done in Millennium.

To Do: Michael Kluk, Victor, and Bulent to meet to discuss the project as it relates to the Molecular (non-IPM) results.

9. Val stated that the PDF report sent to CoPath would need to be embedded/MIME-encoded into the HL7 message itself (as opposed to linked to the message via a path/pointer.) This was communicated to David Artz and the Standard Molecular team.

3/13/2015 - Proposed workflow summary:

1. HL7 order received by CoPath from Epic EMR with a patient identifier, an Epic EMR order number, a referring provider, and the answers to the order questions.

2. Pathologist opens order in CoPath (from a worklist? from a filtered list containing only IPM orders?), and is presented with a list of existing CoPath accession numbers (with tissue tracker column(s)) for this patient, sorted by accession date.

3. Pathologist manually selects an accession number and part(s) upon manually verifying there is enough tissue.

4. CoPath creates a unique PM accession number

5. CoPath issues a reflex order to the IPM system with: the newly generated PM accession number, the "old" accession number + part(s) selected by the pathologist, their barcode strings, the Epic EMR order number, the referring provider, and the order questions and answers received from Epic EMR.

6. "Germ line reference" blood sample tube arrives in the AP laboratory, is accessioned as a part under the same PM number generated in step 4 above and is labeled with the barcoded label produced by CoPath containing the PM number. It is then handed off to the IPM team.

7. The IPM system issues the HL7 message + PDF of the report. The HL7 message of the report contains the PM accession number from CoPath, the "old" CoPath accession number + part(s) selected by the pathologist, the Epic EMR order number (everything COPath sent to the IPM system in step 5 above), and the referring provider.

8. CoPath receives and files the report, associating it with the matching PM accession number.

9. Pathologist signs out the received IPM report in CoPath. If there is an issue preventing signout that is identified with the report content received from IPM, the pathologist will inform the IPM team by phone or in person, and the IPM team would issue a revised report to CoPath which would be signed out in CoPath the Pathologist. The revised report received by CoPath would need to be the one presented by CoPath for signout, overriding the previously received report.

10. CoPath sends an outbound HL7 message to the Epic EMR containing the signed out report, the unchanged PDF received from IPM, the original Epic EMR order number, the PM accession number, the "old" selected accession number, as well as the referring provider.

If the new specimen is collected and being sent to the laboratory (because it is known ahead of time that no existing specimens are suitable or none are available), we have two choices:

i. Always accession such a specimen first as a surgical specimen with an S prefix into CoPath and then proceed with the exact workflow described above.

ii. Accession the newly received specimen into CoPath with a PM prefix directly. To make this possible CoPath would need to simply allow the pathologist in step 3 above to select *no* previous old S-prefixed specimen. The workflow would otherwise remain the same and that field in the messages would be empty.

Rob, Olivier, Andrea, (and Adam) - is the above workflow viable for your teams? Please add your edits or comments as you see fit.

Linda -

A. Is the information above sufficient to include the described workflow into the first quote? This functionality is at the core of this work request, so we do need it as part of the first quote and can not exclude it. Let me know if you need additional information.

B. Please send us the example of the "discrete procedure interpretation" that your development team sends you.

C. What specific variables should the barcode on the "Germ line reference" blood sample tube specimen contain to streamline accessioning into CoPath once the digital order interface from Epic EMR to CoPath is in place? Can you please provide a sample barcode string?

D. Is there a way to generate a worklist for the IPM pathologist containing incoming IPM orders (what is described in step 2 above)? Or would it be possible for the IPM pathologist to simply filter all incoming orders to just IPM orders?

E. During the call, I heard the mentions of the "tissue tracker" functionality. Is this the CoPath functionality we already have? Can you please send me a document that describes this functionality in detail?

Cerner's answers to questions from yesterday's meeting listed below:

Linda: "I’ve attached a subset of the CoPathPlus 2014 release notes that clarifies the ability to enter discrete coded results on procedures, as well as receiving such data over an interface. The structure for the procedure discrete results are a simple name/value pair and are defined in a dictionary.

The CoPathPlus Interface specification for the 2014 release posted on the CoPathPlus uCern page gives an example on page 84 as to what the coded procedure interpretations would be for results sent out. For coded procedure interpretation a new OBX segment of CE can be used...  [The document is now attached to the bottom of this page; see page 84 of CoPathPlus v2014_01_CoPathPlus_Interface_Specifications.pdf ]

The other document defines the bi-directional interface that allows outbound procedure orders and inbound procedure results including text and images."

3/12/2015 - Draft Minutes of Call #2 (please email vib9020@med.cornell.edu with your corrections): 

Thursday, March 12th, 2015 at 1pm Eastern Time

Call in Number: 1-866-921-6338

Call in Code: 01748907

1. The immediately implementable workflow for both having the IPM reports in CoPath and for sending them via HL7 to the Epic Ambulatory EMR would involve:

            a. Creating PM-prefixed cases (Example: PM14-12345) in CoPath and importing the IPM PDF reports into these CoPath cases via PicsPlus. These would then be issued by CoPath upon being signed out via the existing PDF interface to the Epic Ambulatory EMR.

            b. Importing IPM reports into CoPath via PicsPlus and adding them as addendums for existing Surgical Pathology (S-prefixed) reports if existing previously accessioned "old" specimens are involved. The addendums to the old reports would then be issued by CoPath to the Epic Ambulatory EMR via the existing HL7 PDF interface.

                        i. Dr. Chen - do you think this should be done, or should a new PM case always be created in CoPath for every ordered PM report? Adam, Olivier - do you have any preferences? It seems logical to always create a new PM accession number in CoPath for every IPM order, whether it is actually an "old" surgical specimen or a newly accessioned true IPM specimen, but this would mean that old assets (blocks, slides) would have S-prefixed barcoded labels that are not recognized by CoPath as being part of the PM case.

                        ii. Suman - will you be able to add PICSPlus-imported IPM PDF reports as an addendum to the body of any old signed out S-prefixed case report and issue it via the outbound interface? Are there any report template adjustments that are necessary?

                        Answer: "Yes, we can add PicsPlus imported PDF reports as an addendum to signed out S cases.  Template modification is not necessary." - Suman Telesphore

2. Cerner representatives have stated that the announced structured data import and export capabilities of CoPath 2014 will be very limited.

3. Cerner representatives have stated that Cerner CoPath 2014 will not be able to readily receive an HL7 message formatted according to the linked version 2 of the Clinical Genomics HL7 standard.

4. Cerner representatives have stated that CoPath does not have the capability to associate new orders with previously signed out specimens and is thus not currently able to issue addendums with new order numbers. Cerner representatives agreed to incorporate this functionality into the first quote listed below.

5. Cerner representatives have stated that CoPath can issue orders (in this case to the IPM system) and receive and file associated results. In our case a reflex order would be issued by CoPath to the IPM system upon receipt of IPM orders from Epic Ambulatory EMR. This functionality will be incorporated into the first quote listed below.

6. We will disregard the quote already issued and sent to us by Cerner in response to this project request.

7. Cerner representatives have agreed to supply the following documents:

            A. The exact specifications for the "PDF" results  interface for genomic data that has been implemented at one of the client sites with example HL7 messages to illustrate to the IPM team the message format that CoPath 2014 can readily receive and consume.

            B. The exact specifications for the announced CoPath "structured data" import AND export functionality along with example "structured data" inbound AND outbound HL7 messages that CoPath version 2014 is ready to receive, file, display, and issue.

            C. A definitive quote and the associated expected timeline (amount of time for completion) for implementing the workflow that is described in detail here (excluding the referenced Clinical Genomics HL7 standard) but including all other described aspects of orders and results (which would have both HL7 "plain text" reports and an associated PDF version of the same report coming to CoPath from the IPM system and sent by CoPath to Epic Ambulatory EMR upon being signed out):

http://wcmc.pathologysystems.org/ipm-interface-project

            D. An estimated quote and estimated associated timeline (amount of time to completion) for implementing the workflow described by the document linked above that would include receiving, displaying, and filing in CoPath messages as defined by the referenced version 2 of the Clinical Genomics HL7 standard ( http://www.hl7.org/implement/standards/product_brief.cfm?product_id=23 ) from the IPM system. The data elements received via messages formatted using this standard would then be used by CoPath to construct the HL7 message from CoPath to the Epic Ambulatory EMR and include the unmodified PDF received from IPM.

8. Cerner representatives stated that CoPath should not be used to store or collect consents.

9. Cerner representatives mentioned "tissue tracker" as a way to track whether any remaining tissue is available and how much of it is available. This does not solve a problem of automatically matching an order for sequencing that does not specify an accession number of a previously signed out specimen.

3/12/2015 - Agenda and Current (as well as proposed) Workflow along with relevant questions for the Cerner CoPath team:

We are asking for a quote to implement the following scenario:

The assumption is that all orders for sequencing will originate from the Institute for Precision Medicine.  This means that all patients will already be registered into Epic. Three systems are involved: Copath, Epic Ambulatory EMR, and "IPM software" used to analyze the genetic data and output a report. The notes in square brackets [] concern CoPath specifically.

1.     Patient is referred to Institute for Precision Medicine

            a.     Provider will evaluate patient and make a determination if sequencing is appropriate.  If so, the patient will be consented by IPM staff. [What would be a way to store consent in CoPath/should this be done at all in your opinion, Hollie?] Cerner's answer 3/12/2015: CoPath should not be used to store or collect consents.

            b.     Provider will determine if the patient has tissue already collected that could be used for sequencing, if not then new tissue will need to be obtained. [How can CoPath help in this regard? Is there a preferred way to keep track of remaining tissue available for sequencing?] Cerner's answer 3/12/2015: Cerner representatives mentioned "tissue tracker" as a way to track whether any remaining tissue is available and how much of it is available. This does not solve a problem of automatically matching an order for sequencing that does not specify an accession number of a previously signed out specimen.

2.     For tissue that is already collected:

            a.     An order will be placed into Epic that will include order questions to collect enough information to identify the requested sample (Patient ID, surgical date)

            b.     The order will be sent via an orders interface to CoPath. [We need to incorporate the orders interface from Epic Ambulatory EMR to CoPath as a line item into this quote; this order received in CoPath can be treated as the "request for sequencing" for the IPM pathologist described in 2f below - is this the best way to get this done in CoPath?]

            c.      A blood sample will be collected for germ line reference. 

            d.     The blood sample will be sent to Anatomic Pathology with a label from Epic (that will contain more information than the generic label) [What specifically should be in the barcode on this specimen to streamline accessioning if the order interface is already in place?]

            e.     The blood sample will be accessioned by the AP staff into CoPath and then will be directed to the IPM team for sequencing ("germ line reference") in addition to the previously collected tissue with an older CoPath accession number. [We need a way to associate/link the previous accession number for the tissue being sequenced as well as the newer blood sample accession number in a bi-directional user-friendly way since the two CoPath accession numbers refer to parts of a single specimen as far as IPM is concerned]

            f.      The CoPath team will put the request for the sequencing in the work list of the IPM pathologist [Is the order accompanying the "germ line reference" blood sample the best way to achieve this worklist in CoPath?]

            g.     The IPM pathologist will locate the tissue sample and will determine if there is enough tissue

                        i.     If there is NOT enough tissue, the pathologist will create an addendum to the previously filed CoPath result report directly in CoPath and add the requesting provider in the CC field [CoPath needs to add the requesting provider and the order number from the newer order associated with the blood sample to the addendum of the older report].  This will allow Epic to notify the requesting provider. Since we already have the PDF interface, the PDF will be generated by CoPath and link to it transmitted as for other reports.

                        ii.     If there is enough tissue, the tissue will be given to the IPM team for sequencing. [An HL7 order for sequencing will be transmitted from CoPath to the IPM software. Is this possible?] After the sequencing is done, the report will be generated by IPM software and sent to CoPath via the following HL7 standard http://www.hl7.org/implement/standards/product_brief.cfm?product_id=23 . [CoPath will receive the structured data, file it into the database (retaining its structure) and make the IPM report available/viewable/searchable within CoPath. The IPM report will then be combined with the previously finalized and issued AP interpretation report in CoPath [via the Addendum mechanism] and transmitted together from CoPath to Epic Ambulatory EMR as a regular "plain" HL7 lab result message (keeping the separators between discrete values within plain OBXes). The requesting provider and the order number for the non-blood specimen (or both order numbers/both referring providers?) will be transmitted to Epic Ambulatory EMR in this report message. Since we already have the PDF interface, the PDF will be generated by CoPath and link to it transmitted as for other reports.

                        The results will be added to the previously filed AP report as an addendum with the requesting provider added as a CC [CoPath needs to add the requesting provider and the order number from the newer order  associated with the blood sample to the addendum to the older report].  This will allow the results to be routed to requesting provider in Epic. Since we already have the PDF interface, the PDF will be generated by CoPath and link to it transmitted as for other reports.

3.     For tissue that needs to be collected

            a.     The IPM provider will put in the referral order to the appropriate location into Epic Ambulatory EMR (IR, surgical oncology, etc…)

            b.     A separate order will be placed in Epic for sequencing. There will be order questions that will make it clear that the sample is yet to be collected. [This order will be transmitted to CoPath to accompany the blood specimen sent to the lab as a "germ line reference"]

            c.      The patient’s blood will be drawn and this sample will be sent to AP for accessioning with a label from Epic (that will contain more information on it than the generic label).  [What specifically should be in the barcode on this specimen to streamline accessioning if the order interface is already in place?]

            d.     The blood will be sent to the AP lab and will be accessioned and directed to the IPM team for processing.

            e.     The IPM team will monitor the patient’s scheduling and will then have an NP, PA, or MD place another order into Epic Ambulatory EMR on the date that the new sample will be collected, indicating that the sample will be collected that day and that the IPM pathologist should be aware and ready to process the sample as needed for sequencing. [This second order will be transmitted to CoPath to accompany the collected tissue specimen sent to the lab to be accessioned into CoPath]

            f.      Once the sample is accession in the lab, the IPM pathologist (?) will divide the sample and matched sections will be sent to AP for interpretation and to IPM for sequencing.  [An HL7 order for sequencing will be transmitted from CoPath to the IPM software. Is this possible?]

            g.     After the sequencing is done, the report will be generated by IPM software and sent to CoPath via the following HL7 standard http://www.hl7.org/implement/standards/product_brief.cfm?product_id=23 . [CoPath will receive the structured data, file it into the database (retaining its structure) and make the IPM report available/viewable/searcheable within CoPath. The IPM report will then be combined with the AP interpretation of the sample portion [What is the best way of doing this in CoPath? Via the Addendum mechanism?] and transmitted together from CoPath to Epic Ambulatory EMR as a regular "plain" HL7 lab result message (keeping the separators between discrete values within plain OBXes). The requesting provider and the order number for the non-blood specimen (or both order numbers/both referring providers?) will be transmitted to Epic Ambulatory EMR in this report message. Since we already have the PDF interface, the PDF will be generated by CoPath and link to it transmitted as for other reports.

4. If a sequenced gene later proves to be clinically relevant, an unsolicited/unordered amended report will be generated and sent by IPM software to CoPath; CoPath would need to receive it and file it appropriately [What identifiers would the IPM report message need to have in order for it to be filed appropriately?]. The amended report would then need to be re-transmitted from CoPath to Epic to reflect this new observation in the same fashion as described in 3g above.

Appendix:

A. The order information (discrete data elements) currently being captured by IPM in their system upon receipt of the order are as follows:

(The information below gets captured in a custom developed sharepoint-based system to help track the process of the FFPE Pathology review.  If the archival material is acceptable for sequencing, we will schedule the patient for a clinic visit so that blood and consent can be collected.)

Name - Identifier:

Status: Review Blocks/Slides & Deliver to Lab Received in Lab

Date of Request: 2/9/2015

MRN:

Priority: High

Case: S15-

Review Action: Review Blocks/Slides

Report Link: S15-

Initial Review Acceptance: Joanna Cyrta

Initial Accepted Date: 2/9/2015

Attending Reviewed:

Tissue Sufficient: Yes

PMID: PM356

Pathology Notes: Tissue sufficient as per report

Final Review Acceptance:

Initial Review Acceptance (temp): Joanna Cyrta

B. The two order forms attached below are received by IPM. One is the common one, received from the 

clinician/oncologist when they are interested in referring their patient to IPM. 

At this point, if the form is filled out that the tissue is FFPE from past in-house procedures, 

an FFPE Pathology review process gets set in motion. 

The purpose of the FFPE Pathology review is to prevent the patient from having to come in multiple times 

and to avoid additional biopsy/surgery procedures that may not be necessary.  

If no in-house FFPE tissue is available or the in-house FFPE tissue is not sufficient for sequencing, 

IPM will coordinate biopsy/surgery for fresh tissue collection.

The other forms is used rarely. "The only differences between this form and the previous form are the ‘Test Requested Information’ section and the patient’s address is being requested here.  This form is used seldomly when clinicians from external institutions have material for sequencing for their patients.  I envision that in these cases, we would need to register the patient similarly to consults so that they are in the system, which may require us to change this form slightly to capture all the information needed for consult registration. The Test Requested Information section will become obsolete if this information is captured by EPIC when clinicians initially order the test." (Contributed by Rob Kim)

C. The two sample deidentified output reports are attached below.

D. The list of discrete variables that would be sent along with the PDF of the report is attached below. Expression and fusions may apply for RNA-seq. Other variabled that may need to be sent in the HL7 message include:

Point mutations and indels

---

reference genome used

chromosome

genomic coordinate

gene

nucleotide reference

nucleotide allele1 tumor

nucleotide allele2 tumor

nucleotide allele1 normal

nucleotide allele2 normal

VAF normal

VAF tumor

coverage tumor

coverage normal

reference annotation (NP or NM for refseq)

amino acid change and impact

splicing change and impact

frameshift and impact 

interpretation text

Copy number alterations (gene based)

---

reference genome used

gene annotation used 

gene

focal / large-scale

chromosome

genomic coordinate start (breakpoint)

genomic coordinate end (breakpoint)

original log-ratio

estimated copy number

coverage

interpretation text 

E. [Comment from Rob Kim] Do we need to think about scenario’s when things do not work well?  For example, we extract tissue from an archival FFPE block and the quality does not meet our accepted metrics.  We would then need to identify another block and extract again.  If no additional material is available, we’d need to contact the referring clinician and try to schedule the patient to come in for a biopsy.  To my knowledge, this has not yet occurred but these alternative workflows should be considered.

If the test fails at any point, is there ways to report this.  

There are several patients where we’ve consented/enrolled them into IPM and collected the germline control but then we never received tissue.  Patient may have died.  Would there be some mechanism to close out test requisitions?