Attendees

Joginder Madra (Facilitator)
Jean Duteau (Facilitator)
Dave Hamill (HL7)
Karen van Hentenryck (HL7)
Matt Rahn (ONC)
Nagesh Bashyam aka Dragon (ONC)
Dr. Julia Skapic (ONC)
Calvin Beebe (Mayo Clinic)
Ben Flessner and Dave Sundaram-Stukel (Epic)
Dave Camp, Peter James and Kenny Tomlinson (Wellcentive)
Raychelle Fernandez and Ozlem Kurt (Dynamic Health IT)
Dave Carlson (Mieweb)
Lisa Nelson (Life over Time Solutions)
Joe Lamy (Aegis / Sequoia)
Russ Leftwich (Intersystems)
George Cole (Allscripts)
Angel Pinzon (APS Puerto Rico)
Linda Michaelsen (Optum)
Luis Jimenez, Michelet Boursiquot, Clint Walker, and Luis Silva (Document Storage Systems)
Dmitry Shalamov (NextGen)
Dante Hoyte (USF)
Mario Nears (USF)

Agenda

Thursday, January 7, 2016

0900-1000 Introductions
1000-1200 CCD Testing
1200-1300 Lunch
1300-1400 CCD Testing (ctnd)
1400-1700 Discharge Summary Testing

Friday, January 8, 2016

0900-1200 Referral Note Testing
1200-1300 Lunch
1300-1500 Discussion about Testing
1500-1600 Discussion about next IaT

Discussion Items

Item	Notes
Introductions / Housekeeping	Housekeeping (Dave) Introductions (Jean) Some participants could not connect to their proxies or VPNs due to the security setup of the university guest network. Joginder set up a mobile hotspot and that seemed to resolve the issue for those having trouble (God help his data plan).
CCD Scenarios	Worked through CCDS and C-CDA Full Receiver discussion: LOINC codes were not always understood units of measure not UCUM medication codes did not match what was expected Creator discussion: some systems only support a single primary language some systems only allow a binary advanced directives (i.e. yes/no) as opposed of details of the advanced directive examples included concepts not in current vocabulary set "plan of care" not always clear what is supposed to be documented in this section Q: How should text-only narratives (e.g. no medication entries) be handled in the receiving systems UI? A: These should always be included. Certification rules are expected to follow the gold standard set by the Examples Task Force. Certification tests also provide the explicit codes to use...i.e. test data is fixed.
Ask the ONC	Q: What is the source of truth for value sets? Some are obvious, but some are challenging to discern who is the authority. A: If it is in VSAC, certification will use it (approved only). ONC is working with closely with VSAC. Q: Will it be a problem for testing if systems add translations to the expected codes? A: As long as the expected code is sent, sending the translation will not be an issue Q: How is errata handled by the validators - e.g. some validators (i.e. QRDA vs. C-CDA) have incorporated the errata and others have not? A: Send email to ONC inbox to raise the issue. The SITE validator is intended to go above and beyond what is required for certification Q: How are people handling CCDS requirements that are not in the C-CDA document templates? A: People are adding them anyways...as the interpretation is that those sections are always required. Q:What is the CDA extension source of truth? A: The full extension schema is posted to gforge and is accessible via the HL7 wiki. This is no longer shipping with C-CDA...implementers must get the latest from the repository
Discharge Summary Scenarios	Worked through CCDS and C-CDA Full Q: Is there a way to evaluate to how instances have been consumed (e.g. for accuracy, etc.)? A: One option could be to for receiving systems to create a PDF and note any inconsistencies.
Day Two
Receiver Processing	Receivers took turns showing how the samples created during day 1 were consumed by their system. General observations: different vocabulary (i.e. codes, code systems) used for procedures, problems, etc. some variations in some data points (e.g. start/end dates, dosing instructions - i.e. dosage of 5 mg vs 5 mg once daily) some variations in document template choice (e.g. CCD ambulatory vs. CCD) reconciliation will remain an important part of the process schematron bugs were discovered General comments: ID usage is a problem place - e.g. systems assigning their own IDs where IDs already exist or are not present Stylesheet usage can be an issue if receiving systems use vendor stylesheets as they may have errors. Although the base standard indicates that it documents should be stylesheet agnostic. It is difficult to determine if information on a clinical document pertains to the patient or to an encounter (i.e. is this information something that was determined during the encounter or is this information something that was associated with the patient prior to the encounter)
Common Clinical Data Set Discussion	There are multiple places to record medication information (e.g. Medications, Admission Medications, Discharge Medications, etc.). It would be good to clarify which sections are used in which circumstances. Because document templates are open, it appears that one could legally use a CCD and add discharge summary sections (e.g. discharge medications) for discharge summary usage scenarios. It would be good to provide clarification on how discharge summary information should be structured in the Companion Guide. It is important to remember that there is a different context for medications, admission medications, administered medications and discharge medications. Important to have guidance on when to use which and if it is appropriate to "repeat" information for the sake of adding sections to meet certification requirements. Could the use of subsections help this? Unclear as there is no guidance to vendors in terms of how section/subsection information should be consumed or displayed to end users. Common misunderstanding of what information should be in procedures sections vs. what information should be in results section. Vendors support CPT and SNOMED-CT for procedure coding. If implementers have entry information, are they using the entries optional template with entry information or are they using the entries required template. It appears that vendors are using the entries required template...but would change if required by certification (smile). C-CDA is generally very permissive with nullFlavor usage Companion Guide should highlight the necessity to drill down (i.e. ask more than one question) when the patient is a non-smoker as the concept of non-smoker does not exist in the abstract as the existing concepts are "never smoked" and "former smoker". Examples Task Force is leaning towards having UDI's in the medical equipment section. Companion Guide project will need to provide guidance on how to handle medical equipment. ONC has issued clarification that Goals, Plan of Care, Treatment Plans can be narrative-only within the context of transition of care Q: Will HL7 be using the ONC 2016 rules? A: The 2016 rules represent the newest final rules. The Companion Guide project will be issuing guidance as per the 2016 rules.
De-Brief and Wrap Up	Receiver show and share session was valuable Data entry takes a lot of time. It would be helpful to have scenario information in advance. Is virtual an option? It may be an option for future sessions. Participants mentioned collaboration such as skype channels to continue discussion amongst the community. Some of the discussions waded into areas where there may be different stakeholders/participants/SMEs...future IaTs may need to consider different tracks. Having ONC representation was valuable. In-person participation is conducive to collaboration Consider having a common set of samples that are created ahead of time for consumption by receivers during the IaT and review how info was consumed. Strong suggestion to keep approach of taking transport considerations out of the IaT - i.e. continue to exchange documents via shared folders or USB drives. Consider exercise of having a strong, robust narrative for physicians to enter in advance to see how they are created by the physician. Joe Lamy provided a brief update on the Sequoia project