Laboratory

Cerner Corporation USCDI Draft V2 Comments - Laboratory

Provided below are Cerner's comments on the USCDI draft V2 proposals for the Laboratory data class. Cerner's full public comments on the USCDI draft V2 have been posted on the USCDI general comments section.   Cerner supports the proposal to re-classify the Laboratory Report Narrative and Pathology Report Narrative data elements from the Clinical Notes data class to the Laboratory data class. As with the Diagnostic Imaging Narrative data elements (for which we provide recommendations in the Diagnostic Imaging data class section of this comment letter), the data elements do not fit within the intended scope of Clinical Notes and are better classified in a more distinctive data class. That said, there is still significant ambiguity as to the intended definition of those data elements, even with their reclassification. Additionally, we note that if these re-classifications are finalized in USCDI V2, ONC should include the same changes in USCDI V1 via an errata adoption. This is important for consistency as HIT developers and healthcare providers are currently working to align with USCDI V1. If changes other than additions are made for USCDI V2 it may create inconsistencies for exchange of the same data between those versions.   Our first recommendation is to remove “narrative” from the data elements’ naming and adopt them simply as “Pathology Report” and “Laboratory Report.” Including “narrative” in the naming is misleading because it implies the whole of the content is narrative. However, freetext narrative content would be only one component of a full report, which consists of both coded and freetext/narrative data, as well as quantitative and/or qualitative observations. In the case of Laboratory Report, a narrative component may not even be relevant as even textual results are generally encoded as a discrete observation with a specific result “type” classification (e.g., “string”). We recommend accounting for this in adopting the formal definition of the data elements and ensuring that the scope of data to be contained within each is clear so that health IT developers and healthcare providers have a consistent understanding of precisely what is required or expected to be exchanged for the data class.   Additionally, regarding the Laboratory Report data element, we note that there are already well-defined standards established by the regulations of the Clinical Laboratory Improvement Amendments (CLIA) that define what constitutes a laboratory report. This is explicitly the data elements listed below, which are adopted in the CFR at 42 CFR 493.1291(c)(1) through (7).  CLIA laboratory report elements:

• For positive patient identification, either the patient's name and identification number, or a unique patient identifier and identification number.
• The name and address of the laboratory location where the test was performed.
• The test report date.
• The test performed.
• Specimen source, when appropriate.
• The test result and, if applicable, the units of measurement or interpretation, or both.
• Any information regarding the condition and disposition of specimens that do not meet the laboratory's criteria for acceptability.

Furthermore, supporting the ability to represent a laboratory report is already a requirement as part of the ONC’s Health IT Certification Program under the 170.315(e)(1) View, Download, and Transmit to 3rd Party criterion. Guidance in the 170.315(e)(1) Certification Companion Guide states that the laboratory report should be represented with a HL7 CDA C-CDA Result Observation entry template, which HL7 has published examples of (see here). This has similarly been a requirement of past certification of laboratory results reporting in prior editions of ONC’s certification criteria.   Given the precedence for both the established definition of what constitutes a laboratory report and how it should be represented for purposes of ONC Health IT Certification (at least in HL7 CDA C-CDA), the most sensible approach would be to adopt the Laboratory Report data element with a definition aligned with that of CLIA regulations defined above, inclusive of any relevant narrative notes. Following this recommended approach will maintain consistency with long-held industry principles and ensure a common understanding of the scope of the data element.   As a final note, the Laboratory data class is a fitting example for our recommendation under the General Comments section of this comment letter that ONC seek to stratify the USCDI to specify the actors for whom various data classes and/or elements are intended (or required). This is because specific interoperability use-cases for laboratory data may not call for all elements of a CLIA laboratory report to be present.

Common pattern needed for Laboratory and Pathology Data Elements

To present a more consistent way of putting forth testing for Laboratory and Pathology information, we recommend creating a consistent set of data elements and a similar set of definitions and relationships.  The Order expresses the test when it is in an “ordered” state whereas the Test, expresses the test when it is in a “completed” state.  The value/result is the measured value determined by the test as its “results”. The report contains information about the order, the test and the value/result along with other narrative explaining the assessment of the results. Recommended Data Elements: Laboratory Order Laboratory Test Laboratory Value/Result Laboratory Report Pathology Order Pathology Test Pathology Value/Result       Pathology Report

Narrative laboratory and pathology report

• The specification for laboratory results narrative and pathology results narrative are very unclear. They say they contain the consulting specialist’s interpretation of the laboratory or pathology report respectively. Are these talking about consultants who are not responsible for the lab report or pathology report respectively, or the report generated by the lab or pathologist? Realize also that such reports contain a lot more than an interpretation.
• This raises the question of the standard order codes. For laboratory reports, the order and diagnostic report code will usually be a LOINC code. US Core FHIR, which ONC requires all EHRs to support, specifies LOINC codes for the order code when an appropriate LOINC code is available.

• USCDI does specify codes for orders and diagnostic exports of imaging tests but says nothing about order codes or diagnostic report codes for laboratory tests. US Core FHIR which is requires by ONC specifies LOINC codes (when one exists) for this purpose. Why the discordance?

No mention of units of measure...

• The section on laboratory makes no mention of units of measure. Without units of measure, structured results lose most of their value. FHIR requires UCUM units for all quantiles (with a few exceptions) https://www.hl7.org/fhir/datatypes.html#Quantity. C-CDA also specifies UCUM. UCUM is designed for converting values in one unit to another commensurate unit. See https://ucum.nlm.nih.gov/ for tools that validate and convert UCUM codes.
• For vital signs, ONC has explicit requirements for UCUM units of measure and ONC’s standard for delivering reportable disease requires UCUM units. I think a previous  version of USCDI also required UCUM units. FHIR specifies UCUM for all quantitative. Since ONC requires FHIR, USCDI should not have a different approach. Result producers can still deliver the historic units of measure sting with the UCUM units.
• From a sample of 9 billion LOINC mapped test, close to 90% of the total volume of laboratory test are quantitate, so getting the units right is important.

"Report Narrative" Data Elements

Moving the existing Laboratory Report Narrative and Pathology Report Narrative data elements to this section in the Draft USCDI v2 does not resolve the core issue that these data elements are vague and untethered to any structures or bindings to applicable standards that would help clarify their intended meaning. Laboratory Report Narrative Many laboratory test results ARE narrative and are reported as such. Is this for that? If so, is this not already covered in the Values/Tests Data Element? Is it meant for the "interpretation" data element or just general comments/disclaimers/annotations? Is it meant to mean the primary diagnostic report or a secondary consultation note? These kinds of clarifications are necessary for understanding the appropriateness of terminology bindings. I concur with the LOINC Committee recommendation to remove this data element. Pathology Report Narrative Implementers and systems are bit more clear on the notion of a (anatomic) Pathology Report, but it is important to recognize that they exist in various "flavors" from pure narrative, to "structured", to synoptic. See the NAACCR Pathology Laboratory Electronic Reporting, Version 5.0 – May 2020(revised July 2020) standard for more detail on these distinctions, including the reference to the CAP's definition of synoptic reporting. (Note that this standard also specifies relevant structures and terminologies).  In addition to NAACCR's standard, there are other examples of fully defined templates for these structured documents, including IHE's Anatomic Pathology Structured Report. By labeling this data element as Pathology Report Narrative we are confusing implementers about whether only certain styles or formats of pathology reports are required (i.e. only those as text narrative) or certain portions of the whole report. Rather, I believe the intent is more broad...to capture all of these "flavors" and to remain agnostic about the exchange format (HL7v2, CDA, FHIR, etc).  In addition to using to a more precise structure for specifying this as a domain-specific profile (vs a "data element"), I recommend that this USCDI entity be renamed to simply Pathology Report. In the definition and bindings to relevant standards, it would also be helpful to clarify whether this is intended to a) cover anatomic (e.g. surgical) pathology reports only or b) any report by a pathologist (general pathology, also encompassing clinical pathology).