The Optimizing the Use of Data Standards PhUSE CS Working Group: Best Practices for Data Standards Implementation Project has initiated this document to address the challenges faced by industry when implementing standard laboratory units.
The opinions expressed in this document are those of the authors and do not necessarily represent the opinions of PhUSE, members' respective companies or organizations, or regulatory authorities. The content in this document should not be interpreted as a data standard and/or information required by regulatory authorities.
Challenge
When submitting datasets containing clinical laboratory data to regulatory bodies, there is no cross-industry agreement on the units in which lab results should be reported. There may be a lack of agreement within departments of sponsor companies, across companies, and between and within regulatory bodies. Choice of unit can impact familiarity of results in medical review at the regulatory body, and pooling within and across submissions and companies. This Best Practice document provides a recommendation for standardizing units within a sponsor/submission.
Scope
§ Background
§ Process for managing test units within a sponsor/submission
§ Managing character value consistency
§ Defining Internal Units for a Test Concept
§ External Links
§ Glossary
§ Authors and Reviewers
Background
Guidance from both pharmaceutical regulatory authorities such as the FDA, EMEA and clinical data standards organizations such as CDISC call for industry collaboration to report standardized units for test results (see Links section of document). Both groups acknowledge Systems International (SI) units are the worldwide standard, but not necessarily the units commonly used in clinical practice. Laboratories target their unit use to the end user (the clinicians receiving the lab results). US laboratories will typically report results in US conventional units, ex-US laboratories will typically report results in SI units and corporate central laboratories that serve a worldwide audience, are typically able to accommodate both units. In most cases however, for certain tests such as enzymes, all laboratories default to a unit commonly accepted and understood in clinical practice (few laboratories will report enzyme results back to the clinician in the SI unit of katals and will instead report results in enzymatic units). This “regional familiarity” to certain units, and the use of commonly accepted units for certain tests, may also drive unit expectations in submission data between different worldwide regulatory agencies.
Further complicating the issue of the submission data representation, the CDISC implementation guide calls for LBSTRESU to be standardized, and that LBORRESU and LBSTRESU utilize controlled terminology.
The table below provides some examples of the challenges that can be encountered when trying to define a standardized unit for a given test in a given specimen.
Seq / LBTESTCD / LBTEST / LBSPEC / Company A / Company A / Company B / Company B / Company C / Company C / Company D / Company DSource unit reported by laboratory / CDISC submission value (reporting unit) / Source unit reported by laboratory / CDISC submission value (reporting unit) / Source unit reported by laboratory / CDISC submission value (reporting unit) / Source unit reported by laboratory / CDISC submission value (reporting unit)
1 / BASO / Basophils / CSF / /uL / /uL / /mm3 / /uL / 10^6/L / 10^6/L
2 / BASO / Basophils / Blood / 10^3/mm3 / 10^9/L / 10^3/uL / 10^9/L / 10^9/L / 10^9/L
3 / RETIC / Reticulocytes / Blood / 10^3/uL / 10^9/L / 10^3/mm3 / 10^9/L / 10^12/L / 10^12/L / 10^6/mm3 / 10^12/L
4 / SODIUM / Sodium / Serum /
Plasma / mEq/L / mEq/L / mmol/L / mmol/L / mg/dL / mg/dL
Seq 1 – Basophils CSF, all units are synonymous, CDISC supports both /uL and 10^6/L as reporting units
Seq 2 – Basophils blood, all units are synonymous, notice that 10^3/uL and 10^3/mm3 are synonymous and may be used interchangeably by different laboratories to represent the same unit concept. The widely accepted SI unit is 10^9/L and this is the only unit on the row supported by CDISC terminology as a reporting unit
Seq 3 – Reticulocytes blood, has two valid SI unit representations 10^9/L and 10^12/L (both are supported with CDISC controlled terminology as reporting units).
Seq 4 – Sodium (plasma/serum). CDISC supports three unit concept mEq/L, mmol/L and mg/dL as reporting units, with the commonly accepted SI unit being mmol/L
The variability in the possible representation of unit characters can be equally as limiting with respect to standardization. Interchangeable use of Greek letter symbols with short codes is one source of inconsistency in test unit. For example, ‘micro’ can be represented as ‘µ’, ‘u’ or ‘mc’. Inconsistent use of capitalization in units is another cause of inconsistency and possible error.
Process to Manage Test Units, Recommended Solutions
The process for managing lab test units begins with identifying standardized reporting units and documenting these in a Test Unit Plan (TUP), as a necessary component of the Study Data Standardization Plan (SDSP). These units should match the CDISC Controlled Terminology (CT) unit code list (reporting unit) wherever possible. This approach will ensure consistent reporting unit use within a submission. The TUP will allow for early and open internal as well as external discussion on clinical meaning, relevance, and conversion needs/awareness of units.
When providing multiple submissions, it is recommended that companies develop a dictionary of submission test unit standards, again using CT whenever possible. As the industry guidance begins to consolidate and move closer to global standardization of lab units, organizations that have already established internal lab unit standards will be in a better position to convert to the global standard units wherever necessary.
The basic elements for a TUP include for each unique test concept (defined by a combination of LBTESTCD, LBCAT, LBSPEC and/or LBMETHOD), the standardized reporting units in which the lab data for that unique test concept will be reported in the submission datasets. Additional elements for a TUP include but are not limited to, source units and conversion factors, significant digits, rounding off rationale and synonyms. The process for establishing standardized reporting units within a submission utilizing a TUP must incorporate timing to allow for regulatory dialogue, review and agreement. The TUP must be defined and documented as early in the development plan as possible. TUPs should be drafted at the beginning of new drug development with the safety tests planned for early phase. Most early phase lab tests are included within a submission. This is particularly necessary for lab test units which may have a high degree of variance or may not yet be defined within CT. For test units that do not follow CDISC clinical meaning or may vary across clinical application (i.e. sensitivity for platelets for Oncology vs other therapeutic areas), conversion needs, and ability to assess results appropriately to the unit and value should be integrated into the regulatory dialogue.
The table below displays examples and recommendations for handling LBORRESU with respect to whether it should be left “as is” (because it is the original unit) or whether it should be converted to an equivalent CT value (as directed by the SDTMIG). The table provides an example of standardization of ORRESU to available reporting unit in CT.
SEQ / LBTESTCD / LBTEST / LBSPEC / LBMETHOD / Source Unit as provided by the lab (not in SDTM) / LBORRES / LBORRESU (standardized to UNIT codelist) / LBSTRESN / LBSTRESU (standardized to unit codelist)1 / BILI / Bilirubin / SERUM / COLORIMETRY / MG/DL / 1 / mg/dL / 17.1 / umol/L
2 / BILI / Bilirubin / SERUM / COLORIMETRY / mg% / 1 / mg/dL / 17.1 / umol/L
3 / BILI / Bilirubin / SERUM / COLORIMETRY / micromol/L / 10 / umol/L / 10 / umol/L
4 / BASO / Basophils / BLOOD / FLOW CYTOMETRY / /ul / 1 / /uL / 0.001 / 10^9/L
5 / BASO / Basophils / BLOOD / FLOW CYTOMETRY / x10E9/L / 0.001 / 10^9/L / 0.001 / 10^9/L
6 / BASO / Basophils / BLOOD / FLOW CYTOMETRY / 10^3/mm3 / 0.001 / 10^9/L / 0.001 / 10^9/L
7 / BASO / Basophils / BLOOD / FLOW CYTOMETRY / 10^3/uL / 0.001 / 10^9/L / 0.001 / 10^9/L
Only on sequences 6 and 7 is there a situation other than making a standard representation of exactly the same unit. In cases 6 and 7, since 10^3/mm3 and 10^3/uL are synonyms in CDISC Controlled Terminology of 10^9/L, the translation of the source unit (reported by the laboratory) to the CT compliant reporting unit is made in LBORRESU but critically the value LBORRES and LBSTRESN do not change. Traceability can therefore always be mapped back through the synonyms and indeed, teams can put the source unit in a supplemental qualifier if they feel it necessary.
Managing Character Consistency
Character consistency should be managed by following CT, in particular the use of Unit Code List– C71620, whenever possible. Lab tests units not covered by CT (either as a reporting unit or synonym) should be elevated for discussion with the regulatory body. It is recommended that these units are also submitted to CDISC for inclusion in the Unit Controlled Terminology.
Defining Internal Units for a Test Concept
The process for standardizing lab units in a Test Unit Plan can lead to the identification of test concepts that offer the unique challenge of forcing a company to define an internal standard unit for that test concept which does not have a documented CT term. It is recommended that these units are also submitted to CDISC for inclusion in the Unit controlled terminology. To assist in defining the internal standard the following sources are recommended for review; libraries for the National Institute of Health and National Cancer Institute, Clinical Guide to Laboratory Tests, by Norbert Tietz, 4th ed., and the use of peer reviewed publications.
External Links
Standardizing Lab Units
FDA - http://www.fda.gov/ForIndustry/DataStandards/StudyDataStandards/ucm372553.htm)
RAPS - http://www.raps.org/focus-online/news/news-article-view/article/4250/fda-says-it-wants-to-make-it-easier-for-industry-to-submit-international-labora.aspx
CDISC Request New Term
Scroll down to the end of the page at the following URL
http://www.cancer.gov/cancertopics/cancerlibrary/terminologyresources/cdisc
or this might work as a direct link
http://ncitermform.nci.nih.gov/?version=cdisc
Glossary
Source Unit – unit exactly as reported by the laboratory for the given test concept
Original Unit – equivalent to the LBORRESU field in CDISC and a standardized representation of the test unit concept aligned with CDISC controlled terminology (submission value). This unit may differ for a given test concept within a submission
Reporting Unit - equivalent to the LBSTRESU field in CDISC and a standardized representation of the test unit concept aligned with CDISC controlled terminology (submission value). This unit may NOT differ for a given test concept with a submission
Authors and Reviewers
Authors:
Mathew A. Bryant, Nicholas Pemble, John Franchino, Randall K. Carlson,
Sandra Vanpelt, Pawan Gupta, Unnat Patel
Reviewers:
Takuya Kita, Carol Matthews, Karen Curran, Leanne Huang