CDISC QRS: Match the Instrument to the Published Supplement

When a clinical trial protocol specifies an EQ-5D-5L health survey, a standardized six-minute walk test, and an APACHE II disease severity score, the study team faces an operational data-standards decision long before the first subject is enrolled. Clinical operations, biostatistics, and data management teams must determine which Study Data Tabulation Model (SDTM) domain tabulates each instrument, which published Clinical Data Interchange Standards Consortium (CDISC) Questionnaires, Ratings and Scales (QRS) supplement version and controlled terminology package govern the submission, and whether locating the instrument on the public CDISC standards portal confers the legal right to administer it on electronic clinical outcome assessment (eCOA) devices.

A published CDISC QRS supplement settles exactly one of these questions: the tabulation domain. On the CDISC QRS page, questionnaires are stored in the Questionnaires (QS) domain, functional tests in Functional Tests (FT), and clinical classifications and disease-response criteria in Disease Response and Clin Classification (RS). Each SDTMIG supplement provides instrument-specific controlled terminology, an SDTM example, applicable supplemental qualifiers, item-level mapping instructions, and implementation assumptions. On October 18, 2026, the supplements table on the CDISC QRS page displayed 324 records. That count mixes SDTM supplements, ADaM supplements, and rows for instruments with no supplement at all (for example, where the copyright holder denied permission or never responded), so it is not a count of usable tabulation supplements.

Finding a supplement still leaves two technical and legal boundaries unresolved. First is the standards version dependency: the FT domain and the broadened RS domain arrived with SDTM Implementation Guide (SDTMIG) v3.3. Before that, RS was intended only for disease-response classifications, and CDISC advises that submissions using SDTMIG v3.2 or earlier represent ratings and scales other than questionnaires in custom domains that mirror FT and RS. Instrument codelists are not static either. They ride on quarterly CDISC Controlled Terminology releases distributed through the National Cancer Institute Enterprise Vocabulary Services (NCI-EVS); the release dated September 25, 2026 incorporated Review Package 62 and added approximately 102 new QRS terms. Which release a study uses has to be declared and checked against the FDA Data Standards Catalog.

Second is the legal licensing boundary. CDISC must obtain permission from the copyright holder before it creates a supplement for a copyrighted instrument, but CDISC states that granting it that right is not a license for standards users. A trial that administers a copyrighted instrument such as the EQ-5D-5L still follows the owner's process, which for EQ-5D means registering the study with the EuroQol Office and obtaining prior written consent. Treating a supplement as a license, or assuming that domain assignment does not depend on the SDTMIG version, can lead to eCOA specification rework, conformance findings on datasets and define-XML, and gaps in the record an inspector will ask to see.

CDISC develops QRS supplements for named, standalone instruments and stores them in three findings domains. The CDISC QRS page also cross-references the FDA clinical outcome assessment (COA) types: questionnaires can carry clinician-reported (ClinRO), observer-reported (ObsRO), and patient-reported (PRO) outcomes; functional tests correspond to performance outcomes (PerfO); and clinical classifications correspond to ClinRO.

• Questionnaires (QS): Named, standalone instruments made of defined questions with a defined set of potential answers. The items are conceptually related, usually scored, and come with documented administration and analysis methods. Examples in the CDISC table include the EQ-5D-5L, the Patient Health Questionnaire-9 (PHQ-9), and the Geriatric Depression Scale Short Form (GDS SHORT FORM).

• Functional Tests (FT): Named, task-based evaluations of mobility, dexterity, or cognitive ability. CDISC stresses that a functional test is not a subjective assessment of how the subject generally performs a task. It is an objective measurement of the task performed in a specific instance, typically a direct quantitative result such as time or distance. Examples include the 6 Minute Walk Test, the 10-Meter Walk/Run, and the Timed Up and Go (TUG).

• Disease Response and Clin Classification (RS): Clinical classifications are named instruments whose ordinal or categorical output serves as a surrogate for, or ranking of, disease or physiological status, based on a trained professional's observation and clinical-records review. They may be composite scores built from findings that live in other domains, such as labs or vital signs. Disease-response criteria describe changes that result from treatment. Examples include the Acute Physiology and Chronic Health Evaluation II (APACHE II), the Abnormal Involuntary Movement Scale (AIMS), the 17-item Hamilton Depression Rating Scale (HAMD 17), and RECIST 1.1.

Take the opening scenario. The protocol calls all three instruments outcome measures, but they map to three different SDTM findings domains, each with its own variable conventions. The short names and versions below are as listed in the CDISC QRS table on October 18, 2026:

• EQ-5D-5L -> QS: Responses to the five dimensions (Mobility, Self-Care, Usual Activities, Pain/Discomfort, Anxiety/Depression) and the visual analog scale are tabulated as separate QS rows, with QSCAT set to the CDISC short name 'EQ-5D-5L' (supplement v1.3, May 31, 2022). Item codes in QSTESTCD come from the instrument's test-code codelist in CDISC Controlled Terminology rather than being typed from memory. Where the analysis needs an index value derived from a national value set, the SAP and the supplement's assumptions decide whether it is tabulated or derived in analysis datasets.

• 6 Minute Walk Test -> FT: Distance and the other recorded results are tabulated in FT with FTCAT set to 'SIX MINUTE WALK' (supplement v1.0, May 21, 2014). Where the administration method matters, FTMETHOD draws on the QRS method codelist, which keeps the conditions behind the score traceable, as detailed in our guide to standardizing PerfO administration records.

• APACHE II -> RS: The scored classification is tabulated in RS with RSCAT set to 'APACHE II' (SDTM supplement v2.1, May 20, 2025). CDISC describes APACHE II as a composite of laboratory, respiratory, and vital-sign assessments plus the Glasgow Coma Scale. The underlying measurements remain in their topic-based domains, such as Laboratory Test Results (LB) and Vital Signs (VS); only the classification and its scored items populate RS.

flowchart TD
    A["Named instrument or assessment"] --> B{"What is being recorded?"}
    B -->|"Defined questions and answers"| C["QS - Questionnaires"]
    B -->|"Objective task performance"| D["FT - Functional Tests"]
    B -->|"Composite or examiner-rated classification, disease response"| E["RS - Disease Response and Clin Classification"]
    B -->|"Rating of a single body system"| F["Body-system morphology or physiology domain"]
    B -->|"Score based on microscopic tissue data"| G["MI - Microscopic Findings"]
    B -->|"Sponsor-defined single item, VAS or NRS"| H["FA - Findings About"]
    C --> I["Examples: EQ-5D-5L, PHQ-9, GDS Short Form"]
    D --> J["Examples: 6 Minute Walk Test, Timed Up and Go"]
    E --> K["Examples: APACHE II, AIMS, HAMD 17, RECIST 1.1"]
    F --> L["Example: swollen and tender joint counts"]
    G --> M["Example: Gleason score"]
    H --> N["Example: sponsor fatigue VAS"]

Format alone can mislead. HAMD 17 is a clinician-administered rating with scored items, and many practitioner summaries group it with questionnaires, yet the CDISC table lists its SDTM supplement under RS (v2.1, April 21, 2026). This is consistent with the CDISC FAQ's description of examiner-rated scales that cover symptoms across multiple body systems: they are not posed as question-and-answer questionnaires, so they are represented as clinical classifications. Confirm the domain in the live table instead of inferring it from what the form looks like. Forcing functional tests or classifications into QS leaves reviewers with variables whose meaning does not match the data.

A CDISC QRS supplement is not a generic guidance note; it is a version-controlled technical specification. For each named instrument, CDISC issues an SDTMIG supplement containing five essential implementation components: (1) instrument-specific Controlled Terminology codelists, (2) an SDTM example illustrating use, (3) applicable supplemental qualifiers (such as SUPPQS, SUPPFT, or SUPPRS), (4) item-level mapping instructions that dictate how raw responses convert to standardized character and numeric results, and (5) implementation assumptions. Every supplement version carries its own number, and the table records each release date. The EQ-5D-5L supplement, for example, has moved from v1.0 (March 5, 2014) to v1.3 (May 31, 2022), and APACHE II from v1.0 (June 29, 2016) to v2.1 (May 20, 2025).

Versions can also change underneath an old supplement. CDISC reports that many older QRS supplemental-qualifier names were renamed to meet Non-Standard Variable (NSV) Registry naming rules, but older supplements could not all be reissued. Until they are, CDISC recommends that users check its QRS NSV Registry summary spreadsheet on the QRS page and use the current registry names. A supplement PDF saved at protocol design can therefore carry qualifier names that are no longer current.

CDISC also publishes ADaMIG QRS supplements. These describe how to structure the analysis dataset using ADaMIG structures (the CDISC FAQ points to the Basic Data Structure, with ADaM Other for special needs), and they include sample analysis descriptions, scoring for the statistical analysis plan, data checks, metadata examples, and an example final analysis dataset. Practitioner coverage describes the GDS Short Form supplement, listed in the CDISC table with a January 29, 2021 release, as the first ADaM QRS supplement. CDISC also released readme files explaining why four single-item instruments would not receive ADaM supplements. ADaM coverage is still thin: on October 18, 2026, ADaM rows in the table included APACHE II (v1.1, May 20, 2025) and HAMD 17 (v1.0, April 21, 2026). Practitioner guidance flags PARCAT1/PARCAT2 and phantom records for missing items as places where analysis-dataset structure is decided and where reviewers will look.

A common pitfall in multi-year programs is assuming that domain mapping stays fixed across SDTMIG releases. The FT domain and the broadened RS domain are SDTMIG v3.3 features. RS was originally intended for disease-response classifications; at v3.3 it was renamed Disease Response and Clin Classification and revised to include all clinical classifications. Under SDTMIG v3.2 and earlier, QRS supplements were represented in QS or in custom domains. AIMS shows the consequence: a 2022 PharmaSUG paper shows it mapped to QS before SDTMIG v3.3 and as a clinical classification in RS afterwards, and the CDISC table now lists AIMS under RS with a v2.0 released April 3, 2021.

CDISC lists SDTMIG v3.4 as the current implementation guide; it was developed against SDTM v2.0 (November 29, 2021), and the model itself has since advanced to SDTM v2.1 (June 10, 2024). For studies still on SDTMIG v3.2 or earlier, the CDISC QRS FAQ says ratings and scales other than questionnaires should go in custom domains that mirror FT and RS. Those custom domains may reuse the FT and RS terminology and domain codes, and the sponsor should alert regulators to the approach in the data reviewer's guide. When integrated safety or efficacy datasets pool studies run under different SDTMIG versions, programmers must reconcile instruments that sit in QS or a custom domain in older studies and in FT or RS in newer ones. Verify the SDTMIG version in the Study Data Standardization Plan before binding eCOA forms to domain variables.

While the QRS supplement establishes the structural mapping rules, the actual submission values—including category names (--CAT), subcategories (--SCAT), test codes (--TESTCD), full test names (--TEST), and permissible response codelists—are published within CDISC Controlled Terminology. Developed collaboratively by CDISC and the National Cancer Institute Enterprise Vocabulary Services (NCI-EVS), these terminologies are distributed quarterly on the NCI-EVS file repository.

The CDISC QRS FAQ describes how to find an instrument's terms in the SDTM terminology spreadsheet. First, locate the instrument's category in the Category of Questionnaire, Category of Functional Test, or Category of Clinical Classification codelist. Then search for its CDISC synonym to reach the paired test-code and test-name codelists. CDISC's worked example is the Satisfaction With Life Scale, whose QSCAT value is SWLS and whose synonym SWLS01 leads to the SWLS01TC and SWLS01TN codelists. Doing this lookup against the pinned release, not the newest file on someone's desktop, is what keeps eCOA item codes and define-XML in agreement.

At the time of writing, the most recent release is dated September 25, 2026. It replaced the earlier ADaM, CDASH, Define-XML, SDTM, SEND, and other terminology files, includes terms from Review Package 62, and adds approximately 102 new QRS terms, alongside 656 new terms across the ADaM, CDASH, Define-XML, SDTM, SEND, and other files. The same release also updated the QRS Naming and Business Rules document, which matters for any instrument that still lacks published terms.

For FDA submissions, the newest package is not automatically the right one. FDA's technical specifications for COAs that use Item Response Theory direct sponsors to review the FDA Data Standards Catalog so that submissions follow FDA-supported standards, and they point to the Study Data Technical Conformance Guide for general specifications. Declare the terminology version in define-XML and apply it consistently across QS, FT, and RS. A mid-study terminology upgrade should be a documented decision, not a silent refresh.

A frequent operational problem is an instrument that has no CDISC supplement or published terminology. Study teams should not invent arbitrary test codes. FDA's November 2023 technical specifications, Submitting Clinical Trial Datasets and Documentation for Clinical Outcome Assessments Using Item Response Theory, are nonbinding and scoped to covered COAs that use IRT. Even so, their rule is a useful default: where controlled terminology is not available, variables such as QSCAT and QSTESTCD, and item test codes and names, should be constructed according to the CDISC QRS Naming and Business Rules rather than ad hoc. CDISC posts those rules among the QRS implementation documents on its QRS page, and they were updated with the September 2026 terminology release, so check the current file rather than a copy saved at protocol design.

A common misconception among clinical operations and programming teams is that every rating, questionnaire, or scoring algorithm belongs in QS, FT, or RS. The CDISC QRS FAQ and therapeutic area user guides (TAUGs) define explicit cases where assessment data belong in other SDTM domains. Putting these data in a QRS domain is a mapping error that will have to be unwound later.

Clinical development teams must watch for four primary non-QRS categories during endpoint mapping:

1. Single body-system rating and grading scales: Per the CDISC FAQ, a rating or grading scale whose intent is to evaluate a single body system belongs in the morphology/physiology domain for that body system. CDISC points to the swollen and tender joint count example in the rheumatoid arthritis TAUG. Scales spanning multiple body systems, and all composite scores, are clinical classifications in RS.

2. Microscopy-based scores: Scores based on microscopic data belong in the Microscopic Findings (MI) domain. The CDISC FAQ cites the Gleason example in the prostate cancer TAUG, and the QRS page names GHAS and RHI. CDISC's reasoning is that these are assessments of tissue on a slide by a pathologist, not of the person as a whole by a clinician. They may feed a clinical classification as an input but are not classifications themselves.

3. Sponsor-defined single-item assessments: General single-item assessments defined by the sponsor, such as overall quality of life or a symptom rated on a sponsor-defined scale, visual analog scale (VAS), or numeric rating scale (NRS), have no CDISC QRS supplements because they are not standard QRS instruments. CDISC represents them in the Findings About (FA) domain, using FATESTCD, as findings about a disease condition as an event. CDISC adds that such assessments may be public domain or copyrighted, and sponsors are responsible for appropriate use.

4. Composite classifications with distributed source data: Composite instruments draw on data held elsewhere. All source data stay in their topic-based domains (laboratory values in LB, vital signs in VS) while the classification itself is represented in RS, with RELREC used where relationships need defining. For oncology response criteria such as RECIST 1.1, CDISC places the response in RS, tumor data in the TU and TR domains, and imaging procedure information in PR.

Recognizing these exceptions during protocol design prevents costly data re-mapping late in the clinical trial lifecycle. When evaluating an outcome measure, teams should ask: Is this a multi-item standardized instrument, a physical functional test, a composite classification, or a primary physiological observation? The answer dictates whether QRS standards apply at all.

Each row of the supplements table on the CDISC QRS page has a 'Permission' column. Its values include Public Domain, Granted, Exempt from Copyright, Author Permission Required, Pending, Denied, and No Response Received, alongside development statuses such as In Development. A dangerous misunderstanding in trial operations is reading 'Granted' as permission for the sponsor to use the instrument.

CDISC defines 'Granted' as the copyright holder having granted CDISC permission to develop a supplement, and its copyright FAQ is explicit about what that permission does not cover. In CDISC's words:

"Granting CDISC the right to create supplements to represent an instrument in CDISC QRS domains is not a license for CDISC standards users to use the instrument without complying with the copyright holders licensing requirements."

"Sponsors still need to obtain copyright permission from the copyright holder to use the instrument in their clinical trials."

The EQ-5D shows how the two tracks coexist. CDISC lists EQ-5D-5L and EQ-5D-3L supplements in QS with the status Granted, and CDISC states that a QRS supplement does not affect copyright holder licensing fees or procedures. On the owner side, EuroQol manages EQ-5D copyright through the EuroQol Research Foundation and asks users to register their study, trial, or project via its registration form. The EuroQol Office then sets out the terms, conditions, and any licensing fees, which are determined from the registration details. Prior written consent from the EuroQol Office is necessary for using, reproducing, altering, translating, or distributing the EQ-5D. After approval, use is free for non-commercial academic, educational, public-health, and similar purposes. EuroQol also publishes separate digital data collection requirements, which matter for any eCOA build.

Under CDISC's definition, 'Public Domain' means no permission is necessary, and the 6 Minute Walk Test, APACHE II, AIMS, and HAMD 17 all carry that status in the table. Two cautions remain. 'Exempt from Copyright', the status shown for PHQ-9 and RECIST 1.1, is not the same as public domain: CDISC says use is subject to the copyright holder's terms-of-use policy. And a public-domain instrument still needs its administration manual, translations, and electronic version documented, as outlined in our analysis of what to confirm before building an eCOA form.

Distributors can simplify the owner side. Mapi Research Trust, a non-profit, exclusively distributes more than 800 COAs on behalf of their copyright owners through its ePROVIDE platform, which also hosts the PROQOLID database. It says its team can act as a single point of contact for securing licensing agreements, whether or not Mapi distributes the instrument, and that eCOA digitization is tested for equivalence with the original version. CDISC also works with Mapi Research Trust, through PROQOLID, to obtain access to copyrighted instruments for supplement development. That is a separate permission track from the sponsor's license and should not be confused with it.

The submission record should carry the same facts. For covered COAs that use IRT, FDA's technical specifications recommend documentation that includes the COA name, version, and copyright information (if applicable). Outside that scope the recommendation is not binding, but recording the instrument version and license holder next to the supplement version and terminology release keeps one trail for reviewers and inspectors.

To prevent discrepancies before database lock, clinical development teams should assemble a consolidated data-standards mapping table during protocol development. The table bridges clinical operations, eCOA vendors, biostatistics, and statistical programming by linking each instrument to its tabulation domain, supplement version, controlled terminology release, and licensing status.

The table below is an illustrative worked example built from the public CDISC QRS supplements table and owner pages as observed on October 18, 2026. It is not drawn from a sponsor dataset, and versions change, so recheck the live table before locking a specification. The terminology column shows the release a study starting now might pin after checking the FDA Data Standards Catalog; it is not a requirement.

Before an eCOA build is finalized or a clinical database is locked, the trial data management lead and lead biostatistician should audit their data-standards assets against a formal pre-lock reconciliation checklist:

By establishing this alignment early in the trial lifecycle, clinical development teams protect the data chain of custody from the point of participant entry through to submission datasets. Further operational considerations for endpoint data reconciliation are detailed in our checklist for endpoint data before database lock and our guide to selecting COA types for protocol concepts of interest.