| Variable | Comment |
| ASCNSD01 | This variable is only populated when AEACN='MULTIPLE'?. |
| These would be direct copies from SDTM SUPPAE variables, right? Would it be sufficient/better to keep the SUPPAE variables instead of creating an ADaM variable? Could a suggestion be made for the SDTM SUPPAE variable names instead of adding an ADaM variable. e.g. SUPPAE.AEACNSzz. | |
| Would prefer that XX notation be used to allow for multiple drugs to be referenced. | |
| AEACNSD01 | If there are multiple actions taken for a study drug, can all the actions be concatenated? For eg 'Dose reduced' 'Dose interrupted' for SD01.. |
| At GSK, we create a single variable AEACTTX to capture actions taken for each study drug - ACN : SD01 SD02, ACN2 : SD01 etc... | |
| AEACN | Can we put this variable optional when we have several drugs? (meaning we kept only the collected action taken for each drug - present in the SUPPAE dataset). |
AEDCOD AEBODSYS AEHLT AEHLGT | We typically include the numeric code variables for MedDRA terms for easier use in searches. Suggest adding them here. |
| AEDUR | Does the Agency have a duration calculation in mind that could be included? I've typically seen some form of end day - start day + 1 used, but whether to span the gaps between events for the same AE is a question. This may need to take into account the proposed GRPID now, e.g., earliest start date and latest end date. |
| In case of an available GRPID and more than one entry, is it expected to calculate the duration for every single entry or just having the overall duration for the corresponding grouped event?. | |
| AELLT | The MedDRA lower level term isn't included in the dataset. We typically include in the code/decode variables in our ADAE since we remap to later MedDRA versions using the LLT code. Suggest adding them here for completeness. |
AEONGO PREFL (Pre treat flag) FUPFL (Follow up flag) | Some of the flags which we use for analysis are missing. |
| AEOSI Category Flag | Add Safety Topic of Interest (STI) category flag - this is a broader category than just AEs of Special Interest. |
| AEOSIxxFL | AEOSIxxFL is not part of ADaM OCCDS IG v1.0. Could the agency clarify what the expected content of this variable is. Is it related to overdose/misuse, those defined in the protocol or those that might be identified by an SMQ?. |
| How does this flag relate to the ADaM CQzzNAM variable? CQzzNAM is the ADaM variable for AE of special interest category. Would FDA expect to see both, i.e. CQxxNAM containing the name of the AE category, and AEOSIxxFL that holds Y for that same category? if so, would xx in both variables need to match? Or is AEOSIxxFL a higher level categorisation of CQxxNAM?. | |
| The variable name is 9 chars; 'xx' should be 'zz' (per ADaMIG 3.1.1 #2 variable conventions); and the variable label is missing the '01'. Would the name of the category be put in the define comment?. | |
| For oncology, this can be used for immune-mediated AE's. May also consider adding AESIxxGR as perm for the grouping of AE's. For example, all the AE PT's considered immune-mediated pneumonitis will have the AESI01FL of Y, and AESI01GR="Pneumonitis". | |
| According to the spec: Create separate variable for each adverse event of special interest category. This will generate a lot of variables. At Merck, we have two variables to cover this, AEOSICT (AEOSI Category) and AEOSIFL (AEOSI Flag). | |
| AEREL | Note mentions additional variables should be created if this is collected by study drug. Suggest adding the placeholder variables like we have for action taken. |
| AERELS | Can there be only 1 AEREL variable and multiple related study treatments like AERDG1, AERDG2?. |
| This is conditionally required so assume this is only needed in the case the sponsor makes an assessment of relatedness (e.g. adjudication). | |
| AESEV | AESEV variable is missing. |
| AETOXGR | Does the Agency have an expectation regarding the CTCAE version to be used? Including some text may be helpful to sponsors. |
| AETOXGRN | Is this variable needed as this is not from SDTM and the analysis toxicity grade has a numeric equivalent?. |
AOCCFL AOCCRFL AOCCSRFL | Occurrence flags(AOCCFL-1st Occurrence within subject flag, AOCCRFL-1st Occurrence of Trt-Related TEAE within subject flag, AOCCSRFL-1st Occurrence of serious TEAE within subject flag, etc.) in ADAE are often useful to flag a unique record per subject, body system, Preferred Term, ATC code, or any other combination of variables needed for analysis. |
APERSDT APEREDT APERSDTM APEREDTM | These conditional variables should be added if APERIOD is included in ADAE. |
| APHASE | If ADAE has APERIOD added, APHASE should also be added else we would get P21 issue. |
ASTDTF AENDTF | For the partial ASTDT and AENDT, we might need to include imputation flag variables. |
ASTDTF AENDTF | Imputation flags are missing. |
ASTDY AENDY | Relative Day Variables are missing. |
| ATOXGR(N) | Is this required or should it be conditional? What would be the difference with SDTM AETOXGR? Could this be specified in the notes. |
| CONTRSP | How would are multiple treatments for a single Adverse Event expected to be represented as the Specific CMED or Additional Trtmnt Given variable (CONTRSP) only supports a single treatment and adding multiple variables would not be a preferred solution due to the unknown maximum number of treatments. |
| if there are multiple Conmeds given for the same AE, how to display the multiple Conmeds?. | |
| Can this be clarified? Would it contain coded value of conmed (CMDECOD)? When >1 med, should meds be concatenated?. | |
| When we have multiple treatments, do we concatenate the CMDECOD? And which delimiter is preferred? Or would we use CONTRSPx for each CMED?. | |
| There may be multiple treatments given. Suggest using 'zz', e.g. CONTSPzz. | |
| DECDORGw BDSYORGw HLGTORGw HLTORGw LLTORGw LLTNORGw | These original coding variables should be added considering studies which have a very long duration and could update coding over time. Also, studies being integrated will update to be consistent. This helps match analyses – values do differ, so summary counts on tables will be slightly different. It also impacts standardised queries which change with every version of MedDRA. |
| Dictionary Derived Term | Add SMQs, CMQs and other AE groupings. |
| DOSR01FL | Would prefer that XX notation be used to allow for multiple drugs to be referenced. |
| DTHFL | Should this be "ADTHFL" from ADSL instead of DTHFL from SDTM? The description in the last column matches the description of ADTHFL. |
| General | Strongly support this being a minimal set of variables and not to add other content as optional. |
| Does the Agency have an expectation regarding the MedDRA version to be used? In general, we harmonise the coding across studies using the latest version at the time of submission. This means we may be not be able to replicate AE analyses in CSRs that used an earlier version based on the integrated ADAE. Explanation is included in our ISS SAP and ADaM reviewer's guide. Including some text here may be helpful to sponsors. | |
| We typically include the AE coding dictionary name and version in our ADAE. Would it be necessary to add those variables or would documenting in the reviewer's guide be sufficient?. | |
| Occurrence Flags are missing. Please find the below examples for the details. AOCCFL-->1st Occurrence within Subject Flag AOCCSFL-->1st Occurrence of SOC Flag AOCCPFL-->1st Occurrence of Preferred Term Flag AOCC11FL-->1st Occurrence of any SAE Flag | |
| GRPID | How are records that belong to the same AE identified? Can we assume if same AEDECOD, then it's the same AE? In addition to same AEDECOD, the start date of second record should be equal to, or 1 day after the end date of the first record (so no overlap, except for the end/start date). Are those assumption the right ones to apply?. |
| Does the Agency have any guidance on the ideal way to link the AEs? Whether separate AE records are really the same event would require clinical judgment, especially with regards to the timing between events. This may be best handed through an indexing variable on the CRF based on the reported term as opposed to a programming algorithm. Giving sponsors more guidance on this would be helpful. | |
| In our EDC it would be difficult/messy to have the site enter a link number. We are considering deriving this programmatically, by looking for subject records with the same AEDECOD (preferred term) where the dates are continuous, i.e. the end date of one record is the same or one day prior to the start date of the next record. If there was a gap between the AE's, then the AE's would not be linked (because they will be considered separate events). Can you confirm if this logic meets your needs?. | |
| Is this variable required only when a linking variable is collected on the CRF? If sponsors do not collect linking information on the CRF is the expectation/requirement for the linking to be derived? If a derivation is expected, can more details be provided on how to define the "same adverse event" (e.g., AEs with same Preferred Term occurring within X days of one another). | |
| Should the AEs have the same GRPID when for example the same AE reoccurs after it was resolved earlier or should they just be summarised when they are also timely connected. We have for example the convention that only the same AETERM will be grouped and considering dates. Furthermore there is also a new GroupID when the severity of the AE changes. | |
| How is GRPID expected in regard to relatedness or seriousness. Should the same grouping logic apply here if these change or are different GRPIDs expected assuming same episode. | |
STUDYID SUBJID | Suggest adding the study ID and subject ID for completeness. Same comment applies to other datasets. I understand that these specs may not include all core variables, but I think it'd be better to include these as least so someone doesn’t take the specs too literally. |
| TREMzzFL | Does the 'zz' notation refer to treatment periods? If so, should this tie in with the notation 'xx' for e.g 'TRxxSDT', TRxxEDT etc.. |
| Should multi-drug treatment regimens also be included in the definition, e.g., create separate TRTEMFL flags for each component of the regimen?. | |
TRT01P TRT01A | Consider using TRTxxP/TRTxxA instead of a specific number so that it can be used by a number of treatments. General consideration for any variables listed using a zz/xx and in notes can state that at least 1 drug must be identified. |
| TRTxxA | Additional information states: "Required if there are multiple treatment periods; Use TRTA if only one treatment period is relevant for the analyses." The suggestion to use TRTA when there is only 1 treatment period seems questionable. Why not use TRT01A?. |
TRTEMFL TREMzzFL | Can these be combined into one variable with different data values corresponding to the first drug treatment and subsequent therapies? The combine variable will not be a ~FL variable then. For example, the values could be "TRTEM1", "TRTEM2"..."TRTEMxx". |
| TRTEMFL | The FDA has specified a definition of treatment emergent adverse events in this guidance. Would the agency consider alternative definitions of treatment emergent adverse events aligned with the study protocol?. |
| Why in the general rules is SPONSOR defined relatedness mentioned? INVESTIGATOR defined relatedness does not matter?. | |
| In the definition of treatment emergent AEs the text 'up to and including 30 days after last dose of study drug'. Is this fixed or can it be updated to the days as specified in the protocol/analysis plan?. | |
| In the definition of treatment emergent AEs the text up to and including 30 days after last dose of study drug or the day prior to start of subsequent therapy (whichever comes first)'. We do not take into account the further therapy. Is this a fixed definition or can it be updated at sponsor level ? New or Worsening events : Worsening mean increasing NCI grade and/or become serious : Is it ok? | |
| The proposed TEAE definition includes AEs starting >30 days after the last dose that were determined by the sponsor to be related to drug. Currently we only determine TEAEs via programming algorithms. Including sponsor relatedness would require additional clinical review in the process and I'm not sure how easily this could be implemented. | |
| The text in the general rules uses "drug treatment" and "study drug". Suggest we use one term for clarity. | |
| Suggest adding text to clarify the Agency's TEAE definition for multi-drug combinations. Would it be the first dose of any component of a treatment regimen and 30 days after the last dose of any component?. | |
| For studies that require a treatment-emergent definition that differs from the definition in the OOD Standard Data Request document, please clarify if Sponsors should perform additional safety analyses using the OOD standard data request definition and include variables to store both (study specific and OOD standard) treatment-emergent definitions. | |
TRTEMFL GRPID | If I understand this concept correctly, once a record within a grouped AE is flagged as a TEAE using the proposed window, any subsequent record should also be a TEAE regardless of the severity and when it starts relative to the window. Is this correct?. |