FDA’s New Business Rules Q&A

  1. If TS says that study started before the deadline and therefore is not required to have e-data, does this allow the study submission to go forward without DM and Define?
    Currently there is no information if the new eCTD Rejection Criteria will be applicable to all studies or only to studies that started after December 2016.
     
  2. For legacy clinical studies, we create SDTM datasets and ADSL only. Regarding the FDA's technical rejection criteria, is a define.xml for ADSL required in addition to the define.xml for SDTM datasets?
    Yes, the FDA eCTD Rejection Criteria requires a define.xml for both Tabulation (SDTM) and Analysis (ADaM) datasets.
     
  3. Since FDA no longer provides severity, will Pinnacle 21 tool no longer display severity?
    This topic is currently under active discussion. In the meantime, we can explain the source and various interpretations of Severity. FDA and PMDA interprets Errors and Warning as issues that have impact on review process, while Rejects are critical issues that prevent review. Pinnacle 21 has historically interpreted Errors as issues reported with 100% confidence, while Warnings are potential issues that require manual review. CDISC ADaM team has also proposed using Message Type instead of Severity, which aligns closely to Pinnacle 21 interpretation. So it's possible that will need to use both concepts in the future with Severity being simplified to just Reject or Not Reject. Stay tuned for further updates on this topic.
     
  4. For e-submission purpose, is SDTMIG v3.2 recommended? Because it looks like P21 allows to plug in higher version of SDTMIG.
    It's best to submit your data using the latest versions of standards accepted by the regulatory agencies. Currently, SDTMIG v3.2 is the highest accepted version. We recommend that you communicate with FDA/PMDA reviewers assigned to your submission regarding their expectations. Both agencies also encourage that sponsors submit Study Data Standardization Plan (SDSP) [link to https://www.pinnacle21.com/blog/how-prepare-study-data-standardization-plan-sdsp-fda] in advance to discuss details and clarify reviewers' expectations.
     
  5. Is there a document in Pinnacle21 website with a list of FDA validator rules that are not yet implemented into the software?
    It's not possible to have FDA Validator rules that are not yet implemented by Pinnacle 21, since FDA Validator rules are just a subset of Pinnacle 21 Validator rules that specifically implement the FDA Business rules. They are published by FDA for transparency.
     
  6. When will rejection criteria go into effect? Is the DM.xpt check required for all studies, or just studies with start date after 12/17/2016?
    The FDA have not given a date for when the rejection criteria will go into effect. FDA will give the industry 30 days’ notice on the eCTD website prior to the criteria becoming effective. TS will be required regardless of when the study was started and is specifically needed so that FDA can identify the study start date. For other required datasets (DM and ADSL) and define.xml, there is currently no information if the new eCTD Rejection Criteria will be applicable to all studies or only to studies that started after December 2016.
     
  7. When checking the variable names and labels, should these come from SDTMIG or SDTM? e.g. in SDTMIG v3.2 TI.IETESTCD has label='Incl/Excl Criterion Short Name e' whereas in SDTM v1.4 label is 'Inclusion/Exclusion Criterion Short Name'.
    If the variable is shown in the SDTMIG, then the label should come from the SDTMIG, otherwise the label comes from the SDTM. CDISC have addressed this IETESTCD label in the errata section [link to https://wiki.cdisc.org/display/PUB/SDTMIG+v3.2+Errata] on their wiki so that both TI.IETESTCD and IE.IETESTCD variables have the label 'Inclusion/Exclusion Criterion Short Name'.
     
  8. Do we need to include EPOCH in every domain to avoid rejection?
    Failure to include EPOCH is not a rejection criterion. However, it is stated in the Study Data Technical Conformance Guide that it is expected to be included in Observational Class domains. Of course, in some domains, like DM and MH, the EPOCH info is not useful or not applicable. The decision to include EPOCH is also study-specific. For example, if there are several screening events within the same study, EPOCH variable can help separate them and is expected to be included in Inclusion/Exclusion Criteria.
     
  9. FDA Business Rules and Validator Rules, which one is more important or as reference standards for eSUB at this time, year of 2017?
    The FDA Business and Validator rules work hand in hand. FDA Validator rules are just a subset of Pinnacle 21 Validator rules that implement the FDA Business rules. They are published by FDA for transparency.
     
  10. What do you mean users have 'not configured' MedDRA? Our data is 'mapped' to MedDRA. What extra 'configuration' step have we missed?
    In order for Pinnacle 21 Community to validate that your MedDRA variables do in fact hold values only from the version of the MedDRA dictionary you specify, you must configure Pinnacle 21 Community to use MedDRA. Please see further information here: https://www.pinnacle21.com/projects/validator/configuring-opencdisc-validator-external-dictionaries.
     
  11. What is the cost of the enterprise edition? What other services do you offer?
    Please contact us through our website www.pinnacle21.com.
     
  12. Max, what is your thoughts about placement of TS domain in the eCTD datasets folder if for a study legacy data is submitted and not CDISC? Also what if I am using 'DEMO' as the name for demographics dataset - do I need to rename it to 'DM' to be compliant with tech rejection criteria?
    The TS domain should be created and placed with your legacy datasets, even if they are not CDISC compliant. This allows the agency to automatically determine the start date of your study. You will need to change your demographics dataset from DEMO to DM.
     
  13. Approximately how long after CDISC publishes/releases a new CT before the Pinnacle 21 version of the CT file is available?
    This process is highly automated and we often release our version the same day as the CDISC version is released.
     
  14. Does FDA Business rules supersede CDISC Conformance rules?
    Given the fact that data is being submitted to the FDA, we put an emphasis on FDA Business rules over CDISC Conformance rules.
     
  15. What is the governance process for Business Rules at FDA? Is there public input?
    Questions about FDA Business rules and governance process should be sent to edata@fda.hhs.gov.
     
  16. Are all the FDA rules and CDISC rules covered in community version ?
    FDA Business rules are already covered by P21 Enterprise and Community. Many CDISC rules are also already covered by existing rules, and the remaining new rules are in the process of being implemented.
     
  17. Are FDA rejection and business rules already taken care of by Enterprise or are they being implemented in the November release?
    FDA Business rules are already covered in P21 Enterprise.
     
  18. Is SNOMED a requirement for FDA submission?
    From the FDA Study Data Technical Conformance Guide, section 6.6.1 SNOMED CT https://www.fda.gov/downloads/forindustry/datastandards/studydatastandards/ucm384744.pdf The International Health Terminology Standards Organization’s (IHTSDO) Systematized Nomenclature of Medicine – Clinical Terms (SNOMED CT) should be used to identify the medical condition or problem that the investigational product in a study is intended to affect (treat, diagnose or prevent, i.e., the indication). This information should be provided in the SDTM Trial Summary (TS) domain as a record where TSPARM=INDIC and TSPARM=TDIGRP. SNOMED CT was chosen to harmonize with Indication information in Structured Product Labeling (SPL). A reviewer should be able to take the indication term from product labeling and readily search for clinical or nonclinical studies of that indication without having to translate.
     
  19. What takes priority when Pinnacle 21 reports values as rejected in PMDA and error in FDA?
    It depends which agency you are submitting to. If you're submitting your data to FDA, then FDA might accept your study with an issue flagged as an FDA Error and PMDA Reject. If you're submitting to PMDA, then they will not accept your submission if it has any Reject issues until they are corrected.
     
  20. Does FDA require a combined SDTM database if we have one completed SDTM from parent study and ongoing rollover study? Or, we can submit separate SDTM databases if we have parent and rollover studies?
    This is something you will have to discuss with your review team at FDA.
     
  21. You said there is no severity for the validation messages now, so is it ok to drop the severity column in section 4.2 issue summary in the SDRG
    For now, we recommend leaving the severity column in your reviewer's guide as that is how Pinnacle 21 Community and Enterprise currently display the validation issues. Stay tuned for further updates on this topic.
     
  22. From end user perspective how the user will know whether applicable MedDRA versions are configured with P21 Enterprise?
    P21 Enterprise is a managed environment and is automatically configured with the latest versions of MedDRA. When user creates a data package for a study in P21 Enterprise, there is an option to select the applicable MedDRA version.
     
  23. Is PINNACLE 21 compatible for the submission in other countries, like MHRA, CDSCO, or EMA?
    Pinnacle 21 software is used by FDA and PMDA, which are the only regulatory agencies that currently require submission of study data in standardized format. When other countries begin accepting standardized data, Pinnacle 21 plans to provide compatible software.
     
  24. Should TS CTs be changed before submission if NCI CT new version comes or should it stay the same as used at the time of trial initiation?
    Your CDISC CT in the TS domain should be the same version of CDISC CT used throughout your other SDTM domains within a study.
     
  25. If the same human gets in a study, and sponsor assigned different USUBJID within the same study, do you think this conforms to the FDA Study Data Technical Conformance Guide? If not, do you have rule in place currently to report/detect this and if so how do you identify this?
    If a subject enrolls in the same study twice, then he/she must have the same USUBJID. The same is true about the whole submission. USUBJID value must be unique for the same subject across all studies within the submission. A new report in Enterprise is planned for later this year to detect and identify potential subjects that have re-enrolled.
     
  26. is Compatible versions of Community: FDA-2.2.0 comply with the SDTMiG version 3.2 ?
    The latest version of P21 Community is v2.2.0 and it is compatible with the following CDISC standards: SEND IG 3.0 SDTM IG 3.1.1 SDTM IG 3.1.2 SDTM IG 3.1.3 (this configuration also covers SDTM IG 3.1.2 Amend 1) SDTM IG 3.2 SDTM MD 1.0 (Medical Devices) ADaM 1.0 Define-XML 1.0 Define-XML 2.0
     
  27. If the SDTM has a length of USUBJID of 25 , but the maximum length of USUBJID is actually 19 does the validator give error/warning if we use length of 25 ? or it has to be reduced to 19 ?
    This scenario will violate FDA Business rule FDAB018 and will produce an Error. The variable length should be reduced to 19 to correct the issue.
     
  28. Should we be using the Pinnacle 21 version variable at the beginning of the trial or the most recent version prior to submission? obviously if we use the most up to date version once the trial has completed there it no way to fix conformance/business rules.
    You should always be using the most recent version of P21, as that is what the FDA and PMDA will be using to validate your data once it's submitted. After your study has been completed, there still may be SDTM/ADaM migration issues that could be updated. If a study locked 5 years ago but is only just now being migrated to SDTM, then the study should be migrated using the most current SDTM/ADaM standards and CDISC CT. If a study was converted to SDTM/ADaM and locked 5 years ago but is only just being submitted now, then it would be worth running the latest version of P21 to check if there are any updates that should be made to bring it up to the most recent standards. If you cannot fix issues in your data that were not being exposed in older versions of P21, then you should still validate your data using the most recent P21 version and document all findings in the reviewer's guide. Regardless of your studies age, when it locked, when it was migrated to SDTM/ADaM, the agencies will always be using the latest version of P21.
     
  29. Whenever I run Pinnacle 21 for SDTM/DEFINE.XML, I get an warning 'VSSTRESU value not found in Units for Vital Signs Results extensible codelist'. Units are generally extensible. Even then why this warning comes up? Could you clarify please. I get similar warning in Test name and Test code too.
    We cannot provide actual diagnostics without your define.xml file. You can email it to us if needed. However, most likely this message is due to a CDISC CT Codelist (UNIT) incorrectly assigned to VSSTRESU variable instead of expected (VSRESU) Codelist.
     
  30. Pinnacle 21 is quite reliable for SDTM validation but what about ADaM is it that reliable? As few companies are trying to develop SAS programs for ADaM and SDTM conformance is this a good idea? what do think about it?
    Yes, Pinnacle 21 software is reliable for ADaM validation. Both FDA and PMDA are using Pinnacle 21 to validate ADaM submission data. Pinnacle 21 currently support ADaM v1.0 and are working with the CDISC ADaM rules team to develop rules for ADaM v1.1.
     
  31. For nonclinical data it was indicated at the PhUSE CSS that if a study is submitted and it starts before the deadline it does not go to the next gate to check for DM and define as these files are not required for those study starting prior to the deadline.
    Currently there is no information if the new eCTD Rejection Criteria will be applicable to all studies or only to studies that started after December 2016. However, FDA did clarify in March 2017 that TS dataset will be required for all submission, so that they can automatically obtain the study start date.
     
  32. I just googled FDA Technical rejection and found a doc from Mar 2, 2017 with FDA rules code 1734, 1735, 1736, 1737. Your presentation showed 2 codes (1734, and 1736). So do you agree that there are now 4 FDA Technical Rejection Rules, not just two?
    1734 and 1736 have severity descriptions of 'High', while 1735 and 1737 have severity descriptions of 'Medium'. This webinar was to stress that if the two high severity rules were enforced the 46% of all submissions between December 2016 and March 2017 would have been rejected.
     
  33. FDA Business Rules have an ID of FDABxxxx, but in Pinnacle the FDA ID is FDACxxxx and there is no mapping between the two because the numbers do not match. Will the mapping be made available?
    The FDA Business Rules were only release in March 2017. We are currently working on a new version to account for the FDA Business Rules and the CDISC Conformance Rules which were released late January 2017.
     
  34. I am disappointed that a treatment emergent flag is still a SDTM business rule. It's a derived variable that clearly belongs in ADaM, not SDTM. Think about cases were AE start date is missing and needs to be imputed. Is this still a discussion with the FDA and CDISC ADaM team?
    We are not currently involved in discussions between the FDA and CDISC related to the treatment emergent flag. This is something that FDA has requested with FDAB001 business rule.
     
  35. For the CONDITIONAL rules that are not programmable, can P21 publish those under NON-Programmable category in the website ? Eg: V2.3 > 85 are 'Conditional' & the list of those
    The non-programmable checks identified by CDISC can be found on the CDISC website https://www.cdisc.org/standards/foundational/study-data-tabulation-model-implementation-guide-sdtmig/sdtmig-v32
     
  36. How do you configure SNOMED?
    First you'll need to submit a UMLS License request. This is free and will typically be approved on the same day. With a UMLS License you can download SNOMED. From there you can follow the steps on our website for configuring P21 Community and SNOMED. https://www.pinnacle21.com/forum/snomed-install
     
  37. Will you eventually be adding LOINC to the terminology? When?
    Yes, validation of LOINC Terminology will be available in the future. We are currently investigating technical requirements and will provide an update when this feature is added to our roadmap.
     
  38. Is Big Data submission a part of standardized SDTM/ADaM or there are other standards for it? like data from wearable devices, etc thank you
    So far, we are not aware of any initiatives to standardize Big Data in SDTM/ADaM format. However, Pinnacle 21's streaming validation engine is already well positioned to support Big Data, and we are looking forward to supporting it when it becomes available.
     
  39. In Pinnacle 21, the data fitness score will include all the new rules, does FDA look at this score to approve or reject?
    The FDA does not base rejection purely on the data fitness score. The score gives an overall quick indication of data package quality across different functional areas.
     
  40. Can you please elaborate meaning of community in pinnacle 21
    Pinnacle 21 Community is the open-source and freely available version of our software. Pinnacle 21 Community was previously called OpenCDISC.
     
  41. How do we validate FT tests, CC tests, and cats? They are not part of QS?
    These are not standard domains at the moment, and are not currently validated against controlled terminology.
     
  42. For FDA business rule 'FDAB007 All deaths should be an independent row in the adverse event dataset.', does this mean having death as an AEOUT value instead of a standalone 'Death' AE is not sufficient?
    A record with AEOUT = 'FATAL' and AESDTH = 'Y' is sufficient, there is no need to create another AE record with AETERM = 'DEATH'.
     
  43. What role will the FDA continue to have in the acceptance or 'approval' of the CDISC developed Standards Conformance Rules?
    There is currently no process for FDA acceptance of CDISC rules. Per the Study Data Technical Conformance Guide, Standards Development Organizations (e.g., CDISC) should develop and provide access to rules that assess conformance to its published standards.
     
  44. Hi, I just want to confirm, is a trial summary dataset required for all submissions (IND, NDA and so on)? Since when is the TS.xpt required? Is it also required for trials which were finished before Dec 2016? Thank you!
    TS is required for all submissions. The FDA made this a requirement in March 2017. Having the TS dataset allows FDA to automatically obtain the study start date.
     
  45. how will you uniquely identify the 115 validator rules? Will you use a subset of the parent number (FDAB005a, FDAB005b)?
    We'll provide traceability between the rules on our website [link to https://www.pinnacle21.com/validation-rules] and in our configuration files via the Publisher ID.
     
  46. Is there any plan PMDA to use the SDTM conformance rule in near future?
    Yes, PMDA is expected to start using the SDTM conformance rules in the future. Stay tuned for updates on this topic from PMDA.
     
  47. What is the timing for incorporating SENDIG v3.0 rules, since this is the only SEND version currently accepted by FDA?
    The CDISC SEND team is actively working on the conformance rules, however no timeline is currently available. In the meantime, FDA is using the SENDIG v3.0 validation rules that are implemented in Pinnacle 21 Enterprise and Community. Once CDISC SEND team completes their effort and publishes official validation rules, we plan to incorporate them into Pinnacle 21 software.
     
  48. TCG say SEND rules are coming soon. What to do until ? The link to the 'old' validation rules are gone ?
    Use the latest version of Pinnacle 21 software to validate SEND data. The same software and rules are used by FDA on non-clinical submission data.
     
  49. What is the timeline for Beta release?
    Beta is planned for release in August.
     
  50. We consistently get errors for TS, especially one for PCLAS for both TSVAL and TSVALCD. The values and codes we use come directly from NDF-RT. Can you explain why these continuously come up and what we should do to fix this in our TS dataset?
    It is impossible to answer this question without seeing the data. If there is suspected problem with the check, please post the explanation of why it is incorrect to the forum on P21 website.
     
  51. Is custom Domains can be validated through Pinnacle?
    Yes, Pinnacle 21's validation software will identify the class of custom domain and then apply the relevant validation checks to your dataset.
     
  52. What is the latest version for community?
    The latest version of P21 Community is v2.2.0. You can download a copy for free from https://www.pinnacle21.com/downloads
     
  53. With regard to the MedDRA and SNOMED compliance, did you take into account that the SENDIG (Nonclinical) users do not need to use these?
    Before validating your data, you select if you are validating define.xml, SDTM, ADaM or SEND. With SEND selected, the MedDRA and SNOMED validation rules will not be applied to your data.
     
  54. Is it planned to integrate WHODRUG checks?
    Yes, we are currently working on WHODrug integration and plan to have it included in the next major release.
     
  55. You mentioned Enterprise version is hosted and intended for team use. How does one team member access a validation run that was run by another team member?
    All users log onto a website where you can view validation reports for studies you have access to. Users can leave comments for certain issues, assign an issue to themselves or other users, close out issues, rate each issues risk, etc.
     
  56. Is it planned to improve checks for duplicates? E.G. to use a key from define.xml while checking
    Yes, a number of improvements to the duplicate check are under discussion. However, using the keys listed in the define.xml represents problems as many sponsors use sponsor-defined identifiers (that aren't explained in detail, and likely just contain system information), or the domain sequence number. The idea behind the check is to flag data that are duplicate in meaning, and using system variables or sequence numbers hide this issue.
     
  57. Would you please comment on the timing of the validation? Do you have to re-validate before submission if the data was already validated before the final Clinical Study Report?
    Yes, you must validate data with the most current version of the Pinnacle 21 validator. Re-validation of old studies before submission is also required to ensure that there are no Rejection issues. The FDA and PMDA will always use the most current version of the validator. If your data was validated and locked some time ago, then even if you do not intend on making any data updates it would still be benefitial to re-validate the data, so that all issues can be explained in the reviewer's guide. However, any Rejection issues in data must still be fixed.
     
  58. PMDA website says that they are compatible with P21 community v2.1.3. However they do NOT say 2.1.3 or ABOVE. This means folks in Japan think they cannot use community v2.2.0 - can you confirm that 2.1.3 or above is what the PMDA accept?
    We recognize that this has been a source of confusion for the industry and are working with PMDA to clarify the compatibility language. In the meantime, we recommend that sponsors submitting to PMDA use the latest version of both Enterprise and Community.
     
  59. While validating through Pinnacle which one is best either FDA or PMDA version?
    You must validate your data based on the agency you intend on submitting to. If you are submitting to both FDA and PMDA, then we recommend modifying section 4 Data Conformance Summary of your reviewer's guides to suit each agency.
     
  60. How dose Pinnacle 21 ensure the control terminology is correctly used in SDTM and ADaM dataset, how to handle multiple source of CT, SDTM, ADaM CT from CDISC, company defined CT in metadata, study specific CT.in study specification (define file)..
    SDTM and ADaM are validated separately and you enter the version of CDISC CT that you are using for each set of data. FDA and PMDA only recognize CDISC CT, so if you are extending CT and adding in sponsor or even study specific CT, it is up to you to determine if you are doing this correctly.
     
  61. When FDA rules and CDISC rules contradict (e.g. ARM/ARMCD) which should we follow while the issues are being ironed out?
    Given the fact that data is being submitted to the FDA, we put an emphasis on the FDA Business rules over the CDISC Conformance rules.
     
  62. Is define.pdf still a required component of the submission if we have define.xml?
    Define.pdf is only need if you use Define-XML v1.0. With Define-XML v2.0, the pdf version is no longer required.
     
  63. Have you plans to host Enterprise version in the cloud, so small players can subscribe and use the managed environment with reasonable cost?
    Yes, Pinnacle 21 Enterprise is a cloud-hosted application.
     
  64. Does the SNOMED differs from versions of MedDRA? Currently we are using 2016-09-01
    SNOMED and MedDRA are two completely separate dictionaries. There is no relationship between when new versions of each dictionary are released.
     
  65. Where can you find which version of Pinnacle 21 Enterprise FDA/PMDA are using currently?
    PMDA publishes the version of Pinnacle 21 Enterprise on https://www.pmda.go.jp/review-services/drug-reviews/about-reviews/p-drugs/0028.html. FDA does not currently publish the version, but they publicly state that they always use the latest version available.