Je veux savoir à propos de…

jbell on DSKJLSW7X2PROD with RULES

Federal Register / Vol. 86, No. 3 / Wednesday, January 6, 2021 / Rules and Regulations science. This final rule does not preclude the public from engaging subject matter experts to determine whether a study can be independently validated. Also, the definition cannot be considered solely in isolation. The regulatory text in which the term is used informs the extent of the availability of dose-response data underlying studies.
Specifically, 40 CFR 30.5 requires, in part, that the dose-response data underlying studies that the EPA will consider as pivotal science be available in a manner sufficient for independent validation. Scientific information is considered available in a manner sufficient for independent validation when it includes the information necessary to understand, assess, and reanalyze findings. The efficacy of the reanalysis will depend on the expertise of the person conducting the reanalysis.
One commenter noted that the term reproduced in the proposed definition of capable of being substantially reproduced and the use of capable of being substantially reproduced in the proposed definition of independent validation, were inconsistent with the description of reproduce in the 2020 SNPRM preamble and the NAS Workshop Report Ref. 26.
The commenter contended that this adds confusion. Another commenter asserted that there is insufficient guidance or standards for what the term substantially means or who will make the determination e.g., scientific staff with oversight of an EPA scientific advisory panel. Another commenter stated that there were inconsistencies with the proposed definitions for the terms capable of being substantially reproduced and reanalyze.
Commenters asserted that the former proposed definition specifies the use of identical methods, whereas the latter proposed definition specifies the use of the same or different methods.
The EPA finds that these comments have merit. The EPA is modifying the definition of independent validation in the final rule by replacing capable of being substantially reproduced with produced. The EPA will not finalize the proposed 40 CFR 30.2 definition of capable of being substantially reproduced because the term is not used in the final rules definition of independent validation or elsewhere in 40 CFR 30. As a result, substantially will not need to be defined or described in the final rule.
The EPA is also modifying the definition of reanalyze to specify the use of the same methods because as proposed it specified the use of the same or different methods. This change was made so that the definition
VerDate Sep<11>2014

16:19 Jan 05, 2021

Jkt 253001

would be consistent with the final rules definition of independent validation.
2. Data and models. In the 2020
SNPRM, the EPA proposed a definition of data in response to comments on the 2018 proposed rule, contending that a definition for this term was needed to clarify the applicability of the rulemaking. Commenters requested that the EPA clarify which stage of data would need to be available to allow for independent validation. The stage of data that the EPA identified in the proposed 40 CFR 30.2 definition of data is based on the discussion of the different stages of data in the NAS
Workshop Report Ref. 26. The 2020
SNPRM adapted the description of the stage of data from the NAS Workshop Report Ref. 26 that was data at the appropriate level of detail to allow for independent validation via reanalysis.
Several commenters asserted that the proposed definition of data was so broad that it could include potentially any information. One commenter contended that as published scientific results are often the final steps in a process involving several processing and analysis steps, the proposed definition of data definition did not identify what intermediate step of data processing would be subject to this rule.
The commenter noted that determining which of the multiple data processing and analysis steps that should be used would differ from study to study.
Another commenter suggested that the EPA should identify the actual final dataset used in statistical analysis as the appropriate stage of data to be made available.
As the EPA described in the 2020
SNPRM, there are different stages of data. The EPA presented the different stages described in the NAS Workshop Report Ref. 26, There are raw data, which come straight from the survey or the experiment. There are cleaned-up data, which consist of the raw data modified to remove obvious errors.
These are the data that are ready to be analyzed to extract relevant information. There are processed data, which are data that have been computed and analyzed to extract relevant information. There is the final clean data set that is provided with a publication. Since the purpose of 40
CFR 30.5 is to determine the consideration to afford to studies based on, among other factors, the availability of the underlying dose-response data that would support independent validation via reanalysis of the data underlying pivotal science, the appropriate stage of data would not be the processed data data that have been computed and analyzed to extract
PO 00000

Frm 00041

Fmt 4700

Sfmt 4700

479

relevant information or the final clean data set that is provided with a publication. At these two stages of data, the analysis has already been conducted, and the results have already been determined. In order to determine if these results are valid, data that had not already been computed and analyzed are needed.
In this final rule, the EPA is not identifying a specific step in a multistep analysis as the stage of data that would be sufficient for independent validation through reanalysis because this would be overly prescriptive and not informative. As noted by commenters, the step at which the final clean data set will be generated will vary from study to study. The level of detail required would be that needed for a separate party to reanalyze the study.
The appropriate step is where the data are ready to be analyzed to extract relevant information.
One commenter requested that the EPA introduce and define a new term, validated data, which are the data with the proper level of quality assurance. While the EPA routinely conducts quality assurance to ensure that data are acceptable for use, the EPA
does not see the need to create a separate definition. The focus of this rulemaking is the independent validation of the results of studies underlying pivotal science, not the quality assurance of the data itself.
Some commenters contended that the EPA should define data as the raw data in which obvious errors have not been removed. Other commenters stated that raw data in which obvious errors have not been removed would result in skewed analyses for third parties not familiar with the data collection process. Given concerns about potentially skewed analyses, the final definition of data maintains the stage of data in which obvious errors have been removed.
Some commenters also requested that the EPA define model to clarify the applicability of the rulemaking. In the 2020 SNPRM, the EPA proposed a definition of model at 40 CFR 30.2, but the Agency is not finalizing the definition of model because this regulation applies only to dose-response data see Section III.B of this preamble.
3. Dose-response data. In the 2018
proposed rule, the EPA proposed a definition of dose-response data and models. The EPA did not receive significant comment on the definition of dose-response data and models itself.
However, as discussed in Section III.B
of this preamble, this final rule applies to dose-response data, and thus the EPA
is not finalizing a definition for dose-

E:FRFM06JAR1.SGM

06JAR1