Preparing and Submitting RMIP study data to BDC – Phase 5: Study Ingestion and Documentation
C
Caryn Willis
started a topic
40 minutes ago
The final step in the process is study ingestion, in which the prepared study data are made available for responsible data sharing with authorized users. During this step, the data are loaded into the BDC repository, and the relevant documentation and metadata are provided to supply the context needed for findings to be interpreted appropriately.
Final Review
After the study data has been name, organized, and harmonized in alignment with the Data Management Plan, the data is uploaded to an AWS cloud bucket. The DMC has scripts to programmatically transition the data from the investigator’s BDC project to the BDC Data Repository. The DMC will assist with the data submission process when that stage is reached. Following upload, investigators will asked to perform a final review.
Ingestion QC
Once the final review is complete, the BDC ingestion team will perform quality control (QC) checks on the submitted data. If any issues are identified, the study team will need to correct them and reupload any affected files. QC checks include:
Ensure that all file and directory names do not contain spaces or special characters.
Ensuring that all study variables are described within a data dictionary.
Ensuring that all data are associated with a subject ID that is included in the dbGaP sample telemetry report.
Ensuring that any participant ages >89 are reported as the abstracted value “>89.”
Ensure that there are no empty files.
Study Documentation
Supporting documentation is a valuable resource for understanding the data and its generation so that results can be interpretable appropriate and in the appropriate context. It is recommended that any supporting documents be provided as part of the data package including:
A summary document: study overview, components, and the examination and follow-up timeline
Protocol and manual of procedures (MOP), with version history
Annotated data collection forms, for every form version
A README describing dataset organization and the subject ID and age variables in each file, including relationships among subfolders
Description of de-identification methods for clinical and imaging data
The visit schedule table
Algorithms for derived variables and descriptions of any recoding
Outcome definitions and adjudication criteria
Rename maps and release notes
Additional supplemental documentation needed to reproduce study results
A list of study publications
Data or metadata model, such as OMOP, if applicable
Subject consent files (2a/2b in the dbGaP submission process)
Machine-readable BDC-compliant data dictionaries for clinical and image data
Compressed and archive file formats, such as ZIP, GZ, or TAR, with no nested archives. The README may describe the directory structure.
These files should be uploaded to the supporting documentation directory on the BDC project.
Study Metadata
Metadata, or data about data, helps make a study discoverable to the broader research community. As such, the DMC has a defined set of metadata that are requested for each study. These include approximately 10–20 required metadata fields and up to 110 optional fields. The DMC is currently developing a form through which this information can be supplied.
Caryn Willis
The final step in the process is study ingestion, in which the prepared study data are made available for responsible data sharing with authorized users. During this step, the data are loaded into the BDC repository, and the relevant documentation and metadata are provided to supply the context needed for findings to be interpreted appropriately.
Final Review
After the study data has been name, organized, and harmonized in alignment with the Data Management Plan, the data is uploaded to an AWS cloud bucket. The DMC has scripts to programmatically transition the data from the investigator’s BDC project to the BDC Data Repository. The DMC will assist with the data submission process when that stage is reached. Following upload, investigators will asked to perform a final review.
Ingestion QC
Once the final review is complete, the BDC ingestion team will perform quality control (QC) checks on the submitted data. If any issues are identified, the study team will need to correct them and reupload any affected files. QC checks include:
Ensure that all file and directory names do not contain spaces or special characters.
Ensuring that all study variables are described within a data dictionary.
Ensuring that all data are associated with a subject ID that is included in the dbGaP sample telemetry report.
Ensuring that any participant ages >89 are reported as the abstracted value “>89.”
Ensure that there are no empty files.
Study Documentation
Supporting documentation is a valuable resource for understanding the data and its generation so that results can be interpretable appropriate and in the appropriate context. It is recommended that any supporting documents be provided as part of the data package including:
A summary document: study overview, components, and the examination and follow-up timeline
Protocol and manual of procedures (MOP), with version history
Annotated data collection forms, for every form version
A README describing dataset organization and the subject ID and age variables in each file, including relationships among subfolders
Description of de-identification methods for clinical and imaging data
The visit schedule table
Algorithms for derived variables and descriptions of any recoding
Outcome definitions and adjudication criteria
Rename maps and release notes
Additional supplemental documentation needed to reproduce study results
A list of study publications
Data or metadata model, such as OMOP, if applicable
Subject consent files (2a/2b in the dbGaP submission process)
Machine-readable BDC-compliant data dictionaries for clinical and image data
Compressed and archive file formats, such as ZIP, GZ, or TAR, with no nested archives. The README may describe the directory structure.
These files should be uploaded to the supporting documentation directory on the BDC project.
Study Metadata
Metadata, or data about data, helps make a study discoverable to the broader research community. As such, the DMC has a defined set of metadata that are requested for each study. These include approximately 10–20 required metadata fields and up to 110 optional fields. The DMC is currently developing a form through which this information can be supplied.
Additional Resources
NHLBI Instructions for Preparing Clinical Research Study Datasets for Submission
NHLBI Data Sharing FAQs