This summary of the video was created by an AI. It might contain some inaccuracies.
00:00:00 – 00:12:46
The video provides a comprehensive overview of creating and managing datasets essential for pharmacokinetic (PK) analysis, focusing primarily on the DPC (Drug Plasma Concentration) and DPP (Drug Plasma Parameters) datasets. Central to the discussion are the CDISC (Clinical Data Interchange Standards Consortium) standards recommended by the FDA for clinical trial data submission.
The process of deriving the ADPC (Analysis Dataset for Pharmacokinetics Concentrations) from SDTM (Study Data Tabulation Model) domains is elaborated, including the integration of various data sources such as the PC (Pharmacokinetics Concentrations) and ADSL (Analysis Dataset of Subject-Level) datasets. Key variables highlighted include `param` and `paramCD` for analyte identification, as well as several date and time variables critical for calculating PK parameters like relative time to dosing.
The video also delves into the handling of missing data, imputation techniques for concentration values, and the inclusion of flags for identifying records qualifying for analysis. The generation of PK parameters such as CMAX, TMAX, and half-life is facilitated through software tools like Phoenix and results in the creation of the ADPP (Analysis Dataset for Pharmacokinetics Parameters) dataset. This dataset, integrating key variables from ADSL, is essential for statistical analysis and deriving meaningful clinical insights.
Overall, the video emphasizes the meticulous process required to ensure data integrity and regulatory compliance in PK analysis, showcasing the importance of proper dataset creation and variable management.
00:00:00
In this part of the video, the presenter discusses the creation of DPC and DPP datasets essential for pharmacokinetic (PK) analysis. They begin with an overview of these datasets, then explain the PK process flow used at various stages of PK submission. The creation of the DPC dataset is detailed first, followed by the DPP dataset. The importance of specific variables within these datasets for PK analysis is highlighted. The presenter also mentions the FDA’s encouragement of using CDISC standards for clinical trial data submission, where SDTM domains serve as the source for atom datasets. The atom datasets, ADPC and ADPP, are derived from the corresponding SDTM domains and contain both collected and derived data to facilitate PK analysis and creation of tables, listings, and figures. Finally, the process of combining concentration data in CSV format with collected data to create the SDTM PC domain and subsequently deriving the ADPC dataset is explained, emphasizing the integration of treatment and demographic information.
00:03:00
In this part of the video, the process of creating a pharmacokinetics (PK) dataset, specifically the ADPC dataset, is discussed. The ADPC dataset is derived from the SDTM domain with additional information from ATSL, similar to the ADPP dataset. To create an ADPC dataset, the PC and IX domains (exposure data) are merged to retrieve trip, date, and time data, and then integrated with the ADSL dataset for general subject-level information. Derived variables address issues like missing data and values below certain thresholds.
The ADPC dataset is structured to submit all concentration results, with one record per subject, analyte, and time point. It contains variables such as treatment information from ADSL and PK analysis variables like relative time to dosing. Additionally, concentration measurements from the PC domain are included, with missing values imputed as zero. The dataset also includes flags to identify qualifying records for analysis.
00:06:00
In this part of the video, the speaker discusses various variables related to pharmacokinetic analysis within the ADP C dataset. Key variables include `param` and `param CD`, which describe the analytes and their numeric counterparts. Variables like `a DTM`, `a STD TM`, `a en DTM`, and `eval` capture the date and time associated with the analysis. Additionally, `a SD t TM` and `a e TM` are tied to the start and end times of analysis intervals. `a visit` and `a visit n` are derived from `visit` and `visit numb` and relate to PK concentrations referring to the same exposure.
The segment also explains the planned time points represented by `a DP t` and `ATP TN`, emphasizing the adjustment of pre-dose values to zero. It highlights the conversion of planned time points into hours (`PPT` and `a PPT`) and the use of relative time (`a real time`). Relative time calculations are based on a reference time (`e^x STD TC`) and the PC date (`pc dtc`), with relative times for pre-dose typically negative. Finally, `analysis record flags` are mentioned for selecting records for analysis, representing an index range from 0 to 1.
00:09:00
In this part of the video, the speaker discusses the evaluation and handling of pharmacokinetic (PK) parameters within datasets. The main point is that the `eval` variable typically reflects PC stress and is adjusted for various data handling techniques. If sample values fall outside the acceptable range, they can be flagged with a `crit` variable and excluded from the analysis. The `ad PC` dataset can be utilized in software like Phoenix to derive PK parameters such as C MAX, T MAX, and half-life, which populate the `sDTMPP` domain and merge with the `ADSL` dataset to create an `ADPP` dataset. This dataset is crucial for statistical analysis and includes essential variables such as `param`, `paramCD`, `eval`, and `evalC`.
00:12:00
In this part of the video, the speaker discusses the analysis of a flag variable “n LZ z FL,” which can range from 0 to 9. They explain that the analysis value is designated as “PP stres m” and is stored in a numeric format in the “eval” variable, with its character counterpart stored in “eval c.” The inclusion or exclusion of any pharmacokinetic (PK) parameters for the analysis depends on the criteria outlined in the wrapper protocol, managed using trip flags “Wyclef y FL” and “theta y FM,” where Y ranges from 0. The segment concludes with a thank you message for the audience’s attention.