ADaM, short for Analysis Data Model, transforms raw data into analysis-ready datasets, crucial for clinical research. While SDTM standardizes data, ADaM prepares it for analysis. Dive into the world of ADaM with me! Understanding its three key structures: 1. ADSL (Subject Level Analysis Dataset) ADSL: The backbone of analysis. 2. BDS (Basic Data Structure) ADLB, ADEX, ADEG, and ADVS: Building blocks for insightful analysis. 3. OCCDS (Occurrence Data Structure) ADCM, ADAE: Uncover critical occurrences for thorough analysis. To kickstart ADaM programming, essential documents include: 1. ADaM Implementation Guide ( ADaMIG ) 2. Protocol : The blueprint of study design and objectives. 3. CRF : Capturing crucial data points. 4. SAP : Guiding the analysis journey. 5. Mockup Shells : Visualizing dataset structures. 6. ADaM Specification : Navigating dataset creation. 7. Raw SDTM dataset : The foundation of analysis. In this project, armed with these documents, I started the journey by: 1. Prot...
Introduction
In this post, I will elucidate the concept of Date and Time imputation.
Date imputation involves the process of converting a partial date in ADCM, ADAE, and ADMH datasets to complete dates. This is necessary because the data are often collected based on subjects' memory, which can lead to incomplete entries. While in SDTM, the original data are retained, in ADaM, it becomes a critical step in data analysis.
In dealing with Date and Time imputation, there exist numerous rules, particularly with regard to SAP requirements for date imputation in studies. Typically, the start and end dates and times are imputed. Furthermore, We always convert the character date into numeric when programming for ADaM.
Before delving into this topic, I want to clarify that I haven't had exposure to time imputation yet. However, I assure you that I will provide insights on it as I gain exposure. Let's proceed with the discussion.
For Start Date (—STDTC)
- If the day is missing in the date (CMSTDTC⇒ 2007-03), the year and month is present, then impute the 1st day of the month i.e.,= ‘01’ (ASTDT⇒ 2007-03-01) and the imputation flag variable as ‘D’ (ASTDTF= D).
- If the month is missing in the date (CMSTDTC⇒ 2007-__-07), the year and day is present, then impute the first month of the year i.e.,= ’01’ (ASTDT⇒ 2007-01-07) and imputation flag variable as ‘M’ (ASTDTF=M).
- If the month and the day is missing in the date (CMSTDTC⇒ 2007), the year is only present, then impute the first day of the first month of the year i.e.,= ‘01-01’ (ASTDT⇒ 2007-01-01) and imputation flag variable as ‘M’ (ASTDTF= M).
- If the whole date is missing (CMSTDTC⇒ ‘ ’) then keep the treatment start date of the subject (2007-03-07) (depends on the SAP which date to mention) and the imputation flag variable to ‘Y’ (ASTDTF= ‘Y’).
NOTE: In some study, when whole date is missing, they usually give a specific date.
For End Date (—ENDTC)
- If the day is missing in the date (CMENDTC⇒ 2007-03), the year and month is present, then impute the last day of the month i.e.,= ‘either 28, 29, 30, 31’ (ASTDT⇒ 2007-03-31) and the imputation flag variable as ‘D’ (AENDTF= D).
- If the month is missing in the date (CMENDTC⇒ 2007-__-07), the year and day is present, then impute the last month of the year i.e.,= ’12’ (ASTDT⇒ 2007-12-07) and imputation flag variable as ‘M’ (AENDTF=M).
- If the month and the day is missing in the date (CMENDTC⇒ 2007), the year is only present, then impute the last day of the last month of the year i.e.,= ‘12-31’ (ASTDT⇒ 2007-12-31) and imputation flag variable as ‘M’ (AENDTF= M).
- If the whole date is missing (CMENDTC⇒ ‘ ’) then keep the treatment end date of the subject (2007-03-07) (depends on the SAP which date to mention) and the imputation flag variable to ‘Y’ (ASTDTF= ‘Y’).
Important
If this the first post you read, then i highly recommend to read my last blogs on this series, link is below down:
Thank You for Visiting :)
Comments
Post a Comment