“…An automated QC app based on the R Shiny framework can be an agile, versatile, and convenient solution to facilitate QC processes. In order to do so, such an app should perform the following typical tasks: - Matching the dataset to the data specification file 3 , 4 , 12 , 13
- Checking for missing or unavailable (NA) data entries:
- Confirming that the variables have correct data types and units
- Ensuring number of unique data values and their derivation matches data specification
- Data visualization to detect outliers and/or missing values 3 , 4 , 12 , 14 :
- Comparing protocol‐defined nominal times (NTs) with actual observation or event times relative to the first drug administration to detect aberrations
- Reviewing individual profiles to detect missing or anomalous data entries
- Checking covariate distributions and correlations
- Reviewing tabulated summaries to examine attributes of the dataset, such as number of participants per study, amount of missing or unquantifiable samples, and the central tendencies of covariates.
…”