Your Easy Guide to Clinical Studies

A Data Quality Check (DQC), also called statistical data validation, is a process used by statisticians to ensure that data used for the statistical analysis is of high quality.

 

The DQC is analogous to Central Data Monitoring (CDM), and is systematically performed by a statistician prior to data analysis.

 

Depending on the scope of data to be validated, DQC may include checks for:

• Completeness of data (e.g. extent of missing data)
• Consistency of dates (e.g. baseline visits are conducted after informed consents have been signed and prior to 3-month study visits)
• Identification of duplicates
• Range of values (e.g. plausibility checks regarding blood values and age range, identification of unexpected outliers)
• Consistency between variables (e.g. a pregnant men or child).

As a SP-INV:

• Plan together with a statistician (prior to data analysis) the implementation of respective statistical validation checks (i.e. the assessment should be risk-based)
• Ensure potential incorrect data identified during DQCs are properly investigated and corrected (e.g. outliers such as a participant born in 1862, blood pressure of 10/40)

 

DQCs performed by a statistician can also be planned and performed during the conduct phase of a study.

References

ICH Topic E9 – see in particular

• 5.2.1 Full analysis set
• 5.2.2 Per protocol set