Loading Tree…

DQI-3001

Definition

Observed numerical data values are not admissible according to the allowed ranges.

Explanation

This indicator checks for violations of admissibility ranges for numerical variables. In other concepts the applied range limits are also described as “hard limits”. In eCRF commonly such values are not accepted.

Example

In a somatometry examination body weight measures in adults must be positive and no higher than 250kg. Therefore, intervals of admissible values should be defined, e.g. [35; 250] for body weight.

The number of arterial veins is counted based on an eye-background examination. Values below 0 or above 50 are considered impossible and admissibility limits are defined accordingly.

Guidance

The challenge in identifying inadmissible values rests in finding adequate limits for checks for numerical values. In some cases a limit may be set due to a logical reason. For example a count cannot be below 0. Yet, commonly, the cut-off for the upper limit of admissible counts needs to take into account empirical knowledge about what may be possible or expectable within the specific context in which the study takes place.

Any violation of an admissibility rule triggers a data cleaning process with the intention to replace the wrong value by the correct one. If this is not possible affected observations should at least be flagged to be adequately handled during analyses.

Interpretation

The higher the number or percentage of inadmissible numerical values, the lower the data quality.

Implementations

Literature