Skip to main content

Table 4 Types of data extraction errors and their frequency of occurrence in the training group before and after training (for all NICE variables)

From: Training in data definitions improves quality of intensive care data

 

Frequency (%)

Data error type

Before training

After training

Difference

Incomplete

7.4

4.5

2.9*

Nonadherence to data definitions

11.4

7.9

3.5*

   Inclusion of values outside first 24 hours

3.8

1.3

2.5*

   Calculation error

2.8

2.3

0.5

   Other

4.8

4.3

0.5*

Other (errors that could not directly be accounted for)

2.1

1.3

0.8

Total inaccurate data

21.2

13.9

7.3*

  1. Sixty cases were assessed in total in the training group. * P < 0.05.