Skip to main content

Table 2 Missing data pattern in the training dataset. The percentage of missing data for a single proposed prognostic factor within the training dataset can be calculated by summing all the patterns that include that variable

From: The development and validation of prognostic models for overall survival in the presence of missing data in the training dataset: a strategy with a detailed example

Pattern

n (%)

Complete cases

1268 (68.47)

LDH

187 (10.10)

LDH and CRP

107 (5.78)

CRP

74 (4.00)

ISS

46 (2.48)

ISS and LDH and CRP

30 (1.62)

ISS and LDH

28 (1.51)

WHO PS

27 (1.46)

WHO PS and LDH

20 (1.08)

WHO PS and LDH and CRP

16 (0.86)

WHO PS and ISS and LDH and CRP

13 (0.70)

ISS and CRP

9 (0.49)

WHO PS and ISS and LDH

9 (0.49)

WHO PS and CRP

7 (0.38)

L:W and WHO PS and ISS and LDH and CRP

5 (0.27)

WHO and ISS

4 (0.22)

L:W

1 (0.05)

L:W and LDH and CRP

1 (0.05)

Age

0 (0.00)