Table 5 MD and MAD for each imputation method, averaged over 100 simulations

From: Comparison of methods for handling missing data on immunohistochemical markers in survival analysis of breast cancer

 

MCAR

MAR

 

CCA

MS

MI−

MI+

CCA

MS

MI−

MI+

MD

 Grade

−0.06

−0.06

−0.01

−0.03

−0.04

−0.08

−0.02

−0.03

 Nodal status

0.02

0.03

0.18

0.02

−0.02

0.07

0.20

0.02

 Tumour size

−0.01

−0.02

0.04

0.00

0.00

0.02

0.06

0.01

 ER status

0.02

0.14

0.05

0.01

0.01

0.17

0.06

0.02

 PR status

−0.09

−0.20

−0.27

−0.08

0.03

−0.22

−0.30

−0.05

 HER2 status

−0.01

−0.04

−0.03

0.04

−0.01

−0.01

−0.05

−0.01

 BCL2 status

0.02

0.04

−0.02

0.03

0.00

−0.09

−0.06

−0.01

MAD

 Grade

0.27

0.13

0.12

0.14

0.22

0.11

0.11

0.14

 Nodal status

0.20

0.12

0.18

0.11

0.20

0.20

0.20

0.12

 Tumour size

0.16

0.09

0.08

0.09

0.15

0.08

0.11

0.11

 ER status

0.25

0.17

0.12

0.14

0.23

0.20

0.14

0.15

 PR status

0.28

0.24

0.28

0.18

0.29

0.26

0.30

0.18

 HER2 status

0.28

0.14

0.10

0.14

0.22

0.15

0.12

0.16

 BCL2 status

0.24

0.14

0.11

0.15

0.23

0.19

0.14

0.18

  1. Abbreviations: BCL2=B-cell lymphoma 2; CCA=complete case analysis; ER=oestrogen receptor; HER2=human epidermal growth factor receptor-2; MAD=mean absolute difference; MAR=missing at random; MCAR=missing completely at random; MD=mean deviation; MI−=multiple imputation without the outcome; MI+=multiple imputation with the outcome; MS=mean substitution; PR=progesterone receptor.
  2. Numbers in bold indicate method with best result for that variable. Underlined numbers indicate method with worst result for that variable.
Back to article page