Skip to main content

Table 3 Difference analysis of the training set

From: Peritoneal cytology predicting distant metastasis in uterine carcinosarcoma: machine learning model development and validation

Features

0 (N=1930)

1 (N=474)

P

Peritoneal cytology

  

<0.001

 Negative

1562 (80.9%)

143 (30.2%)

 

 Suspicious

75 (3.9%)

21 (4.4%)

 

 Positive

293 (15.2%)

310 (65.4%)

 

Year of diagnosis

  

0.447

 2010-2013

595 (30.8%)

145 (30.6%)

 

 2014-2017

642 (33.3%)

171 (36.1%)

 

 2018-1021

693 (35.9%)

158 (33.3%)

 

Race

  

0.932

 Whites

1284 (66.5%)

308 (65%)

 

 Blacks

455 (23.6%)

117 (24.7%)

 

 Asian or Pacific Islander

178 (9.2%)

46 (9.7%)

 

 Native American/Alaska Native

13 (0.7%)

3 (0.6%)

 

Grade

  

<0.001

 G1

76 (3.9%)

4 (0.8%)

 

 G2

91 (4.7%)

10 (2.1%)

 

 G3

1763 (91.3%)

460 (97%)

 

T-7th

  

<0.001

 T1/T2

1517 (78.6%)

72 (15.2%)

 

 T3/T4

413 (21.4%)

402 (84.8%)

 

N-7th

  

<0.001

 N0

1473 (76.3%)

255 (53.8%)

 

 N1

281 (14.6%)

120 (25.3%)

 

 N2

176 (9.1%)

99 (20.9%)

 

Marital status

  

0.066

 married

1002 (51.9%)

269 (56.8%)

 

 others

928 (48.1%)

205 (43.2%)

 

Median household income

  

0.303

 <50000

126 (6.5%)

39 (8.2%)

 

 50000-90000

1169 (60.6%)

291 (61.4%)

 

 >90000

635 (32.9%)

144 (30.4%)

 

Urban-rural coding

  

0.872

 An urban area with a population of 1 million or more

1137 (58.9%)

290 (61.2%)

 

 Metropolitan areas with a population of 25-1 million

444 (23%)

102 (21.5%)

 

 Metropolitan areas with a population of less than 250,000

144 (7.5%)

31 (6.5%)

 

 Non-metropolitan area and adjacent

120 (6.2%)

31 (6.5%)

 

 Non-metropolitan area and not adjacent

85 (4.4%)

20 (4.2%)

 

 Time from diagnosis to treatment (day)

28(15-42)

23(8-36)

<0.001

 Tumor size

57(40-80)

75(55-100)

<0.001

 Age

67(61-73)

65(59-72)

<0.001