Skip to main content

Table 3 The maximum likelihood estimates (standard errors in parenthesis) of the LC mixture analysis

From: A joint latent class model for classifying severely hemorrhaging trauma patients

Variable

Estimates (SE)

Model 1: linear regression model for latent-class structure

Intercept

−0.791 (0.613)

SBP <90

0.570 (0.618)

HR ≥120

0.554 (0.621)

pH <7.25

0.384 (0.590)

Hgb <9.0

0.574 (0.760)

 

SH

Non-SH

Model 2: class-specific linear models for RBC transfusion

Intercept

2.400 (0.427)

0.767 (0.138)

SBP <90

0.081 (0.261)

0.188 (0.101)

HR ≥120

−0.169 (0.294)

0.089 (0.102)

pH <7.25

0.216 (0.307)

0.132 (0.087)

Hgb <9.0

−0.046 (0.331)

0.105 (0.161)

Plasma:RBC (0 vs. (0,1])

0.682 (0.405)

0.942 (0.104)

Plasma:RBC (0 vs. >1)

−0.145 (0.516)

0.616 (0.111)

Platelet:RBC (0 vs. (0,1])

0.673 (0.313)

0.820 (0.095)

Platelet:RBC (0 vs. >1)

0.125 (0.404)

0.547 (0.174)

 

SH

Non-SH

Model 3: class-specific logistic models for traumatic death

Intercept

−0.797 (1.070)

−1.425 (1.279)

SBP <90

0.169 (0.817)

−1.732 (2.530)

HR ≥120

−0.384 (0.739)

0.451 (1.347)

pH <7.25

1.083 (0.830)

2.027 (1.185)

Hgb <9.0

0.079 (0.938)

−1.850 (3.283)

Plasma:RBC (0 vs. (0,1])

0.727 (1.244)

−1.131 (1.628)

Plasma:RBC (0 vs. >1)

−0.819 (1.468)

−1.837 (2.029)

Platelet:RBC (0 vs. (0,1])

−0.424 (1.056)

−0.044 (1.706)

Platelet:RBC (0 vs. >1)

−1.288 (1.627)

1.969 (1.874)