Skip to main content

Table 3 Binary logistic regression model

From: Drain fluid amylase as a predictor of postoperative salivary fistula in cases with benign parotid tumours

 

β

P

OR

Reconstruction of surgical defect (1) *

 

1.000

 

Reconstruction of surgical defect (2) *

−19.794

0.999

0.000

Reconstruction of surgical defect (3) *

24.441

1.000

41,176,069,965.360

Parotid masseter fascia’s persistence (1) *

 

0.894

 

Parotid masseter fascia’s persistence (2) *

2.661

0.635

14.308

Parotid masseter fascia’s persistence (3) *

−28.638

1.000

0.000

Surgical procedures (1) *

 

0.478

 

Surgical procedures (2) *

−5.313

0.391

0.005

Surgical procedures (3) *

−6.522

0.328

0.001

Surgical procedures (4) *

1.588

0.786

4.895

Drainage volume

−0.011

0.714

0.989

Postoperative pressure time

1.367

0.039

3.923

DFA**

0.129

0.035

1.138

constant

−25.303

0.044

0.000

  1. *: The Numbers in brackets represent dummy variables
  2. **: The unit of DFA is KU/L = 1000 U/L
  3. β: partial regression coefficient; OR: odds ratio; α = 0.05;
  4. The model accuracy is 89%