Table 6 Univariable and multivariable logistic regression.

From: Neurosurgeons’ knowledge, attitudes, and practices regarding 3D-printed guide plate-guided puncture and drainage surgery for intracerebral hemorrhage

Cutoff value: ≥24/<24

No.

Univariable

Multivariable(forward, P<0.1)

Multivariable(forward, P<0.25)

OR(95%CI)

P

OR(95%CI)

P

OR(95%CI)

P

Gender

 Male

92/175

ref.

     

 Female

27/48

1.160(0.610,2.206)

0.651

    

Age groups (years)

 ≤ 35

32/71

ref.

     

 35–40

30/63

1.108(0.561,2.188)

0.768

    

 40–45

20/38

1.354(0.615,2.984)

0.452

    

 > 45

37/51

3.221(1.488,6.974)

0.003

    

Education [adjusted]

 Bachelor’s degree and below

90/160

ref.

     

 Master’s and above

29/63

0.663(0.369,1.192)

0.170

    

Professional title

 None

6/16

0.453(0.152,1.350)

0.155

    

 Junior

14/28

0.755(0.324,1.760)

0.515

    

 Intermediate

46/86

0.868(0.481,1.566)

0.638

    

 Senior

53/93

ref.

     

Hospital grade [adjusted]

 Secondary hospital

20/43

ref.

     

 Tertiary grade A hospital

38/83

0.971(0.464,2.033)

0.938

    

 Tertiary grade B hospital

48/75

2.044(0.954,4.383)

0.066

    

 Other

13/22

1.661(0.587,4.699)

0.339

    

Years of work experience [adjusted]

 ≤ 5 years

15/33

ref.

     

 5–10 years

17/41

0.850(0.337,2.143)

0.731

    

 10–15 years

35/62

1.556(0.665,3.637)

0.308

    

 >15 years

52/87

1.783(0.795,4.000)

0.161

    

Participated in training related to the puncture and drainage surgery for ICH guided by 3D-printed guide plate

 Yes

93/120

ref.

 

ref.

 

ref.

 

 No

26/103

0.098(0.053,0.182)

< 0.001

0.273(0.125,0.598)

0.001

0.273(0.125,0.598)

0.001

Applied puncture and drainage surgery for ICH guided by 3D-printed guide plate

 Yes

73/86

ref.

     

 No

46/137

0.090(0.045,0.179)

< 0.001

    

Number of 3D-printed guide plate-guided surgeries performed [adjusted]

 0

42/132

ref.

 

ref.

 

ref.

 

 1–10

41/52

7.987(3.737,17.072)

< 0.001

3.532(1.405,8.876)

0.007

3.532(1.405,8.876)

0.007

 > 10

36/39

25.714(7.490,88.276)

< 0.001

8.891(2.275,34.740)

0.002

8.891(2.275,34.740)

0.002

Knowledge score

 < 37

32/107

ref.

 

ref.

 

ref.

 

 ≥ 37

87/116

7.031(3.898,12.682)

< 0.001

2.320(1.105,4.872)

0.026

2.320(1.105,4.872)

0.026

Attitude score

 < 37

34/107

ref.

 

ref.

 

ref.

 

 ≥ 37

85/116

5.887(3.301,10.499)

< 0.001

4.475(2.183,9.174)

< 0.001

4.475(2.183,9.174)

< 0.001

  1. Further analysis of the effects between knowledge, attitudes, and practices was conducted using a structural equation modeling (SEM) approach, resulting in a model with a satisfactory fit index [Table 7]. The findings from the goodness-of-fit analysis indicated that the constructed model effectively explains the variation in the data, and there is a high degree of agreement between the model and the actual data. As shown in supplementary table 1, all factor loadings are greater than 0.5, indicating that each observed variable makes a substantial contribution to the corresponding latent variables (Knowledge, Attitude, Practice). The path coefficients among the latent variables (Knowledge, Attitude, Practice) are all greater than 0.3, confirming the strong relationships and validating the correlation analysis results. Bootstrap analysis results demonstrated that knowledge directly impacts attitude (β = 0.379, P = 0.010) and practice (β = 0.462, P = 0.010), attitude directly influences practice (β = 0.383, P = 0.010), and knowledge also has an indirect effect on practice (β = 0.145, P = 0.010) (Tables 8; Fig. 1).
Back to article page