Skip to main content

Table 4 Multivariate regression analysis of factors associated with Knowledge Test %Correct and Confidence Scale Score, and logistic regression analysis or factors associated with the Communications Practices Scale Score.

From: Expertise about herbs and dietary supplements among diverse health professionals

 

Variables in the Final Model for Knowledge Test (%Correct) N = 1268(Score range 0–100)

 

Variables in the Final Model for Confidence ScaleN = 1201(Scale Range 19–95)

 

Variables in the Final Model for Communications Practices ScaleN = 852 *(Scale Range 0–10)

 
 

Coefficient (95% CI)

P value

Coefficient (95% CI)

P value

Odds Ratio (95% CI)

P value

Constant

58.4 (56.6, 60.2)

<.001

60.0 (57.8, 62.2)

<.001

Type of Health Professional‡

.013

Professional Group [Students are reference group]

      

   Physician/PA (yes = 1, no = 0)

6.1 (4.8, 7.5)

<.001

3.6 (1.6, 5.7)

.001

2.6 (1.3, 5.1)

.007

   Nurse (yes = 1, no = 0)

 

NS

 

NS

 

NS

   Nutritionist: (yes = 1, no = 0)

3.8 (2.1, 5.5)

<.001

5.3 (2.4, 8.1)

<.001

3.2 (1.5, 4.4)

.003

   Pharmacist (yes = 1, no = 0)

6.6 (4.0, 9.1)

<.001

    

Had seen patients in past 30 days (yes = 1, no = 0)

2.8 (1.6, 3.9)

<.001

  

[subjects selected on this basis]

 

Enrollment period (spring = 1, fall = 0)

-2.7 (-3.9, -1.4)

<.001

-18.3 (-20.4, -16.3)

<.001

  

Gender (Female = 1, Male = 0)

1.5 (0.24, 2.8)

.002

    

Training Status (Faculty/In practice = 1, In Training = 0)

      

Number of HDS Used in a Typical Week

0.46 (0.38, 0.54)

<.001

0.43 (0.29, 0.58)

<.001

1.11 (1.08, 1.14)

<.001

Age [age ≤ 30; represented by 0 in all other categories]

     

.015

   Age 31–40 years (yes = 1, no = 0)

3.8 (2.3, 5.3)

<.001

  

2.4 (1.3, 4.4)

.006

   Age 41–50 years (yes = 1, no = 0)

2.4 (1.1, 3.7)

<.001

4.8 (2.6, 5.7)

<.001

1.9 (1.0, 3.4)

.048

   Age >50 years (yes = 1, no = 0)

  

2.4 (0.15, 4.7)

.037

  

R-square for the final model

.251

 

.267

 

.155

 
  1. * This analysis was performed for enrollees who had seen patients in the past 30 days.
  2. † Models were determined using backward conditional multiple regression, with P≤ .05 required for variable retention.