Skip to main content
. 2013 Dec 24;14:200. doi: 10.1186/1471-2296-14-200

Table 3.

Table showing mixed regression modelling of factors associated with CARE Measure score (model A) and with interaction effect (IE) of mask wearing on patients’ familiarity with the doctor (model B)

 
Best fitted model A without interaction
Best fitted model B with interaction
Variable Estimate SE P-value* Estimate SE P-value
Intercept#
35.76
1.07
 
35.64
1.06
 
MASK
-0.95
0.44
0.027
-0.76
0.45
0.0836
Age
0.42
0.33
0.201
0.42
0.33
0.1948
General Health
-1.51
0.34
<0.0001
-1.51
0.34
<0.0001
Knowing the doctor well
4.36
1.14
<0.0001
6.94
1.63
<0.0001
Consultation time
0.32
0.05
<0.0001
0.32
0.05
<0.0001
Disease:Old
-0.34
0.60
0.5704
-0.36
0.60
0.5948
Disease:Both
1.92
1.19
0.1130
1.82
1.18
0.1342
MASK-Knowing the doctor well
 
 
 
-4.91
2.22
0.0308
Variance
Between^
 
5.86 (2.42)
 
 
5.81 (2.41)
component (SD) Within   47.00 (6.86)     46.82 (6.84)

#Reference group is patient aged <45 years old, the consulting doctor didn’t wear mask, good self-reported health, didn’t know the doctor well, average consultation time is 7.66 minutes, the consultation was for new (acute) illness.

*P-values were obtained by permutation test on 10,000 permuted sample.

^The variance component of doctor random effect.