Table 4

 Multivariate analysis of factors in association with perceptions and behaviours that may be conducive to the spread of SARS in mainland China

Model 1: dependent varible = would not consult doctors if having influenza-like symptoms in China*Model 2: dependent variable = not wearing masks in public places if having influenza-like symptoms in China†Model 3: dependent variable = wearing masks in Hong Kong but not wearing masks in China if having influenza-like symptoms‡
ORm95% CIp ValueORm95% CIp ValueORm95% CIp Value
ORm, multivariate odds ratio, derived from multiple logistic regression with forward stepwise selection of univariately significant variables listed in table 5; *1 = yes and 0 =  no; †1 = yes and 0 = no; ‡1 = yes and 0 = no.
Education attainment
⩽9 years1.0
10–11 years0.570.29 to 1.110.099
12–13 years0.690.35 to 1.380.293
College or university0.320.16 to 0.670.002
Duration of stay in the travel destination (number of nights)
Day trip7.823.90 to 15.66<0.0010.460.19 to 1.140.094
1–73.642.07 to 6.40<0.0011.370.61 to 3.340.445
Travelled to Shenzhen
Yes0.670.46 to 0.980.040
Travelled to other provinces
Yes0.220.09 to 0.520.001
Perceived risk of Hong Kong residents contracting SARS in Shenzhen
Moderate/low/very low0.520.33 to 0.810.004
High/very high1.0
Perceived risk of Hong Kong residents contracting SARS in Guangzhou
Moderate/low/very low2.351.41 to 3.900.001
High/very high1.0
Availability of medicine to cure SARS
Not available2.461.38 to 4.380.002