Skip to main content
. 2020 Sep 28;8:e10018. doi: 10.7717/peerj.10018

Table 3. Logistic regression models of the age group, gender and MEWS.

Model AIC Estimated coefficients Standard error P
Age Gender Gender* Age MEWS Age Gender Gender* Age MEWS Age Gender Gender* Age MEWS
1 118.01 1.24 0.18 - <0.001
2 116.03 1.16 1.24 0.58 0.18 0.044 <0.001
3 117.68 0.79 1.23 0.51 0.18 0.125 <0.001
4 116.54 0.96 1.24 0.53 0.19 0.071 <0.001
5 118.03 0.04 1.13 1.24 0.82 0.91 0.18 0.964 0.214 <0.001
6 118.52 0.50 0.80 0.34 1.24 0.92 0.66 1.12 0.19 0.584 0.228 0.764 <0.001
7 116.61 0.73 0.91 1.24 0.52 0.54 0.19 0.163 0.089 <0.001
8 116.81 0.65 0.84 1.24 0.59 0.64 0.19 0.272 0.189 <0.001

Note:

MEWS, modified early warning score. Model 1: MEWS; Model 2: Gender*Age + MEWS; Model 3: Age + MEWS; Model 4: Gender + eMEWS; Model 5: Age + Gender*Age + MEWS; Model 6: Age + Gender*Age + MEWS + Gender; Model 7: Age + Gender + MEWS; Model 8: Gender*Age + Gender + MEWS. Logistic regression models were developed against MEWS, age groups (‘1’ for 75 years old or above, and ‘0’ for 60–74 years old), gender (‘1’ for male and ‘0’ for female), and the interaction between age group and gender (age*gender).