Skip to main content
. 2022 Feb 22;11:41. doi: 10.1186/s13756-022-01076-0

Table 2.

Summary of the multivariate log–log models obtained in this study

Model adj.-R2 (%) Method Dependent variable ß0 Nurses DDD CPI log_Acpr intEAPD intEAlos intEAlos_H
M1.M 93.5 Step-wise forward selection log_CRGN (n = 30) 0.238Ø b − 0.085* 0.030* − 0.007** 0.549***
ß − 0.221 0.133 − 0.321 0.462
M2.M 93.8 Step-wise forward selection log_CRGN (n = 30) − 0.150** b 0.034** − 0.008*** 0.432*** − 0.114**
ß 0.148 − 0.350 0.364 − 0.270
M3.M 92.7 Forced entry log_CRGN (n = 30) 0.100Ø b 0.032* − 0.009*** 0.494*** − 0.519*
ß 0.141 − 0.382 0.416 − 0.202
M4.M 93.3 Forced entry log_CRGN (n = 27) 0.567Ø b 0.033* − 0.011*** 0.443*** − 0.730*
ß 0.145 − 0.477 0.372 − 0.156

adj.-R2: adjusted coefficient of determination, ß0: intercept, b: regression coefficient, ß: standardized regression coefficient. Only variables with significant regression coefficients were retained by the respective model through step-wise forward selection (inclusion cut-off p < 0.05; exclusion cut-off p ≥ 0.10) unless otherwise indicated. Ø denotes p > 0.05, * denotes p < 0.05, ** denotes p < 0.01, *** denotes p < 0.001. Information on log-linear models M1.S, M2.S and M3.S are provided in the Supplementary Material, section ‘Model equations’