Table 4.

Piecewise linear regression models with mortality rate in 434 Norwegian municipalities as dependent variable and population density and travel time to hospital care in minutes as predictors

	Univariate models				Multiple model 4
Explanatory variable	Estimate (95% CI)	p Value	AIC	Radj2	Estimate (95% CI)	p Value	AIC	Radj2
log (population density)			3402.2	0.242			3397.1	0.252
Breakpoint	0.81 (0.17 to 1.44)				0.79 (0.08 to 1.51)
Below break	−13.9 (−23.5 to −4.32)	0.005			−12.3 (−21.8 to −2.7)	0.012
Above break	−2.02 (−2.45 to −1.59)	<0.001			−1.77 (−2.33 to −1.20)	<0.001
log(travel time)			3448.4	0.157
Breakpoint	2.32 (2.05 to 2.59)				2.37 (1.93 to 2.80)
Below break	−5.45 (−10.5 to −0.37)	0.036			−4.82 (−9.61 to −0.03)	0.054
Above break	5.80 (4.12 to 7.49)	<0.001			2.14 (0.28 to 4.01)	0.024

Weighted by municipality population.

AIC, Akaike's information criterion.