Development of the probability of return of spontaneous circulation in intervals without chest compressions during out-of-hospital cardiac arrest: an observational study

Table 1 Estimated parameter values of statistical models

Chosen linear mixed-effects model
Parameter	Estimate	95% confidence interval
β ₀	-5.224	[-5.263–5.184]
β ₁	-0.00601	[-0.00712–0.00490]
Standard deviation (U ₀)	0.594	[0.564 0.6324]
Standard deviation (U ₁)	0.00708	[0.00564 0.00889]
Correlation (U ₀, U ₁)	-0.201	[-0.366–0.0416]
Range	3.11	[1.64 5.88]
Nugget	0.520	[0.401 0.629]
Power	-0.598	[-0.805–0.391]
Logistic regression model
Parameter	Estimate	Standard deviation
α ₀	9.28	1.51
α ₁	2.26	0.315

Estimated model parameters of the chosen linear mixed-effects model and the marginal logistic regression model for the logslope to probability of return of spontaneous circulation (P _ROSC) mapping. β ₀ and β ₁ are the fixed-effects parameters. Standard deviation (U ₀), standard deviation (U ₁) and correlation (U ₀, U ₀) give the properties of the zero-mean bivariate Gaussian distribution of the random-effects terms. Range and nugget are parameters in the model for correlation between residuals. Power is the parameter in the variance function. α ₀ is the intercept in the logistic regression models linear function, and α ₁ represents the change in log-odds ratio (ROSC versus no-ROSC) with a unit change in the covariate (logslope).

ISSN: 1741-7015