Pasek boczny

en:statpqpl:survpl:phcoxpl:hrpl

Hazard Ratio

An individual hazard ratio (HR) is now calculated for each independent variable : \begin{displaymath}
HR_i=e^{\beta_i}.
\end{displaymath} It expresses the change of the risk of a failure event when the independent variable grows by 1 unit. The result is adjusted to the remaining independent variables in the model – it is assumed that they remain stable while the studied independent variable grows by 1 unit.

The $HR$ value is interpreted as follows: \item $HR >1$ means the stimulating influence of the studied independent variable on the occurrence of the failure event, i.e. it gives information about how much greater the risk of the occurrence of the failure event is when the independent variable grows by 1 unit. \item $HR <1$ means the destimulating influence of the studied independent variable on the occurrence of the failure event, i.e. it gives information about how much lower the risk is of the occurrence of the failure event when the independent variable grows by 1 unit. \item $HR\approx1$ means that the studied independent variable has no influence on the occurrence of the failure event (1). Note

If the analysis is made for a model other than linear or if interaction is taken into account, then, just as in the logistic regression model we can calculate the appropriate $HR$ on the basis of the general formula which is a combination of independent variables.

EXAMPLE cont. (remissionLeukemia.pqs file)

en/statpqpl/survpl/phcoxpl/hrpl.txt · ostatnio zmienione: 2022/02/16 10:15 przez admin

Narzędzia strony