|
|
 |
| Meditation with Cox Proportional Hazards |
 
|
| Message/Author |
|
|
|
|
Hello,
I used Mplus to analyze a basic mediation model with a survival outcome as Cox proportional hazards (X = continuous, M = continuous, Y = right-censored time-to-event). However, they now want me to provide evidence that proportionality assumptions have not been violated in this analysis and that the approach is valid. Does the standard output contain this information in some form, or is there a way to obtain this test in Mplus?
For more context on the issue, they pointed me to Lapointe-Shaw et al. (2018) https://doi.org/10.1186/s12874-018-0578-7, which states: “Cox Proportional Hazards (PH) regression is commonly used for such analyses, yet its use in mediation analysis poses some important challenges. The semi-parametric Cox model builds on proportionality of the hazards. Proportionality is violated when adding an additional (mediator) variable to a correctly specified Cox regression model. This addition could shift the baseline hazards up or down, rather than only altering the slope of the hazard function [19]. Statisticians term this phenomenon the “non-collapsibility” of the hazard ratio [20]. As a result, parameter estimates obtained with and without a mediator cannot be meaningfully compared as they might be in a linear model [21, 22].” (p. 3)
I’d appreciate any advice you might have. |
|
|
|
|
Take a look at Section 3.1
https://www.statmodel.com/download/lilyFinalReportV6.pdf
if using Model 4 you can formulate Model test:
alpha(i)=alpha(j)
or just follow the approach in Table 6. |
|
|
| Back to top |
|
|
