Sign in to follow this  
Followers 0
DrKkhan

survival study dtatistical tests

3 posts in this topic

What is the difference between log rank, cox proprtional hazard and mantel haenszel and when to use these three as a staistical test in survival studies??

Share this post


Link to post
Share on other sites

log rank test is univariate,non parametric, analysis the difference b/w survival curve of 2groups.

Cox proportional Hazard test semiparametric, assumes effects of multiple variable are contsant over a period of time.

example ;

Statistical analysis

Survival analysis using the Cox proportional hazard regression analysis with time-dependent variable was employed to examine the psychiatric predictors and baseline psychosocial predictors for substance use disorders. Psychiatric predictors included ADHD, conduct disorder, oppositional

defiant disorder, depressive disorders and anxiety disorders. Substance use disorders included misuse of and dependence on nicotine, alcohol, betel nut and illicit drugs. The outcome (survival time) was age at onset of the substance use disorder.

Psychiatric and psychosocial predictors of substance use disorders among adolescents Longitudinal study*

SUSAN S. F. GAU, MIAN-YOON CHONG, PINCHENG YANG,

CHENG-FANG YEN, KUNG-YEE LIANG and ANDREW T. A. CHENG (BJP 2007, 190:42-48 )

Mantel Haennszel Test - can be used to study the effects of two binary independent variables ( eg. male vs. female and treatment vs. placebo)

Share this post


Link to post
Share on other sites

Posted (edited)

Cox's proportional hazards model

- Commonest approach to model survival or time to event data.

- It's analogous to a multiple regression model, and tests the difference between survival times of particular group, while allowing for other factors.

- In this model, the dependent variable is the 'hazard', which is the probability of dying (or experiencing the event in question), given that patients have survived up to a given point in time, or the risk for death at that moment.

- No assumption is made about the probability distribution of the hazard.

- However, it is assumed that if the risk for dying at a particular point in time in one group is, say, twice that in the other group, then at any other time it will still be twice that in the other group. In other words, the hazard ratio does not depend on time. The hazard of failure in one group is a constant ratio (over time) of the hazard of failure in the other group.

Log rank test does not assume proportional hazards per se.

- Used to compare 2 survival curves, and tests whether there's a difference between the survival times of different groups. However, it does not allow other explanatory variables to be taken into account.

- It's used to test the null hypothesis that "there is no difference between the population survival curves" (i.e. the probability of an event occurring at any time point is the same for each population).

- It's the most powerful for detecting alternative hypotheses in which the hazards are proportional.

Quick Reference:

http://thestatsgeek.com/2017/09/21/survival-analysis-power-log-rankcox-versus-rmst/

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065034/

Edited by Alexis

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now
Sign in to follow this  
Followers 0