27.1.2015
We invite you to the mathematical statistics seminar "Delayed entry in Cox proportional hazards model". The Seminar will take place on Wednesday, February 4, 2015, 10:00 in Institute of Measurement Science SAs. Mgr. Silvie Bělašková from Mathematical Institute of Faculty of applied informatics, Tomas Bata University in Zlin, Czech Republic will present a talk.
Delayed entry in Cox proportional hazards model
Silvie Bělašková and Eva Fišerová
Palacky University in Olomouc, Czech Republic
Tomas Bata University in Zlin, Czech Republic
One of the primary goals of analysis of time scale in
medicine is the estimation of treatment effects based on observational
studies. These studies are often based on incomplete observations and it
requires special techniques for analysing. Base methods often used are
survival analysis. Methods of survival analysis (e.g., Cox proportional
hazards model) require that the event time be measured with respect to
some origin time. The choice of origin time is substantively important
because it implies that the risk of the event varies as a function of
time since that origin. Ideally, the origin time is the same as the time
at which observations begin, on the other hand, observations do not
begin until some time after the origin time. These late entries are
treated as left truncated data in the statistical literature. However,
in actual situations, it may be possible that subjects with either
extremely high, or extremely low risk enter the study after the time
origin. There are three tests that are commonly used to test the
hypothesis that a covariate has no effect. These are Wald test, the
score test and the likelihood ratio test. The aim of our contribution is
discussing accuracy of pvalue of these tests in proportional hazards
model.
Keywords
Survival analysis, Cox proportional hazards model, time scale, missing data.
References
 Cox D. R., Regression Models and Life Tables, Journal of the Royal
Statistical Society, Series B,1972, 20, 187220, with discussion.
 Cox D. R., Oakes D., Analysis of Survival Data, London: Chapman & Hall, 1984.
 Thibaut AC., Bénichou J., Choice of timescale in Cox's model
analysis of epidemiologic cohort data: a simulation study, Statistics in
Medicine 2004 Dec 30; 23(24): 380320.
 Chalise P., Chicken E., McGee D., Time Scales in Epidemiological
Analysis. An Empirical Comparison, Statistics in Medicine 2009; 00:113
 Griffin B. A., Anderson G., Shih R., Whitsel E., Use of Alternative
Time Scales in Cox Proportional Hazard Models Implications for
TimeVarying Environmental Exposures, Statistics in Medicine, v. 31, no.
27, Nov. 2012, p.33203327.
