Download presentation
Presentation is loading. Please wait.
1
If we can reduce our desire,
then all worries that bother us will disappear. Survival Analysis
2
Semiparametric Proportional Hazards Regression (Part II)
Survival Analysis Semiparametric Proportional Hazards Regression (Part II) Survival Analysis
3
Inference for the Regression Coefficients
Risk set at time y, R(y), is the set of individuals at risk at time y. Assume survival times are distinct and their order statistics are t(1) < t(2) < … < t(r). Let X(i) be the covariates associated with t(i). Survival Analysis
4
Partial Likelihood Survival Analysis
5
Partial Likelihood The product is taken over subjects who experienced the event. The function depends on the ranking of times rather than actual times robust to outliers in times Survival Analysis
6
Understanding the Partial Likelihood
The partial likelihood is based on a conditional probability argument. The lost information include: Censoring times & subjects in between t(k-1) & t(k) Only one failure at t(k) No failures in between t(k-1) & t(k) Survival Analysis
7
Maximum Partial Likelihood Estimate
An estimate for b is obtained as the maximiser of PLn(b), called the maximum partial likelihood estimate (MPLE). Survival Analysis
8
Score Function Survival Analysis
9
Fisher Information Matrix
Survival Analysis
10
Estimating Covariance Matrix
Let be the MPLE of b, which can be found using the Newton-Rhapson method. The covariance matrix of is estimated by Survival Analysis
11
Ties in Survival Times The construction of partial likelihood is under the assumption of no tied survival times However, real data often contain tied survival times, due to the way times are recorded. How do such ties affect the partial likelihood? Survival Analysis
12
Example Consider the following survival data: 6, 6, 6, 7+, 8 (in months) Survival Analysis
13
Ties in Survival Times When there are both censored observations and failures at a given time, the censoring is assumed to occur after all the failures. Potential ambiguity concerning which individuals should be included in the risk set at that time is then resolved. Accordingly, we only need consider how tied survival times can be handled. Survival Analysis
14
Ties in Survival Times Let Survival Analysis
15
Breslow Approximation
Survival Analysis
16
Breslow Approximation
Counts failed subjects more than once in the denominator, producing a conservative bias. Adequate if, for each k=1,…,r, dk is small relative to size of risk set. Survival Analysis
17
Efron Approximation Survival Analysis
18
Efron Approximation Approximation assumes that all possible orderings of tied survival times are equally likely. Hertz-Picciotto and Rockhill (Biometrics 53, , 1997) presented a simulation study which shows that Efron approximation performed far better than Breslow approximation Survival Analysis
19
Exact Partial Likelihood
Instead of average of all possible orderings of tied survival times, we sum them together, which is the exact conditional probability (the exact partial likelihood component at the tied time). e.g.: Subjects 1, 2, 3 tied at time 6; 6 possible orderings called scenarios A1 to A6 The partial likelihood of Ai is the conditional probability of having Ai, denoted as PL(Ai) The exact conditional probability is therefore: P(A1 or A2 or …A6)=sum of PL(Ai) from i=1 to 6 Survival Analysis
20
Discrete Partial Likelihood
Survival Analysis
21
Discrete Partial Likelihood
Survival Analysis
22
Discrete Partial Likelihood
The computational burden grows very quickly. Gail, Lubin and Rubinstein (Biometrika 68, , 1981) develop a recursive algorithm that is more efficient than the naive approach of enumerating all subjects. If the ties arise by the grouping of continuous survival times, the partial likelihood does not give rise to a consistent estimator of b. Survival Analysis
23
Model Selection In SAS, PROC PHREG, use Efron method to handle ties and conduct backward model selection Once the final model is selected, refit the final model with Exact method to handle ties Survival Analysis
24
Estimation of Survival Function
To estimate S(y|X), the baseline survival function S0(y) must be estimated first. Two estimates: Breslow estimate Kalbfleisch-Prentice estimate Survival Analysis
25
Breslow Estimate Survival Analysis
26
Kalbfleisch-Prentice Estimate
An estimate of h0(y) was derived by Kalbfleisch and Prentice using an approach based on the method of maximum likelihood. Reference: Kalbfleisc, J.D. and Prentice, R.L. (1973). Marginal likelihoods based on Cox’s regression and life model. Biometrika, 60, Survival Analysis
27
Example: PBC Survival Analysis
28
Predicted Curve for a fixed Covariate Value
Survival Analysis
29
Estimation of the Median Survival Time
Survival Analysis
30
Assessing the Proportional Hazards Assumption
By empirical score process/simulations In SAS: add a statement ASSESS PH/ RESAMPLE; A p-value will be given to assess the significance level of deviation from the proportional hazards assumption Survival Analysis
31
Strategies for Non-proportionality
Stratify the covariates with non-proportional effects No test for the effect of a stratification factor (so only for nuisance covariates) How to categorize a numerical covariate? Partition the time axis Add a time-dependent covariate (advanced) Use a different model (such as parametric AFT model) Survival Analysis
32
The End Good Luck for Finals!! Survival Analysis
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.