1. Survival Analysis

2. Key variable = time until some event time from treatment to death time for a fracture to heal time from surgery to relapse

3. Censored observations subjects removed from data set at some stage without suffering an event [lost to follow-up or died from unrelated event] study period ends with some subjects not suffering an event

4. Example PatientTime at entry (months) Time at death/ censoring Dead or censored Survival time 10.011.8D 20.012.5C12.5 * 30.418.0C17.6* 41.26.6D5.4 53.018.0C15.0*

5. Survival analysis uses information about subjects who suffer an event and subjects who do not suffer an event

6. Life Table Shows pattern of survival for a group of subjects Assesses number of subjects at risk at each time point and estimates the probability of survival at each point

7. Motion sickness data N=21 subjects placed in a cabin and subjected to vertical motion Endpoint = time to vomit

8. Motion sickness data 14 survived 2 hours without vomiting 5 subjects vomited at 30, 50, 51, 82 and 92 minutes respectively 2 subjects requested an early stop to the experiment at 50 and 66 minutes respectively

9. Life table SubjectSurvival time (min) Survival proportion 1300.952 2500.905 350 * 4510.855 566* 6820.801 7920.748 8 – 21120*

10. Calculation of survival probabilities p k = p k-1 x (r k – f k )/ r k where p = probability of surviving to time k r = number of subjects still at risk f = number of events (eg. death) at time k

11. Calculation of survival probabilities Time 30 mins : (21 – 1)/21 = 0.952 Time 50 mins : 0.952 x (20 – 1)/20 = 0.905 Time 51 mins : 0.905 x (18 – 1)/18 = 0.854

12. Kaplan-Meier survival curve Graph of the proportion of subjects surviving against time Drawn as a step function (the proportion surviving remains unchanged between events)

13. Survival Curve

14. Kaplan-Meier survival curve times of censored observations indicated by ticks numbers at risk shown at regular time intervals

15. Summary statistics 1.Median survival time 2.Proportion surviving at a specific time point

16. Survival Curve

17. Comparison of survival in two groups Log rank test Nonparametric – similar to chi-square test

18. SPSS Commands Analyse – Survival – Kaplan-Meier Time = length of time up to event or last follow-up Status = variable indicating whether event has occurred Options – plots - survival

19. SPSS Commands (more than one group) Factor = categorical variable showing grouping Compare factor – choose log rank test

20. Example RCT of 23 cancer patients 11 received chemotherapy Main outcome = time to relapse

21. Chemotherapy example

22. No chemotherapy Median relapse-free time = 23 weeks Proportion surviving to 28 weeks = 0.39 Chemotherapy Median relapse-free time = 31 weeks Proportion surviving to 28 weeks = 0.61

23. The Cox model Proportional hazards regression analysis Generalisation of simple survival analysis to allow for multiple independent variables which can be binary, categorical and continuous

24. The Cox Model Dependent variable = hazard Hazard = probability of dying at a point in time, conditional on surviving up to that point in time = instantaneous failure rate

25. The Cox Model Log [h i (t)] = log[h 0 (t)] + ß 1 x 1 + ß 2 x 2 + …….. ß k x k where [h 0 (t)] = baseline hazard and x 1,x 2, …x k are covariates associated with subject i

26. The Cox Model h i (t) = h 0 (t) exp [ß 1 x 1 + ß 2 x 2 + …….. ß k x k] where [h 0 (t)] = baseline hazard and x 1,x 2, …x k are covariates associated with subject i

27. The Cox Model Interpretation of binary predictor variable defining groups A and B: Exponential of regression coefficient, b, = hazard ratio (or relative risk) = ratio of event rate in group A and event rate in group B = relative risk of the event (death) in group A compared to group B

28. The Cox Model Interpretation of continuous predictor variable: Exponential of regression coefficient, b, refers to the increase in hazard (or relative risk) for a unit increase in the variable

29. The Cox Model Model fitting: Similar to that for linear or logistic regression analysis Can use stepwise procedures such as Forward Wald to obtain the best subset of predictors

30. The Cox model Proportional hazards regression analysis Assumption: Effects of the different variables on event occurrence are constant over time [ie. the hazard ratio remains constant over time]

31. SPSS Commands Analyse – Survival – Cox regression Time = length of time up to event or last follow-up Status = variable indicating whether event has occurred Covariates = predictors (continuous and categorical) Options – plots and 95% CI for exp(b)

32. The Cox model Check of assumption of proportional hazards (for categorical covariate): Survival curves Hazard functions Complementary log-log curves For each, the curves for each group should not cross and should be approximately parallel