24-1108 MARVIN GARISSE YOUYATTE
ASSIGMENT SURVIVAL DATA ANALYSIS
Questions
1. What is survival analysis?
It is a collection of statistical procedures for data analysis where the outcome
Variable of interest is time until an event occurs.
2. How do I handle censoring in survival data?
Censoring is a condition in which the value of a measurement or observation is only
Partially known.
3. What is the difference between parametric and non-parametric survival models?
Parametric requires a clear distribution and non-parametric assumes proportional hazard.
4. How do I choose the right survival model for my data?
5. What is the difference between hazard ratio and survival probability
The hazard ratio is simply calculated as the ratio of hazard rates.In contrast ,the survival probability
Estimates the probability of not experiencing the event by a specific time.
10. How do I validate the assumptions of a survival model?
By efficient testing giving right interpreting.
6. Consider the lifetime X, which has the pmf
1
p(xi ) = P(X = xi ) = , i = 1, 2, 3
a simple discrete uniform distribution.
(a) Find the corresponding survival function and hazard function.
QUESTION 3
The graph describes the experience of six persons A, B, C, D, E and F followed over time.
An X denotes a person who got the event.
a) Explain person A, B, C, D, E and F in the graph above (6 marks)
Person A, for example, is followed from the start
of the study until getting the event at week 5; his
survival time is 5 weeks and is not censored.
Person B also is observed from the start of the
study but is followed to the end of the 12-week
study period without getting the event; the survival
time here is censored because we can say only that
it is at least 12 weeks.
X Event occurs Person C enters the study between the second and
third week and is followed until he withdraws
from the study at 6 weeks; this person’s survival
time is censored after 3.5 weeks.
Person D enters at week 4 and is followed for the
remainder of the study without getting the event;
this person’s censored time is 8 weeks.
Person E enters the study at week 3 and is fol-
lowed until week 9, when he is lost to follow-up;
his censored time is 6 weeks.
Person F enters at week 8 .
b) How many persons get the event (1 marks)
Two A and F
c) )How many person censored (2 marks)
B,C,D,E(4)
d) Depicts the information of the survival time data for the six persons in the graph above into
the table below. (3 marks)
A table of the survival time data for the six persons in the graph is now presented as
Person Survival time Failed (1); censored (0)
A 5 1
B 12 0
C 3.5 0
D 8 0
E 6 0
F 3.5 1
QUESTION 4
Consider a clinical trial of 146 patients treated after they had a myocardial infarction (MI) in 37
Military Hospital within one year interval as shown in the table below.
Number alive and under Number
censored
Year since entry observation at the Number dying
into study beginning of interval during interval or withdrawn
[0, 1) 146 27 3
[1, 2) 116 18 10
[2, 3) 88 21 10
[3, 4) 57 9 3
[4, 5) 45 1 3
[5, 6) 41 2 11
[6, 7) 28 3 5
[7, 8) 20 1 8
[8, 9) 11 2 1
[9, 10) 8 2 6
To estimate the 5 year survival rate i.e. S(5) = P(T ≥ 5) if
a) all censoring occurred after 5 years. (5 marks)
b) all individuals were censored immediately upon entering the study. (5 marks)
c) all individuals who are censored and use the remaining ‘complete’ data. (5 marks)
d) Show that S(5)= π5qi (5 marks)
