Scribd
Upload
41 views3 pages

Tutorial 4

This tutorial covers several topics in survival analysis including: 1) Statements about the hazard function h(t) that are correct. 2) Computing the survival and density functions given a hazard function. 3) Identifying censoring in clinical trial examples. 4) Calculating the Kaplan-Meier estimate of the survival function by hand and in R for small datasets. 5) Estimating 5-year and 10-year survival probabilities using actuarial estimates from grouped survival data. 6) Calculating the Kaplan-Meier estimate, its variance, and confidence intervals by hand and in R. 7) Testing for differences in survival distributions between two groups using the log-rank

Uploaded by

Clara Carner
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
41 views3 pages

Tutorial 4

This tutorial covers several topics in survival analysis including: 1) Statements about the hazard function h(t) that are correct. 2) Computing the survival and density functions given a hazard function. 3) Identifying censoring in clinical trial examples. 4) Calculating the Kaplan-Meier estimate of the survival function by hand and in R for small datasets. 5) Estimating 5-year and 10-year survival probabilities using actuarial estimates from grouped survival data. 6) Calculating the Kaplan-Meier estimate, its variance, and confidence intervals by hand and in R. 7) Testing for differences in survival distributions between two groups using the log-rank

Uploaded by

Clara Carner
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

Biostatistics Session 2022–2023 (Semester 2)

Tutorial 4
1. In Survival Analysis, which of the following statements about the hazard function h(t) are
correct? (Choose all that apply.)
(i) The quantity h(t)δt, for small δt, represents approximately the probability that an
individual will die in the interval (t, t + δt).
(ii) If the hazard rate is constant then the survival probability S(t) decays as a linear
function of time t.
(iii) A group with a lower hazard rate will have a higher probability of survival than a
group with a higher hazard rate.
(iv) The cumulative hazard function H(t) is a non-decreasing, unbounded function on
[0, ∞).
(v) In survival modelling, the hazard rate should always be assumed constant over time.
√
2. Suppose that a random survival time T has hazard function h(t) = 1/ t (t > 0).
(a) Obtain the survival and density functions.
(b) Compute the mean and median survival times.

3. In the following examples, determine if there is any censoring and what kind.
(a) Leukemia patients are given a drug or a placebo. Survival time is the duration from
remission to relapse. The study ends at 52 weeks with some patients yet to relapse.
(b) Consider the age at which children are able to count from 1 to 10 at primary school.
Some children may already be able to count before joining school.

4. Consider the following small data set of survival times:

3, 4, 5∗ , 6, 6∗ , 8∗ , 11, 14, 15, 16∗ ,

where ∗ indicates right-censored observations.


(a) Calculate and plot the Kaplan–Meier estimate S(t)
b of the survival function S(t) by
pen and paper.
(b) Using part (a), estimate the median µ of the random survival time.
(c) Calculate and plot the Kaplan–Meier estimate S(t)
b of the survival function S(t) by
using R. Note that you have to type in the survival times and the event indicator see
the lecture notes.

5. In a clinical trial of 146 patients treated after they had had a myocardial infarction, the
following data on their survival were obtained (grouped into one-year intervals).

1
Year from Alive at start Deaths Censored
entry of year during year during year
1st 146 27 3
2nd 116 18 10
3rd 88 21 10
4th 57 9 3
5th 45 1 3
6th 41 2 11
7th 28 3 5
8th 20 1 8
9th 11 2 1
10th 8 2 6

Estimate the 5- and 10-year survival probabilities by calculating the suitable values of the
actuarial (life-table) estimate S(t).
e

6. Consider the following survival data set:

2 10 14 27 27 37 44∗ 46 54 60

Note that at time 44 the observation is censored and not an event!


(a) Calculate the Kaplan–Meier by pen and paper
(b) Calculate the variance of the Kaplan–Meier estimate S(t)
b at points t = 4 by pen and
paper.
(c) Construct symmetric confidence intervals for values of the survival function S(4) by
pen and paper
(d) Calculate the Kaplan–Meier by using R. Use the obtained output to find or calculate
the variance of S(4)?
d

7. You are given a small data set on survival times of subjects in two groups:

Group 1: 1, 1∗ , 2, 2∗
Group 2: 2, 3, 3∗ , 4

where ∗ indicates a censored observation. The corresponding Kaplan–Meier survival plots


are shown below:

2
(a) You wish to compare any differences in the survival distributions in the groups. State
a suitable null hypothesis H0 and an alternative hypothesis H1 .
(b) Conduct the first version of the log-rank test (chi-square) in its to test H0 against H1
at a 5%-significance level.
(c) Now conduct the second version of the log-rank test (normal), also at a 5%-significance
level. Compare the result with part (b).

8. In a clinical study to evaluate the efficacy of maintenance chemotherapy for acute myeloma
leukemia (AML), patients who reached the state of remission were observed until relapse;
Group 1 received a maintenance chemotherapy treatment, while Group 2 (control group)
did not. The data can be found in AML.dat on Brightspace.
(a) Write down the null and alternative hypotheses to test the the survival at t = 10 is the
same. Also write down the corresponding test statistic.
(b ) Write down the null and alternative hypotheses corresponding to a logrank test.
(c) Perform the two logrank tests in R. Formulate your conclusion.

You might also like