Biostats SPSS Cox Proportional Hazards

Part1

Week 9 Step by Step Application Guide 9.1

Cox Proportional Hazards

Problem 1. Data set

a. Access the sample dataset.

Step 1. Open the Week09_PRACTICE_dataset.sav data set

b. Identify the censoring variable, given that you wish to evaluate the event stroke. The censoring variable is “stroke”. For stroke=0, the event did not occur, which makes it censored. A censored event occurs for the stroke variable when stroke=1.

c. Identify the time-to-event variable and produce a frequency table to determine the presence of ties. (10 Points)

The time-to-event variable is “Time followed in Months [followed]”.

Step 2. To produce a frequency table, Analyze Descriptive Statistics Frequencies. Move variable “time followed in months” to Variables box

Based on the frequency table, the only time that had multiple events was 32.02, the longest time followed, with a frequency of 21. Based on a simple frequency table, we do not know whether the ties were a result of multiple events, censoring due to end of study follow up, or a mixture of events and censoring To examine the data more closely, go to:

Step 3: Analyze ( Descriptive Statistics ( Crosstabs

Step 4. Place Time followed in Months in Row(s); place Had stroke in Column(s). Click OK.

In the output Case Processing Summary table, we can see that all tied times were at 32.02, the end of study follow up and all were censored without stroke (stroke = 0). There were no other ties in times.

d. Identify the independent variable for exposure status, given the following research question: Is there an association between hypertension and the time a person was followed before experiencing a stroke?

The independent variable is hypertension.

e. Create a table with the above variables and their role in this analysis.

Variable	Level of Measurement	Values	Role
Stroke	Nominal (dichotomous)	0=no, 1=yes	Dependent variable
Time followed in months	Scale (continuous)	Range: 0.82 – 32.02 months	Censoring/Time to event variable
Hypertension	Nominal (dichotomous)	0=no, 1=yes	Independent variable