Example 1 for PROC ICPHREG

Difficulty Level

Beginner

Published on : 01/01/2023

The script begins by creating a dataset named 'hiv' containing information on disease progression, with left and right-censored observation times. It then uses the ICPHREG procedure in three different ways to model this data: a first basic model, a second model specifying a piecewise constant baseline hazard function, and a third model that calculates Hazard Ratios for the disease stage variable.

Data Analysis

Type : CREATION_INTERNE

The data is created directly within the script via a DATA step and a 'datalines' statement. Values of 0 for the lower bound (Left) and values >= 26 for the upper bound (Right) are converted to missing values to handle data censoring.

1 Code Block

DATA STEP Data

Explanation :
This block creates the SAS table 'hiv' from embedded data ('datalines'). It defines the variables Left, Right, Stage, Dose, CdLow, and CdHigh. The conditional 'if' statements transform certain values into SAS missing data ('.') to correctly represent censored survival data.

Copied!

1	DATA hiv;
2	INPUT Left Right Stage Dose CdLow CdHigh;
3	IF (Left=0) THEN Left=.;
4	IF (Right>=26) THEN Right=.;
5	DATALINES;
6	0 16 0 0 0 1
7	15 26 0 0 0 1
8	12 26 0 0 0 1
9	17 26 0 0 0 1
10	13 26 0 0 0 1
11	0 24 0 0 1 0
12	6 26 0 1 1 0
13	0 15 0 1 1 0
14	14 26 0 1 1 0
15	12 26 0 1 1 0
16	13 26 0 1 0 1
17	12 26 0 1 1 0
18	12 26 0 1 1 0
19	0 18 0 1 0 1
20	0 14 0 1 0 1
21	0 17 0 1 1 0
22	0 15 0 1 1 0
23	3 26 1 0 0 1
24	4 26 1 0 0 1
25	1 11 1 0 0 1
26	13 19 1 0 0 1
27	0 6 1 0 0 1
28	0 11 1 1 0 0
29	6 26 1 1 0 0
30	0 6 1 1 0 0
31	2 12 1 1 0 0
32	1 17 1 1 1 0
33	0 14 1 1 0 0
34	0 25 1 1 0 1
35	2 11 1 1 0 0
36	0 14 1 1 0 0
37	;
38

2 Code Block

PROC ICPHREG

Explanation :
This block executes a proportional hazards regression model for interval-censored data. The response variable is the time interval (Left, Right). The explanatory variables are 'Stage' and 'Dose', treated as classification variables. The 'desc' option on the CLASS statement orders the levels of these variables in descending order.

Copied!

1	PROC ICPHREG DATA=hiv;
2	class Stage Dose / desc;
3	model (Left, Right) = Stage Dose;
4	RUN;

3 Code Block

PROC ICPHREG

Explanation :
This second execution of PROC ICPHREG includes two additional options. 'ithistory' displays the iteration history of the model fit. 'basehaz=pch(intervals=(10))' specifies that the baseline hazard function should be modeled as a piecewise constant function with a breakpoint at 10.

Copied!

1	PROC ICPHREG DATA=hiv ithistory;
2	class Stage Dose / desc;
3	model (Left, Right) = Stage Dose / basehaz=pch(intervals=(10));
4	RUN;

4 Code Block

PROC ICPHREG

Explanation :
This third model focuses on the effect of the 'Stage' variable. The 'hazardratio Stage' statement is added to request the calculation and display of Hazard Ratios for the different levels of the 'Stage' variable, allowing quantification of the effect of this variable on the risk.

Copied!

1	PROC ICPHREG DATA=hiv;
2	class Stage / desc;
3	model (Left, Right) = Stage / basehaz=pch(intervals=(10));
4	hazardratio Stage;
5	RUN;

This material is provided "as is" by We Are Cas. There are no warranties, expressed or implied, as to merchantability or fitness for a particular purpose regarding the materials or code contained herein. We Are Cas is not responsible for errors in this material as it now exists or will exist, nor does We Are Cas provide technical support for it.

Difficulty Level

Published on : 01/01/2023

Data Analysis

1 Code Block

2 Code Block

3 Code Block

4 Code Block

Related Documentation