Usage Note 37665: The PROC PHREG BASELINE statement default METHOD= has changed in SAS® 9.2 TS1M0
 |  |  |  |
Beginning in SAS® 9.2 TS1M0, the default method for estimating the survival function has changed. The method is specified using the METHOD= option in the BASELINE statement and its default has changed to the cumulative hazards method (METHOD=CH) from the product-limit method (METHOD=PL). The CH estimate is known as the Breslow, Nelson or empirical cumulative hazards estimate and can be requested by specifying any of METHOD=CH, METHOD=BRESLOW, METHOD=EMP, or METHOD=NELSON. Beginning in SAS 9.3 TS1M2 the Fleming-Harrington (METHOD=FH) estimator is also available. The Fleming-Harrington estimator is a tie-breaking modification of the Breslow estimator. If there are no tied event times, the FH estimator reduces to the Breslow estimator
In Cox regression, the Breslow method for estimating the survival function is much more common than the product-limit method, although these two estimates are asymptotically equivalent. Other than as presented in Kalbfleish and Prentice (2002, first edition: 1980), the product-limit method is seldom used in the literature to estimate the survivor function for Cox regression. In PROC PHREG, there are situations where only the Breslow method or Fleming-Harrington method can be used. For example, METHOD=BRESLOW is enforced when using the counting-process MODEL statement syntax — MODEL (start, stop) — which is typically used with time-dependent covariates or in repeated events scenarios. Alternatively, METHOD=FH is also supported for the MODEL (start, stop) counting-process syntax.
The product-limit estimate is more often used when trying to estimate the survivor functions of homogeneous samples, such as when there are no covariates (the null model). With a null model, the METHOD=PL survival estimates in PROC PHREG will match those of the default METHOD=KM|PL in PROC LIFETEST. However, in PROC PHREG you also have the ability to get METHOD=PL estimates for left-truncated data (late entry into the risk sets).
_____
Kalbfleisch, J.D. and Prentice, R.L. (2002), The Statistical Analysis of Failure time Data, Second Edition, New York: John Wiley & Sons, Inc.
Operating System and Release Information
| Product Family | Product | System | SAS Release | |
| Reported | Fixed* | |||
| SAS System | SAS/STAT | Microsoft® Windows® for 64-Bit Itanium-based Systems | 9.2 TS1M0 | |
| Microsoft Windows Server 2003 Datacenter 64-bit Edition | 9.2 TS1M0 | |||
| Microsoft Windows Server 2003 Enterprise 64-bit Edition | 9.2 TS1M0 | |||
| Microsoft Windows XP 64-bit Edition | 9.2 TS1M0 | |||
| Microsoft® Windows® for x64 | 9.2 TS1M0 | |||
| Microsoft Windows Server 2003 Datacenter Edition | 9.2 TS1M0 | |||
| Microsoft Windows Server 2003 Enterprise Edition | 9.2 TS1M0 | |||
| Microsoft Windows Server 2003 Standard Edition | 9.2 TS1M0 | |||
| Microsoft Windows XP Professional | 9.2 TS1M0 | |||
| z/OS | 9.2 TS1M0 | |||
| Windows Vista | 9.2 TS1M0 | |||
| 64-bit Enabled AIX | 9.2 TS1M0 | |||
| 64-bit Enabled HP-UX | 9.2 TS1M0 | |||
| 64-bit Enabled Solaris | 9.2 TS1M0 | |||
| HP-UX IPF | 9.2 TS1M0 | |||
| Linux | 9.2 TS1M0 | |||
| Linux for x64 | 9.2 TS1M0 | |||
| OpenVMS on HP Integrity | 9.2 TS1M0 | |||
| Solaris for x64 | 9.2 TS1M0 | |||
| Type: | Usage Note |
| Priority: | |
| Topic: | Analytics ==> Regression Analytics ==> Survival Analysis SAS Reference ==> Procedures ==> PHREG |
| Date Modified: | 2017-08-01 11:06:10 |
| Date Created: | 2009-10-30 15:10:10 |