Propensity score methods effectively balance the distribution of measured covariates that are associated with an exposure, but their effect on the balance
of unmeasured variables is less clear. We undertook an analysis of this by examining variables
that were measured but un-used- “pseudo-unmeasured variables” -in a large study of
the effect of right heart catheter (RHC) use in seriously ill patients. Subjects included
5,735 seriously ill adults admitted to an ICU in one of 5 U.S. teaching hospitals
(1989–1994; SUPPORT study). We used the published model of propensity for RHC that
used 88 variables to generate 1008 matched pairs. Matching effectively balanced the
measured covariates across exposure groups (RHC+ or RHC−), yielding estimates of RHC's
effect on outcomes, assuming no hidden bias.
To read this article in full you will need to make a payment
Purchase one-time access:
Academic & Personal: 24 hour online accessCorporate R&D Professionals: 24 hour online accessOne-time access price info
- For academic or personal research use, select 'Academic and Personal'
- For corporate R&D use, select 'Corporate R&D Professionals'
Subscribe:
Subscribe to Journal of Clinical EpidemiologyAlready a print subscriber? Claim online access
Already an online subscriber? Sign in
Register: Create an account
Institutional Access: Sign in to ScienceDirect
Article info
Identification
Copyright
© 2003 Elsevier Inc. Published by Elsevier Inc. All rights reserved.
