Format

Send to

Choose Destination
See comment in PubMed Commons below
BMC Health Serv Res. 2008 Aug 13;8:176. doi: 10.1186/1472-6963-8-176.

Impact of date stamping on patient safety measurement in patients undergoing CABG: experience with the AHRQ Patient Safety Indicators.

Author information

1
Department of Anesthesiology, University of Rochester School of Medicine and Dentistry, Rochester, New York, USA. laurent_glance@urmc.rochester.edu

Abstract

BACKGROUND:

The Agency for Healthcare Research and Quality (AHRQ) Patient Safety Indicators (PSIs) provide information on hospital risk-adjusted rates for potentially preventable adverse events. Although designed to work with routine administrative data, it is unknown whether the PSIs can accurately distinguish between complications and pre-existing conditions. The objective of this study is to examine whether the AHRQ PSIs accurately measure hospital complication rates, using the data with present-on-admission (POA) codes to distinguish between complications and pre-existing conditions

METHODS:

Retrospective cohort study of patients undergoing isolated CABG surgery in California conducted using the 1998-2000 California State Inpatient Database. We calculated the positive predictive value of selected AHRQ PSIs using information from the POA as the gold standard, and the intra-class correlation coefficient to assess the level of agreement between the hospital risk-adjusted PSI rates with and without the information contained in the POA modifier.

RESULTS:

The false positive error rate, defined as one minus the positive predictive value, was greater than or equal to 20% for four of the eight PSIs examined: decubitus ulcer, failure-to-rescue, postoperative physiologic and metabolic derangement, and postoperative pulmonary embolism or deep venous thrombosis. Pairwise comparison of the hospital risk-adjusted PSI rates, with and without POA information, demonstrated almost perfect agreement for five of the eight PSI's. For decubitus ulcer, failure-to-rescue, and postoperative pulmonary embolism or DVT, the intraclass-correlation coefficient ranged between 0.63 to 0.79.

CONCLUSION:

For some of the AHRQ Patient Safety Indicators, there are significant differences in the risk-adjusted rates of adverse events depending on whether the POA indicator is used to distinguish between pre-existing conditions and complications. The use of the POA indicator will increase the accuracy of the AHRQ PSIs as measures of adverse outcomes.

PMID:
18700979
PMCID:
PMC2529290
DOI:
10.1186/1472-6963-8-176
[Indexed for MEDLINE]
Free PMC Article
PubMed Commons home

PubMed Commons

0 comments
How to join PubMed Commons

    Supplemental Content

    Full text links

    Icon for BioMed Central Icon for PubMed Central
    Loading ...
    Support Center