2006 HSR&D National Meeting Abstract
3043 — Patient Safety: How Do VA Hospitals Compare to Non-VA Hospitals on Rates of Patient Safety Events?
Christiansen CL (Center for Health Quality, Outcomes, and Economic Research and Boston University SPH)
Elixhauser A (Agency for Healthcare Research and Quality)
Loveland SA (Center for Health Quality, Outcomes, and Economic Research and Boston University SPH)
Rivard P (Center for Health Quality, Outcomes, and Economic Research and Boston College)
Rosen AK (Center for Health Quality, Outcomes, and Economic Research and Boston University SPH)
Zhou S (Center for Health Quality, Outcomes, and Economic Research)
A specific aim for QUERI is to use risk-adjusted outcome measures to compare VA quality of care with care delivered by non-VA healthcare systems. These measures can be used to identify opportunities for quality improvement (QI) in the VA. This study aims to identify differences in risk-adjusted rates of patient safety indicators (PSIs) between veterans discharged from VA hospitals and individuals discharged from non-VA hospitals using a convenience sample of 994 hospitals from the Nationwide Inpatient Sample (NIS) developed by the Agency for Healthcare Research and Quality (AHRQ).
We estimated risk-adjusted rates of 14 Patient Safety Indicators (PSIs) using AHRQ’s PSI software V2.1R03a for 427,718 FY2004 VA hospitalizations. Rates were also calculated for 2,381,353 unweighted hospitalizations at non-VA hospitals from calendar year 2003. To ensure comparability, these VA and NIS samples included only hospitalizations of males age 18 and older for conditions other than mental health and substance abuse. VA stays were restricted to acute care hospitalizations; the NIS is limited to short-term acute care hospitals. The PSIs are risk-adjusted using age, comorbidities, and modified DRGs. Rates were compared using 95% confidence intervals; statistically significant differences were determined by t-tests (alpha = .01).
VA risk-adjusted hospitalization rates were lower (p<0.01) than unweighted NIS rates for some PSIs and higher for others. For example, decubitus ulcer rates per 1000 discharges (VA 20.4 versus NIS 25.3) and postoperative sepsis rates (9.4; 12.7) were lower in VA hospitals. On the other hand, VA rates were higher for indicators such as foreign body left during procedure (0.14; 0.09), iatrogenic pneumothorax (1.3; 0.8), and wound dehiscence (4.8; 1.5).
These preliminary analyses reveal that for some PSIs, rates were lower in VA hospitals than in non-VA hospitals and some were higher.
This study explores the use of administrative data to identify patient safety events in VA hospitals. Further analyses will refine these estimates and will provide insights into those patient safety problems the VA has addressed effectively as well as others where more work is needed.