Talk to the Veterans Crisis Line now
U.S. flag
An official website of the United States government

VA Health Systems Research

Go to the VA ORD website
Go to the QUERI website

IIR 03-207 – HSR Study

IIR 03-207
Estimating the Magnitude of Unmeasured Risk in VA Patients
Mary S. Vaughan-Sarrazin, PhD MA
Iowa City VA Health Care System, Iowa City, IA
Iowa City, IA
Funding Period: May 2004 - April 2006
Numerous past studies have compared patient outcomes in VA and private sector hospitals. While VA patients have unique socioeconomic and clinical characteristics, it is unclear whether current methods appropriately adjust for differences between VA and private sector patients. Thus, it is it difficult to discern whether observed differences in risk-adjusted outcomes between VA and private sector hospitals are due to unmeasured illness or to differences in the quality of care. This study estimates unmeasured severity of illness for users of VA services by comparing outcomes of VA users and non-users in private-sector hospitals.

Determine whether VA utilization is an independent risk factor for poor outcomes. Specifically, we compared outcomes of VA users and other patients who receive care in private sector hospitals for 10 high-volume medical and surgical diagnoses, adjusting for severity of illness. We also identified characteristics of VA users that are associated with poor outcomes.

The two-year study employed a retrospective cohort design to identify male VA users and non-users age 67 years and older who received treatment for 10 medical or surgical conditions in private sector hospitals during the period 1996-2002. The study used Medicare administrative data (MedPAR and Denominator files), VA Administrative data (Patient Treatment File, Outpatient Care File, and the Compensation and Pension Minifile), and the US Census Summary Data File. Patients who used VA services during the two years preceding the index private hospital admission were identified using the VA administrative files. Endpoints (mortality after 30, 90, and 365 days) were compared in patients who did and did not use VA services prior to the index admission, adjusting for observed patient demographic and clinical risk factors using hierarchical models.

Surgical cohorts included: abdominal aortic aneurysm (n=130,746), coronary artery bypass graft (n=610,773), colectomy (n=292,331), lower extremity bypass (n=153,375), valve replacement surgery (n=128,298). Medical cohorts included: acute myocardial infarction (n=866,068), chronic heart failure (n=1,059,807), COPD (n=557,373), pneumonia (n=1,503,553), and stroke (n=728,738). Overall 8.5% of private sector patients used VA services during the 2 years prior to admission. VA users were more likely (p<.001) to have diabetes, heart failure, cerebro-vascular, peripheral vascular, obstructive lung, and chronic renal disease than patients who did not use VA services, and also resided in zipcodes with lower median household incomes. Of the five surgical conditions, only CABG had higher adjusted odds of death (p<.001) for VA users relative to non-users (30-day mortality Odds Ratio = 1.07; 95% CI, 1.03-1.11). For medical conditions, differences in mortality between VA users and non-users were generally small. Within each patient condition, VA users with low incomes or with 50% or greater service-connected disabilities generally had increased risk of death, compared to non-VA users. The relative risk of death for VA users also increased slightly with decreasing age. Results using propensity-matched samples were similar.

This study provides new insights into the magnitude of risk in VA populations that has been unmeasured in prior studies, and, in so doing, adds to our understanding of the relative quality of care in VA hospitals. This study found slightly higher risk of death for VA users undergoing CABG in private sector hospitals. However, the difference was smaller than the increased risk of death after CABG or AMI for VA hospital patients that has been estimated in previous studies. Overall, VA users with service-connected disabilities or low incomes had higher risk-adjusted mortality than the average private-sector patient. Furthermore, the relative risk of death appears to be higher for younger VA users. Current risk-adjustment methods may not adequately control for severity of illness in VA patients with these characteristics.

External Links for this Project

Dimensions for VA

Dimensions for VA is a web-based tool available to VA staff that enables detailed searches of published research and research projects.

Learn more about Dimensions for VA.

VA staff not currently on the VA network can access Dimensions by registering for an account using their VA email address.
    Search Dimensions for this project


Journal Articles

  1. Katz DA, Graber M, Birrer E, Lounsbury P, Baldwin A, Hillis SL, Christensen AJ. Health beliefs toward cardiovascular risk reduction in patients admitted to chest pain observation units. Academic emergency medicine : official journal of the Society for Academic Emergency Medicine. 2009 May 1; 16(5):379-87. [view]
  2. Vaughan-Sarrazin MS, Wakefield B, Rosenthal GE. Mortality of Department of Veterans Affairs patients undergoing coronary revascularization in private sector hospitals. Health services research. 2007 Oct 1; 42(5):1802-21. [view]
Conference Presentations

  1. Schelbert EB, Vaughan-Sarrazin MS, Welke KR, Rosenthal GE. Renal disease, valve type, and outcomes after aortic valve replacement in older patients. Paper presented at: American Heart Association Annual Scientific Sessions; 2005 Nov 15; Dallas, TX. [view]
  2. Schelbert EB, Vaughan-Sarrazin MS, Welke KR, Rosenthal GE. Valve type and long-term outcomes after aortic valve replacement in older patients. Paper presented at: Duke University Clinical Research Institute Cardiology Fellows' Clinical Research Annual Conference; 2006 Mar 10; Atlanta, GA. [view]

DRA: Health Systems, Aging, Older Veterans' Health and Care
DRE: none
Keywords: Patient outcomes, Risk factors, Utilization patterns
MeSH Terms: none

Questions about the HSR website? Email the Web Team

Any health information on this website is strictly for informational purposes and is not intended as medical advice. It should not be used to diagnose or treat any condition.