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

Health Services Research & Development

Go to the ORD website
Go to the QUERI website
FORUM - Translating research into quality health care for Veterans

» Back to Table of Contents

What are the Quality Consequences of Medicare-VA Dual Use?

Over half of VA enrollees are dually enrolled in Medicare.1 Veterans who are dual Medicare enrollees include about 40 percent of Veterans in the highest priority groups—those with service-connected disabilities—and higher percentages in lower priority groups. Dually-enrolled Veterans frequently obtain health services from VA and Medicare in the same year, typically relying on VA for less than 40 percent of their outpatient care. When they need inpatient care, these Veterans are more than four times as likely to rely on Medicare as they are to rely on VA.2 These basic facts imply that VA costs are much lower than they would be if dual enrollees relied exclusively on VA for care. They also imply that the quality of VA care might be suffering because VA clinicians may have difficulty coordinating care and exchanging information with non-VA providers across the boundary between VA and non-VA networks.

To investigate the impact of dual use on quality of care, we combined VA utilization records with Medicare claims data and studied the relationship between fragmentation of care across the two systems and the likelihood of experiencing a hospitalization for an ambulatory care sensitive condition (ACSC).3 These hospitalizations have been widely used to assess the quality of outpatient care in geographic regions. Because inpatient admissions are costly, they are also an important measure of inefficient resource use.

Our research database contained 288,000 observations on dually-enrolled Veterans with Medicare and/or VA outpatient use in 1999 and 2000. We counted outpatient visits in each system by six-month periods and then characterized the degree of fragmentation of care for each patient in each period as one minus the percentage of the patient's outpatient care provided by the VA or the percentage provided by Medicare, whichever was larger. This formula implies that a Veteran relying exclusively on either VA or Medicare would have a fragmentation measure equal to zero (1-1=0). The maximum degree of fragmentation possible is 0.5, arising if the Veteran evenly divided his or her outpatient visits between the two systems (1-0.5=0.5).

We related this measure of fragmentation to the probability of experiencing a hospitalization for an ACSC, as defined by the Agency for Healthcare Research and Quality. ACSC hospitalizations consist of 13 types of adult admissions thought to be potentially preventable through high quality outpatient care. The most common ACSC admissions in the VA population are those for congestive heart failure (CHF), chronic obstructive pulmonary disease (COPD), dehydration, urinary tract infection (UTI), long-term complications of diabetes, and pneumonia. We risk adjusted this relationship by controlling for the age and priority status of each Veteran as well as 30 co-morbidities defined by ICD-9-CM codes recorded in the period before measuring for fragmentation and outcomes.

This analysis was complicated by the fact that patient health or behavior that is unobservable to the researcher may affect fragmentation and ACSC hospitalization. For example, a patient with poor self-care skills might need frequent outpatient visits in both systems to address minor problems; this patient might also experience ACSC hospitalizations when more serious problems develop. In this example, fragmentation of care does not necessarily cause hospitalizations, instead, poor self-care causes both fragmentation and hospitalizations.

To address this methodological challenge, we used an instrumental variables statistical model constructed in two stages. The first stage predicted the degree of fragmentation of care for each Veteran as a function of the distance between that Veteran's residence and the nearest VA Medical Center. The second stage estimated the relationship between the predictable component of fragmentation and the probability of ACSC hospitalization.

Our estimates indicated that the degree of fragmentation of outpatient care between VA and Medicare had a strong and statistically significant effect on the probability of hospitalization for an ACSC. A one-standard deviation change in fragmentation was associated with a 20 percent change in hospitalization rates. The instrumental variables statistical technique allows us to infer a causal relationship more confidently than we could from a simpler observational design, but the inference is not as strong as it would be with a randomized experiment.

These results imply that VA and Medicare are spending substantial resources on inpatient care for Veterans who could be managed more efficiently and more effectively if their care were better coordinated. Several initiatives are currently being implemented by VA management to facilitate sharing of electronic medical records between VA and non-VA health plans and facilities. These efforts have the potential to reduce the harm from fragmentation. In addition, the reorganization of VA primary care to emphasize continuity and comprehensiveness of care from a patient-centered team has the potential to reduce fragmentation itself. With initiatives like these, VA managers have a rare opportunity to save money while simultaneously improving the quality of care for a particularly vulnerable population of Veterans.

  1. 2010 Survey of Veteran Enrollees' Health and Reliance Upon VA. Assistant Deputy Under Secretary for Health for Policy and Planning. Washington, DC: U.S. Department of Veterans Affairs, July 2011.
  2. Quality Initiative Undertaken by the Veterans Health Administration. Washington, DC: Congressional Budget Office, August 2009.
  3. Pizer, S.D. and Gardner J.A. "Is Fragmented Financing Bad for Your Health?" Inquiry 2011; 48:109-122.

Questions about the HSR&D 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.