We postulated that associations between two specific provider characteristics, class (nurse practitioner relative to physician) and primary care providers who are proficient and interested in women's health (designated women's provider relative to nondesignated) and overall satisfaction with provider, were mediated through women veterans' perception of enough time spent with the provider. A national patient experience survey was administered to 7,620 women veterans. Multivariable models of overall patient satisfaction with provider were compared with and without the proposed mediator. A structural equation model (SEM) of the mediation of the two provider characteristics was also evaluated. Without the mediator, associations of provider class and designation with overall patient satisfaction were significant. With the proposed mediator, these associations became nonsignificant. An SEM showed that the majority (>80%) of the positive associations between provider class and designation and the outcome were exerted through patient perception of enough time spent with provider. Higher ratings of overall satisfaction with provider exhibited by nurse practitioners and designated women's health providers were exerted through patient perception of enough time spent with provider. Future research should examine what elements of provider training can be developed to improve provider-patient communication and patient satisfaction with their health care.

Background: In 2010, the Department of Veterans Affairs Healthcare System (VA) implemented policy to provide Comprehensive Primary Care (for acute, chronic, and female-specific care) from designated Women's Health providers (DWHPs) at all VA sites. However, since that time no comparisons of quality measures have been available to assess the level of care for women Veterans assigned to these providers.Objectives: To evaluate the associations between cervical and breast cancer screening rates among age-appropriate women Veterans and designation of primary-care provider (DWHP vs. non-DWHP).Research Design: Cross-sectional analyses using the fiscal year 2012 data on VA women's health providers, administrative files, and patient-specific quality measures.Subjects: The sample included 37,128 women Veterans aged 21 through 69 years.Measures: Variables included patient demographic and clinical factors (ie, age, race, ethnicity, mental health diagnoses, obesity, and site), and provider factors (ie, DWHP status, sex, and panel size). Screening measures were defined by age-appropriate subgroups using VA national guidelines.Results: Female-specific cancer screening rates were higher among patients assigned to DWHPs (cervical cytology 94.4% vs. 91.9%, P < 0.0001; mammography 86.3% vs. 83.3%, P < 0.0001). In multivariable models with adjustment for patient and provider characteristics, patients assigned to DWHPs had higher odds of cervical cancer screening (odds ratio, 1.26; 95% confidence interval, 1.07-1.47; P < 0.0001) and breast cancer screening (odds ratio, 1.24; 95% CI, 1.10-1.39; P < 0.0001).Conclusions: As the proportion of women Veterans increases, assignment to DWHPs may raise rate of female-specific cancer screening within VA. Separate evaluation of sex neutral measures is needed to determine whether other measures accrue benefits for patients with DWHPs.

Background: Women veterans comprise a small percentage of Department of Veterans Affairs (VA) health care users. Prior research on women veterans' experiences with primary care has focused on VA site differences and not individual provider characteristics. In 2010, the VA established policy requiring the provision of comprehensive women's health care by designated women's health providers (DWHPs). Little is known about the quality of health care delivered by DWHPs and women veterans' experience with care from these providers.Methods: Secondary data were obtained from the VA Survey of Healthcare Experience of Patients (SHEP) using the Consumer Assessment of Healthcare Providers and Systems (CAHPS) patient-centered medical home (PCMH) survey from March 2012 through February 2013, a survey designed to measure patient experience with care and the DWHPs Assessment of Workforce Capacity that discerns between DWHPs versus non-DWHPs.Findings: Of the 28,994 surveys mailed to women veterans, 24,789 were seen by primary care providers and 8,151 women responded to the survey (response rate, 32%). A total of 3,147 providers were evaluated by the SHEP-CAHPS-PCMH survey (40%; n = 1,267 were DWHPs). In a multivariable model, patients seen by DWHPs (relative risk, 1.02; 95% CI, 1.01-1.04) reported higher overall experiences with care compared with patients seen by non-DWHPs.Conclusions: The main finding is that women veterans' overall experiences with outpatient health care are slightly better for those receiving care from DWHPs compared with those receiving care from non-DWHPs. Our findings have important policy implications for how to continue to improve women veterans' experiences. Our work provides support to increase access to DWHPs at VA primary care clinics. Published by Elsevier Inc.

Background: Surveys are increasingly used to assess patient experiences with health care. Comparisons of hospital scores based on patient experience surveys should be adjusted for patient characteristics that might affect survey results. Such characteristics are commonly drawn from patient surveys that collect little, if any, clinical information. Consequently some hospitals, especially those treating particularly complex patients, have been concerned that standard adjustment methods do not adequately reflect the challenges of treating their patients. Objectives: To compare scores for different types of hospitals after making adjustments using only survey-reported patient characteristics and using more complete clinical and hospital information. Research Design: We used clinical and survey data from a national sample of 1858 veterans hospitalized for an initial acute myocardial infarction (AMI) in a Department of Veterans Affairs (VA) medical center during fiscal years 2003 and 2004. We used VA administrative data to characterize hospitals. The survey asked patients about their experiences with hospital care. The clinical data included 14 measures abstracted from medical records that are predictive of survival after an AMI. Results: Comparisons of scores across hospitals adjusted only for patient-reported health status and sociodemographic characteristics were similar to those that also adjusted for patient clinical characteristics; the Spearman rank-order correlations between the 2 sets of adjusted scores were >0.97 across 9 dimensions of inpatient experience. Conclusions: This study did not support concerns that measures of patient care experiences are unfair because commonly used models do not adjust adequately for potentially confounding patient clinical characteristics.

Both government and private health care systems have engaged in efforts to improve quality, but the effect of these initiatives on racial and ethnic disparities has not been well studied. In the decade following an organizational transformation, the Veterans Affairs (VA) health care system achieved substantial improvements in quality of care with minimal racial disparities for most process-of-care measures, such as rates of cholesterol screenings. However, in our study we observed a striking disconnect between high levels of performance on widely used process measures and modest levels of improvement in clinical outcomes, such as control of blood pressure, blood glucose, and cholesterol levels. We also observed a gap in clinical outcomes of as much as nine percentage points between African American veterans and white veterans. Almost all of the disparity in outcomes in the VA was explained by within-facility disparity, which suggests that VA medical centers need to measure and address racial gaps in care for their patient populations. Moreover, because cardiovascular disease and diabetes are major contributors to racial disparities in life expectancy, the findings of this study and others underscore the urgency of focused efforts to improve intermediate outcomes among African Americans in the VA and other settings.

The Department of Veterans Affairs (VA) and other federal agencies require funded researchers to include women in their studies. Historically, many researchers have indicated they will include women in proportion to their VA representation or pointed to their numerical minority as justification for exclusion. However, women's participation in the military-currently 14% of active military-is rapidly changing veteran demographics, with women among the fastest growing segments of new VA users. These changes will require researchers to meet the challenge of finding ways to adequately represent women veterans for meaningful analysis. We describe women veterans' health and health-care use, note how VA care is organized to meet their needs, report gender differences in quality, highlight national plans for women veterans' quality improvement, and discuss VA women's health research. We then discuss challenges and potential solutions for increasing representation of women veterans in VA research, including steps for implementation research.

Context: Cross-sectional studies have identified rural-urban disparities in veterans' health-related quality-of-life (HRQOL) scores. Purpose: To determine whether longitudinal analyses confirmed that these disparities in veterans' HRQOL scores persisted. Methods: We obtained data from the SF-12 portion of the veterans health administration's (VA's) Survey of Healthcare Experiences of Patients (SHEP) collected between 2002 and 2006. During that time, the SHEP was randomly administered to approximately 250,000 veterans annually who had used VA outpatient services. We evaluated 163,709 responses from veterans who had completed 2 or more surveys during the years studied. Respondents were classified into rural-urban groups using ZIP Code-based rural-urban commuting area designations. We estimated linear regression models using generalized estimating equations to determine whether rural and urban veterans' HRQOL scores were changing at different rates over the time period examined. Findings: After adjustment for sociodemographic differences, we found that urban veterans had substantially better physical HRQOL scores than their rural counterparts and that these differences persisted over the study period. While urban veterans had worse mental HRQOL scores than rural veterans, those differences diminished over the time period studied. Conclusions: Rural-urban disparities in HRQOL scores persist when tracking veterans longitudinally. Reduced access among rural veterans to care may contribute to these disparities. Because rural soldiers are overrepresented in current conflicts, the VA should consider new models of care delivery to improve access to care for rural veterans.

VA Greater Los Angeles HSR&D Center of Excellence, Sepulveda, CA, USA; Department of Health Services, UCLA School of Public Health, Los Angeles, CA, USA; U.S. Department of Veterans Affairs, Women Veterans Health Strategic Health Care Group, Office of Public Health & Environmental Hazards, Washington, DC, USA; Office of Quality & Performance, Veterans Health Administration, Providence, RI, USA; White River Junction VA Medical Center, National Center for PTSD, White River Junction, VT, USA; Dartmouth Medical School, Hanover, NH, USA; VA HSR&D Service, Office of Research & Development, Veterans Health Administration, Washington, DC, USA; Department of Medicine, UCLA School of Medicine, Los Angeles, CA, USA.

We compared demographic profiles across two rural urban classification schemes to determine whether rural urban disparities in health status persisted among Veterans Administration (VA) users over time. Using demographic and SF-12 survey data collected from 2002 to 2006, we conducted serial cross-sectional analyses of demographic variables and health status for veterans residing in VA- and rural urban commuting area (RUCA)-defined rural urban groups. VA and RUCA definitions yielded similar results for the "urban" population; however. VA- and RUCA-defined "rural" categories represent dissimilar populations. Compared to earlier years, the VA user population in 2006 was younger, more educated, wealthier, and more likely to be employed and privately insured. For all years and using both VA and RUCA rural urban definitions, physical component summary (PCS) scores were lower but mental component summary (MCS) scores were slightly higher for more rural compared to urban veterans. Anticipating and meeting the needs of rural VA users will require accurate identification of those who lack access to services and therefore defining "rural" appropriately.

BACKGROUND:Recent studies have suggested that there is a positive impact of patient-centered care (PCC) on both the patient-physician relationship and subsequent patient health-related behaviors. One recent prospective study reported a significant relationship between the degree of PCC experienced by patients during their hospitalization for acute myocardial infarction (AMI) and their postdischarge cardiac symptoms. A limitation of this study, however, was a lack of information regarding the technical quality of the AMI care, which might have explained at least part of the differences in outcomes. The present study was undertaken to test the influence of both PCC and technical care quality on outcomes among AMI patients.METHODS:We analyzed data from a national sample of 1,858 veterans hospitalized for an initial AMI in a Department of Veterans Affairs medical center during fiscal years 2003 and 2004 for whom data had been compiled on evidence-based treatment and who had also completed a Picker questionnaire assessing perceptions of PCC. Cox proportional hazards models were used to estimate the relationship between PCC and survival 1-year postdischarge, controlling for technical quality of care, patient clinical condition and history, admission process characteristics, and patient sociodemographic characteristics. We hypothesized that better PCC would be associated with a lower probability of death 1-year postdischarge, even after controlling for patient characteristics and the technical quality of care.RESULTS:Better PCC was associated with a significantly but modestly lower hazard of death over the 1-year study period (hazard ratio 0.992, 95 percent confidence interval 0.986-0.999).CONCLUSIONS:Providing PCC may result in important clinical benefits, in addition to meeting patient needs and expectations.

BACKGROUND:We compared risk-adjusted mortality rates between Medicaid-eligible patients in the Medicare Advantage plans ("MA dual enrollees") and Medicaid-eligible patients in the Veterans Health Administration ("VHA dual enrollees"). METHODS:We used the Death Master File to ascertain the vital status of 1912 MA and 2361 VHA dual enrollees. We used Cox regression models to estimate hazard ratios (HRs) with 95% confidence intervals (CIs). RESULTS:The 3-year mortality rates of VHA and MA dual enrollees were 15.8% and 19.0%, respectively. The adjusted HR of mortality in the MA dual enrollees was significantly higher than in the VHA dual enrollees (HR, 1.260 [95% CI, 1.044-1.520]). This was also the case for elderly patients and those from racial/ethnic minority groups. CONCLUSIONS:The VHA had better health outcomes than did MA plans. The VHA's performance is reassuring, given its emphasis on equal access to healthcare in an environment that is less dependent on patient financial considerations.

Objective: To determine, for Veterans Health Administration (VA) enrollees who lived and were hospitalized in New York State between 1998 and 2000, the primary payers for their non-VA admissions, whether the primary payer mix varied by condition treated, and whether the Medicare claims data that VA acquired on its Medicare-enrolled patients captured all or most of their non-VA inpatient care. Methods: Using VA and New York State administrative and clinical databases, we conducted a retrospective study examining 75,046 residents of New York State who were enrolled in the VA and had 159,843 inpatient admissions in New York hospitals not in the VA system. For each admission, we determined the major diagnostic category, the primary payer for the admission, and whether the patient was Medicare-enrolled. Our analyses separated veterans into those younger than age 65 and those ages 65 or older. Results: The payer mix for younger veterans' non-VA admissions varied considerably by major diagnostic category. Among veterans who also were Medicare enrollees, Medicare did not pay for 10% of the non-VA hospitalizations of older patients or 20% of those for younger patients. Conclusions: Using only Medicare claims data may significantly underestimate VA patients' reliance on non-VA inpatient care. To better inform planners about VA's service market and diagnosis-specific service utilization patterns across VA and non-VA providers, VA should work with states to develop comprehensive inpatient datasets.

Background: Electronic medical records systems (EMR) contain many directly analyzable data fields that may reduce the need for extensive chart review, thus allowing for performance measures to be assessed on a larger proportion of patients in care.Objective: This study sought to determine the extent to which selected chart review-based clinical performance measures could be accurately replicated using readily available and directly analyzable EMR data.Methods: A cross-sectional study using full chart review results from the Veterans Health Administration's External Peer Review Program (EPRP) was merged to EMR data.Results: Over 80% of the data on these selected measures found in chart review was available in a directly analyzable form in the EMR. The extent of missing EMR data varied by site of care (P < 0,01). Among patients on whom both sources of data were available, we found a high degree of correlation between the 2 sources in the measures assessed (correlations of 0.89-0.98) and in the concordance between the measures using performance cut points (kappa: 0.86-0.99). Furthermore, there was little evidence of bias; the differences in values were not clinically meaningful (difference of 0.9 mg/dL for low-density lipoprotein cholesterol, 1.2 mm Hg for systolic blood pressure, 0.3 mm Hg for diastolic, and no difference for HgbA1c).Conclusions: Directly analyzable data fields in the EMR can accurately reproduce selected EPRP measures on most patients. We found no evidence of systematic differences in performance values among these with and without directly analyzable data in the EMR.

Objectives. Influenza and pneumococcal vaccinations reduce morbidity, mortality, and health care costs, but their use lags behind goals set by public health experts. We evaluated the effect of a performance measurement program instituted by the Veterans Health Administration in 1995 to improve vaccination rates. Methods. We used cross-sectional chart-abstracted data to calculate influenza and pneumococcal vaccination rates among eligible patients, and administrative data to calculate pneumonia admission rates. We compared vaccination and hospitalization rates before and after the institution of the performance measurement program with rates outside the Veterans Health Administration. Results. Influenza and pneumococcal vaccination rates for eligible patients rose from 27% and 28% during 1994 to 1995 to 70% and 85%, respectively, by 2003 (P for trend<.001). Geographic and other variations were substantially reduced. During this time, pneumonia hospitalization rates decreased by 50% among elderly Veterans Health Administration enrollees but increased among Medicare enrollees by 15% (P for differences in trend<.001). Conclusions. The performance measurement program was associated with increases in vaccination rates, reduced variation, and reduced pneumonia admissions. Health systems instituting similarly effective programs may substantially improve the quality of their clinical health care.

OBJECTIVE:We sought to quantify Veterans Health Administration (VA) patients' utilization of coronary revascularization in the private sector and to assess the potential impact of directing this care to high-performance hospitals.METHODS:Using VA and New York State administrative and clinical databases, we conducted a retrospective cohort study examining residents of New York State who were enrolled in the VA and underwent either coronary artery bypass graft (CABG) surgery or percutaneous coronary intervention (PCI) in 1999 or 2000 (n=6562) in either the VA or the private sector. We first calculated the proportion of revascularizations obtained in the VA and the private sector. We then identified the private sector hospitals in which these men obtained revascularizations and determined potential changes in mortality and travel burden associated with directing private sector care to high performance hospitals.RESULTS:VA patients in New York were much more likely to undergo revascularization in the private sector than in VA hospitals: 83% of CABGs (2341/2829) and 87% of PCIs (4054/4665) were obtained in the private sector. Private sector utilization was distributed evenly across high- and low-mortality hospitals. Directing private-sector CABG surgery to high-performance hospitals could have reduced expected mortality by 24% (from 2.3% to 1.7%) and would only increase median travel time from 21 to 30 minutes. The benefit of redirecting PCI care is minimal.CONCLUSIONS:For high-mortality procedures that veterans frequently obtain in the private sector, like CABG, directing care to high-performance hospitals may be an effective way to improve outcomes for veterans.

OBJECTIVES:The goal of this study was to assess racial differences in process of care and outcome for acute myocardial infarction in the VA health care system. DESIGN:Retrospective cohort study using clinical data. SETTING:Eighty-one acute care VA hospitals. PATIENTS:Four thousand seven hundred sixty veterans discharged with a confirmed diagnosis of acute myocardial infarction. The analysis was restricted to 606 black and 4005 white patients. MAIN OUTCOME MEASURES:Comparison of use of guideline-based medications, invasive cardiac procedures, and all-cause mortality at 30 days, 1 year, and 3 years. RESULTS:Black patients were equally likely to receive beta-blockers, more likely than white patients to receive aspirin (86.8% vs. 82.0%; P <0.05), and marginally more likely to receive angiotensin converting enzyme inhibitors (55.7% vs. 49.6%; P = 0.07) at the time of discharge. In contrast, black patients were less likely than white patients to receive thrombolytic therapy at the time of arrival (32.4% vs. 48.2%; P <0.01). There was no significant difference in refusal of angiography or percutaneous transluminal coronary angioplasty between black patients and white patients, or in crude rates of either of these procedures. There was also no difference overall in the percentage of patients who refused coronary artery bypass graft surgery. However, black patients were less likely than white patients to undergo bypass surgery (6.9% vs. 12.5% by 90 days; P <0.001). Black patients remained less likely to undergo bypass surgery even when high-risk specific coronary anatomy subgroups were examined. There was no difference in mortality in the two groups. CONCLUSIONS:In this integrated health care system, no significant racial disparities in use of noninterventional therapies, diagnostic coronary angiography, or short- or long-term mortality was found. Disparities in use of thrombolytic therapy and coronary artery bypass surgery existed, however, even after accounting for differences in clinical indications for treatment and patient refusals. Further work should assess the role of the medical interaction and physician behavior in racial disparities in use of health care.

One way to monitor patient access to emergent health care services is to use patient characteristics to predict arrival time at the hospital after onset of symptoms. This predicted arrival time can then be compared with actual arrival time to allow monitoring of access to services. Predicted arrival time could also be used to estimate potential effects of changes in health care service availability, such as closure of an emergency department or an acute care hospital. Our goal was to determine the best statistical method for prediction of arrival intervals for patients with acute myocardial infarction (AMI) symptoms. We compared the performance of multinomial logistic regression (MLR) and discriminant analysis (DA) models. Models for MLR and DA were developed using a dataset of 3,566 male veterans hospitalized with AMI in 81 VA Medical Centers in 1994–1995 throughout the United States. The dataset was randomly divided into a training set (n = 1,846) and a test set (n = 1,720). Arrival times were grouped into three intervals on the basis of treatment considerations: <6 hours, 6–12 hours, and >12 hours. One model for MLR and two models for DA were developed using the training dataset. One DA model had equal prior probabilities, and one DA model had proportional prior probabilities. Predictive performance of the models was compared using the test (n = 1,720) dataset. Using the test dataset, the proportions of patients in the three arrival time groups were 60.9% for <6 hours, 10.3% for 6–12 hours, and 28.8% for >12 hours after symptom onset. Whereas the overall predictive performance by MLR and DA with proportional priors was higher, the DA models with equal priors performed much better in the smaller groups. Correct classifications were 62.6% by MLR, 62.4% by DA using proportional prior probabilities, and 48.1% using equal prior probabilities of the groups. The misclassifications by MLR for the three groups were 9.5%, 100.0%, 74.2% for each time interval, respectively. Misclassifications by DA models were 9.8%, 100.0%, and 74.4% for the model with proportional priors and 47.6%, 79.5%, and 51.0% for the model with equal priors. The choice of MLR or DA with proportional priors, or DA with equal priors for monitoring time intervals of predicted hospital arrival time for a population should depend on the consequences of misclassification errors.

The Veterans Health Administration is the largest integrated health care system in the United States. Approximately 3 million veterans use this system in any given year.1 Many veterans also have access to health care outside this system through private or public health insurance plans such as Medicare or Medicaid,2,3 and there is increasing awareness of the dual use by veterans of both VA and non-VA health care services.4–9 Research in this area is of interest from an efficiency point of view, as some have proposed that the taxpayer is “paying twice” for care provided to veterans who are enrolled in Medicare HMOs and also use the VA system for care.5 In this issue, Borowsky and Cowper contribute to our understanding about dual use of VA and non-VA primary care services.10 Their examination is important because most people agree that having a single primary care provider leads to better continuity and coordination of health care.11 Borowsky and Cowper found that 28% of the veterans in their sample who reported a relationship with a VA primary care provider were “dual users.” The data for this study came from a telephone survey of a random sample of primarily white, male Minnesota veterans. Two thirds of respondents were users of the Minneapolis Veterans Affairs Medical Center. Strikingly, 50% of primary care visits by dual users were to non-VA providers. Not surprisingly, the odds of dual use were increased for those who had insurance, were more educated, or were less satisfied with VA care. Though type of insurance was not reported, the proportion of veterans with dual use was the same for those over or under age 65, suggesting that Medicare coverage was not the sole explanation for this non-VA utilization. As is common in this type of research, these findings raise more questions than they answer. It would be interesting to know why veterans sought dual primary care, but this study was not designed to answer this question. It also would be interesting to know which provider veterans considered their “primary” primary care provider, especially since the VA has recently reorganized to a primary care delivery model.12 Are taxpayers really paying twice for primary care services?5 Because we do not know which primary care services were received by the patients in this study, we cannot know whether services were duplicated. We need to know because the possibility of fragmented care increases when patients get care from multiple providers without common medical records or mechanisms for the transfer of information. Fragmented care interferes with the delivery of good primary care, whose cardinal features are comprehensiveness, continuity, and coordination.11 Conversely, it is possible that dual use occurs because non-VA plans are providing more convenient access to outpatient care and to services such as ophthalmology, dermatology, and urology, but the VA provides easier access to prescription drugs, mental health care, acute inpatient services, and long-term care. This situation might allow veterans to patch together the spectrum of services they need. Therefore, dual use might be beneficial. Also, dual use may enhance satisfaction by providing choice. An important next step in this type of work will be to determine which services are used by dual users and to compare the outcomes of dual users with those of single users. Of those in this study with at least one primary care visit to a VA facility, 76% reported having either private or public insurance coverage. Yet only 28% were dual users. These data suggest that those who were most dissatisfied and had insurance were the most likely to use non-VA primary care services. Though few VA users were dual users, Borowsky’s study may actually underestimate the percentage of veterans who are dual users for several reasons. First, veterans may have been hesitant to participate in the survey because of worries that dual use might disqualify them from VA services. Similarly, the investigators were not able to confirm the patients’ self-reported utilization of primary care, and patients may have underreported their use of primary care services or may not have understood the definition of “primary care.” Conversely, this study was done in 1993–94, before the implementation of meaningful primary care reorganization in the VA. If this reorganization has been effective in improving patient satisfaction, perhaps dual use has declined. The VA’s capitation-style resource allocation system does not account for utilization of services outside the system. In other work, we have shown that the use of services by elderly veterans under fee-for-service Medicare financing differs significantly by VA service network.13 In another study, VA costs in fiscal year 1996 were $3,118 per patient for those continuously enrolled in Medicare HMOs and $5,547 for patients of the same age who had no HMO coverage.14 Importantly, the numbers of VA outpatient visits per capita were nearly identical for the two groups. What is not clear from these studies is whether veterans are receiving duplicate services or using whatever means necessary to obtain the services they need.15 In any case, these data indicate substantial incentives for managed care plans and the VA to encourage out-of-system use. What should the policy response be to dual use of health care services? It might seem desirable to have a “primary payer” to ensure that there would be no payment for duplicate services. For example, in the private sector, dual eligibility occurs when a subscriber and spouse are both employed and each has health insurance with a different employer. Benefits are coordinated so that both plans do not pay for the same service. No such arrangement exists in the VA. Yet, without understanding the nature of services actually being used by these dual users, it is not clear what the policy should be. Veterans may be behaving in a very rational way, given the constraints of the current health care environment and the fact that many VA users are poor, disabled, and ill. We must ensure that any primary payer policy does not prevent veterans from receiving care to which they are legally entitled. Borowsky and Cowper go so far as to suggest that dual use may be a marker for dissatisfaction with VA services. Yet, in the current health care environment, the demand for VA services can change dramatically and unexpectedly because of changes in market forces such as Medicare HMO capitation payments, managed care penetration, patient selection by HMOs,16 or patient dissatisfaction with HMO care.17 Given that the data in Borowsky and Cowper’s study are limited to white, male, Midwesterners who primarily used a single VAMC, any change in veterans’ health care benefits should await confirmation of these results in a more generalizable sample. Changes also should be delayed until we know more about which services are involved in dual use and what incentives lead to dual use.