Abstract
Background:
Enactment of the Veterans Choice Act (VCA) in 2014 and expansion of Medicaid programs led to greater options for Veterans to receive health care outside of the U.S. Department of Veterans Affairs (VA). However, little is known about how much care women Veterans receive in VA or non-VA settings paid by VA or other insurance types. We explore trends in inpatient care received by women Veterans before and after implementation of the VCA, focusing on hospitalizations, as these patients are often higher risk.
Methods:
Data for women Veterans were linked to hospital discharge data for 2013–2017 from 11 diverse states. Patient and hospital characteristics by VA and non-VA payer and age group (less than 65 years and 65 years and older) were compared for acute hospitalizations in Pearson’s chi-square and analysis of variance tests. We compared the odds of a hospitalization occurring before and after VCA implementation for each specific setting/payer using separate logistic regressions.
Results:
Our sample included 117,177 acute medical/surgical hospitalizations. For younger women, the most common hospital setting/payer was VA hospitals (36%); older women most frequently received hospital care in non-VA hospitals covered by Medicare (61%). After the VCA, the probability of receiving VA-purchased care increased significantly for older women only. Younger women saw a significant decrease in VA hospitalizations post-VCA.
Conclusions:
We found that after multiple expansions in non-VA health care options, women Veterans sought more VA-purchased community hospital care.
Introduction
After widely publicized problems in access to care in the Veterans Affairs (VA) Health Care System, the Veterans Access, Choice, and Accountability Act (Choice Act) was passed in 2014 to allow more Veterans to obtain care in the community if they met certain wait time and travel distance criteria. In addition, the Affordable Care Act’s (ACA) Medicare expansion became effective beginning in January 2014, extending coverage for low-income adults and providing additional options for Veterans to receive care outside of the VA. 1
One prior study of women Veterans’ use of VA and private sector inpatient services in New York state between 1998 and 2000 found that women who used the VA for inpatient services were more likely to be Black, younger, have lower comorbidity scores, have a service-connected disability, and live in urban settings. 2 Of the 2,913 hospital stays analyzed, they found that 44% of admissions took place at the VA versus 56% at non-VA settings. A more recent study of five states found that the Choice Act and ACA led to decreased use of VA hospitals and increased use of non-VA hospital care covered by the VA and Medicaid; however, the study looked at the Veteran population as a whole and did not explore the effects of these recent policies on women Veterans. 3 Women are the fastest-growing group of Veterans using VA services and comprise 30% of all new VA patients. 4 Many women also obtain VA-purchased care in the community when gender-specific services cannot be provided at a VA clinic: in fiscal year 2010, one of every three women Veterans received care purchased by the VA. 5 With recent access expansions, it is expected that women Veterans have increased their use of non-VA hospitals and decreased their use of VA hospitals, similar to all Veterans; nevertheless, there may be differences by type of care and condition treated.
In this article, we seek to describe trends in care before and after the implementation of the Veterans Choice Act (VCA) and explore how extended coverage affected women Veterans’ use of inpatient care in 2013–2017. We build on prior research by using a unique dataset of VA and non-VA hospitalizations for Veterans living in 11 diverse states that represent approximately 40% of VA enrollees. 6 We include VA care and non-VA care through VA-purchased, Medicare, Medicaid, commercial, and other payers to provide a comprehensive picture of where women Veterans received care before and after implementation of the VCA.
Methods
Cohort and Data Sources
We identified all women Veterans enrolled in VA health care during calendar years 2013–2017 in 11 geographically diverse states (AZ, CA, CT, FL, IL, LA, MA, MO, NY, PA, SC) and obtained their non-VA hospital records from all-payer discharge data. We linked VA enrollment data to the 11-state discharge data using personal identifiers, and our subsequent sample included all utilization of acute inpatient services within and outside the VA for women Veterans enrolled in the VA. Women Veterans’ VA utilization and cost records were obtained from Inpatient Encounter Files and Managerial Cost Accounting files in the VA Informatics and Computing Infrastructure.
We identified VA acute hospital stays from medicine and surgery bed sections and diagnosis-related groups. Stays within 30 days of discharge from a prior admission and stays longer than 180 days were excluded since they were not considered index or acute stays. VA and non-VA hospitalizations were assigned to diagnosis groups using principal diagnosis codes and the Agency for Health care Research and Quality Clinical Classifications Software. 7 Hospital stays were grouped into 11 categories: reproductive (e.g., pregnancy/delivery, ovarian cysts, fibroids), cancer, cardiology, endocrinology, gastroenterology, infectious disease, musculoskeletal, neurology, respiratory, urinary, and other conditions. Elective and nonelective admission information for non-VA stays was obtained from state discharge records. As VA data does not specify whether admissions were elective or not, VA admission information was populated using a probabilistic method based on non-VA admissions. 8 We measured patients’ comorbidity scores for each stay using the Elixhauser-van Walraven index calculated from all recorded diagnosis codes. 9
Patient sociodemographic characteristics, including age, gender, race/ethnicity, marital status, priority group, and rurality were obtained from VA Health Enrollment Files, VA Observational Medical Outcomes Partnership Files, and Assistant Deputy Under Secretary for Health Enrollment Files. 10 VA priority groups impact Veteran copays and are assigned based on military service history, disability rating, income level, Medicaid eligibility, and receipt of other VA benefits. 11 Area-level median household income was obtained from the U.S. Census. Veteran and VA hospital addresses were obtained from the VA Geospatial Services Support Center Files; non-VA hospital addresses were obtained from the Centers for Medicare and Medicaid Services Provider of Service Files; and Veteran death information was obtained from the VA Vital Status File. The post-VCA period was defined as beginning in 2015, the first full implementation year of the program that began in November 2014.
Analytic Methods
All patient and hospital characteristics by VA and non-VA hospitals and age group were compared in Pearson’s chi-square and analysis of variance tests. In an analysis of the use of VA or non-VA hospitals, we conducted a series of logistic regressions of all women Veterans enrolled in the VA and their likelihood of using a VA or non-VA hospital (by payer). For non-VA hospitals, we estimated their likelihood of using a non-VA hospital covered by VA-purchased care, Medicare, Medicaid, or private insurance in separate models. We adjusted for factors potentially influencing use of VA and non-VA hospitals that included post-VCA, linear yearly trend, patients’ age, marital status, rurality, priority for VA care, and area-level income. The study was approved by a large private university in the U.S. and two public U.S. university IRBs with a waiver of consent granted by the IRBs.
Results
Patient Characteristics
Characteristics of the study sample population are shown in Table 1; our sample included a total of 117,177 acute medical/surgical hospitalizations from 2013 to 2017. Across all six setting/payer categories, White women were the largest racial/ethnic group among Medicare-covered patients at 75%, followed by 63% of VA-purchased care, 61% of VA, 57% of other insurance types, 55% of commercial care, and 47% of women receiving care through Medicaid. Black women comprised 33% of Medicaid-covered patients, compared to 26% of VA patients, 23% of other-coverage patients, 22% of VA-purchased and commercial insurance patients, and only 12% of Medicare-covered patients. Women were more likely to be divorced, widowed, or separated in most payer categories except for commercial insurance, where 42% of women were married compared to 31% divorced/widowed/separated. Median area-level income was similar for hospitalizations of all payers, ranging from $53,058 (SD 21,144) for VA patients to $59,475 (SD 21,115) for commercially-insured patients. Mean age and comorbidities differed among women Veterans; patients admitted to VA hospitals were older (57 years [SD 16]) than those hospitalized outside of the VA paid by other types of insurance, with the exception of Medicare (74 years [SD 15]). Women receiving community inpatient care purchased by the VA were, on average, 10 years younger (47 years [SD 17]) than those who obtained care directly through the VA. Women with Medicare-covered hospitalizations had the highest comorbidity scores, with a mean Elixhauser-van Walraven score of 6.5 (SD 7.4); second highest were VA-hospitalized women with a mean score of 2.4 (SD 5.7), followed by VA-purchased (2.1 [SD 5.9]), Medicaid (1.2 [SD 5.6]), commercial (1.4 [SD 5.2]), and 1.3 (SD 5.0) for other payer-covered hospitalizations.
Unweighted Patient and Hospital Characteristics of Women Veterans Hospitalized in VA and Non-VA Settings, 2013–2017
p-Values reported for Pearson’s chi-square tests for categorical variables and analysis of variance tests for continuous variables.
There was variation in the proportion of women with service-connected disabilities across VA and non-VA hospitals and payers. Over half (51%) of women with VA-purchased and other hospitalizations had the highest service-connected disabilities (priority groups 1 and 2) compared to 45% of women hospitalized in the VA, 38% in commercial, 30% in Medicaid, and 23% in Medicare-purchased hospitalizations. All results p < 0.001.
Location played a role in how Veterans received care. Women in VA hospitals were least likely to live in rural areas (16.4%). Women covered by Medicaid (21.1%) and other insurance (22.3%) were most likely to live in rural areas.
Type of admission also differed across payers: hospitalization for elective inpatient care was more common in commercial patients (36%), followed by 28% of Medicaid and 25% of VA-purchased patients, 18% of VA patients, and 15% of Medicare patients (all p < 0.001). Patients were hospitalized for nonelective care most commonly among Medicare patients (80%), followed by VA-purchased (70%), VA (69%), Medicaid (68%), and commercial patients (60%) (all p < 0.001).
Comparison of Hospital Admission by Setting/Payer and Diagnostic Category
When considering hospitalizations by diagnostic category, VA hospitalizations compared with VA-purchased community hospitalizations had a higher proportion of admission due to cancer (4.4 vs. 2.8%), cardiovascular (15.2% vs. 12.9%), endocrine (5.3% vs. 2.7%), gastrointestinal (16.5% vs. 9.7%), musculoskeletal (13.3% vs. 9.8%), neurological (5.5% vs. 4.7%), respiratory (9.7% vs. 7.1%), and urinary-related (5.4% vs. 3.1%) issues (Fig. 1). The greatest difference in diagnostic category between VA and VA-purchased care was for reproductive health: these hospitalizations only accounted for 7% of VA hospitalizations but over 34% of VA-purchased hospitalizations.
Average Percentage of Hospitalizations in Each Diagnostic Category by Setting/Payer for Women Veterans, 2013–2017.
For most conditions, the most frequent setting/payer of hospital care was either VA or Medicare (Fig. 2). The VA provided care most frequently for cancer (38%), gastrointestinal (34%), and endocrine-related hospitalizations (33%). Medicare provided the most care for infectious disease (53%), urinary (46%), cardiology (45%), respiratory (45%), musculoskeletal (38%), and neurology-related (31%) issues. The exception to this was reproductive health; these hospitalizations were more likely to be covered by commercial insurance (32%), other insurance (23%), VA-purchased care (19%), and Medicaid (16%).
Average Percentage of Hospitalizations Provided by Setting/Payer for Each Diagnostic Category for Women Veterans, 2013–2017.
Comparison of Hospital Admission by Age Group
Overall, women younger than 65 years were significantly more likely to receive inpatient care at the VA, with 36% of hospitalizations among younger women Veterans occurring in VA hospitals compared to 29% of hospitalizations of older women during our study period (Fig. 3). Younger women had 18% of stays covered by commercial insurance, followed by other insurance (15%), VA-purchased (13%), Medicare (10%), and Medicaid (7%). For Veteran women older than 65, these patients overwhelmingly received inpatient care at hospitals paid by Medicare (61%). Relatively few hospitalizations of older women were covered by VA-purchased care (5%), commercial insurance (3%), other insurance (2%), and Medicaid (0.2%).
Hospitalizations of Women Veterans by Setting/Payer and Age Group, 2013–2017.
In adjusted analyses of all women enrolled in the VA, younger women Veterans experienced a decrease in the probability of VA hospitalization before and after the implementation of the Veterans Choice Act (VCA) from 2.5% (CI: 2.4, 2.6) to 2.3% (CI: 2.2, 2.4), but older women did not see a significant change after the VCA, from 1.7% (CI: 1.6, 1.8) to 1.8% (CI: 1.7, 1.9) (Fig. 4
Probability of Hospitalization by Payment Source Pre- and Post-Veterans Choice Act, by Age Group.
We also found in adjusted analyses that rural Veteran women were significantly less likely to receive VA hospital care than urban women (OR: 0.73, p < 0.001). There was no significant difference for rural patients’ likelihood of VA-purchased, Medicare, and Medicaid-covered hospitalizations compared with urban patients.
Discussion
This is the first study to compare women Veterans’ utilization of inpatient care in VA and non-VA hospitals before and after implementation of the VCA. In our longitudinal study using a comprehensive, population-based dataset of 11 diverse states, we found small but significant differences in younger and older women’s utilization of VA and non-VA care after the VCA. Only older women experienced a statistically significant increase in VA-purchased care after the VCA. However, only younger women saw a significant decrease in VA hospitalizations post-VCA. The small number of VA-purchased hospitalizations for younger women may have limited our ability to detect significant differences in VA-purchased care post-VCA. Overall, utilization for Medicare- and Medicaid-covered hospitalizations remained similar for both groups throughout the study period. Our results demonstrating women Veterans’ increased use of VA-purchased care in the community are consistent with previous evidence on all Veterans.3,12 In addition, we found that VA hospitals provided care more often than non-VA hospitals for issues related to cancer, endocrine, and gastrointestinal systems. When considering total hospitalizations during the 2013–2017 study period, VA and VA-purchased care provided the largest share of inpatient care for women Veterans compared with other payers.
Our findings differ from a prior study that found women Veterans who used the VA tended to be younger, Black, and more likely to be service-connected; we found that women using the VA tended to be older, White, and less likely to be service-connected, especially compared with those who sought VA-purchased care. The prior study utilized data from New York state for only VA and private hospital admissions between 1998 and 2000, prior to VCA expansion of community care when overall VA community care spending was significantly lower. It is likely that our multi-state dataset and recent study period explain the shift in demographics.
Our findings contribute new information not included in a previous study of five states that looked at the entire Veteran population. We found changes in VA-purchased hospitalizations and VA hospitalizations associated with the VCA that differed for older and younger women. Our data suggests that Medicaid covered more care for women than men Veterans, at an average of 6% of hospitalizations for women versus 2% of all Veterans in the prior study. Medicare also covered fewer hospitalizations among women (30%) than the entire Veteran population (54%). The younger average age of women Veterans compared with men Veterans likely accounts for the difference in results.
There are several limitations to our study. Our data does not extend past 2017, so our results may not accurately represent the population of Veteran women currently enrolled in the VA. As our dataset only includes hospitalizations from 11 states, our findings may not be representative of women Veterans nationwide. In addition, our data only includes 2 years before and after the implementation of the VCA, and it is therefore difficult to attribute the observed hospitalization patterns solely to the effect of the VCA. Furthermore, the Medicaid expansion was implemented at a similar time frame as the VCA so it is possible that observed effects may be associated with this expansion.
Our results have important implications, especially with the implementation of the 2018 Maintaining Internal Systems and Strengthening Integrated Outside Networks (MISSION) Act, which further opened the doors for Veterans to receive community care by significantly expanding access to VA-purchased care. As the shift toward VA-purchased community care continues, our findings are especially relevant for practice and policy. Dual coverage by VA and other health insurance and expansions of VA-purchased care may lead to greater utilization of non-VA care for VA enrollees that may increase the risks of care fragmentation. Studies suggest that Veterans receiving care from both VA and non-VA providers have an increased likelihood of rehospitalization.13–15 In addition, it is possible that Veterans may experience worse health outcomes without well-developed community care networks and quality standards. 16 Strengthening structural resources and coordination around community inpatient services to ensure smooth transitions between VA and non-VA care may be a valuable step toward mitigating these risks.
Conclusions
Our study evaluated the effects of multiple expansions in access to non-VA health care for women Veterans and found a trend towards increases in VA-purchased community care. To prevent adverse health outcomes, it is important for the VA to monitor the usage of non-VA forms of insurance to maintain smooth care coordination and high quality of care.
Footnotes
Acknowledgments
The authors acknowledge the Arizona Department of Health Services, California Department of Health Care Access and Information, Connecticut Department of Public Health, Florida Agency for Health Care Administration, Illinois Department of Public Health, Louisiana Department of Health, Massachusetts Center for Health Information and Analysis, Missouri Department of Health and Senior Services, New York Statewide Planning and Research Cooperative System, Pennsylvania Health Care Cost Containment Council, and South Carolina Revenue and Fiscal Affairs Office for providing data for this study.
Authors’ Contributions
The authors confirm contribution to the article as follows: E.P.W., J.Y.—study conception and design; J.Y.—data collection; E.P.W., J.Y.—analysis and interpretation of results; E.P.W., J.Y.—draft article preparation. All authors reviewed the results and approved the final version of the article.
Disclaimer
The views expressed in this article are those of the authors and do not necessarily reflect the position or policy of the Department of Veterans Affairs or the United States government or any of the state agencies or universities listed above. The Pennsylvania Health Care Cost Containment Council (PHC4) is an independent state agency responsible for addressing the problem of escalating health care costs, ensuring the quality of health care, and increasing access to health care for all citizens regardless of ability to pay. PHC4 has provided data to this entity in an effort to further PHC4’s mission of educating the public and containing health care costs in Pennsylvania. PHC4, its agents, and staff, have made no representation, guarantee, or warranty, express or implied, that the data—financial, patient, payor, and physician specific information—provided to this entity, are error-free, or that the use of the data will avoid differences of opinion or interpretation. This analysis was not prepared by PHC4. This analysis was done by Emily Wong and her team. PHC4, its agents and staff, bear no responsibility or liability for the results of the analysis, which are solely the opinion of this entity.
Author Disclosure Statement
The authors declare that they have no conflicts of interest.
Funding Information
Funding for this work was provided by Health Services Research and Development, Veterans Health Administration, Department of Veterans Affairs, Award #IHX002653-01A1. The funding organization had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the datal preparation, review, or approval of the article; or decision to submit the article for publication.
Abbreviations Used
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
