Perceptions of U.S. Veterans Affairs and community healthcare providers regarding cross-system care for heart failure

Abstract

Objectives

This study explores perceptions of US Veterans Affairs (VA) and non-VA healthcare providers caring for Veterans with heart failure (HF) regarding Veteran knowledge and motivations for dual use, provider roles in recommending and coordinating dual use, systems barriers and facilitators, and suggestions for improving cross-system care.

Methods

Twenty VA and 11 non-VA providers participated in semi-structured interviews, which were analyzed using parallel qualitative content and discourse analysis.

Results

VA and non-VA providers described variable HF knowledge and self-management among Veterans, and both groups described the need for improved education addressing medication adherence, self-care, and management of acute symptoms. Both groups described highly limited roles for providers in shaping choices surrounding dual use. VA and non-VA providers had significantly different perceptions regarding the availability, quality, and effectiveness of VA HF services. Multiple non-VA providers expressed frustration with and difficulty in contacting VA providers, accessing records, and making referrals into the VA system. Suggestions for improved care focused on patient education and care coordination.

Discussion

Dual healthcare system use for Veterans is increasingly common. Similarities and contrasts in perceptions of VA and non-VA providers are instructive and should be incorporated into future policy and program initiatives.

Keywords

Veterans decision making heart failure qualitative research dual use

Introduction

Heart Failure (HF) is a highly prevalent chronic medical condition prone to frequent acute exacerbation that currently affects an estimated 6.5 million U.S. adults.¹ Heart failure causes approximately 1.1 million hospitalizations and 3.4 million ambulatory care visits each year at an estimated annual cost of $30.7 billion.² Among patients hospitalized for HF, almost one quarter (i.e., 23.8%) are seen in an emergency department (ED) within 14 days of hospital discharge, and one-fifth (20.7%) are readmitted within 30 days of hospital discharge.³ Approximately half of all HF patients die within five years of diagnosis, and one in one deaths have included HF as a contributing cause.⁴ Heart failure is also highly prevalent in the Veteran population, where it is a frequent cause for hospital admission and one of the most frequent causes of unplanned hospital readmission.⁵

For most Americans, health care is scattered across multiple providers and health care systems. While care across settings is often necessary and desirable, emerging evidence suggests that such fractured care can be less efficient and less safe.^6–9 Within the U.S. Veterans Health Administration (VA), dual healthcare system use (dual use) occurs when Veterans enrolled for VA care also receive care from non-VA providers or healthcare facilities, and it is particularly common. Weeks and colleagues examined dual use for cardiovascular disorders among Veterans enrolled in VA care in New York State and found that younger Veterans relied solely on VA hospitals for 41% of their care, while older Veterans relied solely on VA hospitals for only 27% of their care.¹⁰ Among Medicare HF patients rehospitalized within 30 days, studies estimate that 19% are readmitted to a different hospital the second time around.¹¹ Also, a recent study of Veterans with HF found significantly higher rates of ED visits, hospitalizations for HF, and hospital readmissions among dual users as compared to single-system users.¹² Thus, the VA population is a reasonable one in which to study the phenomenon of cross-system care using HF as a model chronic disease.

The reasons for cross-system care, or dual use, are complex and include patient, provider, and systems-level drivers. Dual use is important in the VA given recent passage of the Veterans Access, Choice, and Accountability Act of 2014 (VACAA), which has significantly increased the care of Veterans in non-VA systems, including an expected expansion of non-VA hospitalizations.¹³ This line of investigation is also timely given active Center for Medicare and Medicaid Services (CMS) programs focused on improving hospital care quality and to promoting population health management. These actions include penalties related to all-hospital 30-day readmission rates and ongoing Accountable Care Organization (ACO) demonstration projects.¹⁴ Thus, examination of cross-system care is relevant both in VA and also outside the Veterans Health Administration as well.

To date, most studies of dual use have been retrospective studies using administrative data.^6,15–19 While a few studies have assessed drivers of dual use from the patients' perspective,²⁰ we are aware of no studies that assess drivers of dual system use from the perspective of providers in both VA and non-VA settings. In order to address this gap in knowledge, we performed a qualitative research study with a group of VA and non-VA providers as a source of information to better understand different providers' perspectives regarding Veteran knowledge and motivations for dual use, provider roles in recommending and coordinating dual use, systems barriers and facilitators of dual use, and suggestions for improving systems of care.

Methods

Ethics approval

Initial human subjects' protection approval was obtained from the Department of Veteran Affairs (VA) Central Institutional Review Board (CIRB), followed by review from the local VA Research and Development (R&D) Committee. The first CIRB approval (12 January 2013) was followed by approved Continuing Reviews (10 March 2014, 10 March 15 and 10 March 16) as part of a larger mixed methods study.

Study participants

VA providers were recruited from three VA medical facilities in two Southern states, including inpatient, primary care, specialty care and emergency department services. Non-VA providers were recruited from similar geographic and hospital-based facilities in the same states using a similar purposive sampling methodology. The size of our qualitative sample was informed by several practical factors including our specific aims, the specificity provider experiences, repetitions in interview dialogue, and our use of parallel analysis strategies.²¹ The participants included 20 VA and 11 non-VA providers, who are described by demographic category in Table 1. The smaller non-provider sample reflects a more rapidly reached consensus of perceptions, with no new concepts emerging. The use of two persons coding the interviews separately while entering codes and categorizing in an electronic spreadsheet that was then compared and negotiated functioned as a saturation grid, as Fusch and Ness describe as a means of confirming saturation has been achieved in justifying sample size.²²

Table 1.

Characteristics of study participants.

	VA Providers	Non-VA Providers
Occupation
Cardiologists	3	1
Emergency Department Physicians	5	5
Hospitalists	3	1
Nurse, Nurse-practitioner or Nurse
Manager	3	2
Pharmacists	1	0
Physician Assistant/ Primary Care	1	0
Primary Care Physicians	4	2
Age
25–34	3	0
35–44	4	3
45–54	6	2
55–64	6	5
65 and older	1	1
Race
American Indian and Alaska Native	0	0
Asian	4	0
Black or African American	0	1
Native Hawaiian and Other Pacific
Islander	0	0
Other	1	0
White	15	10
Ethnicity
Hispanic	1	0
Non-Hispanic	19	11
Gender
Female	7	6
Male	13	5
Total	20	11

Semi-structured interviews

In semi-structured interviews lasting 30–60 min in duration, VA and non-VA providers were asked 11 open-ended questions prompting reflection on how Veterans make decisions to seek care for their HF exacerbations and what Veterans understand about their condition and other topics (Table 2). All interviews were done in one-to-one format and were recorded and transcribed for coding, analysis and interpretation. The interview guide was first developed using recommendations from the Qualitative Research Guidelines Project of the Robert Wood Johnson Foundation to use concepts from a focused research question, experience from preliminary research and team qualitative expertise. Questions were developed iteratively with feedback from persons similar to study participants.²³ Members of the research team with current and past experience as clinical providers reviewed and provided feedback on the questions for revision.

Table 2.

Interview guide.

How do Veterans usually wind up in your facility or emergency room with HF problems?

Describe a typical example.

What do the Veterans commonly reflect what they know about their HF or call their condition?

What is your impression about how prepared Veterans are to manage their heart failure at home?

What has been your experience with how Veterans with HF make decisions about when and how to seek care for HF? At your facility? At (other: VA or non-VA)?

What role does primary care (actual and perceived) play on their care choice?

How do you usually decide which HF patients to hospitalize and which to send home?

How do you perceive the quality of heart care for Veterans in VA versus non-VA facilities?

How does use of a non-VA hospital by a Veteran during an acute episode of HF exacerbation affect the use of hospital services or the longer term outcome of the HF?

What has been your experience communicating with or coordinating care with the VA (or non-VA facilities) for a Veteran with HF?

What kind of intervention would it take to decrease avoidable hospitalizations for Veterans with HF?

What additional comments would you like to share with us at this time?

VA: Veterans Affairs; HF: heart failure.

In keeping with an approach developed for sociological methods, inter-rater agreement was negotiated by consensus using a matrix that standardized the units of text between the two parallel qualitative methods used by the coders: qualitative content analysis and discourse analysis.²⁴ This approach to inter-rater agreement emerges from an approach to qualitative data analysis that uses principles of inductive reasoning that support the parallel development of code types for a negotiated consensus for data analysis and interpretation.²⁵

Data analysis

Each cohort of provider interviews was coded separately into perceptions of VA providers and non-VA providers by a coder using qualitative content analysis and a second coder using discourse analysis. Our choice to combine content analysis in parallel with discourse analysis warrants some discussion. Hsieh and Shannon explain that qualitative content analysis “is defined as a research method for the subjective interpretation of the content of text data through the systematic classification process of coding and identifying themes or patterns.”²⁶ In that classification, our approach is considered summative content analysis (pp.1283–1284), as it goes beyond quantifying content in the text to examining its context as part of its interpretation. As summarized by Schreier, content analysis in quantitative research is usually considered a method for data collection to be counted, while in qualitative research content analysis counts as a method for data exploration and analysis.²⁷ This method organizes and elicits meaning from the data collected and draws conclusions from a process of examining the context of what was said, establishing categories and compiling results.²⁸

Though there are a variety of approaches to discourse analysis, we used the type of discourse analysis that focuses on ways of speaking and that incorporates pragmatics, or an emphasis which takes into consideration the contexts surrounding the speakers from which inferences and references are drawn.²⁹ In this type of discourse analysis, coding identifies categories of language used in interviews that denote how the speakers think about their world, how they position themselves, the footing or stance they take in relation to themselves, others and actions they take, and choices in speaking that signal identity, beliefs, and attitudes. As an example, for the code “positioning”, in examining transcripts line by line, pronoun use and exclusions frequently identified how providers positioned perceptions about themselves and the Veterans. In referring to Veterans as “they,” Veteran HF decisions were portrayed as internally related to listed symptoms, with little reference to conditions at home, resources, family, or contextual factors in which symptoms escalated and decisions were made. By contrast, in referring to Primary Care or hospital services, providers used “we” to describe the array of services offered, even when the services were inaccessible or not their service, positioning themselves as extensions of the institution.

Table 3 shows a sample of speech acts, first derived from Searle (#1–5) and some others described subsequently in discourse analysis, sought in text.³⁰

Table 3.

Examples from discourse analysis.

Selected types of speech acts	Examples (Quotations in italics)
1. Expressives: Something that signals the speaker's emotions	Non-VA cardiologist expressing frustration: “When we have prescriptions, it's very hard to get the prescriptions and recommendations for medicine changes from their primary care so they can get written for. It's very, very difficult. It's hard to get information from the records as far as what… ”
2. Adverbs and adjectives: Grammar that marks attitudes, viewpoints, and perceptions	VA ED Physician: “I would say they're a pretty educated population. Most of them have graduated high school so I think they have all of the tools to actually manage their heart failure because we can provide a lot at the VA. So there's really nothing lacking for them to do it.”
3. Genre: A metaphorical frame that identifies discourse associated with a particular sociocultural role or convention	Quote from VA ED physician: “I typically don't ask them their knowledge base.”
4.Narrative: A story in conversation with a purpose	Quote form VA Primary Care Physician: “So, for me one of the base issues—one of the most helpful ways for them to control their heart failure symptoms at home is to weigh themselves daily cause it's a simple thing to do if they got an active—you know a scale at home. They can step on it; they weigh themselves daily; they track their weight and if their weight's going up, unless they've been a lot, just like the fluid accumulation and if they are able to contact their doctor or adjust their meds a little bit they might be able to avoid hospitalization or maybe avoid a ED visit. So I think that's the easiest thing for them to do, but it seems like I'd say very often most of them are not doing that.”
5. Register: Discourse features that signal social class, age, gender, profession, culture, etc.	Quote from VA ED physician: “The primary care is just overwhelmed. They're understaffed, underpaid, overworked, and there's only so much they can do. You can't drag them in by the collar and say “It's your turn to come to clinic.”
6. Stance-taking: A speaker's choice of words, silences, hedges, stressors or tone that create a position or attribute a position for others	Quote from VA Nurse Practitioner: “I think it can be managed at home. I think the VA provides them with the necessary… Whatever they need, home oxygen a scooter….” [Stance-taken identifies VA perspective, with listing of things]

VA: Veterans Affairs; HF: heart failure; ED: emergency department.

Grouped by questions, codes from the qualitative content analysis were compared to emerging codes for discourse analysis and a consensus agreement was negotiated. First, two electronic spread sheets (VA and non-VA providers) listed the content categories that emerged from each question, and a coding structure was developed, consistent with what has been described as an integrated approach to developing a code structure with both inductive and deductive components.³¹ Then, the responses were examined using discourse analysis to classify ways of speaking that marked concepts, contexts, positioning, inferences, and references. The two coders met to compare codes and distribute codes into common categories and create a finalized coding structure, negotiating consensus agreement, as Bradley and colleagues describe.³¹ These steps created a taxonomy or system for classifying the categories into four emerging themes: patient knowledge and motivation; provider roles in dual use; systems aspects of dual use; and, opportunities for improvement. These themes are compared between VA and non-VA providers in Table 4, with excerpts from representative quotations that illustrate the perceptions of providers.

Table 4.

Key themes related to dual system use as articulated by VA and non-VA healthcare providers.

	VA	Non-VA Providers
Patient knowledge and motivation	Patient HF knowledge varies depending on multiple factors, including education, family support systems, and literacy Variable views on Veteran compliance with HF medications and self-care Low literacy a common problem Veterans with multiple comorbidities poorly equipped to manage HF care at home	Variable estimates of Veteran HF knowledge Variable estimates of adherence to HF medications Generally negative perceptions of regarding HF self-management including weight/fluid monitoring Perception that many Veterans not actively involved in their own healthcare Belief that many Veterans are not prepared to manage HF at home
Provider Role in Dual Use	Uncertainty surrounding role of VA providers in dual use Triage to ED for escalating symptoms Referral for specialty care Belief PCPs do not have much influence regarding dual use Desire by PCPs that VA and non-VA ED doctors should consult with them PCP plays role in giving patients criteria for ED visits	View limited role for Primary Care in decisions regarding dual use Belief that PCP does not play a role in the veterans ER choice. Believe veterans tend to choose Non-VA Primary Care and tend to use VA for medications and specialty care Belief that Primary Care should play a significant role and could manage HF better
Systems Aspects of Dual Use	Perception that the healthcare system is set up for frequent ED use Generally positive perceptions regarding access to Primary Care helps veterans to manage Positive views of VA HF programs including HF clinics and telemonitoring	Perceived barriers in timely access to same-day care and procedures for HF Belief that the VA system not set up to help Veterans with escalating HF symptoms Attempts to communicate with VA described as difficult and time consuming Perceived difficulty accessing VA records Frustration with obtaining financial reimbursed from VA
Opportunities for Improvement	Improved education of Veterans regarding HF More frequent follow up, especially after ED visits or hospitalization. Veterans needs more frequent follow-up Suggestions for improved care coordination across care settings Suggestions for home-based HF care	More education with Primary Care doctor Improved access to VA Primary Care Better teamwork and communication Develop a HF care coordinator Medication delivery to homes, increased follow-up

VA: Veterans Affairs; HF: heart failure; ED: emergency department.

Results

Overall, VA and non-VA providers had similar ranges of perceptions regarding Veteran knowledge of HF and their motivations for dual use. Providers in both groups endorsed that the status of Veterans' understanding of their HF condition was highly variable. For instance, a VA primary care physician stated, “I would say they're a pretty educated population. Most of them have graduated high school so I think they have all the tools to actually manage their heart failure….” By contrast, a VA nurse practitioner stated, “Well, they are typically well under-educated… They need lots of education, reality checks on what's actually happening to you, and the intervention to manage it.” Providers in both groups pointed to differences in Veteran levels of education, the social factors in their lives, health literacy, resources available, and family support systems as factors that affected both the capacity to understand and manage HF as well as choose sources of health service. Veterans with more formal education were assumed to be more likely to manage their HF. This variability led to conclusions that some Veterans were poorly equipped to manage their heart failure or cope at home after hospital discharge, with one estimate as high as only 50% capable of HF self-management. There was a wide spectrum of belief among VA and non-VA providers regarding compliance with medications, monitoring weight and fluid retention, and response to escalating symptoms. Some providers saw Veterans as compliant, with one non-VA provider suggesting their military experience accustomed them to following orders. Others saw emergency department use and readmissions as evidence of Veteran non-compliance, a more dominant belief among VA providers.

Both VA and non-VA providers expressed uncertainty in identifying the role they or their colleagues played in initiating or coordinating dual system use. For instance, when asked what role primary care providers played in choice of site of care for Veterans with heart failure, several interviewees asked for clarification or stated they did not understand the question. One typical VA provider responded, “…I don't think primary care plays a lot of roles.” Once clarified, several VA and non-VA primary care providers described examples limited to triage in acute exacerbation and referral for specialized procedures (regardless of site). However, one non-VA emergency provider noted that Veterans themselves seemed to use services across systems selectively:

When it comes to needing prescription refills and things like that they're going to go back to the VA so they can get the medications paid for, of course. And probably more serious problems that require subspecialty care, they'll probably go back into the VA, but the [community] primary care physicians play a significant role in their care.

This role uncertainty regarding dual use suggested areas of communication between services managing the same Veterans with HF that would benefit from support and enhanced coordination. Though non-VA providers also saw little current role for primary care in improving the process of HF care, they were more likely to recommend that primary care providers should be more involved. For the most part, both VA and non-VA providers perceived the decision to hospitalize a Veteran with HF with escalating symptoms to be a matter of clinical symptoms and lab test results, with little reference to the Veteran's role in decision-making, resources, their fears or feelings.

A third area explored in interviews with VA and community providers related to systems aspects of dual use. VA providers voiced overall satisfaction with care coordination, the availability of cardiology and nurses available for education, and appreciation for the VA Care Coordination Home Telehealth (CCHT) service that is frequently used to monitor Veterans with HF.^32,33 For instance, one VA primary care physician said:

We also have Telehealth here at the VA which is pretty responsive, if they get patients enrolled quickly, and [we] monitor them through the phone system… I think that when patients get to be end-stage or need palliative care, I think the VA is good about getting them on hospice when needed. I think, for the most part, I haven't had a lot of problems with my heart failure patients in the VA.

While VA providers tended to believe that the VA system offered a variety of resources to meet the needs of Veterans with HF, non-VA providers offered a different perspective about VA care. Non-VA providers often felt Veterans accessed community healthcare because of difficulties accessing VA primary care or getting timely appointments for related services. One community emergency physician opined, “I think they need to have efficient, easy access to see their doctor in clinic. A nurse to follow them would be outstanding.” To some degree, these concerns unmasked lack of awareness of available VA services since the VA Medical Center serving the catchment area for this provider's ED actually did have a nurse-practitioner-led heart failure follow up clinic. Other community providers described difficulty communicating with the VA as time-consuming and frustrating. One community nurse-practitioner working in primary care stated:

It is extremely frustrating. You can call, and they might call you back. Or they might not call you back. If they're on vacation, no one's covering them. And the Veterans unfortunately are staying in the hospital for many more days than what might be needed just because we can't get hold of someone to do the care that's needed or we can't get a hold of someone. So, we really struggle with contacts.

Some community providers described difficulty coordinating medications and other treatments with VA. “When we have prescriptions, it's very hard to get the prescriptions and recommendations for medicine changes from their primary care so they can get written for. It's very, very difficult,” referring to difficulty managing without access to VA medical records. Additional comments focused on difficulty obtaining VA medical records, arranging follow up testing in VA, and occasionally in obtaining reimbursement from VA for services.

Finally, patterns emerged regarding providers' recommendations for changes to improve cross-system healthcare. The most common recommendations from both VA and community providers focused on improving patient education for HF management. According to one VA emergency physician, “Well, compliance. Anything that would improve compliance and education so the patient does understand.” Many VA providers also recommended better follow-up and use of current services, such as the care Home-based Primary Care which features visiting nurses,³⁴ and more use of the CCHT. Both groups of providers called for more care coordination for Veterans receiving cross-system care. A comment by one community nurse-practitioner summarizes this theme: “I definitely think if we could work together as a team and do more of a team approach for the patient it would be much better. More clear communication between providers would be better.”

Discussion

Our study is one of the first to explore the perceptions of healthcare providers who collaborate in the care of patients receiving treatment across healthcare systems, and it is the first to specifically explore HF care. We observed several areas of concordance in views expressed by VA and non-VA providers. For example, both sets of providers described varied experiences in the level of HF knowledge and self-management among Veterans. Providers in both groups felt that improved patient education addressing medication adherence, self-care, and management of acute symptoms is warranted, but there was no consensus on who should provide this education. Both VA and non-VA primary care providers perceived they had limited roles in shaping choices or Veteran decision making surrounding dual use. Seemingly missing were descriptions of advanced planning or intentional counseling regarding site of care and acknowledgement of the role to be played in care coordination. Our interviews uncovered striking differences in perception between VA and non-VA providers regarding the availability, quality, and effectiveness of VA HF services. Finally, multiple non-VA providers expressed frustration with and difficulty in contacting VA providers, accessing records, and making referrals into the VA system.

As discussed above, few prior studies have qualitatively analyzed provider perceptions of cross-system care, although a few studies designed to evaluate aspects of cross-system care within the context of the Veterans Choice Program were recently commissioned by the VA Quality Enhancement and Research Initiative (QUERI), which are ongoing. Previously, Kramer and colleagues conducted focus groups with providers in the Veterans healthcare system and the Indian Health Service (IHS) who shared care for Veterans dually enrolled in both systems.³⁵ In patient interviews, they observed that, in addition to financially based decision making, patients deliberatively chose healthcare services in given systems based on available services matched to their health needs. Similar to our findings, both VA and IHS providers described poor communication and coordination of care across systems. Providers in both systems noted that dual use prolonged visits given the need to identify potential conflicts in treatment and/or duplication of services, perceptions that we also captured. Authors described the need for care coordination and a corresponding need to identify which primary care provider was “primary.” Such determinations can be even more complex in VA-community interactions given the wealth of community providers without linkages to each other or to the VA.

This dilemma is pertinent given that both groups of providers in our study described patient education as desirable, but frequently inadequate. The question becomes, “Who should provide patient education including guidance on heart failure self-management and sick day care?” A fairly robust literature exists documenting the value of personalized care planning for a variety of chronic diseases. Coulter and colleagues analyzed 19 studies enrolling over 10,000 patients with diseases including diabetes, hypertension, mental health conditions, and one study in heart failure patients, and their meta-analysis concluded that personalized care planning leads to improvements in physical and psychological health status and in self-management when compared to usual care.³⁶ In heart failure, Shearer and colleagues observed improved ability for heart failure self-management among patients receiving a nurse-led, telephone-delivered intervention that included personalized care planning and patient empowerment.³⁷ Similar interventions have been found effective as well, though more intensive care management interventions may be more potent in reducing hospitalization and hospital readmission in HF.³⁸ Our findings highlight this area of confusion among providers, and the need to reach clarity about care leadership and patient education in future programs designed to improve the safety of cross-system care. It may also be useful to explicitly solicit patient preferences and discuss site of care decisions as a component of patient education and care planning.

Comments by both VA and community providers regarding barriers in access to information across systems are particularly relevant. While electronic systems for health information exchange (HIE) show great promise, they also face significant barriers and have been largely underutilized in the U.S. to date.³⁹ Many with VA have called for informatics-based approaches, such as HIE to gaps in care.⁴⁰ Within the past five years, the VA piloted and disseminated consumer-initiated health information exchange within My Health eVet, the VA's patient-facing electronic health record (EHR) through a feature called the Virtual Lifetime Electronic Record.⁴¹ In a small pilot trial, Veterans trained to access a “Continuity of Care Document” (CCD) and prompted to share their CCD with non-VA care providers were significantly more likely to do so as compared to control patients.⁴² However, the uptake of this form of HIE nationally and its effectiveness on patient outcomes are as-yet unclear. Far more complex than information exchange is the issue of care coordination across systems. Several studies of dual use document duplicative testing and prescribing practices as well as higher rates of hospitalization and even higher mortality among dual users.^43–45 Many of these risks could be mitigated through improved care coordination. Recognizing the potential negative unintended consequences of dual use, the VA Office of Community Care is currently piloting initiatives specifically designed to coordinate care when Veterans are seen by community providers.⁴⁶ Such efforts are needed and should be rigorously evaluated to ensure efficacy, scalability, and spread.

Certain limitations should be acknowledged regarding our study. First, we interviewed healthcare providers from three separate VAMCs and healthcare organizations across two Southern states, so our observations may not be generalizable nationally. Second, our sample lacked significant racial/ethnic diversity relevant to the population at large, and it is unclear the extent to which our sample reflects predominant racial/ethnic distributions among physicians practicing in our geographic region. Third, because we interviewed a variety of healthcare providers in VA and non-VA systems, we may not have achieved full thematic saturation for any given discipline, so view these perspectives as an initial exploration. We were also not able to get perspectives of all potentially relevant provider disciplines. Instead, we were able to garner a broad-based set of perspectives from key provider groups in VA and the community, an approach meant to generate hypotheses rather than demonstrate significance. Fourth, our study focused on dual system care for Veterans with heart failure. Thus, while there may be parallels to dual use in other chronic conditions, such as asthma, COPD, diabetes, or cancer, care should be taken in extrapolating our observations to other populations. Finally, it is important to note that qualitative themes emerging from this work will benefit from further empirical testing using quantitative methods to further validate the observations.

In summary, findings from this study have clear implications within the VA system where a priority has been placed on better understanding the dynamics of VA care provided in community settings. In August 2017, the VA will pass an important milestone with the sunset of the Veterans Access to Care and Accountability Act (VACAA) which authorized marked increases in non-VA care among Veterans. It is not yet clear which programs will modify or replace VACAA, but possibilities include extension of contract care similar to the Veterans Choice Program, VA-coordinated care that more closely resembles earlier VA programs, or possibly voucher systems as proposed by some advocacy groups.⁴⁷ In every case, patient and provider input should be key factors in designing policies and systems of care. Our findings have implications outside the VA as well. Seventy-eight percent of hospitals received penalties for higher than expected hospital readmission rates for one or of five target conditions in 2016 under the CMS Hospital Readmission Reduction Program.⁴⁸ Among primarily outpatient primary care and specialty providers and groups, the recent Medicare Access and Chip Reauthorization Act (MACRA) mandates expanded merit-based incentive payments (MIPS) and alternative payment models (APMs) that will focus both groups more heavily on episodes of care and/or longitudinal care across settings.^49–51 These imminent program changes would benefit from attention to the perspectives of VA and community providers as well as the Veterans making decisions about care.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by VA Health Services Research and Development (IIR 12-331, Axon-PI).

References

Benjamin

Blaha

Chiuve

et al.

Heart disease and stroke statistics- 2017 update: a report from the American Heart Association. Circulation 2017; 135: e146–e603.

Heidenreich

Trogdon

Khavjou

et al.

Forecasting the future of cardiovascular disease in the United States: a policy statement from the American Heart Association. Circulation 2011; 123: 933–944.

Goodman DC, Fisher ES and Chang C-H. After hospitalization: a Dartmouth Atlas report on post-acute care for Medicare beneficiaries. The Dartmouth Atlas of Healthcare. 2011 Contract No.: December 15.

Kochanek

Murphy

et al.

Deaths: final data for 2009. National Vital Statistics Reports Centers for Diseases Control and Prevention 2011; 60: 1–116. .

. Guidelines for assessing and managing the perioperative risk from coronary artery disease associated with major noncardiac surgery. Ann Intern Med 1997; 127: 309–312.

Helmer

Sambamoorthi

Shen

et al.

Opting out of an integrated healthcare system: dual-system use is associated with poorer glycemic control in veterans with diabetes. Primary Care Diab 2008; 2: 73–80.

Maciejewski

Wang

Burgess

Jr. et al.

The continuity and quality of primary care. Med Care Res Rev: MCRR 2013; 70: 497–513.

Thorpe

Gellad

et al.

Dual health care system use and high-risk prescribing in patients with dementia: a national cohort study. Ann Intern Med 2017; 166: 157–163.

Gellad

Zhao

Thorpe

et al.

Overlapping buprenorphine, opioid, and benzodiazepine prescriptions among veterans dually enrolled in Department of Veterans Affairs and Medicare Part D. Substance Abuse 2017; 38: 22–25.

10.

Weeks

West

Wallace

et al.

Comparing the characteristics, utilization, efficiency, and outcomes of VA and non-VA inpatient care provided to VA enrollees: a case study in New York. Med Care 2008; 46: 863–871.

11.

Nasir

Lin

Bueno

Normand

et al.

Is same-hospital readmission rate a good surrogate for all-hospital readmission rate?

Med Care 2010; 48: 477–481.

12.

Axon RN, Gebregziabher M, Everett CJ, et al. (2016) Dual health care system use is associated with higher rates of hospitalization and hospital readmission among veterans with heart failure. Am Heart J 174: 157–163.

13.

West

Charlton

Vaughan-Sarrazin

. Dual use of VA and non-VA hospitals by Veterans with multiple hospitalizations. BMC Health Services Res 2015; 15: 431.

14.

Chen

Ackerly

. Beyond ACOs and bundled payments: Medicare's shift toward accountability in fee-for-service. J Am Med Assoc 2014; 311: 673–674.

15.

Carey

Montez-Rath

Rosen

et al.

Use of VA and Medicare services by dually eligible veterans with psychiatric problems. Health Services Res 2008; 43: 1164–1183.

16.

Liu

Bolkan

Chan

et al.

Dual use of VA and non-VA services among primary care patients with depression. J Gen Intern Med 2009; 24: 305–311.

17.

Petersen

Byrne

Daw

et al.

Relationship between clinical conditions and use of Veterans Affairs health care among Medicare-enrolled veterans. Health Service Res 2010; 45: 762–791.

18.

Zhu

Penrod

Ross

et al.

Use of Medicare and Department of Veterans Affairs health care by veterans with dementia: a longitudinal analysis. J Am Geriatrics Soc 2009; 57: 1908–1914.

19.

Ajmera

Wilkins

Sambamoorthi

. Dual medicare and veteran health administration use and ambulatory care sensitive hospitalizations. J Gen Intern Med 2011; 26: 669–675.

20.

Nayar

Apenteng

et al.

Rural veterans' perspectives of dual care. J Commun Health 2013; 38: 70–77.

21.

Malterud

Siersma

Guassora

. Sample size in qualitative interview studies: guided by information power. Qual Health Res 2016; 26: 1753–1760.

22.

Fusch

Ness

. Are we there yet? Data saturation in qualitative research. Qual Rep 2015; 20: 1408–1416.

23.

Cohen D and Crabtree B. Interviewing. The Qualitative Research Guidelines Project, http://www.qualres.org/HomeInte-3595.html (30 June 2017).

24.

Campbell

Quincy

Osserman

et al.

Coding in-depth semistructured interviews: problems of unitization and intercoder reliability and agreement. Sociol Meth Res 2013; 42: 294–320.

25.

Bradley

Curry

Devers

. Qualitative data analysis for health services research: developing taxonomy, themes, and theory. Health Services Res 2007; 42: 1758–1772.

26.

Hsieh

Shannon

. Three approaches to qualitative content analysis. Qual Health Res 2005; 15: 1277–1288.

27.

Schreier

Qualitative content analysis. In: Flick

(ed). Sage handbook of qualitative analysis, Thousand Oaks, CA: SAGE, 2013, pp. 170–183.

28.

Bengtsson

. How to plan and perform a qualitative study using content analysis. NursingPlus Open 2016; 2: 8–14.

29.

Johnstone

Intention and interpretation. In: Malden

(ed). Discourse analysis, 2nd ed. Malden, MA: Blackwell Publishers, 2007. .

30.

Strauss

Feiz

. Discourse analysis: putting our world into words, NY: Routledge/Taylor & Francis Group, 2014.

31.

Bradley

Curry

Devers

. Qualitative data analysis for health services research: developing taxonomy, themes, and theory. Health Serv Res 2007; 42: 1758–1772.

32.

Darkins

Ryan

Kobb

et al.

Care Coordination/Home Telehealth: the systematic implementation of health informatics, home telehealth, and disease management to support the care of veteran patients with chronic conditions. Telemedine J E-Health 2008; 14: 1118–1126.

33.

Luptak

Dailey

Juretic

et al.

The Care Coordination Home Telehealth (CCHT) rural demonstration project: a symptom-based approach for serving older veterans in remote geographical settings. Rural Remote Health 2010; 10: 1375.

34.

Hughes

Weaver

Giobbie-Hurder

et al.

Effectiveness of team-managed home-based primary care: a randomized multicenter trial. J Am Med Assoc 2000; 284: 2877–2885.

35.

Kramer

Vivrette

Satter

et al.

Dual use of veterans health administration and Indian Health Service: healthcare provider and patient perspectives. J Gen Intern Med 2009; 24: 758–764.

36.

Coulter

Entwistle

Eccles

et al.

Personalised care planning for adults with chronic or long-term health conditions. Cochrane Database Syst Rev 2015, 3: pp. CD010523, .

37.

Shearer

Cisar

Greenberg

. A telephone-delivered empowerment intervention with patients diagnosed with heart failure. Heart Lung: J Crit Care 2007; 36: 159–169.

38.

Feltner

Jones

Cene

et al.

Transitional care interventions to prevent readmissions for people with heart failure, Rockville, MD: Agency for Healthcare Research and Quality, 2014. .

39.

Eden

Totten

Kassakian

et al.

Barriers and facilitators to exchanging health information: a systematic review. Int J Med Inform 2016; 88: 44–51.

40.

Dixon

Haggstrom

Weiner

. Implications for informatics given expanding access to care for Veterans and other populations. J Am Med Inform Assoc 2015; 22: 917–920.

41.

Bouhaddou O, Bennett J, Teal J, et al. Toward a virtual lifetime electronic record: the Department of Veterans Affairs experience with the Nationwide Health Information Network. In: AMIA annual symposium proceedings, Chicago, IL, 4 November 2012, pp.51–60.

42.

Turvey

Klein

Witry

et al.

Patient education for consumer-mediated HIE. A pilot randomized controlled trial of the Department of Veterans Affairs Blue Button. Appl Clin Inform 2016; 7: 765–776.

43.

Gellad

Zhao

Thorpe

et al.

Dual use of Department of Veterans Affairs and medicare benefits and use of test strips in veterans with type 2 diabetes mellitus. JAMA Intern Med 2015; 175: 26–34.

44.

Gellad

Cunningham

Good

et al.

Pharmacy use in the first year of the Veterans Choice Program: a mixed-methods evaluation. Med Care 2017. 55(Suppl 1–7): S26–S32.

45.

Jia

Zheng

Reker

et al.

Multiple system utilization and mortality for veterans with stroke. Stroke 2007; 38: 355–360.

46.

Personal Communication, Leo Greenstone. November 30, 2016.

47.

Stefos T and Burgess JF, Jr. Health Affairs Blog. Health Affairs, 2014, http://healthaffairs.org/blog/2014/10/02/establishing-vouchers-for-veteran-health-care/ (accessed 20 March 2017).

48.

Boccuti

Casillas

. Aiming for fewer hospital U-turns: The Medicare Hospital Readmission Reduction Program, Washington, DC: The Kaiser Family Foundation, 2017.

49.

Clough

McClellan

. Implementing MACRA: Implications for physicians and for physician leadership. J Am Med Assoc 2016; 315: 2397–2398.

50.

Williams

Sr. Casale

Oetgen

. A micro view of MACRA: how the ACC and NCDR will help members navigate radical changes ahead. J Am College Cardiol 2015; 66: 2675–2677.

51.

Farmer

Darling

George

et al.

Existing and emerging payment and delivery reforms in cardiology. JAMA Cardiol 2017; 2: 210–217.