Abstract
Background:
The Veterans Health Administration (VA) has implemented the largest shift to a patient-centered medical home (PCMH) model of care in the United States to date.
Objective:
We interviewed veterans about their experiences of primary care to understand whether they observed changes in care during this period as well as to learn which characteristics of care mattered most to their experiences.
Method:
Qualitative interviews were conducted with 32 veterans receiving primary care at 1 of 8 VA clinics in the northwest United States. Interviews were analyzed using an inductive–deductive hybrid approach by an interdisciplinary team that included a veteran patient.
Result:
Participants noticed recent positive changes, including improved communications and shorter waits in clinic, but rarely were aware of VA’s PCMH initiative; a strong relationship with the primary care provider and feeling cared for/respected by everyone involved in care delivery were key components of quality care. The needs of the veteran community as a whole also shaped discussion of care expectations.
Conclusion:
The PCMH model may provide benefits even when invisible to patients. Veteran awareness of population needs suggests a promising role for veteran involvement in further PCMH transformation efforts.
Keywords
Introduction
Over the last decade, the patient-centered medical home (PCMH) model of care has been widely adopted as part of a transformative agenda to improve primary care in the United States. The PCMH core concepts include continuous care by a designated provider as part of a team, with strong care coordination, expanded care access, and effective use of communication, data, and information technology tools to improve patient outcomes (1), but individual PCMH projects vary widely and have multiple outcomes of interest, complicating evaluation of the model’s success (2). To date, the largest system transition to a PCMH model has taken place within the Veterans Health Administration (VA), which provides care to over 5.8 million military veterans every year (3). The PCMH models differ in the details of their team structures and desired staffing ratios; in VA’s model, each patient has an assigned core care team—their Patient Aligned Care Team (PACT)—comprised of a primary care provider (provider), nurse care manager, clinical associate, and administrative associate. Because PACT has been conceptualized as being driven by team-based care, the way that the “team” is defined has implications for workflow in the clinic and, potentially, for how patients experience communication and coordination of care.
Similar to findings from a systematic review of PCMH care for low-income patients (4), strong implementation of PACT has been associated with higher patient satisfaction, better clinical quality measure scores, lower staff burnout, lower emergency department use, and lower rates of avoidable hospitalizations (5 –7). The most comprehensive economic evaluation of PACT implementation to date found that high initial investments meant the initiative had not, in its first 2 years, generated a positive financial return in terms of reduced expenditures (8), but trends in that study’s data, as well as results from other PCMH projects (9,10) suggest that PCMH models like PACT may prove cost-effective as they mature.
Studies of PACT and adult PCMH have most often assessed patient experience using survey data, with results that have included positive (11), negative (1), mixed (12), and no associations between PACT or PCMH implementation and patient experience (1,13,14). Despite the intense investment of resources, PACT does not appear to have shifted patient experience, at least not as captured in one regional survey of Veteran satisfaction (14). The authors of that study pointed out the limitations of VA’s current patient survey methods for assessing how PACT may have impacted experience of care and concluded with a call for “ways to incorporate the patient’s voice into these transitions”(14).
We interviewed veterans about their experiences of primary care delivery as part of a mixed-method evaluation of PACT in primary care clinics within a rapidly growing VA system in the Pacific Northwest. During the early phase of PACT rollout, significant effort was made to introduce veterans to the new model of care. For example, posters and other signage throughout participating facilities announced the new PACT model, and PACT teams were encouraged to send their patients letters explaining the new model, naming the care team members, and giving guidance on how to contact the team directly. Our primary goal in this study was to understand which aspects of veterans’ health-care experiences most strongly informed their general perceptions of quality care and how key concepts from the PACT model were or were not relevant to veteran experience, as well as whether veterans had perceived changes during PACT implementation.
Method
In 2014, we conducted qualitative semi-structured interviews with patients receiving care from clinics that had been participating in PACT implementation. The study received approval from the primary VA site’s institutional review board. We used administrative records from 8 participating primary care clinics to identify veterans who met our inclusion criteria, namely, they had accessed primary care at least once in the last 30 days and had either a chronic health condition or had been recently hospitalized. We hypothesized that these patients would be most likely to be exposed to changes in care delivery resulting from PACT implementation and that those seen in the last 30 days would be likely to have sufficiently detailed recall of recent care experiences. Patients were excluded if their records indicated serious mental illnesses (not including post-traumatic stress disorder, depression, or anxiety) or cognitive impairment. Potential participants were sent letters of invitation to the study, and those who did not respond were contacted by phone for follow-up until we had enrolled at least 30 participants, which is a reasonable sample size consistent with common practice for in-depth qualitative analysis and also was the sample size we determined our research team capacity and time line could accommodate. We employed a purposive sampling method in order to ensure inclusion of different genders, races, ethnicities, and usual care clinics.
We designed a semi-structured interview guide focused on care experiences since PACT implementation. We allowed phone or in-person interviews to accommodate veteran preference and make participation easier for those who could not travel to an in-person interview. In-person interviews were conducted in a private office setting at the VA facility where the research team was based. Interviews used a funnel format starting with open-ended questions about care experiences within the VA and then moved to questions exploring core PACT concepts of timeliness, access, continuity of care, communication, patient-centered care, and team-based care (see Online Appendix A). Patients were given definitions of the PACT concepts being discussed, for example, in relation to team-based care, participants were asked: “An important part of the PACT model is that your primary care provider is supposed to be working as part of a team. Does your provider have a team? If so, who is on your provider’s team?” and, as one of several follow-up probes, “What do you like or not like about having a team of people take care of you?” Interviews lasted 50 to 90 minutes and were audio-recorded, transcribed, and deidentified prior to analysis.
Transcripts were coded in ATLAS.ti software (15) by 2 of 3 coders with a social science background, including 1 veteran patient with previous training in qualitative research and thematic coding. Analysis was led by an investigator team that included 2 physicians and a sociologist. Our team employed a hybrid inductive–deductive thematic approach to analysis. We first inductively analyzed interviews attending to salient patient perspectives on care experience, followed by deductive coding focused on indications of change. Our analysis was conducted primarily at the semantic level (16), meaning with a focus on what veterans said explicitly. However, identity and experience as a veteran seemed to shape participant responses in important ways, in that participants often responded to questions about what they valued or desired by spontaneously switching to a discussion of what veterans as a group needed or wanted. This led our team to conduct a latent analysis (16) of the role of veteran identity, meaning we looked for underlying assumptions and patterns, not always directly stated, about the influence of this identity.
Results
Administrative records were pulled for 259 veterans meeting the “within 30 days of visit” inclusion criterion; 114 either met exclusion criteria or did not have current contact information. The remaining 145 veterans were invited by letter to participate. Of these 145, 23 called the research team to request an interview, and additional 17 veterans were called by the research team, and 9 of these agreed to an interview, for a total of 32 veterans interviewed (10 via telephone and 22 in person) between March and July 2014. One additional veteran scheduled an interview but then canceled, saying he was no longer interested. Participant demographics are listed in Table 1.
Participant Characteristics.a
aAge range 43-88 years (median: 65 years).
Patients did observe changes since PACT began but rarely linked these changes to the PACT model. In addition to describing observed changes, we identified 2 themes illuminating what participants valued in defining excellent care. The first of these was the importance of a strong relationship with the provider; the second was a sense of feeling cared for by and respected by everyone involved in their care within the health-care system. This respect and caring mattered to patients both as individuals and as part of a veteran community. We looked for, and did not find, variation in the presence of these themes related to veteran age, gender, race, or number of comorbidities.
Implementation of PACT Was Largely Invisible to Patients, but Changed Practices Were Noticed
Most participants were unaware of the PACT initiative; as one commented, “I didn’t even realize that they had put into effect any type of patient care specific, you know?” (G, male in his 60s). Nevertheless, when asked about changes in recent years in the VA health system, approximately two-thirds (n = 21) of participants reported overall positive changes in care. Seven participants reported no specific changes and consistent satisfaction before and after PACT, and 2 participants who had been with their clinics for <2 years did not comment on change. Two participants had observed a negative change in access (“it’s harder to get in”) and said this was caused by more veterans seeking care (“there’s more of us”). Older participants saw positive changes in the VA health system over a longer time frame since the 1970s in addition to noticeable changes over the last 3 years; some had difficulty distinguishing this longer term trend from recent changes.
The specific recent positive changes that patients most often recalled were decreased waiting times upon arrival/check-in for appointments and improved direct communication with their clinic. With some exceptions, patients noticed and appreciated that appointments tended to run on schedule. One patient felt that the “ability to keep appointments on schedule” (C, male in his 50s) was perhaps the most important measure of care quality. Less crowded waiting rooms, fewer unhappy people in the waiting rooms, and shorter lines for check-in were also mentioned frequently.
In contrast, when asked about changes in how quickly they could get a primary care appointment (the VA’s primary measure of access), nearly all participants felt there had been no or only slight positive change but did not perceive this as a problem. One participant who observed negative change had experienced a delay in receiving primary care, but otherwise concerns with access related to specialty services, surgery, and the emergency department, which had not changed over time. Patients at smaller clinics reported satisfaction that more community-based outpatient clinics were opening and expanding the services they offered and felt this improved access by reducing driving times.
Improved communication and continuity of care were also frequently noted as recent positive changes. Some participants linked more providers and staff to this change: “I have noticed an improvement…[in the past] they had so many people to work with, they didn’t really have time to talk to you about it [care options]” (B, male in his 60s). Being able to call one’s care team directly was an appreciated change for some, and follow-up calls and appointment reminders were described as having become more consistent.
While some participants associated these changes with the PACT model—“since they’ve gone to the PACT team, you do have better communication” (B, male in his 60s)—most did not attribute changes to any specific initiative. Similarly, and contrary to our expectations, when asked about team-based care, most participants did not talk about their PACT members but described coordination between specialty and primary care or between inpatient and outpatient services. Most could not name the nonprovider members of the team—“I would not recognize them if they were sitting there in your chair…[but] it works for me” (J, male in his 80s)—and did not describe having strong relationships with them, one even stating “that’s not the nature of their function” (C, male in his 50s). Some patients admitted concerns that the new model of care might put the nurse care manager in a gatekeeper role, but for some at least these concerns had been allayed by experience: “I thought something was going to get lost from, between her and the doctor. But I found that was not the case” (J, male in his 80s).
A Strong Relationship With the Provider Is Highly Valued
A strong, trusting, individual relationship with the primary care provider was nearly universally cited by participants as necessary for quality care and central to rating care positively. A few patients described establishing strong relationships very quickly with a provider, but most felt building a strong relationship with a single provider took time. [I] have a doctor that cares…I’ve been hospitalized several times in the last 2 or 3 years, ah, thinking that maybe I was very close to death and first thing [my provider] said when she saw me was, “if you were to die I would’ve cried.” And so I knew then that we had become very good friends over the year….and that is not replaceable. (A, female in her 50s) They knew my history…I’d be able to say, “Well, this is a new problem or this is an ongoing problem, you may remember I spoke of this or this,” and they’d…acknowledge when they come across it…So I really think that’s a big benefit. (C, male in his 50s) It is about trust, which is real important to me. It is about knowing that my primary care doctor is gonna see me through when I have a problem, is gonna see that I get what I need to the best of her ability. (D, female in her 50s)
Veterans Value Feeling Cared for/Respected by Everyone Involved in Their Care (Not Just the PACT Team)
A second theme identified the importance for veterans of feeling cared for by and respected by everyone involved in their care within the health-care system. More than 1 patient observed that care “starts at the front desk.” Some participants observed that the provider’s behavior could influence this broader culture of caring: “It kinda seems like the nursing staff is really affected by the doctor. If the doctor really cares then they really care, you know” (F, male in his 70s). Many highlighted that they felt respected when they were not kept waiting for appointments.
Another commonly cited example of the caring system “team” was getting phone calls to relay information or get help with scheduling. Notably, while calls from providers were usually described as personal caring, calls from others within the VA system, whether they were designated PAC T members or not, were valued “not as care, but…updates” (G, male in his 60s) or in other ways that suggested their value was informative rather than emotional. Good communication shaped some patients’ sense of being “taken care of” even if they did not see the same provider each time: “as long as the nurse informs…[the provider] of the patient’s needs and, you know, their past history, I think most any doctor would be fine.” (E, male in his forties)
The importance of this respectful, caring approach on the part of the all-inclusive VA team often led participants into discussions about the needs of the veteran community and how addressing these collective needs was a component of good care, delivered not just by individuals but by the system as a whole. No interview guide questions asked about veteran identity or experience, yet participants commonly brought up the needs of veterans as a population. Sometimes participants generalized their priorities to the larger veteran population, as when one stated that “Vietnam vets, and probably all vets, the one thing that’s important to you is you know you matter a little bit, somebody cares” (H, male in his 60s). One participant suggested providers and staff should have ongoing training on “how to deal with veterans,” because “they wasn’t in the military, so they might not be able to…use terminology or relate to their needs, and that’s where the misconnect comes in” (I, male in his 40s).
Many participants emphasized the significant care needs in the veteran population and accepted that their own access to care might be limited by this need. I have to understand that…there’s veterans out there that are in a whole lot worse shape than I am that need care, and it’s gotta be prioritized. And I think the teams, the people have been doing [a] pretty reasonable job in that aspect. (G, male in his 60s) People will get back to me and that’s all I can ask—especially when you are dealing with thousands and that’s what the VA is attempting to do. This is a large job and I am aware of that. And I try to give them consideration for that fact. (J, male in his 80s)
Discussion
We found that, in the context of the nation’s largest PCMH implementation initiative to date (PACT), nearly all participants in our small study perceived positive changes in their primary care, even though most had little awareness of the change in care model that had been occurring during this time. This points to a distinction between patient awareness of the process of PACT implementation—something that was stressed in early rollout of the model—and patient perceptions of outcomes that may be the result from PACT implementation. The importance of short waiting times in the clinic and the perceived improvement in this area, despite no specific initiative to do so, were surprising and, we feel, important findings of this study. We hypothesize that improved team coordination and staffing were behind this improvement, suggesting that changes “behind the scenes” can impact patient experience even when patients do not see the mechanism for improvement. Another qualitative study recently investigated non-veteran adult patients’ experiences of PCMH implementation and also found the provider relationship was of central importance and could outweigh other negative perceptions of care (17). Consistent with our findings, that study found that many patients were unaware of the model or concepts underpinning it, leading the authors to suggest that use of a PCMH model might better target only patients most likely to benefit. In contrast, we interpret our results to suggest that patients may perceive changes facilitated by the model (such as more efficient and better organized care) without necessarily endorsing the model itself. This is particularly relevant in relation to the concept of team-based care. Patients might only see results of the team’s activities, rather than the team itself, or the team as defined by a particular implementation model, but this does not invalidate the contribution of team-based care. Previous research has suggested that factors unobserved by patients still impact quality of care and health outcomes (18); we would argue that they may also potentially impact staff satisfaction and burnout that affect overall functioning of the clinic.
PACT’s envisioned role of the nurse care manager as a key provider of patient education and counseling remains largely unrealized. Some have suggested that the PCMH model may be challenging to adopt because of the way that it upends traditional expectations of the independent provider operating over rather than with other care team members (19); our findings suggest the PCMH model may challenge patients’ expectations as well. The question is to what extent these expectations should be addressed through dialogue and education and to what extent the PCMH model should be modified to acknowledge patients’ continuing emphasis on strong provider relationships.
Our findings highlight that the PACT model’s concept of a “team,” while important for organizational needs such as determining and tracking sufficient staffing commitments, is largely irrelevant to veterans’ broader experience of the care team as incorporating everyone involved in their care, from the moment they walk in the door. Participants did not limit their assessment of care to their PACT but most often spoke in terms of “everyone” they encountered in the course of their VA care—schedulers, pharmacy staff, front desk attendants, and others. This resonates with earlier research looking at PACT implementation from the care team’s perspective, where the ability to implement PCMH was found to be influenced by systemic factors (20). Considering all interactions veterans have with the health-care system, not just their PACT team, may more accurately assess total quality of care delivery. As others have also noted (21,22), “fixing” primary care is only one element needed to address the systemic crisis in health system access and affordability in the United States.
Our findings are interesting to compare with those of Wagner and colleagues (23), the only other qualitative study of veteran patient experience since PACT implementation that were able to identify. In that study of 25 women veterans receiving care in a different regional system, participants found barriers to accessing primary care to be common, which is in contrast to our results and is not surprising, given regional variations in access as well as potential gender-specific disparities in access to care. Women veterans in that study valued continuity of care and “personal connection” with their providers but also valued a broader “personalized context” such as having a separate women’s health clinic, suggesting how even the use of space may convey caring and respect. The nuanced differences in our themes may be the result of different analytic interpretations or the influence of gender on attitudes toward care, but it is significant, we feel, that both studies’ findings reflect patients’ giving importance to both the personal relationship with the primary care provider and the tone or quality of the broader context of care delivery.
The influence of veterans’ military experiences on their individual care experience has been examined elsewhere (24), but the implications of veteran identity for system redesign remain underconsidered. As reported in our findings, participants in this study often spontaneously recognized that complex and increasing demand for care often challenged VA’s capacity to meet that demand. The frequency and richness of comments framing individual care needs in view of the needs of other veterans indicated to us that the VA’s patient population is possibly unusually well prepared to be involved in systems redesign efforts, because so many patients already consider needs at the population level. When PACT began, there was little formal involvement of veterans in the planning and implementation process. Recently, many VA facilities have formed patient and family advisory councils to bring patient perspectives into system redesign efforts (25,26). Our results suggest this approach should continue and expand.
Our study possessed several limitations. Our sample was small compared to the large number of veterans affected by the rollout of PACT and could not capture the many different characteristics such as age, race, and length of military experience likely to influence veteran perceptions of care. Like most qualitative studies, our intent was to obtain in-depth insights from a purposeful sample of participants representative of a particular population; results may not be generalizable to other populations. Our selection criteria (a chronic health condition or recent hospitalization) skewed the age distribution of our sample toward those older than 40 years. Our findings therefore do not capture the experiences of younger veterans. Participants who had received VA care for many years tended to discuss change over a long time frame, which may have introduced a bias to underreport more recent changes.
Conclusion
Since PACT began, veterans in our study observed improvements in their primary care experiences, particularly related to improved communication and shorter wait times. These changes were arguably made possible by team-based primary care, suggesting the PCMH model can have “behind-the-scenes” value. Veterans in our study strongly valued a trusting, individualized provider relationship and feeling cared for and respected by everyone encountered in their process of care as key elements of quality care, characteristics that align with PACT’s focus on patient-centered care but which define that concept, and the idea of the team, through a different lens. These different perspectives add evidence to support continued calls for the need to involve veterans in ongoing system redesign.
Supplementary Material
Supplemental Material, VET_EXP_APPENDIX_A - Veteran Patient Perspectives and Experiences During Implementation of a Patient-Centered Medical Home Model
Supplemental Material, VET_EXP_APPENDIX_A for Veteran Patient Perspectives and Experiences During Implementation of a Patient-Centered Medical Home Model by Anaïs Tuepker, Summer Newell, Christina Nicolaidis, Marie-Elena Reyes, Maria Carolina González-Prats, Eleni Skaperdas, and Devan Kansagara in Journal of Patient Experience
Footnotes
Authors’ Note
The views expressed in this article are those of the authors and do not necessarily represent the views of the Department of Veterans Affairs.
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: Funding for the PACT Demonstration Laboratory initiative was provided by the VA Office of Patient Care Services.
Supplementary Material
Online Appendix for this article is available online.
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.
