INTRODUCTION

The Veterans Health Administration (VA) contributes to the education of at least 75 % of the nation’s physicians at either the undergraduate or graduate stages, and most VA tertiary care centers are affiliated with academic medical centers.1 While the presence of trainees yields benefits, it also leads to challenges in longitudinal care delivery. This is particularly true for primary care. Graduate medical education is finite and fragmented. Residents divide their time between education, training, and patient care; that division can undermine patient care continuity. Residents may spend only 1 or 2 days weekly in clinic, with no continuity clinic during certain months. In programs with ambulatory block scheduling, several weeks may elapse between clinic assignments. Further, every year, approximately one-third of residents will complete their training and leave, creating disruptions in care for Veterans receiving primary care from residents. Despite this, the Accreditation Council for Graduate Medical Education (ACGME) in Internal Medicine requires that “…residents develop a continuous, long-term therapeutic relationship with a panel of general internal medicine patients.”2 For the VA, as a health care system that has integrated trainees across the spectrum of care delivery, these provider transitions present a challenge that must be addressed.

Transitions of care have been identified as a significant source of medical errors across settings. For example, omission of key information during sign-outs results in adverse events.3 As many as 60 % of post-discharge adverse drug events have been deemed preventable with adequate communication.4 Patients in a VA general medical clinic staffed by discontinuous providers experienced an almost two-fold increase in emergency admissions and hospital days.5 In malpractice claims involving trainees, inadequate hand-offs were among the most prevalent contributing factors.6 Finally, in two separate studies,7,8 a significant number of high risk patients were lost to follow-up during the end-of-residency (EOR) transition period, despite interventions to prevent this attrition.

To improve continuity and coordination in primary care, the VA implemented the Patient Aligned Care Team (PACT) in 2010. The core PACT principle is to deliver Veteran-centered care that is team-based, efficient, comprehensive, continuous, and coordinated.1 Over 900 VA primary care clinics have been transitioning to the PACT model of care. Many resident clinics are included in this transformation. In theory, PACT implementation in resident clinics has the potential to ameliorate the inherent discontinuity that results from the trainee presence. With a PACT approach, when one resident graduates, the remainder of the PACT is intact. Other PACT members could be involved in bridging the EOR transition, extending the one-on-one resident–patient relationship to that of the team working together with the patient.

The success of the PACT approach is dependent on fostering the relationships among clinic members that lead to the formation of effective teams. However, little has been reported with regard to the role of non-physician health care team members in supporting resident transitions. Additionally, physician and non-physician attitudes towards the EOR transition or the potential roles of team members in supporting this transition are unknown. The purpose of this study was to assess the knowledge, attitudes, and perceptions of non-physician PACT members and resident physicians in a VA PACT continuity clinic, so that we may explore the potential value of a team approach to EOR transitions. We also sought to obtain potential ideas for new approaches to managing the EOR transition that leverage the PACT model.

METHODS

General Design

We conducted a cross-sectional survey of non-physician and resident PACT members’ knowledge, attitudes, and perceptions regarding EOR transitions in a single VA continuity clinic. This project was approved by the Institutional Review Board of the University of Texas Health Science Center at San Antonio (UTHSCSA) and the VA Research & Development Committee.

Setting

This study was conducted at the Internal Medicine Clinic (IMC) of the Audie L. Murphy VA Hospital in the South Texas Veterans Health Care System. The IMC serves as one of three sites for the UTHSCSA Internal Medicine (IM) residency program. Approximately 30 of the 76 categorical residents have continuity clinic at IMC. The IMC also contains full-time attending physicians who are not connected to the IM residency.

There are three IMC PACT groups: one containing the full-time attending physicians and two resident PACT groups. Each resident PACT has two part-time attendings who are also UTHSCSA faculty and supervise residents in the clinic, 15 resident physicians, one nurse practitioner (NP) or physician assistant (PA), one registered nurse (RN), one licensed vocational nurse (LVN) and one clerk. Within each resident PACT, the residents are assigned to one of the attendings, creating two teamlets comprised of one attending and seven to eight residents. Each resident is assigned to the same PACT and attending for all 3 years in continuity clinic. PACT members are physically co-located within the clinic, along with extended non-physician PACT members, including a nurse manager, assistant nurse manager, social workers, pharmacist, doctors of pharmacy, dietician, psychologists, and registered nurse specialists. These extended PACT members interact with all three PACTs in the IMC. Communication between PACT members occurs primarily via written and in-person communication.

Patients are assigned an attending primary provider and a resident associate provider per ACGME expectations.2 Each new post-graduate year (PGY) 1 resident is assigned a panel of 30 patients. The panel is increased to approximately 45 patients in the second year and 60 patients in the third year. At the end of the third year, approximately 600 patients from the third-year panels are redistributed to new PGY-1s and PGY-2s in the same PACT. Thus, resident primary care patients may be reassigned to other residents after 1, 2, or 3 years, but other PACT members remain the same.

The PACT model was implemented in the IMC in July 2011. Prior to implementation, graduating residents’ patients were divided and transferred to the panels of incoming PGY-1s, and current PGY-2s and PGY-3s. There was no formal evaluation of patient panels and no formal EOR checkout process. In addition, patients with resident primary care providers were not assigned to attendings.

Survey Development and Administration

We developed a survey to assess non-physician and resident PACT members’ knowledge, attitudes, and perceptions regarding the EOR transition. We first searched the literature related to EOR transitions to identify interventions that had already been reported. We then developed basic survey questions relevant to both groups. These questions were reviewed by faculty in the Divisions of General and Hospital Medicine and refined based on group consensus. We then created two survey versions, making questions relevant to non-physician PACT members and to residents, respectively. These questions also underwent review and refinement by the larger faculty group. Final surveys are shown in Appendices 1 and 2 (available online).

The non-physician PACT member survey contained four different sections; the resident version contained three. In the non-physician PACT member survey, the first five questions pertained to the members’ roles in clinic and background knowledge of residents and their training. These questions were not included in the resident survey version. The next section contained statements regarding perceptions and attitudes on EOR transitions using a Likert format. Likert scale responses ranged from 1 (strongly disagree) to 5 (strongly agree). The following section asked for respondent perceptions of patient awareness and experience of EOR transitions using the same Likert format. The final questions asked respondents to rate the potential utility of reported approaches to EOR transitions, including discussion of patient feelings regarding transfer with physician,9 face-to-face provider–patient discussions and additional appointments with the physician prior to transfer if necessary,10 personal notification letter from new PCP,10 and structured written checklist conveying concerns to new PCP.11 A free-response question was included for respondents to suggest other potential strategies to improve EOR transitions.

We administered the surveys to all 24 non-physician PACT members (including PAs, NPs, RNs, LVNs, pharmacists, doctors of pharmacy, psychologists, dietician, social workers and clerical staff) and 30 residents between August 2012 and January 2013. Participation was voluntary, and responses anonymous. Union approval was obtained for the non-physician PACT member survey.

Data Analysis

We calculated frequencies of responses to each question by respondent type, and compared responses between non-physician PACT members and residents. Data were analyzed using SAS© version 9.3 (SAS Institute, Cary, North Carolina).

Write-in responses were reviewed separately by two authors (MVC, LKL) and categorized into themes based on content. Responses could be categorized into more than one theme if the response fit into multiple categories. These categorizations were compared and combined, with any disagreements resolved by consensus.

RESULTS

Twenty-seven residents and 24 non-physician PACT members completed the survey, with response rates of 90 % and 100 %, respectively. Residents included nine PGY-1s, ten PGY-2s, and eight PGY-3s. Years of IMC experience among non-physician PACT members varied from less than one to 25 years (mean = 6.5; s = 6.7). Non-physician PACT respondents represented thirteen different occupational titles, the largest group comprising nurses (LVN, RN, NP) who made up slightly more than half of the non-physician PACT member respondents (54 %). Nearly two-thirds of non-physician PACT members (n = 15, 63 %) were core resident PACT members; the rest were extended PACT members.

Responses related to non-physician PACT members’ knowledge of residency training and EOR transitions are shown in Table 1. The majority of this group was aware that they worked in a teaching clinic with residents who spent 3 years in the clinic. Interestingly, 21 % of respondents did not realize that residents were medical school graduates.

Table 1 Non-Physician PACT Members’ Self-Reported Knowledge of Resident Education Practices (n = 24)
Full size table

Residents and non-physician PACT members agreed or strongly agreed that the resident should be responsible for informing patients about EOR transitions (100 % and 96 % respectively). When asked whether non-physician PACT members should be responsible for these transitions, results were mixed. About half of the residents (52 %), but only 38 % of non-physician PACT members, indicated that they should be responsible.

Similar proportions of residents and non-physician PACT members (81 % and 79 %) agreed or strongly agreed that there should be a formalized approach to EOR transitions. Sixty-seven percent of non-physician PACT members were willing to support the EOR transition, with 29 % of non-physician PACT members expressing a neutral position.

Non-physician PACT member and resident responses regarding strategies for improving EOR transitions reported in the literature are shown in Table 2. Sending letters to patients was the most endorsed approach, followed by discussing the transition with patients at the beginning of an appointment. Most respondents endorsed more than one approach. There was no clear preference for a particular set of interventions over another.

Table 2 Attitudes Regarding End-of-Residency Transitions by Resident (n = 27) or Non-Physician PACT Member (n = 24) Position
Full size table

Write-in responses are summarized in Table 3. Reviewers agreed on categorization of individual write-in items 89 % of the time. Increased communication among providers and patients and between providers was a recurring theme in responses.

Table 3 Summary of Themes in Write-In Responses on Strategies to Improve End-of-Residency Transitions
Full size table

DISCUSSION

We assessed non-physician staff and resident knowledge, attitudes and perceptions of the EOR transitions using survey methodology in a single VA teaching clinic. Resident physicians and non-physician PACT members viewed the resident primary care provider as primarily responsible for the EOR transition. However, non-physician PACT members reported being willing to help support this transition.

Despite the willingness of most respondents to support the EOR transition, responses suggest areas of poor understanding among PACT members’ roles that could be a barrier to effective coordination during the EOR transition. Not all non-physician PACT members had an understanding of resident physician training, and approximately half of residents did not see a role for the non-physician PACT member in the EOR transition. These findings may reflect larger issues of PACT communication and coordination. While we assume that collaboration among PACT members is fostered through the day-to-day PACT functioning, residents’ competing demands and transitory clinic presence provide barriers for them to regularly engage in activities that would foster PACT relationships, such as team meetings or huddles.

The emphasis on resident responsibility in our responses may also suggest an incomplete implementation of the PACT model, as our results indicate that PACT members do not yet view themselves as having collective responsibility for the EOR transition. Overcoming this barrier is important to realizing PACT potential. The Patient Centered Medical Home (PCMH) transformation has faced barriers in “shifting of mental models at the individual level and culture change at the practice level.”12 The culture change away from physician-centered care to team-based primary care is crucial to PACT success. With mounting evidence that primary care clinics that deliver team-based care manage chronic diseases more effectively and lower risk of avoidable hospitalizations, it is critical that PACT members are, in fact, able to create effective teams to provide coordinated care and accommodate team member transitions.13–15

Insights relevant to developing effective team-based care are found in complexity theory.16,17 It highlights the interdependency among processes of care, system infrastructure, and provider relationships. While the PACT model has changed the infrastructure and processes of care, such as post-discharge calls within 48 h, the optimal approaches to changing team-based relationships have been less clearly addressed. Relationships among team members are increasingly recognized as important to patient outcomes.18–21 To most effectively develop the PACT model, we must pay increased attention to the relationships among PACT members.18

As part of the PACT transformation, roles are being redefined. This re-definition also influences the relationship structure amongst PACT members. Attendings have moved from having no official relationship with resident patients to having oversight responsibility for a defined group of residents and patients, and greater integration into the PACT. Similarly, the role of other providers has evolved. For example, nurses have more of a coordination responsibility. However, changing the roles of individual providers may or may not have the intended effect of creating team-based care without a change in the relationships. For care to become more shared among PACT members, responsibility, trust, mindfulness, and heedfulness must be present. In addition, research on teams has shown there is a decrement in function when team members change, even if only a single member leaves,22 underscoring the harm that can potentially occur in resident clinics with EOR transitions and disrupted relationships.

Thus, the EOR transition is a microcosm of larger issues with role re-definition and relationship building. Because most respondents identify the resident as having responsibility for the EOR transition, our results suggest that we have not yet developed the PACT required for more effective coordination.

Interestingly, many of the write-in approaches reflected this direction, referring to the need for more communication among the team members. One theme that emerged supported development of a PACT visit dedicated to the introduction and departure of residents in continuity clinics. Based on this finding, and the willingness of the non-physician PACT members to be more involved, we propose the idea of “goodbye” and “hello” PACT visits. At the shared-team departure, or “goodbye” visit, PACT members can explicitly identify and discuss issues that need to be followed through during the transition, explain how responsibility for monitoring those issues will be shared, and discuss the transition between outgoing and incoming residents. Subsequently, at the initial visit with the incoming resident, or “hello” visit, the key issues can be reviewed with the patient, orienting the new team member and patient to how these issues will be addressed moving forward. These visits would not create additional visits for the patient or the resident, but rather would include additional members of the PACT in place of the one-on-one visit the patient would have had with the resident. These visits would not only ensure greater continuity and PACT communication, but also would provide the patient with reassurance that issues are being addressed and continuity ensured.

The communication among all PACT members and with the patient and caregivers throughout the “goodbye” and “hello” visit process would ensure that important concerns are addressed, and could be used as a mechanism to further engage the patient in his or her care. It would also foster characteristics of effective relationships, using rich in-person communication, leveraging diverse viewpoints of the PACT members, and engendering trust, respect, heedfulness, and mindfulness among the PACT. This would also serve as an educational opportunity for residents in team-based care and PACT principles. We recognize the logistical and time constraints inherent in incorporating this type of approach. It may not be effective to incorporate these visits for every patient. However, identification of high-risk patient registries may assist in targeting patients who may most benefit from this type of intervention. Understanding which patients would be best served by this facilitated intervention may be an area of further investigation.

Our study is limited in its focus on a single institution. However, while our institution may not be representative of every resident PACT clinic in specifics of staffing and processes of care, the general PACT principles are shared. Additionally, the challenges of fostering effective relationships and implementing team-based care are universal. Our study was also exploratory and descriptive, without testable hypotheses. We did not frame our initial survey in the lens of complexity theory generally or relationships specifically, but these ideas emerged from our analysis. Nevertheless, this framework complements our results, and provides universal insights into creating potentially effective EOR transition interventions.