A Failure to Communicate: A Qualitative Exploration of Care Coordination Between Hospitalists and Primary Care Providers Around Patient Hospitalizations

Abstract

Background

Care coordination between adult hospitalists and primary care providers (PCPs) is a critical component of successful transitions of care from hospital to home, yet one that is not well understood.

Objective

The purpose of this study was to understand the challenges in coordination of care, as well as potential solutions, from the perspective of hospitalists and PCPs in North Carolina.

Design and Participants

We conducted an exploratory qualitative study with 58 clinicians in four hospitalist focus groups (n = 32), three PCP focus groups (n = 19), and one hybrid group with both hospitalists and PCPs (n = 7).

Approach

Interview guides included questions about care coordination, information exchange, follow-up care, accountability, and medication management. Focus group sessions were recorded, transcribed verbatim, and analyzed in ATLAS.ti. The constant comparative method was used to evaluate differences between hospitalists and PCPs.

Key Results

Hospitalists and PCPs were found to encounter similar care coordination challenges, including (1) lack of time, (2) difficulty reaching other clinicians, (3) lack of personal relationships with other clinicians, (4) lack of information feedback loops, (5) medication list discrepancies, and (6) lack of clarity regarding accountability for pending tests and home health. Hospitalists additionally noted difficulty obtaining timely follow-up appointments for after-hours or weekend discharges. PCPs additionally noted (1) not knowing when patients were hospitalized, (2) not having hospital records for post-hospitalization appointments, (3) difficulty locating important information in discharge summaries, and (4) feeling undervalued when hospitalists made medication changes without involving PCPs. Hospitalists and PCPs identified common themes of successful care coordination as (1) greater efforts to coordinate care for “high-risk” patients, (2) improved direct telephone access to each other, (3) improved information exchange through shared electronic medical records, (4) enhanced interpersonal relationships, and (5) clearly defined accountability.

Conclusions

Hospitalists and PCPs encounter similar challenges in care coordination, yet have important experiential differences related to sending and receiving roles for hospital discharges. Efforts to improve coordination of care between hospitalists and PCPs should aim to understand perspectives of clinicians in each setting.

INTRODUCTION

Patients are particularly vulnerable during transitions in care from hospital to home. Care coordination efforts during transitions become more critical as patient complexity and care fragmentation increase.¹ The Agency for Healthcare Research and Quality (AHRQ) describes care coordination as “deliberately organizing patient care activities and sharing information among all of the participants concerned with a patient’s care to achieve safer and more effective care.”²

Inadequate care coordination between hospitalists and primary care providers (PCPs) is a widespread problem that can lead to medication errors, missed test results, and patient harm.^3–5 In a study of recently discharged older patients and their PCPs, at two weeks post-discharge, only 30 % of PCPs were aware of the patient’s hospitalization, and patients whose PCPs were not aware of their hospitalization were more likely to report problems following discharge (e.g., medication problems, readmissions).⁶ In another study, PCPs reported that direct communication with hospitalists occurred only 23 % of the time, and discharge summaries were available to PCPs only 42 % of the time.⁷

Challenges to care coordination include difficult interprofessional communication, lack of information feedback loops between settings, and lack of clarity regarding accountability for tests pending at discharge.^5,8,9 Although health information technology offers great potential to improve coordination of care, this potential has yet to be realized despite the growing prevalence of electronic medical records (EMRs).^10,11

To improve our understanding of care coordination between adult hospitalists and PCPs around hospitalizations, we conducted a qualitative study to explore current processes, barriers, and potential solutions from the perspective of hospitalists and PCPs practicing in North Carolina. Although the study was developed to explore care coordination barriers and solutions equally, the discussions—and therefore our results—focused more on barriers than solutions in the coordination of transitional care.

METHODS

Study Design

We conducted an exploratory qualitative study of hospitalists and PCPs recruited from practices in North Carolina. Eight focus groups—four groups of hospitalists, three groups of PCPs, and one hybrid group of both hospitalists and PCPs—were convened from February through May 2013. Participants completed informed consent and a demographic survey at the beginning of each session. Each group discussion lasted approximately 45 minutes; discussions were audiotaped on digital recorders and transcribed verbatim. This protocol was reviewed and granted exemption by the Institutional Review Boards at the University of North Carolina at Chapel Hill (13–0019) and Duke University (00045149).

Population

The sampling and composition of the focus groups was based on the assumption that clinicians who deliver care in similar practice settings have common experiences. A purposeful sampling strategy was employed to ensure that practices represented diverse settings. Primary care practices were recruited from a collaborative of 10 primary care practices in North Carolina that were working to improve care transitions through an AHRQ grant supporting primary care transformation (U18 HS020940). Two hospitalist practices were recruited from 32 hospitals participating in a North Carolina Preventing Avoidable Readmissions collaborative. Two additional hospitalist practices were recruited that had admitting relationships with two of the recruited PCP practices. We purposefully recruited such that three PCP–hospitalist dyads had patients in common (i.e., patients from the PCP practice were known to be admitted to that hospitalist practice). We recruited clinicians from practices active in quality improvement of transitional care, as they were actively considering strategies to improve coordination of care.

Leadership from nine eligible hospitalist and primary care practices were first approached by phone or e-mail to determine interest in participation. All agreed to participate, with the exception of one private rural hospitalist group, who initially agreed but then did not participate due to scheduling difficulty. Within interested practices, depending on the wishes of practice leadership, e-mail invitations were sent to clinicians either by the study team or by leadership. In the e-mails, all clinicians active in patient care were encouraged to participate in focus groups that were held at the practices. Due to variability in this recruitment process, we did not systematically collect information about clinicians who may have received an e-mail invitation but did not participate in a focus group.

Framework

We started with the conceptual framework of care coordination activities included in the AHRQ Care Coordination Measurement Framework.¹ We broadly mapped themes from the framework for our interview guide to represent activities that hospitalists and PCPs could use to coordinate care (Table 1). We developed modified themes because we found that items in the AHRQ framework could not be narrowed to a specific activity, but rather represented multiple potential activities for both hospitalists and PCPs. The interview guide and analysis were structured within a conceptual framework that included accountability, care coordination, information exchange, medication management, and follow-up care.

Table 1.

Interview Guide, Organized by Conceptual Theme

Care coordination	Tell me about how you coordinate care with PCPs/hospitalists.
	- Probe(s): Give me an example of a time that coordination with a PCP/hospitalist did/did not work well.
Care coordination/information exchange	In an ideal world, how would you coordinate care with a PCP/hospitalist?
	- Probe: What method do you think is best to communicate with PCPs/hospitalists?
Follow-up care	How is follow-up care with a PCP arranged following discharge?
	- Probe: How do you determine how soon a patient needs to be seen?
Information exchange	In general, what information is most critical to provide to PCPs at discharge?
	In your opinion, how soon after discharge should the PCP receive this info?
Medication management	How do you support medication management for patients around discharges?
	- Probe: How do you feel about the PCP being involved in decisions about changing medications for patients upon discharge?
Accountability	How do you establish accountability for tests pending at the time of discharge (i.e., PCP or hospitalist)?
	- Probe: How about tests that are recommended following dc (e.g., repeat CT scan in 6 months)?
	How do you determine the primary clinician who interacts with home health to clarify questions about the patient’s care?

Analysis

Initially, the research team developed a code book based on the conceptual framework shown in Table 1. Codes were applied to the data to develop an initial set of themes. One main coder, an internal medicine physician active in both inpatient and outpatient settings and trained in qualitative methods, moderated the focus groups (CDJ) and was involved in coding all manuscripts. At least one additional team member (CO, SP, MA) independently coded each transcript and met with the main coder to reconcile code disparities after each transcript was completed. Additional team members were all board-certified internal medicine physicians; one had prior qualitative coding experience (SP), and all were trained by the main coder using the framework from Table 1. Discrepancies were resolved through team consensus, where the main coder facilitated discussions. Additional codes were added based on emergent findings, which were unexpected findings that we discovered through the coding process. Analysis was facilitated using ATLAS.ti, version 7 (ATLAS.ti GmbH, Berlin, Germany).

A summary of themes and emergent findings was prepared and reviewed by investigators (DD, CDJ, MV) after every one or two focus group sessions until we reached thematic saturation as a means of ensuring that content was building iteratively (i.e., themes were being repeated).¹² Key discussion themes of the interviews were sent to members of the focus groups to perform member checks.

For analysis, comparisons were made based on the predominant clinician type in the focus group (hospitalist/PCP) to compare care coordination experiences and barriers among groups. Statements from the hybrid group of clinicians were designated as either “hospitalist” or “PCP” based on the context of the comment. We analyzed across transcripts to identify the most common barriers as well as successes in coordination of care. The constant comparative method was used to evaluate findings across clinician type in order to identify common and disparate experiences in care coordination. Because many of the identified themes did not fit exclusively into one facet of our conceptual framework, results are presented by perspective rather than within the framework.

RESULTS

A total of 58 clinicians participated in eight focus groups; one PCP practice administrator also participated but was excluded from the analysis. The four hospitalist groups comprised 32 of the 58 clinicians, the three PCP groups comprised 19 clinicians, and the hybrid group consisted of seven physicians from various settings (three PCPs, two hospitalists, and two physicians that practiced equally in both settings). One hospitalist group and two PCP groups were primarily academic; one hospitalist and two PCP groups were private practices. The remaining two hospitalist groups were at community hospitals with academic affiliations. Three hospitalist groups and two PCP groups were in urban settings; the remaining three groups were in rural settings.

Demographic and practice-level data for 56 of the 58 clinicians are shown in Table 2. Overall, more hospitalists than PCPs were male, and hospitalist groups were generally larger than PCP groups. More PCPs than hospitalists practiced in rural settings. All clinicians used electronic medical records (EMRs) in their practices.

Table 2.

Characteristics of Clinicians Participating in Focus Groups

	Hospitalist (4 groups; 30 of 32 clinicians completed survey)	PCP (3 groups; 19 of 19 clinicians completed survey)	Hybrid Hospitalist/PCP (7 of 7 clinicians completed survey)
Male (n, %)	19 (63 %)	8 (42 %)	2 (29 %)
Number of full-time employees in practice (mean, SD)	20.0 (6.4 SD)	9.5 (5.0 SD)	3.5 FTEs in clinic
			23 FTEs in division
Clinician type (n, %)
Physician	28 (93 %)	10 (53 %)	7 (100 %)
Physician assistant	2 (7 %)	8 (42 %)	0 (0 %)
Pharmacist	0 (0 %)	1 (5 %)	0 (0 %)
Years since completion of clinical training (mean, SD)	8.9 (6.8 SD)	10.4 (7.7 SD)	9.9 (7.0)
Hours/week at practice (mean, SD)	47.9 (10.9 SD)	40.1 (8.6 SD)	41.4 (14.9)
Rural practice (n, %)	4 (13 %)	11 (58 %)	0
Use of EMRs at practice (n, %)	30 (100 %)	19 (100 %)	7 (100 %)

Although we asked general questions about care coordination, we found that discussions in all groups gravitated toward barriers to coordination of care rather than solutions to address these barriers. When discussing challenges in care coordination, hospitalists and PCPs generally related experiences that were more similar than different. In the following sections, we discuss (1) care coordination challenges, (2) accountability challenges, and (3) successes and ideals for care coordination and accountability. For each theme, we highlight first the shared perspectives, followed by the hospitalist and PCP perspectives (Table 3). Each quote is identified by participant and practice setting.

Table 3.

Summary of Results by Theme and Perspective

Theme	Shared	Hospitalist Only	PCP Only
Care coordination challenges	Lack of time	Difficulty obtaining PCP appointments	Unaware patients were hospitalized
	Difficulty reaching other clinicians		No hospital records for appointment
	Lack of personal relationships		Difficult to navigate hospital records
	Lack of information feedback loops		Feeling undervalued
	Medication list discrepancies
Accountability challenges
Pending tests	Unclear accountability for pending tests and home health	PCPs accountable for tests recommended after discharge	Hospitalists accountable for tests until results shared with PCP
	Concern about missed test results		PCPs uncertain about responsibility for follow-up on tests
	Specialists accountable for tests they order	Accountability for test results varies by test	PCPs often do not receive information about pending tests
	Hospitalists accountable for sharing information with PCPs
Home health care (HHC)	Hospitalists responsible for initial HHC orders	PCPs responsible for HHC once patient discharged	PCPs have difficulty addressing HHC issues prior to follow-up appointment
		Unwelcome receipt of HHC documents beyond initial contact
Care coordination solutions and ideals	Greater efforts for “high-risk” patients	Centralized scheduling system for PCP appointments	Short, structured, timely summaries
	Direct phone access to other clinicians	Hospitalist-run follow-up clinics	Follow-up appointments made prior to discharge
	Shared EMRs		Outpatient-based transitional care innovations
	Enhancing interpersonal relationships
Accountability solutions and ideals	Defined accountability for pending tests and home health

Care Coordination Challenges

Shared Perspectives

Multiple individuals in both hospitalist and PCP groups described having little time for coordination of care around patient hospitalizations, which compounded the frustration they felt when they had difficulty reaching each other by phone. PCPs described uncertainty in how to contact the hospitalist and having to speak with multiple persons before reaching the correct hospitalist. Hospitalists also described frustration about not having access to direct phone lines for PCPs:

"There’s one office in town…that has this incredible fortress that keeps you from speaking to the actual docs." (hospitalist, urban)

Both hospitalists and PCPs cited lack of personal relationships with clinicians in the other setting as contributing to care coordination challenges:

"When they [clinicians] know you face-to-face, probably things are much easier the next time you call them… at this moment my perception is there is a big barrier." (hospitalist, rural)

With regard to exchanging information such as discharge summaries, individuals described a lack of feedback loops between clinicians in different settings to confirm receipt of information. Both hospitalists and PCPs noted that hospitalists were often unaware of what information PCPs needed following a hospitalization:

"I guess…the magnitude of needing it [a discharge summary] doesn’t necessarily always strike people who aren’t trying to pick up the pieces on the back end." (PCP, rural)

Hospitalists expressed uncertainty about PCP receipt of information and described situations in which they attempted to send or share information that was not received by the PCP:

"There’s one doc [PCP] we cover…he complains all the time he never gets his H&Ps [History and Physicals]…but he turns his fax off at five o’clock when he leaves, and he doesn’t leave it on over the weekend…He has access to the EMR but he doesn’t go there, so I’m, like, come on…." (hospitalist, urban)

Hospitalists described having systems in place for delivery of discharge summaries to PCPs, but not having confidence in the system:

"I perhaps overly rely on the idea that the EMR discharge [summary] gets there. I mean, I really try to make sure that I get the name of the PCP [on the summary]…but I have no real confidence that it gets there…." (hospitalist, urban)

No group described systematically establishing feedback loops to ensure that clinicians in the other setting were receiving necessary information.

Both hospitalists and PCPs expressed having uncertainty about the medications that patients should be taking because of discrepancies in medication lists from different sources. While both groups described making telephone calls to other clinicians for rare or exceptional cases to resolve questions about medications, hospitalists described frequently making changes perceived as necessary without involving PCPs due to time constraints.

Hospitalist Perspectives

Hospitalists described difficulty in arranging follow-up for patients who were discharged after-hours or on weekends, as PCP offices were generally closed during these times. In these cases, arranging follow-up care was frequently delegated to the patient, although some hospitalists described keeping a list of weekend discharges and making the appointments themselves on weekdays. Many hospitalists also described having difficulty establishing a follow-up plan for patients without PCPs or insurance.

PCP Perspectives

PCPs noted that they were frequently unaware that patients from their practices were hospitalized. They also described lacking important information at a patient’s follow-up appointment; this was more frequent in rural practices that did not share EMRs with the discharging hospital. PCPs also noted that when records were received from the hospital, important information was often difficult to locate:

"I don’t need necessarily to know the entire timeline of when did they check the blood pressure and when did they recheck it again, what time was the Tylenol given. I really don’t need all that. I need to know what do you need me to do and what are we looking at today." (PCP, rural)

The perception of feeling undervalued by hospitalists was also described in some of the PCP groups, particularly with regard to medication decisions:

"There can be a lot of disrespect between the providers who are taking care of the patient, which confuses the patient a lot…like sometimes assuming that patients are on certain medications, because their doctor [PCP] is an idiot…but the thing is the patients have trust in us, and they’re going to come back. You’re only going to take care of them for a week…at most, and so we have to…communicate better." (PCP, rural)

ACCOUNTABILITY CHALLENGES

Accountability for Pending and Abnormal Tests

Shared Perspectives

PCPs and hospitalists both expressed uncertainty with regard to responsibility for following up on specific tests pending at discharge. Both groups brought up concerns about missing test results, and even litigation:

"Well, one time, just through the digging—it was not even in the discharge summary—I just happened to find a new lung nodule on CT that needed follow-up with primary care, but it wasn’t said in the discharge. It was just in the mix of all that stuff, and that was scary." (PCP, rural)

Although concerns about missed test results were frequently mentioned in both groups, there was a lack of clarity around the legal responsibility for pending and abnormal tests among both hospitalists and PCPs.

Both hospitalists and PCPs expressed the opinion that specialists should be accountable for the results of tests that they had ordered or performed, such as biopsy results from a procedure. In general, hospitalists were viewed as accountable for sharing information with PCPs regarding pending or abnormal test results, most commonly through discharge summaries. Both groups believed that once information regarding pending or abnormal tests was shared with the PCP, the responsibility for the test results was then shared between clinicians.

Hospitalist Perspectives

In all hospitalist groups, individuals stated that PCPs were accountable for lab and imaging tests that were recommended during the weeks to months following hospital discharge; this viewpoint was expressed in only one PCP group. The belief that patients shared some accountability for pending or recommended tests was expressed only by hospitalist groups. In addition, hospitalists expressed feeling greater accountability for certain tests, such as culture results, compared with other test results:

"Well, it depends on the test…If I run a tuberculosis culture, and I thought there was a chance it was going to turn positive, you know, and it comes back three weeks later…I would feel more responsible for that than a ferritin [blood test] that hadn’t come back…" (hospitalist, urban)

PCP Perspectives

PCPs expressed the view that hospitalists were accountable for tests pending at the time of discharge unless the hospitalist provided that information to the PCP via a discharge summary or phone call. PCPs also stated that they typically followed up on tests that they knew were pending at discharge, but that they were uncertain whether this should be their responsibility:

"We end up doing it [following up on tests] because it’s our job to follow up the loose ends, but sometimes you’re like, well, you ordered that test…that’s not a test I would have gotten…please do something with it." (PCP, urban)

Individuals in PCP groups expressed frequent difficulty in assuming accountability for pending or abnormal tests due to the absence of a handoff from hospitalists:

"…if I don’t happen to get that piece of paper [about a test], I have no idea. So, to me, the medical responsibility would be on the provider who ordered the test, read it, and found it abnormal, to make sure that the follow-up is taken care of, and once they notify us, it’s off of them." (PCP, rural)

Accountability for Home Health Care

Shared Perspectives

Both clinician types generally viewed hospitalists as accountable for initial home health care (HHC) certification at discharge, but hospitalists and PCPs held divergent views regarding the responsibility for HHC orders following the initial certification. Both PCPs and hospitalists noted that hospitalists often avoided interactions with HHC agencies following initial certification.

Hospitalist Perspectives

Avoidance of HHC interactions on the part of the hospitalist was partially explained by their view that once a patient was discharged from the hospital, HHC was primarily the responsibility of the PCP:

"I think once the patient goes home, it should be their primary care provider. I mean…as hospitalists, I think we need to focus when the patient is in the hospital, and once they leave, they should be under the care of their primary care provider." (hospitalist, rural)

Individuals in all hospitalist groups expressed frustration about continuing to receive documents to sign from HHC agencies for weeks to months after a patient encounter. Hospitalists expressed uncertainty about whether they should be signing such documents and even frustration at the volume of documents they received for months beyond their perceived accountability for HHC orders:

"I see patients all the time in the hospital who I’ve discharged and…I get, like, twenty sheets about them every month. And it’s like random questions about do they still need home oxygen, do they still need this, and I’m like, I saw them for two days in February of 2011. I do not know. Please find out who the PCP is." (hospitalist, urban)

PCP Perspectives

In all PCP groups, individuals expressed difficulty in addressing requests from HHC agencies when the PCP had not yet seen the patient:

"One of my patients had a stroke, and I have not seen her since she got out of the hospital. But I did have the records, so they wanted to know…what…criteria…does she meet…It’s a stretch for me to say…what physical ailments specifically she has because I have not seen her yet, so it just—it puts us in a bind…" (PCP, rural)

Successes and Ideals for Coordination of Care

Shared Perspectives

Hospitalists and PCPs both described engaging in more intensive communication such as personal telephone calls placed from hospitalists to PCPs for “high risk” patients (e.g., patients taking warfarin). Clinicians in both settings noted that access to direct phone lines for clinicians in the other setting made it easier to complete these calls.

EMR-based solutions were also identified as contributors to successful communication between hospitalist and PCP, including having access to the same EMR and/or healthcare system. Specific EMR functions, such as direct messaging capabilities between the two groups, were noted to promote better communication:

"I actually prefer the [EMR] message. I just think it’s most efficient because…I’m doing stuff in the [EMR] system, and it’s easier just…to hit that tab, open that up, type it in, send it…I also feel like it’s efficient for the folks on the other side…instead of that three-page summary, I’ve hit the one or two points that I think are issues, and they can write back with questions…" (hospitalist, urban)

In the settings in which clinicians did not share the same EMRs, they expressed a desire for shared EMR access and development of EMR functions such as sending and receiving auto-alerts about hospitalizations and the ability to search and pull data from the EMR.

Enhanced interpersonal relationships were frequently identified as an important contributor to successful coordination of care. For example, if a hospitalist and PCP had worked together previously, they were more likely to have access to direct phone lines for each other. As a result, they described making more concerted efforts to coordinate care:

"[A] recently graduated resident was the hospitalist who took care of her [my patient], and as soon as she hit the door in the ER, my previous resident e-mailed me right away and said, '…I just saw your patient. I’m picking her up as the hospitalist, and I will forward [the] H +'P”…I got the admit notes right away. I got the discharge note right away…she [the patient] came back for a hospital follow-up in the next week and… did great…" (PCP, urban)

Improving relationships with clinicians in the other setting was also mentioned as an aspiration for practices in which such relationships did not exist:

"Probably having frequent meetings and gatherings where we mingle and... see each other and... formally or informally communicate. We don’t know what their problems are—I mean, in regards to communicating with us, and we know what problems we have but we haven’t had the chance to connect, so that would be a good way…" (hospitalist, urban)

Solutions for establishing accountability for pending and abnormal tests included increased use of EMR alerts to prompt clinicians to follow up on pending tests and automatic forwarding of alerts to PCPs for test results. To clarify accountability for home health care, one proposed solution was to develop a formal system with specific expectations and timing for hospitalists and PCPs:

"… [the home health] qualification is done by the hospitalist and the discharge planner, and that is in service until they [patients] see their follow-up, or give a set amount of time and then they [patients] have to bring the pink sheet or whatever and get it done. I mean, you’d have to change the whole system, but I mean…that way there’s a cutoff…" (PCP, rural)

Hospitalist Perspectives

Hospitalists expressed the desire for access to a centralized or EMR-based scheduling system to enable easier booking of PCP appointments. A few hospitalist groups also discussed the possibility of creating a hospital follow-up clinic for patients who had no access to insurance or PCPs. However, only one group from this sample had a functioning hospitalist-run follow up clinic.

PCP Perspectives

Multiple individuals in PCP groups suggested that receiving a short, structured, timely summary regarding a patient hospitalization was an attractive alternative to the multiple pages of data they currently receive. In addition, PCPs expressed that they would like PCP follow-up appointments for patients to be scheduled by hospitals prior to discharge.

Some PCPs noted positive experiences with outpatient-based transitional care innovations, including standardized readmission risk assessments to determine follow-up appointment timing, structured hospital follow-up appointments, and the development of infrastructure to support Medicare transitional care billing. In a few practices, specific staff members (e.g., social workers, pharmacists, nurses) were designated to help patients with post-hospitalization needs.

DISCUSSION

In this exploratory qualitative study, we found that hospitalists and PCPs both identified many similar challenges in coordination of care. When differences were noted between groups, they were often issues that clinicians in the other setting were not aware of. Solutions for improving care coordination included better interpersonal and informational access through both direct telephone communication and shared EMRs. Building and sustaining personal relationships was perceived as a key component for improving coordination of care and establishing accountability between groups.

Communication between hospitalists and PCPs around patient hospitalizations occurs infrequently and is associated with scenarios involving more serious patient issues, including readmissions, following discharge.^6,7 The reasons for suboptimal communication between hospitalists and PCPs, however, are not fully understood. Hospitalists and PCPs in our study identified some of the same challenges as those of participants in prior qualitative studies of hospital discharges, including difficult interprofessional communication, suboptimal information transfer between settings, lack of information feedback loops between settings, and lack of clarity with regard to accountability for pending tests.^8,9,13 We also identified emergent themes that we had not anticipated, including a lack of clarity regarding accountability for home health care services and the importance of personal relationships, both as contributors to challenges as well as keys to improving the coordination of transitional care.

Our study also found possible solutions to the challenges described, such as creating formal systems to establish accountability between inpatient and outpatient settings for tests, imaging, and home health care. Another proposed solution, enhancement of interpersonal relationships, included frequent gatherings of inpatient and outpatient clinicians.

Further research would be of value in order to investigate whether establishing accountability for pending tests and home health care via formal service agreements between hospitalists and PCPs results in fewer missed test results and/or hospital readmissions. It would also be of great interest to evaluate whether mandatory attendance at gatherings of inpatient and outpatient clinicians would contribute to improved interprofessional communication and coordination of care. Successful strategies might be developed and disseminated through practice collaboratives.

Limitations of this study include the narrow generalizability of our findings, as we recruited participants from practices or hospitals within the State of North Carolina that were already actively working to improve care transitions. The experiences of these focus group members may not represent the global experience of hospitalists and PCPs. All participating clinicians were using EMRs in their practices, yet they still described challenges in coordination of care, which suggests that the mere presence of an EMR does not remedy all issues that may arise with regard to coordination of transitional care.

CONCLUSIONS

Hospitalists and PCPs encounter many similar challenges in the coordination of care, yet their experiences differ in light of their distinct roles in the sending and receiving of information around patient discharges. Although clinicians were asked general questions about care coordination, responses centered around care barriers rather than solutions, which is a reflection of their current experience with the process of coordinating care. Efforts to improve the coordination of transitional care between hospitalists and PCPs should focus on understanding the perspectives of clinicians in each setting and implementing improvement strategies that engage both groups.