Understanding how health systems facilitate primary care redesign

Abstract

Objective

To understand how health systems are facilitating primary care redesign (PCR), examine the PCR initiatives taking place within systems, and identify barriers to this work.

Study Setting

A purposive sample of 24 health systems in 4 states.

Study Design

Data were systematically reviewed to identify how system leaders define and implement initiatives to redesign primary care delivery and identify challenges. Researchers applied codes which were based on the theoretical PCR literature and created new codes to capture emerging themes. Investigators analyzed coded data then produced and applied a thematic analysis to examine how health systems facilitate PCR.

Data Collection

Semi‐structured telephone interviews with 162 system executives and physician organization leaders from 24 systems.

Principal Findings

Leaders at all 24 health systems described initiatives to redesign the delivery of primary care, but many were in the early stages. Respondents described the use of centralized health system resources to facilitate PCR initiatives, such as regionalized care coordinators, and integrated electronic health records. Team‐based care, population management, and care coordination were the most commonly described initiatives to transform primary care delivery. Respondents most often cited improving efficiency and enhancing clinician job satisfaction, as motivating factors for team‐based care. Changes in payment and risk assumption as well as community needs were commonly cited motivators for population health management and care coordination. Return on investment and the slower than anticipated rate in moving from fee‐for‐service to value‐based payment were noted by multiple respondents as challenges health systems face in redesigning primary care.

Conclusions

Given their expanding role in health care and the potential to leverage resources, health systems are promising entities to promote the advancement of PCR. Systems demonstrate interest and engagement in this work but face significant challenges in getting to scale until payment models are in alignment with these efforts.

1. What we already know on this topic

• Primary care redesign (PCR) could shift the orientation of the US health system toward one that emphasizes comprehensive and coordinated care.

• Substantial efforts have been made to redesign the delivery of primary care in the United States.

• Primary care practices often lack the infrastructure and resources to transform care delivery.

2. What this study adds

• This study provides a look at 24 health systems to understand the changes they are making and barriers to PCR.

• PCR is currently not widespread across our sample of health systems, suggesting that there has not been wholesale redesign.

• Health systems are promising settings to promote PCR, but significant barriers, including being out ahead of the pace of payment change, must be overcome.

1. INTRODUCTION

Primary care is the foundation of an effective health care system. ¹ , ² , ³ When primary care practices foster prevention and support disease self‐management, health care quality improves along with the cost and efficiency of care. ⁴ , ⁵ Substantial investments have been made to transform the delivery of primary care including coordinating care for those with chronic disease, improving patient access, and implementing alternative payment models. ⁵ , ⁶ , ⁷ , ⁸ , ⁹ , ¹⁰

Despite some improvements, substantial barriers remain, and primary care practices often lack the infrastructure to transform care delivery. ⁷ , ¹¹ Additionally, as practices acquire an increasingly complex caseload, primary care providers (PCPs) may struggle to keep abreast of the latest innovations. ¹² , ¹³ Financial limitations may restrict practices seeking to adopt the processes and teams necessary to manage chronically ill patients. ¹⁴ Furthermore, in an era of measurement and data, the cost of investing in and maintaining an electronic health record (EHR) with robust decision support capabilities may be out of reach for small practices, and documentation and analytic requirements may add to PCP workloads. ¹²

There are several frameworks that document the components of primary care transformation. ⁸ , ¹¹ , ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ Looking across the frameworks, common elements include engaged leadership, data‐driven improvement, empanelment, and team‐based care. Additional components include population management, continuity of care, and care coordination. The ultimate goal is a payment system that rewards improvement, and high‐quality patient care that is offered in formats that meet the patient's needs, including telehealth, group visits, and appointments with team members. ⁷ , ¹¹ The term Primary Care Redesign (PCR), coined as early as 1993, ¹⁶ refers to the transformation of primary care through systematic changes to improve quality, reduce costs, improve patient satisfaction, and improve the work life of providers. ⁷ , ¹¹

While the patient‐centered medical home (PCMH) and other quality improvement initiatives predate the concept of PCR, some professionals view PCMH as only scratching the surface. For example, Porter, Pabo, and Lee contend that to fundamentally transform primary care, it must be deconstructed to reflect not a single set of services, but rather “a group of services delivered to meet the different needs of multiple subgroups of patients.” ¹¹ PCR could shift the orientation of the US health care system toward one that emphasizes comprehensive and coordinated primary care. ¹ , ² , ³ , ⁴ , ⁵ , ⁶ , ⁷ , ¹¹

Many freestanding primary care practices may encounter substantial barriers when redesigning care delivery. Given their expanding role in health care and potential to leverage resources, we theorize health systems could be better positioned to overcome barriers and support PCR. ¹¹ , ¹⁸ , ¹⁹ , ²⁰ For this study, we define a health system as at least one hospital and one physician organization, where affiliation may occur through shared ownership or a contractual relationship for payment and service delivery. ²⁰

Adding to growing pressure to increase efficiency and control costs, capital constraints, regulatory changes, and value‐based payment are driving the consolidation of health care providers into vertically integrated health systems, thereby significantly increasing the number of physicians who fall under the influence of a health system. ²¹ , ²² , ²³ Health systems may offer greater access to capital and economies of scale in information technology and data analytics, and operational support services, which can assist providers with synthesizing evidence, training, or specialized tasks such as quality improvement. ²³ Health systems have contract management expertise ¹⁶ and care for larger patient populations which can mitigate financial risk and help achieve cost and care efficiencies, which are critical to earning incentive payments under value‐based payment models. ²² , ²³ Health systems provide a centralized leadership and management structure and employ case managers who play essential roles on care teams, manage care across the continuum, and support patient‐centric care. ²⁴ Lastly, health systems have partnerships through contracting or ownership with physician organizations, and networks of multispecialty and primary care physicians where care is delivered.

To our knowledge, no studies have looked across a large number of health systems to understand how systems are facilitating PCR, what PCR initiatives are taking place within clinics, and the challenges faced.

2. METHODS

This analysis is part of a larger qualitative study to identify the characteristics of high‐performing health systems and examine how health systems promote evidence‐based practice in care delivery. ²⁰ , ²⁵ Twenty‐four health systems in four states recruited from a purposive sample agreed to participate in the study. The primary source of data was semi‐structured interviews with 162 respondents. Interviews were conducted with 5‐10 leaders associated with each system and physician organization (PO) [ie, chief executive officer (CEO), chief financial officer, chief medical officer (CMO), chief quality officer (CQO), chief information officer]. Tailored interview instruments were created for each type of respondent; topics included the origins, organization, and governance of the health system; market context; care structures; and relationships among system affiliates and providers (full description of data and methods is available in the Appendix S1). ²⁰

Interview transcripts were systematically reviewed to identify how system leaders define and implement initiatives to redesign care delivery and identify challenges and lessons learned. Researchers applied 17 codes, based on interview topics, via Dedoose software. There were a total of 16 063 code applications across 2970 pages of transcripts.

Investigators analyzed coded data, then produced in‐depth memos on care delivery redesign (CDR), detailing and organizing what was learned about how each system uniquely defined, structured, integrated, and supported care redesign. Each system's CDR memo was between 7 and 13 pages long, for a total of 228 single‐spaced pages across the 24 systems.

The CDR memos ²⁶ and transcripts from the CEO interviews formed the primary data for this analysis. Theoretical definitions of PCR initiatives from the literature guided initial code development; new codes were added as themes emerged. A comparative approach was used to identify themes related to how health systems may facilitate PCR. The PCR analysis enables the identification of both cross‐cutting and unique factors by analyzing each system independently before comparing the results across systems. This study was approved by the Pennsylvania State University Institutional Review Board.

3. RESULTS

Our sample included academic, nonprofit, and quasi‐public systems and an associated physician organization for each system. The health system respondents described complex organizational structures between the system, PO (eg, medical groups, medical foundations, independent practice associations [IPAs], and faculty practices), and primary care clinics, resulting in complicated governance structures (Table 1). Nearly all the health systems employ PCPs either through the health system or a medical foundation and nearly half also work with affiliated IPA physicians through a Clinically Integrated Network (CIN).

TABLE 1.

Health System Characteristics (n = 24)

Health system characteristics			Physician organization characteristics					Payment models
Organization type	Region served	Number of hospitals	Length of time integrating with system (y)	Practice type	Physicians employed	Clinically integrated network	Other affiliated physicians (ACO, IPA, or JVs)	Health plan	ACO participation
Academic	Multiple counties across several states	17 owned	>20	Employment	Yes	No	Yes	No	No
	Multiple counties within a state	5 owned	<5	Faculty practice and medical group	Yes	Yes w/partners	No	Employees only	Yes
	Multiple counties within a state	3 owned, 1 JV, 7 affiliated	>20	Faculty practice	Yes	Yes	Yes	No	Yes
	Multiple counties within a state	1 owned	>20	Faculty practice and employment	Yes	No	Yes	Yes (JV)	No
	Single county	2 owned	>20	Medical foundation (medical group and IPA)	Yes	No	No	No	Yes
Academic AA	Multiple counties within a state	13 owned	>20	Faculty practice and employment	Yes	Yes	Yes	Yes	Yes
	Multiple counties within a state	3 owned 2 affiliated	11‐20	Medical group and faculty practice	Yes	No	Yes	Yes (JV)	Yes
Non‐profit	Multiple counties across 3 states	49 owned 2 JV	5‐10	Medical foundation (medical group and IPA)	Yes	Yes w/partners	Yes	Yes (JV)	Yes
	Multiple counties across 3 states	2 owned 1 managed	>20	Employment	Yes	No	Yes	No	Yes
	Multiple counties across 2 states	12 owned	>20	Employment	Yes	Yes	Yes	Yes, JV (launching)	Yes
	Multiple counties across 2 states	8 owned	11‐20	Employment	Yes	Yes	No	Yes (JV)	No
	Multiple counties across 2 states	5 owned, 1 affiliated 1 JV	>20	Employment	Yes	No	Yes	Yes	Yes
	Multiple counties across 2 states	2 owned	>20	Employment	Yes	Yes w/partners	No	Yes, JV (launching)	Yes
	Multiple counties within a state	8 owned	>20	Employment	Yes	Yes	Yes	No	Yes
	Multiple counties within a state	7 owned	>20	Employment	Yes	Yes	Yes	No	Yes
	Multiple counties within a state	5 owned	5 to 10	Medical foundation (medical group and IPA)	Yes	No	No	Yes	Yes
	Multiple counties within a state	4 owned, 1 JV, 3 affiliated	>20	Employment	Yes	No	Yes	Yes	Yes
	Multiple counties within a state	2 owned 1 JV	>20	Medical foundation (medical groups)	No	Yes w/partners	Yes	No	Yes
	Multiple counties within a state	2 owned	>20	Employment	Yes	No	Yes	No	Yes
	Multiple counties within a state	2 owned	11‐20	Medical foundation (medical groups and IPA)	Yes	No	Yes	Employees only	No
	Single county	3 owned 2 affiliated	>20	Medical foundation (medical groups and IPA)	Yes	No	Yes	Yes	Yes
	Part of one county	3 owned	<5	Medical group and IPA	Yes	No	No	No	No
Quasi‐public	Two counties within a state	2 owned	11‐20	Employment	Yes	Yes w/partners	Yes	No	Yes
	Single county	1 owned	11‐20	Employment	Yes	No	Yes	No	Yes

Academic affiliation agreement (AA) refers to a health system (HS) that incorporates parts of a university health system and a nonprofit health system operated under a joint governance arrangement. A quasi‐public corporation is a private company that is supported by the government with a public mandate (and funding) to provide a service. Hospitals is a count of acute, general hospitals and excludes behavioral health facilities. (For multistate HS, some of the counted hospitals are located outside of the four focal states of the study). Owned refers to hospitals owned and operated by the HS. Affiliated refers to hospitals managed by the HS under affiliation agreements (such as MOUs, management contracts). JV refers to jointly owned and operated or managed. IPA refers to an Independent Physician Association. Clinically integrated network w/partners refers to participation in a CIN sponsored by two or more HS. Health plan is a health insurance company either owned by a health system or through a JV. ACO refers to whether the health system offers an ACO product to payers or participates in an ACO network offered by another entity.

3.1. Key findings

Each of the 24 systems described engaging in PCR initiatives, yet the scope and scale across systems were heterogeneous. Even though health systems have substantial history with public reporting and working with POs, nearly all respondents described PCR as new and small‐scaled. For health systems, PCR not only involves transformation within the primary care practices but also includes a culture shift, new organizational goals, and change management for the health system and PO. In some systems, there is tension between the historical emphasis on acute care and transitioning to PCR. According to one CEO:

When I get into a capital discussion, I’m looking at how do I build in stability that can provide better primary care and have some resources available in the clinic that help patients remain healthy so they stay out of the hospital. That takes capital. Well, over here, I need more robotic surgeries. I need more hospital beds, or I need more infrastructure, technology, equipment that's going to take care of those patients. That's really expensive… I’ll do that [provide resources to enhance primary care] eventually. I just don't want to do it right now.

Only two health systems described a “large and organized effort” around PCR, both beginning approximately 10 years ago. Not surprisingly, these systems had the most advanced and widely disseminated interventions. These health systems have spread PCR to nearly all of their primary care practices and appear to have several characteristics in common. First, these health systems prioritized a centralized approach with overarching and aligned strategic and operational plans. They worked to overcome silos and established systems to promote structure and accountability, with formal processes and clear roles and expectations. Transparency and shared decision making were prioritized and bolstered by data and reporting structures. At the practice level, physicians and front‐line workers were involved in workflow development, as each practice implemented PCR according to the needs of their specific patient populations. Respondents described allowing practices additional time and incentives for planning as essential to successful implementation.

Four key themes were identified and provide organization for the results: (1) centralized health system resources to facilitate PCR; (2) PCR initiatives taking place within health systems; (3) payment mechanisms to support PCR; and (4) motivating factors.

3.2. Centralized health system resources to support PCR

A significant benefit of health systems facilitating PCR is centralized resources that may not otherwise be available to clinics. Centralized resources may relieve some of the administrative burden from PCPs and in turn promote increased focus on patient care.

3.2.1. Leadership and governance

Health systems provided centralized leadership, decision making, governance, and strategic planning to support PCR. This was viewed as a benefit for PCPs, allowing them to focus on patient care. As described by a health system CEO:

Being a part of a larger, integrated system, quite frankly, in my opinion, allows them [primary care physicians] —gives them the freedom to be able to actually do what they want to do, which is provide care and have the safety and resources of being able to take some of the administrative burden off, standardize support and systems, and let them focus on patient care and let the administrative stuff be taken care of by people like me.

The importance of leadership and buy‐in from the health system, PO, and collaborating practices was described as essential to implementing successful PCR initiatives. For example, a respondent described the development of the team‐based models as “both top‐down and bottom up” and noted “primary care owns [team‐based care development], but with a lot of support, guidance and nurturing from the health system.” Acknowledging the considerable time commitment for busy providers who were participating in 2‐hour meetings every 4 to 6 weeks, the system granted relative value unit (RVU) credit for the time spent developing the team‐based model.

3.2.2. EHR & data‐driven approaches to PCR

The EHR is viewed as essential to PCR. For example, health systems use EHRs to identify high‐risk patients with chronic conditions in order to better coordinate care and can supplement practice's EHR data with data from external sources (eg, state‐level quality organizations) in order to obtain a more comprehensive view of patients. In particular, one health system is integrating clinical data from its EHR, health plan claims data, and county‐level social services data to evaluate population needs, stratify by risk, and assess what programs would be most effective for supporting at‐risk patients. According to one CEO, an integrated EHR can be a valuable PCR tool:

Our physicians, whether the patient is being seen at [clinic name] or here, can look at the patient's record and know exactly what happened at the last site of service, wherever that may be.

3.2.3. Technology

The majority of health systems are investing in centralized technology to support care management within primary care practices, often cost prohibitive for small practices to procure and manage. For example, one health system initiated a remote monitoring program which is geared toward complex chronic disease patients who are high risk for hospital readmissions. As part of the program, patients send data (eg, blood pressures) to a panel of nurses who then interact with the patients, troubleshoot, and make adjustments as needed. When patients stabilize, they “graduate” from the program and transition back to their own primary care teams.

3.2.4. Centralized staff to support primary care

Health systems have the ability to centralize staff (eg, patient navigators, case managers) who supplement care in the primary care setting, enabling PCPs to “take care of quote/unquote bigger things.” One health system leader described deploying a coordinated approach with the CIN by purchasing a tool and partnering with contracted payers to identify high‐risk patients. Nurse case managers, employed by the CIN, help patients manage chronic conditions and improve medication adherence and diet. The nurses follow up with patients who may otherwise “bounce in and out of the hospital for various reasons” and help monitor weight and vital signs. If a high‐risk patient arrives at the emergency department, the nurse case managers are alerted.

3.2.5. Standardized processes and guidelines

Several respondents described the importance of standardized processes in order to enhance patient experience, reduce inefficiency and augment providers’ and staff members’ ability to effectively work in multiple clinics. Health systems also are instrumental in assisting primary care practices overcome a common barrier to evidence‐based care through centralized physician working groups that support decision making and upkeep of practice guidelines. As one CMO explained:

[W]e do have a primary care guidelines committee that we created just to look at new recommendations that come out and look at changes to existing guidelines, and decide what we think is important and do we want to build decision support around that, or do we need to modify our existing decision support when it comes to things relevant to primary care.

In most cases, leaving workflows as customizable was viewed as important, as many primary care practices were previously independent and providers seek to maintain some individuality and autonomy.

3.3. PCR initiatives taking place within health systems

We identified three common PCR initiatives across the majority of health systems (team‐based care, population management, and care coordination). In the most advanced systems, these initiatives are well integrated; for example, the most advanced primary care teams also include nurses and clinicians who manage populations and coordinate care. However, in the majority of health systems, these initiatives were siloed and not yet disseminated across all practices. In some cases, the initiatives do not appear to fully leverage the systems’ resources, and some PCR initiatives have yet to advance beyond what practices could do independently.

3.3.1. Team‐based care

The use of teams was the most developed PCR initiative and sometimes was described as part of the system's PCMH effort. A health system medical director for quality noted that the system's team‐based approach standardizes care in primary care practices, allowing it to systematically address care delivery and quality.

Among the 22 health systems that described a focus on team‐based care, there were differences in how the teams were structured (Table 2). Ten of the primary care teams were designed to help the PCPs address complex patient needs, including combinations of nurses, case managers, patient navigators, social workers, and sometimes pharmacists. Seven health systems also include advanced practitioners, an approach that was described as ensuring “that everyone is practicing at the top of their license and we're maximizing the effectiveness of health care.” Some respondents described using a more traditional team that utilizes a combination of medical assistants, nurses, and/or scribes to enhance a physician's ability to efficiently and effectively care for patients. Although this level of team fails to leverage the resources of the health system, it was described as beneficial. According to a vice president of medical affairs, with little investment, adding scribes to the team was “unequivocally” effective in “delivering better, more prompt care accessible to more people” and “you're making money within a week.”

TABLE 2.

Health systems stratified by self‐reported structure of initiatives (n = 24)

	Number of health systems
	Some (1‐3)	Several (4‐10)	Many (11+)
Structure of teams
Expanded care teams including nurses, case managers, patient navigators, social workers, and sometimes pharmacists		X
Addition of APPs to care teams		X
Small care teams including MAs and nurses	X
Front office and nonmedical staff	X
Team structure not discussed	X
Population management initiatives
Complex care management
Super utilizers and chronic disease management			X
Care guides and navigators		X
Utilizing pharmacists/others to enhance care		X
Panel management
Identifying patients in need of screenings, preventive care, and disease management		X
Assessing social determinants of health and connecting patients with services	X
Health coaching
Coaches/disease management education	X
Preventative lifestyle coaching	X

Abbreviations: APPs, advanced practice providers; MAs, medical assistants.

Most team‐based PCR initiatives were described as being in pilot or implementation stages with significant variation in how many clinics within the system had fully implemented the interventions. The strategy for one health system that launched team‐based care in 2014 was to spread its team‐based model to 3 to 4 practices per month, with a goal of full implementation by the end of 2018. A different health system disseminated its team‐based care model across its entire community practice in stages. Stage 1 implementation began in 2013 and included infrastructure changes, training, and development of tools and dashboards. Stage 2 included restructuring primary care practices into teams with advance practice providers, nurses, panel managers, social workers, and pharmacists. In Stage 3, multiple disciplines were integrated into a single visit and complex care coordination was added for high‐risk patients. In Stage 4, which was under development at the time of the interview, specialists (eg, cardiologists) will be colocated with or physically come to the primary care practice to facilitate comanagement of patients. Approximately 250 teams are in Stage 2 or 3, with about one‐third of practices operating in Stage 3.

3.3.2. Population management

While population health management is not exclusive to primary care settings, respondents describe population management as beginning with primary care and extending out to the larger group of community patients who are attributed to the system. A benefit of health systems facilitating population management through the primary care infrastructure is the ability to develop an overarching patient management strategy, as described by one CMO:

There are many things in population health management that fall to the primary cares’ and to the physician offices, but we also have worked on transitions of care from the acute side to primary. When patients are discharged, working that out with our primary care clinics, so our case managers on the acute side working with the care coordinators on the ambulatory side, so it does affect the entire organization.

Population management approaches in primary care traditionally include 3 mechanisms: complex care management, panel management, and health coaching to facilitate behavioral changes for chronic disease. ⁷ Respondents from all 24 systems described PCR initiatives related to complex care management (Table 2). The next most common approach was utilization of panel management where gaps in care are identified and patients are directed to preventive care or screenings. Through empaneling patients in subgroups, health system respondents described the POs, and in turn primary care practices, as having an enhanced ability to provide the right level of care and to utilize a wider array of tools and technologies. Only a handful of leaders discussed health coaching.

Almost all of the respondents reported that centralized population management was relatively new to the health systems, initiatives were small‐scale, and the standardization of services was evolving. As one CQO explained:

I can tell you that as recent as yesterday, we're starting to shift some of that thought process about how do we look at this more population‐based because we recognize that the tertiary and quaternary care should be a blip in the patient's health needs over time, that the real care and the real interventions to keep people well or healthy need to happen outside of the four walls of the hospitals. So we're beginning to shift some of that perspective.

3.3.3. Coordinated and Integrated primary care

To meet patient needs, primary care includes the coordination of care for services that fall outside of what is provided within the practice. ⁷ Respondents described initiatives to better manage patient care, including care coordinators in primary care practices to assist patients with medication management, scheduling specialist appointments, arranging for transportation to medical appointments, and coordinating follow‐up, if needed.

All CEOs described an integrated delivery system, which provides services and maintains relationships across the continuum of care. These relationships can facilitate care coordination and were described as a benefit to PCPs by one CMO:

I talk a lot with primary care physicians, and I think it is becoming extremely difficult to practice independently, completely independently. To keep up with all the regulatory burdens and run the business model I imagine must be increasingly difficult. I think there is some value to being part of the system for support and help along that vein to help take care of patients in a safe, satisfying manner. From that aspect, I think it's good to be part of an integrated system.

Some respondents acknowledged that without a health plan they could not fully meet the definition of an integrated delivery system (Table 1).

3.4. Payment mechanisms to support PCR

Alternative payment systems were frequently mentioned as a necessary component of PCR. Specifically, for health systems PCR was described as an avenue to manage complex patients outside of the emergency department, and to manage low‐acuity conditions at a lower cost. In addition, 79% of the health systems participate in an accountable care organization [ACO (Table 1)], and those systems placed stronger emphasis on the relationship between financial risk, population management, and care coordination. Systems with a health plan placed increased emphasis on the importance of care coordination and population management, according to one CEO:

Only now are we getting an insurance arm back associated with our organization. That will drive us obviously more to think about population health. I would say our organization has built the capacity and the care coordination and registries and that sort of thing.

Across PCR initiatives examined in this study, return on investment (ROI) was a concern for respondents. Moreover, the slower than anticipated rate in moving from fee‐for‐service to value‐based payment was noted by multiple respondents as a challenge. As one executive stated, under the current reimbursement system, the only financial benefit of this work “might be cost avoidance.” Without a payment model that supports PCR, many health systems are limited in how quickly they can transition:

As we moved into the [team‐based care] structure, we found that we were increasing our infrastructure cost to deliver care in a different way, but we weren't necessarily being reimbursed for that kind of care.

3.5. Motivating factors

Nearly all of the health systems established relationships with a PO more than 10 years ago, yet the majority of respondents describe a delivery model that is still largely focused on acute care. When asked about key investments for the future, respondents describe a necessary shift in priority from acute care to primary care, in some cases as a way to transform a broken system:

I’d say the primary care model itself…Number one, physicians are already burning out trying to deal with the 2,000 [patients] that they have and dealing with the electronic medical record. It's not good for them. It's not good for patients. It's hard to get into, there's not enough access, it takes weeks to get new appointments…We know that's [primary care] the model to invest in. We're investing heavily into redesigning the model. We've got to fix it.

In other cases, as a way to remain viable in competitive markets:

I think the biggies are building out a scalable infrastructure, primary and specialty infrastructure, to feed our tertiary and quaternary services. Here, we're thinking about how do we create a primary care network that is self‐sustaining and then can drive a lot of—preserve the funnel to our tertiary and quaternary specialists and—in combination with the CIN and the Medicare ACO, and our regional growth with our affiliations.

When examining the specific PCR initiatives, we found different factors reported to motivate initiation of team‐based care compared to population management and care coordination (Table 3). Respondents articulated multiple factors that motivated team‐based care such as increasing physician/staff job satisfaction, better managing high‐risk/complex patients, and improving patient access. The goal for implementing team‐based care within primary care was described as “bringing back joy to medicine” since primary care providers were spending an exorbitant amount of time documenting, taking work home, staying up late, and ultimately getting burned out.

TABLE 3.

Health systems stratified by factors motivating team‐based care and population health initiatives (n = 24)

Motivating factors	Number of health systems
	Some (1‐3)	Several (4‐10)	Many (11+)
Team‐based care
Practicing at top of one's license		X
Increasing physician/staff satisfaction		X
Increasing efficiency		X
Increasing standardization of care		X
Better managing high‐risk/complex patients		X
Enhancing PCMH efforts		X
Becoming more patient‐centric	X
Increasing access to care	X
Team structure not discussed	X
Population management and care coordination
Changes in payment, risk assumption, participation in ACOs, value‐based reimbursement			X
Needs of community/SDH			X
Reduce readmissions/utilization/costs		X
Specific initiatives motivated by EHR data		X
Enhance value/reputation	X

Abbreviations: ACOs, accountable care organizations; EHR, electronic health record; PCMH, patient‐centered medical home; SDH, social determinants of health.

Health systems are motivated to engage in population management and care coordination through primary care practices because of the financial risk associated with readmissions and other outcome penalties. According to one CMO:

We spend a lot of time and money on working with primary care physicians and patients to make sure that they're seen after discharges, their medication reconciliation happens, and they get the care they need so they don't come back to the hospital.

Other motivating factors for population management work and the selection of corresponding initiatives included meeting local and health system needs. For example, one health system's metrics revealed substantial care disparities between different races and spoken languages. The CQO stated, “It's clear that our health system is working better for some than for others…[population management reporting] is to say not just how we're doing overall, but how we're doing for specific populations, and then target some of our interventions to specifically address if there are areas that show those disparities.” In this case, the health system developed a program within primary care clinics, which focuses on assessing social determinants of health for Medicare and Medicaid patients and connecting patients with the appropriate community resources.

4. DISCUSSION

We explored the role of health systems in PCR to better understand their resources, initiatives, motivations, and associated barriers. Respondents at all 24 health systems described initiatives to redesign the delivery of primary care, but there was significant heterogeneity across the systems, and many were in the early stages of PCR. This result may not be surprising since each health system is unique and began its PCR journey with differing resources, context, and population needs. This finding is supported by quality improvement theory, which states that improvement work is an iterative process, and delivery systems should learn from best practices but make adaptations to fit local needs. ²⁷

The literature suggests that PCR can be time‐ and resource‐intensive for primary care practices. ⁷ , ¹⁴ , ²⁸ , ²⁹ , ³⁰ We found time and resources to be a barrier in this study, even with the enhanced resources of the health system. The most notable challenge across systems with robust PCR initiatives was the substantial length of time they dedicated to PCR. This is consistent with quality improvement and public health literature, where it can take 3‐10 years to see an improvement in outcomes. ³¹

There does appear to be some value added for health systems to facilitate PCR. For example, PCR initiatives require connections beyond the walls of any one health care organization. Advanced team‐based care, population management, and care coordination should connect primary care with acute care and specialty care. This is viewed as more efficient when done in the context of a health system where resources can be centralized.

4.1. Limitations

The study has several limitations. First, although we utilized a purposive sample of 24 health systems based on attributes such as size and performance, our results may not be generalizable to other systems or states. However, the selection of a small but diverse group of health systems did allow for the collection of rich data that facilitated a deep dive into PCR initiatives and barriers, allowing us to formulate hypotheses for future research.

Second, interviews were conducted with leaders to obtain high‐level information on how the organizations promote evidence‐based practice in care delivery. We did not interview all executives or any front‐line staff whose views may vary from those we interviewed. We did, however, look across health system and PO respondents for areas of agreement/disagreement within a system.

4.2. Implications

Primary care practices have historically been underfunded and undervalued. ¹¹ Value‐based payment and alternative reimbursement models that shift care from acute settings may further incentivize provision of coordinated care rather than volume of services. These models were a strong motivator for health systems to begin PCR. Currently, systems are caught between two payment models, and therefore, a majority of health systems struggle to move their efforts beyond the pilot or implementation phase.

By utilizing centralized infrastructure and leveraging resources, health systems may be positioned to assist physician organizations in PCR efforts. However, our data point to significant challenges that must be overcome. The majority of health systems were devoting considerable resources to PCR, yet the work is not very far along and requires more time and investment. Next steps should include the exploration of payment models that support PCR.

We found substantial variation in how health systems are facilitating PCR and in the structure of PCR initiatives taking place within practices. This exploratory analysis identifies PCR initiatives that may be enhanced by the centralized resources of a health system (eg regionalized staff, standardized processes), yet we were unable to categorize systems by their approach to PCR. While there were similarities in the types of initiatives, in many ways we observed 24 different ways health systems can facilitate PCR. Additional measurement and evaluation to measure and categorize health system involvement, and the role of contextual factors (contractual relationships, structural factors, etc) in PCR is warranted. This step is necessary before examining the relationship between PCR initiatives and improved outcomes. As health systems and practices continue to dedicate substantial time and resources to PCR, future research also should identify which initiatives are most cost‐effective.

Supporting information

Author matrix

Appendix S1

Harvey JB, Vanderbrink J, Mahmud Y, et al. Understanding how health systems facilitate primary care redesign. Health Serv Res. 2020;55:1144–1154. 10.1111/1475-6773.13576