SUSTAINING THE PRIMARY CARE WORKFORCE

Early in the COVID pandemic, primary care practices rapidly switched to offering most visits by video or telephone.1 Clinicians and patients experienced virtual approaches without compromising substantially on values of relationship and technical excellence. In response to financial incentives, many practices have slipped back into old patterns as COVID restrictions ease. Nonetheless, the momentum behind virtual-first care has created a window to accelerate value-based changes to clinical practice.

The case for primary care as the foundation of high-value healthcare has grown stronger.2 Unlike other medical specialties, increased access to primary care, which is “integrated, accessible health care services by clinicians who are accountable for addressing a large majority of personal health care needs, developing a sustained partnership with patients, and practicing in the context of family and community,” improves life expectancy.3,4 Compared to areas with more than one primary care physician (PCP) per 3500 individuals, life expectancy is nearly 1 year shorter in areas with fewer PCPs.5 Findings like these led the National Academy of Science, Engineering and Medicine to designate primary care as a “common good.”6 Continued analysis of accountable care organizations’ impact on cost showed that integrated systems built around primary care perform better than those built around hospitals.7,8 Meanwhile, large companies have accelerated investments in primary care and new market entrants have introduced models of “virtual-first” primary care.9,10,11,12

Given such evident societal value, accelerating threats to the primary care workforce suggest the need for urgent changes.13 During COVID, increasing numbers of independent practices suffered financial challenges and closed or were acquired.14 Clinician burnout worsened and, consequently, physicians are seeking care for anxiety and depression in record numbers.15,16 Fewer trainees than needed are choosing primary care, while PCPs are leaving practice.17,18 As our population continues to age and require more medical services (the average Medicare patient has nearly five specialist visits annually19), the need for primary care grows ever greater.

RESTRUCTURING PRIMARY CARE WORK

The transition to virtual-first care creates opportunities to extend innovations from the patient-centered medical home (PCMH), which achieved better quality and workforce satisfaction and, in some cases, lower costs of care.20 The foundational PCMH concept of team-based care can be extended to redefine the role of PCPs as advanced practice practitioners (APPs)—nurse practitioners and physician assistants—provide care for common conditions with at least equivalent quality to physicians, at lower cost.21,22 Pharmacist-led teams can attain higher rates of hypertension and hyperlipidemia control.23 Health coaches and technology can support patients in building healthy habits. Licensed clinical social workers can partner to manage behavioral health diagnoses.

Physicians increasingly are receiving unique training to care for patients with medically complex conditions.24 Patients with complexity—defined as those with multiple chronic conditions, complicated by behavioral and social complexity—account for a disproportionate amount of healthcare spending.25 Given these exigencies, we need to accelerate changes enabling PCPs to primarily consult and collaborate with APPs in the care of patients with complexity, with APPs serving as the locus of continuity. Although prior research has focused on physician continuity and has shown that continuity improves outcomes, we also know that physicians and nurses working together and APP-led care, supported by physicians, improves outcomes.26,27,28 Physicians, working on teams with multiple APPs, are able to partner in providing high-quality, continuity care to larger panels of patients. Table 1 shows one example of care team roles and responsibilities in a virtual-first model.

Table 1 Team Member Roles on a Virtual-First Primary Care Team
Full size table

Reinforcing differentiated roles, while enabling effective real-time collaboration across primary care teams (PCTs), requires redistributing patient interactions from synchronous to asynchronous interactions. Yet fee-for-service (FFS) payments and scheduling and workflow software designed for revenue management in an FFS environment fill most physicians’ clinical time with synchronous visits, with little regard to types of patients scheduled. Nearly 5% of physician visits are filled by low-acuity conditions that could be handled by an APP without face-to-face or synchronous visits; fewer than 50% of visits focus on patients with two or more chronic conditions.29 Synchronous care for patients with chronic conditions overwhelms the schedule of most PCPs; physicians could spend 3.6 h a day caring for patients with chronic illness.30,31 Lacking timely access, patients with urgent complex needs are often seen in far less optimal settings, such as the emergency room.

The work of PCTs should be restructured to achieve optimal outcomes, rather than visit quantity, for patient populations, while reducing the total cost of care. Physicians should have time reserved for patients with acute or evolving complex issues. PCT members in this model have at least 50% of their time allocated to asynchronous work, including online chat, building care plans, and remotely consulting specialists. PCTs supported with real-time, population-level outcomes data effectively manage population health, eliminating external costs to close gaps in care.

Shifting to asynchronous patient interactions unlocks opportunities for efficiency and value. Data for clinical decision making can be collected and automated via chat or digital forms. Condition management can be standardized and templated, and flow into evidenced-based clinical workflows, ensuring closed-loop follow-up. This enables PCTs to use synchronous interactions to build rapport with patients and families, partnering on complex decisions and personalizing care plans.

Optimizing utilization requires attention to triage and routing of clinical scenarios to the appropriate clinician and care location. Synchronous visits should not be booked more than 4 weeks in advance (a variation of the “open access” model32), with some appointments held for low-lead time booking, to ensure urgent access. Rather than defaulting to a visit, patients should be responded to over chat, email, or telephone when a new health-related need arises, ideally within minutes. This rapid response positively reinforces the behavior of patients reaching out with health concerns rather than self-triaging. Registered nurses (RNs) can then route patients to the optimal clinician and setting—to a synchronous visit, an asynchronous chat, or a direct referral to the optimal site of care.

When RNs determine a presenting complaint should be managed internally, they choose among three options: (1) managing routine conditions through patient education about next steps and red flags for returning to care; (2) having patients schedule a visit with their PCT clinician; or (3) having patients schedule a visit with a clinician outside of their team if one of their PCT clinicians are not available. Table 2 shows one example of patient triage in a virtual-first model.

Table 2 Triage and Routing of Clinical Scenarios in a Virtual-First Primary Care Model
Full size table

ADVANCING PRIMARY CARE PAYMENT

Transitioning to outcomes-based care requires three changes in financing. Firstly, payments must be structured as per-member, per-month (PMPM) fees. We demonstrated that when more than two-thirds of practices’ patients are paid a PMPM fee, the practice will change from a visit- to an outcomes-based approach.33 Secondly, to ensure practices have incentive to care for complex patients, payments must be risk-adjusted.

Thirdly, payment levels should increase to cover the cost of advanced primary care and reward PCTs for the value they create by managing total medical expense (TME). For decades, PCTs have labored under an FFS incentive structure that generates costs outside of primary care for managing conditions that are efficiently managed within primary care. Examples of these costs include visits to specialists, utilization of urgent care, the emergency room and hospital, and health plan care management programs. When actuaries run an analysis to calculate a PMPM payment for PCTs, they quantify the expected spending on primary care for a population based on current practice patterns where complex cases are cared for in sub-optimal settings outside of primary care. The US primary care spend as a proportion of TME, viewed narrowly, is typically 5–6%, and only about 3.6% for patients with Medicare.34,35 Converted to a PMPM fee, typically in the range of $20–40 PMPM, this amount is insufficient to cover the resources needed for advanced primary care.

There has been progress in moving to enhanced bundled or PMPM payments for primary care. Nearly 20 states have initiatives in place to support increases in primary care spending.36 In Massachusetts, MassHealth (Medicaid) has received a waiver that will enable value-based payment for all MassHealth patients and increased support for practices providing advanced services.37,38 Rhode Island had a mandate to double primary care spending between 2010 and 2015 from 5–6% to 10–12% of TME. The increase in spending in primary care lowered overall costs during the first few years of the investment, largely due to a cap in cost increase for hospitals.39 Several states working on similar mandates have shown reductions in TME.40,41,42 Finally, some national systems have increased primary care spend (e.g., the Veterans Health Administration spends 9% of TME on PCMHs43).

TECHNOLOGY FOR OUTCOMES-BASED CARE

EHRs, built to support billing for visits rather than continuous, accountable care, do not support outcomes-based care. Patient relationship management systems (PRMs) are newer technology platforms that enable teams to communicate with one another and patients to continuously coordinate care and manage risk. PRMs typically sit on top of an EHR with integrations that allow data to flow between the two platforms.

In addition to chat threads, PRMs have features that support task management between patients and PCTs and forms that allow teams to automatically collect structured data via patient-facing screens. Unlike patient portals, which predominantly enable patients to access raw data (e.g., laboratory values) and send messages that are responded to after several days, if ever, the PRM is optimized for real-time interaction between patients and PCTs with closed loops for communication and care. With recent advances in artificial intelligence that enable real-time translation of verbal patient-clinician interactions into appropriate structured fields and clinical notes, when providers deliver care over PRMs, documentation can be produced as a byproduct of delivering care.

After decades of experience with a clinical software market dominated by technologies for transactional and hospital-based care, clinicians may be skeptical about the near-term availability of PRMs in their practice. Nonetheless, Epic announced the launch of a similar product in 2022. And we may be nearing a tipping point where frontline clinicians and patients have more agency in the software products they use.

IMPLICATIONS FOR WORKFORCE SATISFACTION AND EQUITY

Evolving towards the outcome-based model addresses the core drivers of clinician burnout. Studies have found that agency, co-management support, quality improvement activities, and reducing waste all help address burnout.44,45,46 When clinicians see real-time data on clinically important outcomes, they feel confident they are providing high quality care and/or can team with cross-disciplinary colleagues to make iterative changes to improve outcomes. Well-coordinated teamwork, schedules with adequate time for asynchronous work and supportive technology also make it possible for clinicians to complete tasks, including documentation, during work hours. Enhanced payments for appropriately structured roles make it possible to increase the financial compensation of PCTs, countering the payment disparities among specialties.

Outcomes-based approaches also hold promise for reducing health disparities. Tracking clinical outcomes enables teams to respond to gaps in care not only within populations, but also between populations. Assuming continued progress in broadband coverage, uncoupling a significant proportion of care from the need for in-person visits, removes barriers to access for people who live in rural areas and/or struggle to leave their homes. Individuals whose access is constrained by literacy, numeracy, language, or culture can be connected via technology with PCT members, including peers, positioned to provide them with effective assistance.

Finally, outcomes-based care holds potential beyond countering burnout and reducing disparities. Practicing in this way can drive joy in clinical practice. Uncoupling decision-making from the time pressure and interpersonal dynamics of synchronous visits, and alleviating administrative and documentation burden, creates the bandwidth for clinicians to think through complex diagnostic workups and treatment decisions in consultation with the literature and colleagues. PCTs can longitudinally partner on caring for patients with complexity with specialists of all types, including psychiatrists, without the need for external referrals. Patients can weigh difficult decisions in the comfort of their homes and communities, asking clarifying questions as needed over chat. Partnering across time, space, and clinical disciplines with patients and families to overcome health-related challenges—celebrating wins, big and small, and showing solidarity when challenges cannot be overcome—spark the moments of joy that reconnect us with why PCT members chose primary care in the first place.