From Health Care to Population Health: Retooling Legal Structures for a New Paradigm

In recent years, health policy thought leaders have urged refocusing our health care system from provision of episodic, reactive care to “population health” — a concept previously identified with the field of public health. As more Americans are burdened with chronic conditions, government officials and healthcare organizations increasingly realize that population health is as dependent on health care interventions as it is on traditional public health activities. The advent of universal coverage under the Patient Protection and Affordable Care Act1 (ACA) in 2014 offers an opportunity to diagnose and treat chronic conditions on a population-wide scale. But several likely barriers present challenges, including workforce shortages;2 health behaviors, medication adherence,3 and the fragmented nature of America's fee-for-service based health care structure.

Now, policymakers are exploring initiatives that would promote an increased population health orientation in health care. This article discusses changes to three areas of legal regulation of health care — drug prescribing, professional licensing, and accountability — that may be necessary to implement certain aspects of this vision. Specifically, it describes an alternative course that policymakers could chart — one that devolves certain responsibilities to lay health workers while encouraging physicians to assume new ones.

A convergence of events and prominent publications provide the impetus for this article:

• In 2008, Dr. Donald M. Berwick introduced his “Triple Aim” concept — improving the experience of care, reducing costs of care, and “improving the health of populations.”4 Upon his appointment as Administrator of the Centers for Medicare and Medicaid Services, these three aims were established as official Department of Health and Human Services (HHS) policy5, and he tasked the agency with thinking in terms of improving population health in administering its programs.

• In 2009, Kaiser Permanente researchers published the results of a study of its initiative to make bundled cardioprotective therapy (a statin for cholesterol and an ACE inhibitor for hypertension) available to all Kaiser Permanente members with diabetes over the age of 55 years and all members with coronary artery disease (CAD). Treatment with these inexpensive generic medications reduced the risk of hospitalization for myocardial infarction (MI) or stroke by 26 events per 1,000 person-years. The research team estimated that nationally implementing a simple treatment regimen of these medications, plus aspirin — a so-called “ALL” bundle — Aspirin, Lisinopril, and Lipid Lowering therapy — would result in 32,000 fewer MIs and strokes per year.6 

• In September, 2011, HHS, in conjunction with several private sector partners, launched Million Hearts, an initiative to prevent one million heart attacks and strokes over five years.7 The campaign targeted focus on what it called “ABCS” — Aspirin for people at risk, Blood pressure control, Cholesterol management and Smoking cessation — closely paralleling the Kaiser Permanente formula for addressing major risk factors for cardiovascular disease and preventing heart attacks and strokes.

• There has been an increasing focus on the use of lay health workers — a category that encompasses such job titles as community health workers, community health aides, patient navigators and patient coaches — to reach out to underserved and vulnerable populations. Denver Health, a safety-net health system, received grant funding to provide patient navigation for men — a population that is disproportionately uninsured — in its service area in 2002. “By the time the grant ended, the health system was convinced the program was worth paying for with internal funds,” having determined that it helped them avoid almost $300,000 in costs per year.8 

• A number of hospitals have established emergency department frequent users programs to address the needs of persons in their catchment areas who are regular users of their emergency rooms.9 Perhaps the most prominent of these is the University of Chicago Hospital's Urban Health Initiative, 10 which employs lay health workers that it calls Patient Advocates to educate patients and “connect them to clinics and primary care doctors so that they receive the right care at the right time and place.”11 For these hospitals, the obligation under the Emergency Medical Treatment and Active Labor Act12 (EMTALA) to provide emergency care to all comers has given rise to a de facto assumption of responsibility for the health of a regional sub-population, simply because the provision of primary care and case management to the hospital's most expensive “frequent flyers” is more cost effective than providing emergency room treatment to uninsured or underinsured patients.

• Dr. Atul Gawande, whose New Yorker articles tend to be influential with federal policymakers, in January 2011 published an article highlighting “hot spotters”13 — initiatives that direct intensive resources on patients “with the highest medical costs — the people cycling in and out of the hospital — [who are] usually the people receiving the worst care.” 13 

• A U.S. Food and Drug Administration (FDA) staff proposal published in 2012 requested public comment on changing the status of drugs for certain common medical conditions such as high cholesterol and high blood pressure from requiring a prescription to dispense to over-the-counter availability.14 The proposal, dubbed the “New Paradigm,” was premised on the possibility that these conditions were being undertreated owing to the “cost and time required to visit a health care practitioner.”14 

Three common threads run through these developments. First is a recognition that promoting population health ought to be an explicit aim of the health care system. Second is the insight that simple, if non-traditional, interventions may be able to provide great gains in health at a relatively low cost. Third is the realization that the existing, physician-centric legal regime may constitute a barrier to a greater population health focus.

With regard to this last thread, the FDA proposal would address the fact that of the three cardioprotective drugs common to the Kaiser Permanente ALL and HHS ABCS campaigns, only one, aspirin, is available without a prescription. The proposal also acknowledges that merely easing access to the drugs is not sufficient — patients would still need to be screened and given advice on whether and when to use the drugs, and monitored for proper dosing and side effects. The FDA suggested that a non-medical practitioner could assume this role.

The final legal question is what incentives would spur the greater assumption of responsibility for population health. This article begins and ends by briefly addressing this last point, but the primary focus will be on how the regulatory regime surrounding health care could be retooled so the system would proactively address chronic illness in the population.

We begin with a working definition of population health in the health care context offered by Donald Berwick and colleagues:

Berwick et al point to the need for an “integrator” to assume responsibility for all components of the Triple Aim for a specific population. They note that a Kaiser Permanente, with its integrated financing and exclusive relationships with delivery structures, is best able to accept such responsibilities, but argue that “other models can also take on a strong integrator role, even without unified financing or a single delivery system.” 4 Developing other such models that can operate within the current, fragmented delivery system will be important to achieving swift, incremental gains.

Synthesizing the seven developments recounted earlier, we can surmise that an integrator might want to increase utilization of inexpensive generic drugs to treat latent conditions, and deploy lay health workers to identify enrollees with latent conditions and initiate drug regimens, as well as promote increased patient engagement in management of so-called “super-utilizers.” We can also suggest that policymakers will want to encourage physician participation in integrative mechanisms that address population health.

One of the subjects of Gawande's “Hot-Spotters” article was Dr. Rushika Fernandopulle, director of the Special Care Center in Atlantic City, New Jersey. Speaking of the customer service focus needed to avoid complications in hot spots, Fernandopulle told Gawande, “We recruit for attitude and train for skill … This kind of care requires a very different mind-set from usual care. For example, what is the answer for a patient who walks up to the front desk with a question? The answer is ‘Yes.’ ‘Can I see a doctor?’ ‘Yes.’ ‘Can I get help making my ultrasound appointment?’ ‘Yes.’ Health care trains people to say no to patients.”13 

There are at least two problems with bringing Fernandopulle's “yes” to scale. One is that if the clinician workforce is inadequate, an affirmative answer to the question “Can I see a doctor?” may be out of reach. Another is, if the follow-up question is “Can somebody else resolve my problem?” a negative answer may be required by the legal regime. If the likely result of a “no” is a preventable complication, then restrictions on dispensing medications or giving basic advice on treatment of chronic conditions will undermine the efficacy of ACA's coverage expansion. For people who have been screened for and identified as having an asymptomatic condition, a negative or inconvenient answer to “Can I see a doctor?” may mean that treatment is never initiated. The next two sections of this article discuss how the legal regime might be adjusted to help facilitate Fernandopulle's “yes.”

In its notice of public hearing, the FDA observed that “Undertreatment of many common diseases or conditions in the United States is a well-recognized public health problem. Increasing the number of people who are able to obtain for the first time and those who continue on necessary drug therapy could provide improved health outcomes.”14 The FDA notice suggested that “the requirement to obtain a prescription for appropriate medication (and to make one or more visits to a practitioner) may contribute to undertreatment…”14 

The FDA paper floated a “New Paradigm” under which visits could be eliminated by making certain prescription medications available “without a prescription but with certain other conditions of safe use.” Benefits would include “improved health outcomes” and reducing the time that practitioners “expend on routine tasks related to prescription refills” so they would have more time to attend to more seriously ill patients.14 

The FDA paper posited two alternative means to prescriptions that could be employed to ensure safe use: “requiring pharmacist intervention to ensure appropriate nonprescription use” and “the use of innovative technologies… that consumers could use to self-screen for a particular disease or condition and determine whether a particular medication is appropriate for them. For example, kiosks or other technological aids in pharmacies or on the Internet could lead consumers through an algorithm for a particular drug product.”14 

With regard to the latter, the FDA paper suggested that the algorithm “could consist of a series of questions that help consumers properly self-diagnose certain medical conditions, or determine whether specific medication warnings contraindicate their use of a drug product.”14 

The FDA's “technological aid” option is very much reminiscent of the sort of decision support tools embedded in electronic health records, except that in this case, the tool would be used by the consumer rather than the physician. Certainly the potential benefits of placement of appropriate candidates on an ALL/ABCS regimen merit consideration of expediting the process. But is the use of a form of health information technology by consumers a wise and realistic means of achieving this goal?

The wisdom of removing prescription-only status of cholesterol modifying agents was questioned by Consumers Union in 2007, when Merck applied to the FDA for permission to sell an over-the-counter (OTC) version of the medication Mevacor. Consumers Union argued that “high cholesterol can't be self-diagnosed, many people with elevated cholesterol can bring it under control with diet and lifestyle changes, and statins are potent drugs that can have serious side effects.”15 

Consumers Union's arguments primarily addressed the question of a patient losing the benefits of the prescriber's advice and judgment that are conveyed during a consultation, but they also alluded to the marketing of drugs by manufacturers and the possibility of consumers being unduly influenced by advertising. When we try to imagine the circumstances that will spur the consumer to go to the kiosk or website envisioned by FDA, we must consider the likelihood that someone with an interest in selling drugs is encouraging their use.

Indeed, when the FDA held its hearing on the proposal in March 2012, the kiosk concept was presented by the manufacturer of Mevacor, which by then was GlaxoSmithKline (GSK). The lobbyist for the trade association representing manufacturers of over-the-counter drugs acknowledged that the kiosks would be “sponsor-driven”16 — which is to say, the manufacturer of a particular (and presumably branded) drug would be promoting its product. Moreover, the GSK presentation did not follow the FDA staff proposal's legal concept of a “new paradigm” premised upon an intermediate status between prescription-only and over-the-counter (“condition of safe use”), but rather tried to convince the expert panel that the deployment of a kiosk or other application using the algorithm, including smartphones, justified a switch to OTC status and for sales at any retail outlet, not just a pharmacy.16 Industry representatives offered no alternative vision to sponsor-driven technology, such as a mechanism more directly aimed at improving population health.

In the author's view, a kiosk in a drug store could not plausibly be used to guide consumers to a drug regimen on the model of ALL or ABCS. After all, even aspirin, with its ubiquity and long history of efficacy, has yet to be the subject of an interactive decision aid that any but the highest-risk consumers can use to self-screen for its use to prevent heart attacks and strokes. The American Heart Association 17and Mayo Clinic18 refer consumers to physicians. WebMD has a decision aid on its website, but first refers consumers to physicians and, after leading the user through questions and answers, leads the user only to an “overall impression” of “leaning toward taking a daily aspirin” or “leaning NOT toward taking a daily aspirin.”19 

Potential algorithms are complicated in that even the questions for aspirin presuppose some prior interaction with medical professionals, e.g.: in the statement “I am at risk for a heart attack or a stroke because I have diabetes, high cholesterol, high blood pressure, or a family history of heart disease or stroke (yes/no),” three of the four conditions named would likely have been diagnosed by at least one visit to a doctor, who presumably would have offered advice on the subject.

Gathering of facts for blood pressure or cholesterol therapy is typically longitudinal in nature. Since blood pressure readings vary, determining the existence of hypertension requires more than one reading at more than one time. Even a single test for cholesterol requires advance planning, since fasting is necessary for an accurate reading of LDL. The American Heart Association is guarded about the benefits of screenings that take place outside a physician practice, suggesting that they be limited to worksites or, in the case of “low-income, low-education level groups… not connected to traditional healthcare and information systems,” at schools, churches, community centers or neighborhood clinics. It notes that “Getting blood cholesterol and blood pressure measurements should be part of an overall plan of medical care. Physicians or para-professionals working at the direction of the physician should order the cholesterol tests and interpret the result for the patient. The physician should then direct any follow-up care if needed, such as prescribing a diet, exercise program or drug therapy.”20 

An example of an appropriate screening and medical referral program is one sponsored by the University of Texas Southwestern Medical Center that targeted African American men in Dallas for blood pressure (BP) checks at their barbershops. “The African American-owned barbershop holds particular appeal as a cultural institution that draws a loyal male clientele and provides an open forum for discussion of any number of topics—including health—with influential peers,” according to the project team.21 Moreover, the longitudinal nature of the customer's interaction, as with worksites, schools or churches, facilitates the ability to obtain multiple readings. For the Dallas project, barbers were “trained, equipped, and paid to conduct the intervention protocol [i.e.,] to offer a BP check with each haircut, to measure and interpret BP, to complete written encounter forms,” as well as discuss role model stories aimed at persuading male customers to adopt changes in health behavior.

The project team's evaluation concluded that an “enhanced intervention program—in which barbers continuously monitored BP and actively promoted physician follow-up with personalized sex-specific messages—resulted in improved BP control among black male barbershop patrons with HTN [hypertension;].”22 The evaluation found that blood pressure control even improved in the comparison group, which received only testing and “standard written information about high BP” without the role-model messaging, indicating that perhaps the “Hawthorne effect” of experiencing medical screening in a community setting surrounded by peers motivated follow-up or medication adherence.

Other concerns about the FDA's New Paradigm concept are that consumers might not find their way to the technological aid in the first instance, or stick with the regimen once initiated. If the FDA's vision is reminiscent of decision support tools, it is also reminiscent of the highly touted remote health monitoring technology. One evaluation of that innovation in a randomized clinical trial for chronic heart failure patients found “14% of our patients who were randomly assigned to undergo telemonitoring never used the system. Moreover, by the final week of the study period, only 55% of the patients were still using the system at least three times per week… Thus, the adherence rates in this trial most likely represent the best-case scenario and are in fact similar to previously documented rates of medication adherence.”23 

Shrank et al contend that increasing value in the area of prescription medications takes place along three dimensions: “(1) use of cost-effective medications, (2) timely initiation of appropriate medication therapy, and (3) adherence to that therapy.”3 The New Paradigm, as envisioned by FDA staff and pharmaceutical manufacturers, addresses but one of these dimensions.

The success of the barbershop program suggests an alternative formulation for the New Paradigm: If a decision support tool that can walk consumers through drug therapy is impracticable, but there is merit to the idea of making these drugs available without a prescription, would it be possible to train lay health workers to initiate drug therapy for some patients that they would otherwise only be helping to “navigate” the medical system?

The following discussion is informed by the “value-based competition” approach articulated by Harvard Business School scholar Michael E. Porter. In a sense, the rationale underlying the New Paradigm is Porter's argument that tasks be devolved to “appropriately skilled but lower-cost health care professionals” when feasible.24 Obviously, a physician consultation, culminating in issuance of a prescription and dispensing by a pharmacist, is the highest cost activity through which an ALL/ABCS drug regimen could be initiated. Making the drugs available after just a pharmacist consultation would still be costly, as pharmacist and pharmacy representatives noted at the FDA New Paradigm hearing. But imagine if the barbers or field interviewers involved in the UT-Southwestern trial could have initiated treatment rather making referrals to practitioners? If treatment could, as in Porter's formulation, be decentralized, and guided not by the intuition of experts but by laying out rules for non-experts,25 it might be possible to improve population health and also achieve increased economic welfare through initiating appropriate drug regimens for more individuals, lowering the cost of initiating each treatment, and eliminating the opportunity cost of a physician conducting routine activities at the expense of more complex evaluations when it is expected that universal coverage will lead to physician shortages.

Porter argues that we must analyze health care processes by measuring the costs of treating “patients with specific medical conditions over their full cycle of care.”24 In the case of cardiprotective medications, we know from the Kaiser Permanente study that treatment of eligible patients reduces the costs incurred by acute CAD events. The question becomes how to reduce costs in the cardioprotective medication regimen cycle. The FDA New Paradigm paper perhaps jumped to the conclusion that patient-initiated therapy would reduce costs. This seems unlikely in that consumers — guided presumably through direct-to-consumer advertising campaigns mounted by sponsors of the kiosks or other technology — might not always be those for whom the drugs are appropriate; would be directed to expensive brand drugs without first undergoing “step therapy” with inexpensive generics; and might not be followed to determine whether the therapy was effective, was being adhered to, or was causing side effects or dangerous drug interactions.

In contrast, a New Paradigm implemented by community health workers under the guidance of physicians could be longitudinal in nature, to ensure accurate screening, diagnosis, initiation of therapy and monitoring of therapy. Lay health workers could be proactive in reaching out to the underserved members of their community, identifying those that are likely to require medical attention, and assisting them in navigating the health care system. They could help absorb the overhang of illness among the currently uninsured, performing a triage function that could alleviate access problems owing to a practitioner shortage. They could also assume a “coaching” function that encourages adherence to medication and other healthy behaviors.

“Health coaching is a relatively new behavioral intervention that has gained popularity in health promotion, public health, and disease management because of the ability to address multiple behaviors, health risks, and self-management of illness in a cost-effective manner.”26 Health coaching often utilizes motivational interviewing to persuade patients to embrace behavioral changes necessary for improved health.

As such, heightened deployment of lay health workers could not only save money by promoting cardioprotective medication regimens, but add value above and beyond that, by motivating adherence to the regimens and to other healthy behaviors.

There are at least three obstacles to such a paradigm. The first is practical, and no small hurdle: formulating decision rules that a layperson can apply in recommending any part of an ALL/ABCS regimen. Credible researchers must overcome the scientific barrier of producing a valid and reliable instrument. This is not impossible; as Geersing et al note, diagnostic clinical decision rules have been developed for physicians to “guide their decision making and reduce unnecessary patient burden and costs,”27 citing rules for the diagnosis of deep venous thrombosis (DVT) and pulmonary embolism that have reduced further diagnostic testing from between 25% and 50% over physicians' judgment alone. But, they note various forms of resistance, including “a physician's disdain for ‘cookbook medicine’ or the unpractical use of the decision rule instrument itself in busy daily clinical practice.”27 

These considerations are all the more daunting in the context of guiding laypersons. But one must wonder whether the failure of the medical establishment to develop guidelines for the use of aspirin is truly driven by impracticability, or rather by an unwillingness to challenge the norm of superiority of clinical judgment. There appears to be an absence of any empirical evidence as to the safety of trained laypersons proffering basic advice on medications for chronic conditions. This suggests that efforts to devolve such responsibilities to lay health workers must begin with well-supervised experiments.

In a 2008 report, the Institute of Medicine (IOM) released recommendations from a Committee on the Future Health Care Workforce for Older Americans. The impetus for its study was projected needs resulting from the aging of the baby boom cohort, but its work was equally, and perhaps more immediately, applicable to the needs expected from the 2014 expansion. A key element of the IOM recommendations was an increased role for “trained laypersons”28 in managing chronic conditions. The IOM concluded that “Cascading various patient-care responsibilities — shifting them from one type of worker to another — will be an essential tool in efforts to alleviate the projected shortages in the numbers of providers available to meet the care needs of an aging society.”28 (The reference in the title of this article to “retooling” alludes to the name of this report.) The IOM noted that “In some cases, the assumption of these new responsibilities will require regulatory changes, most often through the expansion of state-based definitions of scopes of work.” The report issued a recommendation that “Health care disciplines, state regulators, and employers should look to expand the roles of individuals who care for older adults with complex clinical needs at different levels of the health care system beyond the traditional scope of practice.”

As the IOM noted, a further obstacle to a new paradigm engaging lay health workers in chronic care is the legal regime: in addition to federal laws that restrict prescribing to eligible clinicians, there are state laws governing practitioner licensure. While the FDA could loosen the prescription-only status of relevant medications, the recommendation of the medications would fall within licensed practitioners' scope of practice under state laws. Currently, community health workers “do not provide clinical care or replace other health care providers. Instead, they complement services delivered through the more formal health care network ‘to provide more comprehensive and supportive care.’”29 

The National Conference of State Legislatures (NCSL) notes that “As the role of community health workers grows, other health professions may question these increasing roles and responsibilities, particularly if they perceive that the workforce is encroaching on their scope of practice. Some states place health workers under the authority of their nursing practice act and require oversight by a registered nurse.”29 But, says NCSL, “States also may consider developing a standard scope of practice for community health workers.”29 

Creating a scope of practice is politically complex because it requires carving out, and awarding to a new group, activities for which higher-prestige professions enjoy protection from competition. It is scientifically complex because it involves an assessment of the safety of performance of a medical function by a person with less training than the professionals performing those activities heretofore.

A monograph developed through a collaborative effort by representatives of six organizations of health professions regulators—National Council of State Nursing Boards, Federation of State Medical Boards, Association of Social Work Boards, Federation of State Boards of Physical Therapy, National Association of Boards of Pharmacy, and the National Board for Certification in Occupational Therapy — provides a basic framework for objectively analyzing proposed changes in statutory scope of practice. The regulators' purpose was to articulate neutral principles that legislators could use to decide the fiercely contested and often emotional disputes that they are often called upon to referee. According to their consensus statement, arguments for scope of practice changes “should have a foundational basis within four areas: 1) an established history of the practice scope within the profession, 2) education and training, 3) supportive evidence, and 4) appropriate regulatory environment.”30 

Obviously there is no established history of lay health workers in providing advice on initiating a regimen of cardioprotective medications. According to the National Community Health Advisor Study, their “core roles” are mostly non-clinical and include activities such as “providing cultural mediation between communities and health and human services systems” and “ensuring that people obtain necessary services.”31 

Only the last of seven enumerated roles, “providing basic screening services,” provides a clinical springboard from which a New Paradigm approach would proceed, albeit in a broader scope somewhat attenuated from these roots. As the National Community Health Advisor Study found, “in the U.S., the CHA role in providing clinical services is minimal compared to CHA roles in the developing world. Yet, especially in remote areas, CHAs in the U.S. do provide needed basic services, thus making them accessible.”31 The Bureau of Labor Statistics' standard occupational classifications list as among Community Health Workers' responsibilities “provid[ing] services such as first aid and blood pressure screening.”32 Typically, blood pressure readings are taken by medical assistants — medical office employees whose duties are primarily clerical33 — and it would not seem unreasonable to make blood pressure screening a defined responsibility of CHWs.

Minnesota has taken an unusual approach to legal recognition of a role for lay health workers. The Minnesota Community Health Worker Alliance, described as a “broad-based, statewide consortium of community health workers, public agencies, and nonprofits,”34 promulgated what it calls a “scope of practice” for community health workers. Ordinarily, in a legal sense, “scope of practice” implies licensure, and, in the context of allied health professions, permission for a claimant occupation to perform duties hitherto reserved to one or more senior licensed occupations, beginning with physicians. As NCSBN explains, “Physicians were the first health professionals to obtain legislative recognition and protection of their practice authority. The practice of medicine was defined in broad and undifferentiated terms to include all aspects of individuals' care. Therefore, when other healthcare professions sought legislative recognition, they were seen as claiming the ability to do tasks which were already included in the universal and implicitly exclusive authority of medicine.”30 

But the “scope of practice” delineated by Minnesota's alliance does not appear to lay any claim of permission to perform health services assigned to any licensed occupation.35 Instead, a Minnesota statute confers authority for that state's Medicaid program to cover “the care coordination and patient education services provided by a community health worker” when the worker meets certification, experience, or supervision requirements36 that mimic those often found in allied health professions' licensure statutes.

Discussions of delineating a community health worker “scope of practice” in New York also appeared to proceed from a presumption that no legal claim was to be made on activities assigned to other professions, nor for an exclusive privilege to perform its unique services.37 This approach — building consensus for pronouncements on appropriate services and gaining public program coverage and payment, while declining to seek licensure — avoids the acrimony and political controversy that surround scope of practice claims in the licensing context. It helps to develop a collaborative political environment that complements the collaborative clinical environment in which lay health workers are meant to function. However, without legal recognition connected to specific under-delivered clinical services, the full potential of CHWs as clinician extenders cannot be fulfilled.

A possible template for credentialing lay health workers with expanded responsibilities might be the list of qualifications for a “patient coach” at HealthCare Partners Medical Group (HCP). HCP employs a Care Management Team that includes patient coaches that are analogous to community health workers but who appear to have duties in a grey area between the strictly advisory and the clinical. The team includes a physician, care manager, and social worker intended to “leverage” the primary care practitioner to “improve efficiencies” and “deliver patient-focused care.”38 

Patient coach functions at HCP include patient care coordination, assisting with screening and managing using protocols, and confirming a patient's understanding of disease process or treatment by reviewing information that was shared with the patient. They also conduct motivational interviewing.39 

Patient-coach job requirements at HCP include some post-high school education, e.g., a degree from a two-year college and Medical Assistant, Phlebotomy, and first-aid credentials.

NCSBN argues that in creating new scope of practice:

The foregoing suggests that it would be premature to delineate a permanent scope of practice for lay health workers to dispense advice on medication. But it does not rule out a role for “safe harbors” from prosecution for unauthorized practice of medicine during supervised trials, intended to generate evidence, conducted in collaboration with licensed professionals. The FDA could set the stage for such activities by loosening the prescription-only status of appropriate medications so that they may be dispensed by trained lay health workers participating in collaborative arrangements.

Care management teams within a large group practice such as HCP would be relatively easy to accommodate within such an arrangement. But most physicians are not lodged in large group practices. One model for sharing support infrastructure among multiple small physician practices is the State of Arkansas' Patient Centered Medical Home (PCMH) project, in which the state government has organized Care Coordination Support Vendors. The PCMH is intended to coordinate care for patients who have previously been identified as having chronic conditions. This has considerable value, but has the disadvantage of acting off patient registries based upon previous diagnoses—it cannot proactively screen a large population and initiate first-line interventions for latent conditions. The broad-based support infrastructure envisioned here would incorporate such proactive measures.

Let us designate such a support arrangement a Community Health Worker Care Organization, or CHWCO. The CHWCO could be a unit of a health plan, or perhaps an organization convened by a pharmaceutical benefit manager or community- or faith-based group under contract to and with sponsorship by one or more health plans. They could be a natural outgrowth of a hospital patient navigator program or even of outreach assisters employed by insurers to enroll members newly eligible for Medicaid in expansion populations. The purpose of the CHWCO would be to serve populations that cannot be enrolled with a provider like HealthCare Partners Medical Group, which is a superior model for integrating such support services.

The FDA could authorize certified paraprofessionals employed by a CHWCO to initiate drug regimens for common chronic conditions such as hypertension. A prerequisite to this would be development of decision rules to guide lay health workers, which could perhaps be undertaken by a consortium led by the medical directors of health plans and disease associations, with assistance from academicians and officials of relevant professional societies. The CHWCO would be required to maintain a collaborative practice agreement with one or more physicians and perhaps other NPPs as well. At the state level, the CHWCO, its sponsor, and supervising professional(s) could apply for a safe harbor agreement with relevant licensing agencies. The safe harbor agreement would set forth the treatment protocols, the collaborative practice agreements, financial arrangements with the sponsoring organization such as shared savings or per-member per-month fees, and an agreement with a principal investigator to evaluate the CHWCO. At the close of the trial period, evaluations would form the evidentiary basis for delineating a legal scope of practice to the extent, of course, that the trial resulted in successful outcomes.

The third barrier to a new paradigm guided by community health workers is the ability of those workers to make referrals to physicians when first-line therapies prove ineffective. While the regime proposed here envisions supervising professionals monitoring the work of CHWs, those professionals probably would lack the time to visit all patients within the population their CHWCO serves who require follow-up. In an era in which physicians limit their encounters with Medicaid patients,40 and in some cases, limit their encounters to patients who have paid an additional annual “concierge care” fee,41 it may be necessary to bind physicians to pools that are held accountable, to some degree, for population health and the total cost of care for an imputed population.

Just as fee-for-service health care is reactive and episodic, so too is accountability for outcomes. We usually think of physician accountability in terms of tort liability or professional regulation, which focus on individual incidents in the care setting, not on the longitudinal care experienced by patients. A patient dying due to a misdiagnosis in an acute care setting may give rise to a malpractice suit or a disciplinary inquiry, but ordinarily no entity is held accountable for not identifying or initiating treatment for the latent chronic condition that gave rise to the acute event.

As Berwick et al note, Kaiser Permanente would be in a position to take overall responsibility for the patient's health, reaching out to him for screening, and providing the drug regimen, as well as evaluating him during an acute episode. But physicians in fee-for-service medicine assume responsibility only for the patients who make appointments. They are not bound together, as are Permanente Medical Group physicians, into a pool that rewards them for preventing acute episodes.

The nature of a policy intervention that would incentivize greater physician involvement in ensuring population health is beyond the scope of this article, which focuses on regulation. But it is worth mentioning that the accountable care organization (ACO), a “virtual” integrated delivery system that attempts to duplicate the financial incentives inherent in a staff model HMO such as Kaiser Permanente without requiring physicians to be staff members of a corporate entity,42 may form the basis for future efforts in this direction. As originally proposed by Fisher et al, the ACO was to be a mandatory payment pool created by devolving Medicare's nationwide sustainable growth rate (SGR) target to the local level.42 

The idea of mandatory membership in ACOs for doctors participating in Medicare never took hold, presumably because it encroached on physicians' autonomy. Instead, ACA established a mechanism for creating voluntary ACOs for Medicare.43 In Massachusetts, a special commission44 and, thereafter, legislators, conclud ed that adoption of ACOs statewide, coupled with “global payment” targets, were necessary to control costs and incentivize care coordination. As was the case at the federal level, Massachusetts declined to mandate physician membership in an ACO. Nevertheless, as health care costs continue to rise, policymakers may take further steps in this direction.

The Medicaid coverage expansion under ACA represents an opportunity arising amidst various looming crises, which include health care costs “on an unsustainable financial path,”45 discouraging population health indicators, and an expected lack of access to care for the newly insured. As the vignettes at the beginning of this article illustrate, providers, regulators, insurers, and health policy thought leaders have been creative and proactive in suggesting, and in some cases testing, possible solutions to these problems. The FDA and Institute of Medicine have, among others, correctly in the author's view, identified the legal regime surrounding healthcare as a barrier to innovation in addressing these problems. Policymakers should consider regulatory relief options that permit flexibility in designing new systems to deliver care oriented to population health.