SPECIAL COMMUNICATION

How Evolving United States Payment Models Influence Primary Care and Its Impact on the Quadruple Aim Brian Park, MD, MPH, Stephanie B. Gold, MD, Andrew Bazemore, MD, MPH, and Winston Liaw, MD, MPH

Introduction: Prior research has demonstrated the associations between a strong primary care founda- tion with improved Quadruple Aim outcomes. The prevailing fee-for-service payment system in the United States reinforces the volume of services over value-based care, thereby devaluing primary care, and obstructing the health care system from attaining the Quadruple Aim. By supporting a shift from volume-based to value-based payment models, the Medicare Access and Children’s Health Insurance Program Reauthorization Act may help fortify the role of primary care. This narrative review proposes a taxonomy of the major health care payment models, reviewing their ability to uphold the functions of primary care, and their impacts across the Quadruple Aim.

Methods: An Ovid MEDLINE search and expert opinion from members of the Family Medicine for America’s Health payment and research tactic teams were used. Titles and abstracts were reviewed for relevance to the topic, and expert opinion further narrowed the literature for inclusion to timely and relevant articles.

Findings: No payment model demonstrates consistent benefits across the Quadruple Aim across a limited evidence base. Several cross-cutting lessons from available payment models several recommen- dations for primary care payment models, including the following: implementing per member per month– based models, validating risk-adjustment tools, increasing investments in integrated behavioral health and social services, and connecting payments to patient-oriented and primary care-oriented met- rics. Along with ongoing research in emerging payment models, data systems integrated across health care and social services settings using metrics that can capture the ideal functions of primary care will be critical to the development of future payment models that most optimally enhance the role of pri- mary care in the United States.

Conclusions: Although the ideal payment model for primary care remains to be determined, lessons learned from existing payment models can help guide the shift from volume-based to value-based care. To most effectively pay for primary care, future payment models should invest in a primary care infra- structure, one that supports team-based, community-oriented care, and measures the delivery of the functions of primary care. ( J Am Board Fam Med 2018;31:588 – 604.)

Keywords: Delivery of Health Care, Family Medicine, Health Expenditures, Primary Health Care

Forty years ago, in the milestone “Declaration of Alma Ata,” all member nations of the World Health Organization declared that achieving health for all was dependent on a foundation of primary

care.1 A quarter century later, Dr. Barbara Starfield added to the evidence base, demonstrating that primary care produces higher quality of care, im-

This article was externally peer reviewed. Submitted 26 September 2017; revised 11 March 2018;

accepted 13 March 2018. From the Department of Family Medicine, Oregon

Health & Science University, Portland, OR (BP); Eugene S. Farley, Jr. Health Policy Center, University of Colorado School of Medicine, Denver, CO (SBG); Robert Graham

Center for Policy Studies in Family Medicine and Primary Care, Washington, D.C. (AB, WL).

Funding: none. Conflict of interest: none declared. Corresponding author: Brian Park, MD MPH, Department

of Family Medicine, Oregon Health & Science University, 3181 SW Sam Jackson Pk Rd, Mailcode FM, Portland, OR 97239 �E-mail: parbr@ohsu.edu).

588 JABFM July–August 2018 Vol. 31 No. 4 http://www.jabfm.org

o n 3

1 M

a y 2

0 2 1

b y g

u e st. P

ro te

cte d b

y co p yrig

h t.

h ttp

://w w

w .ja

b fm

.o rg

/ J A

m B

o a

rd F

a m

M e

d : first p

u b

lish e

d a

s 1 0

.3 1

2 2

/ja b

fm .2

0 1

8 .0

4 .1

7 0 3 8 8 o

n 9

Ju ly 2

0 1 8 . D

o w

n lo

a d e d fro

m

 

 

proves health outcomes, increases access, lowers costs, and attenuates disparities.2,3,4 She attributed the positive impact of primary care on health sys- tems to the “4 Cs,” which define its function: first contact, continuity, comprehensiveness, and coor- dination (Figure 1).4 Subsequent research has dem- onstrated that supporting these 4 Cs are the ele- ments of primary care that help health systems achieve the Quadruple Aim of improving patients’ experience of care, population health, and physi- cian satisfaction, while reducing costs.5,6,7,8

Starfield’s work and the healthcare system’s longstanding inattention to primary care may ex- plain the ongoing failure of the United States to achieve its Quadruple Aims, given the inadequate system level support for primary care.9,10,11,12,13,14

Its predominant fee-for-service (FFS) payment model has long been thought to undermine or insufficiently support the 4 Cs that explain primary care’s positive effects.15,16,17 Under pure FFS pay- ment models, clinicians are reimbursed retroac- tively for services, incentivizing higher volume, treatment rather than prevention, and fragmenta- tion of care without regard for quality or cost. Such models reward greater numbers of services ren- dered (ie, volume) rather than the quality and cost of care provided to patients (ie, value).18,19

Payers, public and private, are experimenting with shifting from paying for volume to paying for

value. The Affordable Care Act included provisions that advance primary care and value-based pay- ment, including the creation of the Center for Medicare and Medicaid Innovation (CMMI), which has tested innovative payment and delivery system models aimed at improving value.20,21,22 Five years after the Affordable Care Act, the Medicare Access and Children’s Health Insurance Program CHIP Re- authorization Act (MACRA) passed. Under MACRA, providers1 will select 1 of 2 incentive tracks: the al- ternative payment model (APM; see Table 1) or the Merit-Based Incentive Payment System (see Table 2).23 Both programs provide incentives for improving quality and reducing costs.

As value-based payment spreads, better under- standing of existing models can guide which ap- proaches deserve ongoing implementation and re- search efforts. This narrative review of the literature proposes a taxonomy of the major health care pay- ment models, highlights their distinguishing charac- teristics (Table 3), and reviews their impacts across the Quadruple Aim (Table 4). We also discuss the impact of each payment model in supporting the 4 Cs of primary care; given the lack of widespread use and standardized metrics in measuring these pri-

1Eligible clinicians provide care for at least 100 Medicare patients and bill for greater than $30,000 of Medicare Part B services.

Table 1. Scheduled Adjustments in APM Eligibility Criteria under Medicare Access and Children’s Health Insurance Program Reauthorization Act

Year Eligibility

2019 and 2020 �25% of total Medicare revenue is from a qualified, eligible APM 2021 and 2022 �50% of total Medicare revenue OR

�25% of total Medicare revenue and 50% of all-payer revenue (eg, Medicaid, private insurers) is from a qualified, eligible APM

2023 and beyond �75% of total Medicare revenue OR �25% of total Medicare revenue and 75% of all-payer revenue is from a qualified, eligible APM

APM, alternative payment model; OR, odd ratio.

Figure 1. The 4 Cs of Primary Care.

• Contact: Accessibility as the first contact with the health care system • Comprehensiveness: Accountability for addressing a vast majority of personal health

care needs, • Coordination: Coordination of care across settings, and integration of care for acute

and (often comorbid) chronic illnesses, mental health, and prevention, guiding access to more narrowly focused care when needed,

• Continuity: Sustained partnership and personal relationships over time with patients known in the context of family and community.

doi: 10.3122/jabfm.2018.04.170388 U.S. Payment Models’ Impact on the Quadruple Aim 589

o n 3

1 M

a y 2

0 2 1

b y g

u e st. P

ro te

cte d b

y co p yrig

h t.

h ttp

://w w

w .ja

b fm

.o rg

/ J A

m B

o a

rd F

a m

M e

d : first p

u b

lish e

d a

s 1 0

.3 1

2 2

/ja b

fm .2

0 1

8 .0

4 .1

7 0 3 8 8 o

n 9

Ju ly 2

0 1 8 . D

o w

n lo

a d e d fro

m

 

 

mary care attributes24, when relevant, we consider the hypothetical impacts of each model when for- mal metrics were not used. Based on these findings, we provide policy and research recommendations for payment reform to best advance primary care.

Methods Starfield Summit I: Advancing Primary Care Research, Policy, and Patient Care The first iteration of this narrative review was con- ducted before the inaugural Starfield Summit (http://www.starfieldsummit.com) on April 24 to 26, 2016, in Washington, D.C. It was intended to inform and capture informant input from the Sum- mit’s nearly 150 invited primary care leaders (PCPs), researchers, and health care leaders to dis- cuss and enable research and policy agenda-setting around primary care payment, measurement, and teams.25

Literature Review We first conducted a literature search26 on primary care payment, enriched through expert consulta- tion before, during, and after the Summit. In March 2016, an Ovid MEDLINE search was con- ducted using the search terms “payment” and “pri- mary care.” The search was limited to articles pub- lished in English since 2010, yielding a total of 391 results2, with 97 articles ultimately included in the review. Exclusion criteria included the following: inclusion in a subsequent systematic review, up- dated evidence available (ie, more recent article from the same demonstration), not focused on pay- ment models, not focused on Quadruple Aim and/or the 4 Cs, and non-US evaluations that were subnational. Additional articles and gray literature were identified from the expert opinions of mem- bers of the Family Medicine for America’s Health payment and research tactic teams and a “snowball”

method of reviewing the references of the search results. The literature was summarized for each model, and key demonstrations or projects were selected, with agreement from at least 2 authors from the writing group, to highlight examples.

Results Fee-For-Service Under FFS, a provider is retrospectively paid a predefined amount for each service. Consequently, providers are incentivized to increase volume with- out bearing financial risk for quality or costs; in- surers bear high financial risk in this arrangement. In 1992, the Centers for Medicare and Medicaid Services (CMS) began using the Resource-Based Relative Value Scale to set a fee schedule for dif- ferent services, which has been criticized for dis- proportionately weighing specialist care and proce- dures over primary care.27,28 Despite concerns over the limitations of FFS, its inclusion in a payment model may enhance the use of services that are low-cost and underutilized29, such as vaccines in low immunization areas, where increased volume is desirable for population health.

Traditional (Or Full-Risk) Capitation In response to rising costs from FFS, health main- tenance organizations (HMOs)3 emerged in the 1980s to coordinate care and reduce use30 by capi- tating payments.26 In traditional capitation, provid- ers are paid a prospective amount to cover all ser- vices within a specific period of time, most often as a per member per month (PMPM) fee. Payments vary by age-group and sex and are determined based on prior average costs of care under FFS.31,32

A capitated fee can cover all primary care services, all outpatient services, or all health care services,

2In the case that a more recent report on a demonstration project was published between the time of the initial litera- ture search and submission of this manuscript, we replaced the prior report with the most up-to-date evidence.

3HMOs and other managed care models also include other mechanisms for cost control (e.g., narrow provider networks and pre-authorization of services). For the pur- poses of this paper, we have examined this model as a surrogate for capitated payment, though we acknowledge other mechanisms were in place to contribute to outcomes.

Table 2. Scheduled Payment Adjustments in Merit-Based Incentive Payment System

Adjustment 2019 2020 2021 2022 and beyond

Baseline payment adjustment �4% �5% �7% �9% Maximum payment adjustment for high performers �12% �15% �21% �27%

590 JABFM July–August 2018 Vol. 31 No. 4 http://www.jabfm.org

o n 3

1 M

a y 2

0 2 1

b y g

u e st. P

ro te

cte d b

y co p yrig

h t.

h ttp

://w w

w .ja

b fm

.o rg

/ J A

m B

o a

rd F

a m

M e

d : first p

u b

lish e

d a

s 1 0

.3 1

2 2

/ja b

fm .2

0 1

8 .0

4 .1

7 0 3 8 8 o

n 9

Ju ly 2

0 1 8 . D

o w

n lo

a d e d fro

m

 

 

Ta bl

e 3.

O ve

rv ie

w of

Pr im

ar y

Ca re

Pa ym

en t

M od

el s

D es

cr ip

ti on

P ro

sp ec

ti ve

vs re

tr os

pe ct

iv e

F in

an ci

al ly

di sc

ou ra

ge s

vo lu

m e

of se

rv ic

es ?

F in

an ci

al ly

en co

ur ag

es hi

gh qu

al it

y of

ca re

?

P ar

ty th

at pr

im ar

ily be

ar s

th e

fi na

nc ia

l ri

sk ?

R is

k ad

ju st

s fo

r pa

ti en

t co

m pl

ex it

y? K

ey E

xa m

pl e

F ee

-f or

-s er

vi ce

(F F

S) P

ai d

fo r

ea ch

in di

vi du

al se

rv ic

e re

nd er

ed R

et ro

sp ec

ti ve

N o

N o

In su

re rs

N o

M ed

ic ar

e P

at ie

nt s

(v ia

co st

-s ha

ri ng

: co

– pa

ys ,

de du

ct ib

le s)

T ra

di ti

on al

ca pi

ta ti

on (f

ul l-

ri sk

ca pi

ta ti

on ,

gl ob

al pa

ym en

t)

P ai

d to

co ve

r al

l se

rv ic

es w

it hi

n a

sp ec

ifi c

pe ri

od of

ti m

e

P ro

sp ec

ti ve

Y es

N o,

ex ce

pt fo

r ou

tc om

es re

la te

d to

us e

P ri

m ar

y ca

re pr

ac ti

ce s

N o

M ed

ic ar

e A

dv an

ta ge

H M

O s

P ay

-f or

-p er

fo rm

an ce

(P 4P

) ex

is ts

in ad

di ti

on to

un de

rl yi

ng m

od el

(g en

er al

ly F

F S

or ca

pi ta

ti on

)

P ai

d fo

r ac

hi ev

em en

t of

(o r

im pr

ov em

en t

in )

a qu

al it

y m

ea su

re

B ot

h ex

is t

(m os

t m

od el

s re

tr os

pe ct

iv el

y; ho

w ev

er ,

ca n

be pa

id pr

os pe

ct iv

el y

an d

su bs

eq ue

nt ly

re co

nc ile

d)

P ot

en ti

al ly

(d ep

en ds

on qu

al it

y m

et ri

cs )

Y es

, fo

r se

rv ic

es be

in g

m ea

su re

d vi

a qu

al it

y m

et ri

c

D ep

en ds

on un

de rl

yi ng

pa ym

en t

m od

el

P ot

en ti

al ly

M ed

ic ar

e P

hy si

ci an

G ro

up P

ra ct

ic e

D em

on st

ra ti

on P

ro je

ct P

ri m

ar y

ca re

pr ac

ti ce

s, if

ta rg

et s

no t

m et

B un

dl ed

pa ym

en t

(e pi

so de

– of

-c ar

e) P

ai d

fo r

al l

se rv

ic es

re nd

er ed

fo r

a gi

ve n

ep is

od e

of ca

re

M ix

ed (g

en er

al ly

re tr

os pe

ct iv

el y

tr ig

ge re

d an

d pr

os pe

ct iv

el y

pa id

)

Y es

(b ut

do es

no t

di sc

ou ra

ge vo

lu m

e of

ep is

od es

)

N o,

ex ce

pt fo

r ou

tc om

es re

la te

d to

ut ili

za ti

on

P ri

m ar

y ca

re pr

ac ti

ce s,

or ga

ni za

ti on

s

N o

C M

M I’

s B

un dl

ed P

ay m

en ts

fo r

C ar

e Im

pr ov

em en

t

Sh ar

ed sa

vi ng

s P

ai d

ba se

d on

sp en

di ng

be lo

w a

pr ed

et er

m in

ed be

nc hm

ar k

ov er

a pe

ri od

of ti

m e

(c on

ti ng

en t

on m

ee ti

ng ce

rt ai

n qu

al it

y ta

rg et

s)

M ix

ed (p

ro sp

ec ti

ve at

le ve

l of

th e

A C

O ,

bu t

pr ov

id er

s of

te n

st ill

pa id

vi a

F F

S)

Y es

Y es

A C

O s

P ot

en ti

al ly

M ed

ic ar

e Sh

ar ed

Sa vi

ng s

P ro

gr am

A C

O s

B le

nd ed

F F

S an

d ca

pi ta

ti on

P ai

d a

pr ed

et er

m in

ed am

ou nt

in te

nd ed

to co

ve r

m ed

ic al

ho m

e se

rv ic

es fo

r a

sp ec

ifi c

pe ri

od of

ti m

e in

ad di

ti on

to F

F S

M ix

ed N

o (t

o th

e ex

te nt

th at

F F

S is

th e

pr ed

om in

an t

pa ym

en t

m ec

ha ni

sm )

N o

D ep

en ds

on un

de rl

yi ng

pa ym

en t

m od

el

P ot

en ti

al ly

M ed

ic ar

e C

om pr

eh en

si ve

P ri

m ar

y C

ar e

In it

ia ti

ve

C om

pr eh

en si

ve (p

ri m

ar y)

ca re

pa ym

en t

P ai

d a

ri sk

-a dj

us te

d am

ou nt

to co

ve r

al l

pr im

ar y

ca re

se rv

ic es

fo r

a sp

ec ifi

c pe

ri od

of ti

m e;

in cl

ud es

co m

po ne

nt of

P 4P

P ro

sp ec

ti ve

Y es

Y es

P ri

m ar

y ca

re pr

ac ti

ce s

Y es

Io ra

H ea

lt h

C on

ti nu

ed

doi: 10.3122/jabfm.2018.04.170388 U.S. Payment Models’ Impact on the Quadruple Aim 591

o n 3

1 M

a y 2

0 2 1

b y g

u e st. P

ro te

cte d b

y co p yrig

h t.

h ttp

://w w

w .ja

b fm

.o rg

/ J A

m B

o a

rd F

a m

M e

d : first p

u b

lish e

d a

s 1 0

.3 1

2 2

/ja b

fm .2

0 1

8 .0

4 .1

7 0 3 8 8 o

n 9

Ju ly 2

0 1 8 . D

o w

n lo

a d e d fro

m

 

 

including inpatient and outpatient. In contrast to FFS, capitation incentivizes cost control. Capita- tion may also exist as part of blended models with mixed PMPM payments and FFS, or in a further risk-adjusted form mixed with pay-for-perfor- mance in comprehensive primary care payment; these models are discussed in a later section. In contrast to FFS, capitation shifts financial risk to the provider, while the payer has lower risk.

One study examined the impact of capitation on one of the 4 Cs and finding capitated models was associated with decreased first contact (access).33

This may reflect the incentive for providers to avoid sicker patients (termed adverse selection or “cherry-picking”) to reduce costs. Another possible negative impact on the 4 Cs is a financial incentive to inappropriately underdeliver services, leading to decreased comprehensiveness.34 The prospective element of capitation could benefit primary care by enabling upfront investments in practice compo- nents that enhance the 4 Cs (eg, care coordination) and providing flexibility for practices to determine how finances are spent.

Traditional capitation has demonstrated mixed effects on cost and quality35,36,37, although most evidence suggests a decreased use of hospitals and other expensive resources and worse patient satis- faction, consistent with the backlash toward HMOs in the 1990s.38

Pay-For-Performance (P4P) P4P supplements an underlying payment model, most often as a bonus on top of FFS. P4P refers to payment based on the achievement of a quality target (eg, hemoglobin A1c [HbA1c] level �8 for diabetic patients or delivery of cancer screening) or improvement in performance (eg, change from baseline for HbA1c); the latter approach may at- tenuate variation in quality across providers, and provide incentives for both high-performing and low-performing practices.39

Limited evidence exists for the impact of P4P on the 4 Cs. The United Kingdom’s Quality and Out- comes Framework (QOF) found decreased conti- nuity rates and no differences in patient-reported perception of coordination, when compared with preintervention periods.40 Incentivized metrics tended to improve, whereas nonincentivized met- rics demonstrated unchanged or worsened rates of improvement; a limited set of targeted metrics could thus inhibit the comprehensive function ofTa

bl e

3. Co

nt in

ue d

D es

cr ip

ti on

P ro

sp ec

ti ve

vs re

tr os

pe ct

iv e

F in

an ci

al ly

di sc

ou ra

ge s

vo lu

m e

of se

rv ic

es ?

F in

an ci

al ly

en co

ur ag

es hi

gh qu

al it

y of

ca re

?

P ar

ty th

at pr

im ar

ily be

ar s

th e

fi na

nc ia

l ri

sk ?

R is

k ad

ju st

s fo

r pa

ti en

t co

m pl

ex it

y? K

ey E

xa m

pl e

D ir

ec t

pr im

ar y

ca re

P ai

d ou

ts id

e of

th ir

d- pa

rt y

in su

re rs

(o ft

en di

re ct

ly fr

om pa

ti en

ts )

a pr

ed et

er m

in ed

am ou

nt to

co ve

r al

l pr

im ar

y ca

re se

rv ic

es fo

r a

sp ec

ifi c

pe ri

od of

ti m

e

P ro

sp ec

ti ve

Y es

N o

P ri

m ar

y ca

re pr

ac ti

ce s

fo r

pr im

ar y

ca re

ex pe

ns es

N o

Q lia

nc e

P at

ie nt

s fo

r ot

he r

as pe

ct s

of ca

re (a

nd in

su re

rs if

pa ti

en ts

ha ve

th ir

d pa

rt y

in su

ra nc

e)

A C

O ,

ac co

un ta

bl e

ca re

or ga

ni za

ti on

; H

M O

, he

al th

m ai

nt en

an ce

or ga

ni za

ti on

; C

M M

I, C

en te

r fo

r M

ed ic

ar e

an d

M ed

ic ai

d In

no va

ti on

.

592 JABFM July–August 2018 Vol. 31 No. 4 http://www.jabfm.org

o n 3

1 M

a y 2

0 2 1

b y g

u e st. P

ro te

cte d b

y co p yrig

h t.

h ttp

://w w

w .ja

b fm

.o rg

/ J A

m B

o a

rd F

a m

M e

d : first p

u b

lish e

d a

s 1 0

.3 1

2 2

/ja b

fm .2

0 1

8 .0

4 .1

7 0 3 8 8 o

n 9

Ju ly 2

0 1 8 . D

o w

n lo

a d e d fro

m

 

 

Ta bl

e 4.

Im pa

ct of

Pr im

ar y

Ca re

Pa ym

en t

M od

el s

on th

e Q

ua dr

up le

Ai m

an d

Te ne

ts of

Pr im

ar y

Ca re

P ay

m en

t M

od el

Q ua

dr up

le A

im A

llo w

s P

ro ac

ti ve

In ve

st m

en t

in P

ri m

ar y

C ar

e

T he

4 C

s of

P ri

m ar

y C

ar e

E le

m en

ts A

ss oc

ia te

d w

it h

Su cc

es sf

ul P

ro gr

am s

H ea

lt h

O ut

co m

es E

xp er

ie nc

e of

C ar

e C

os t

C on

tr ol

P ro

vi de

r Sa

ti sf

ac ti

on C

on ta

ct (A

cc es

s) C

on ti

nu it

y C

oo rd

in at

io n

C om

pr eh

en si

ve ne

ss

F ee

-f or

-s er

vi ce

(F F

S) 2

2 2

2 ✕

2 4 3

2 1 2

* B

ill in

g m

ec ha

ni sm

s av

ai la

bl e

th at

re co

gn iz

e pr

im ar

y ca

re te

ne ts

an d

no n-

fa ce

-t o-

fa ce

se rv

ic es

T ra

di ti

on al

(f ul

l- ri

sk )

ca pi

ta ti

on

4 3

M os

tl y 2

M os

tl y

1 2

2

In su

ff .

ev id

en ce

In su

ff .

ev id

en ce

1 2

R is

k lim

it ed

to pr

im ar

y ca

re se

rv ic

es P

M P

M de

te rm

in at

io n

ba se

d on

an ti

ci pa

te d

ne ed

ra th

er th

an F

F S*

P ay

-f or

– pe

rf or

m an

ce (P

4P )

2 1

2 1

2 1

2 ✕

2 1

2 4 3

2 A

pp ro

pr ia

te ,

al ig

ne d

m ea

su re

s fo

r us

e in

pr im

ar y

ca re

* B

un dl

ed pa

ym en

t (e

pi so

de -o

f- ca

re )

4 3

(w ea

k) In

su ff

. ev

id en

ce In

su ff

. ev

id en

ce In

su ff

. ev

id en

ce

/✕ (r et

ro ac

ti ve

ly tr

ig ge

re d)

In su

ff .

ev id

en ce

In su

ff .

ev id

en ce

1 (w

ea k)

In su

ff .

ev id

en ce

N /A

, m

ay no

t ap

pl y

to pr

im ar

y ca

re gi

ve n

di ffi

cu lt

y de

fi ni

ng an

d as

si gn

in g

bu nd

le s

Sh ar

ed sa

vi ng

s 1

1 2 1

In su

ff .

ev id

en ce

/✕

(p ro

vi de

rs of

te n

pa id

F F

S)

In su

ff .

ev id

en ce

In su

ff .

ev id

en ce

1 *

In su

ff .

ev id

en ce

T ar

ge t

hi gh

-n ee

ds pa

ti en

ts A

dd re

ss ps

yc ho

so ci

al ne

ed s

A pp

ro pr

ia te

ri sk

– ad

ju st

m en

t N

on -F

F S

in ce

nt iv

es at

pr ov

id er

le ve

l* P

hy si

ci an

-l ed

or in

te gr

at ed

A C

O

C on

ti nu

ed

doi: 10.3122/jabfm.2018.04.170388 U.S. Payment Models’ Impact on the Quadruple Aim 593

o n 3