PE R S PE C T IV E
variably refer patients with dense most efficacious. For example,
breasts for whole-breast ultra-
sound screening, with some prac-
tices referring 100% of such
women and others referring none.
Furthermore, only 45% of Con-
necticut women who were re-
ferred for follow-up ultrasonog-
raphy actually received it.5 Still,
breast-density legislation provides
an opportunity to strengthen
patient–provider relationships by
encouraging physicians to engage
women in a conversation about
the risks and benefits of screen-
ing, regardless of breast density.
In this era of cost contain-
ment, and given the limited data
supporting screening ultrasonog-
raphy, a rational and cost-effec-
tive approach to screening is
needed. So how should the medi-
cal community address the grow-
ing concern over breast density
and breast-cancer detection? It is
An audio interview critical that radiolo-
with Dr. Slanetz gists work with
is available at NEJM.org other specialists
and primary care physicians to
develop evidence-based recom-
mendations regarding situations
in which supplemental screening
is advisable and which method is
Should We Practice What We Profess? Care near the End of Life
Philip A. Pizzo, M.D., and David M. Walker, B.S.
P hysicians should be in a bet-
ter position than ­ people
without medical training to judge
the likely value of health care ser-
vices available near the end of
life. Yet ­several studies have re-
vealed a disconnect between the
way physicians themselves wish
to die and the way the patients
they care for do in fact die.
A 1998 survey of participants
in the Precursors Study, which en-
rolled 999 physicians who gradu-
n engl j med 372;7 nejm.org february 12, 2015
The New England Journal of Medicine
Downloaded from nejm.org at UNB on March 10, 2015. For personal use only. No other uses without permission.
Copyright © 2015 Massachusetts Medical Society. All rights reserved.
Disclosure forms provided by the authors
some practices now use digital
breast tomosynthesis, which
leads to increased cancer detec-
tion while limiting the need for
additional imaging in women
with dense breast tissue, accord-
ing to preliminary data.
Having dense breast tissue
does increase a woman’s lifetime
risk of breast cancer, but it’s im-
portant for providers to place
this risk in perspective for each
patient. Risk stratification will
be an essential tool in determin-
ing the best screening plan for
each woman. It would be helpful
if the medical community could
reach a consensus on how best
to advise women with dense
breasts with regard to the limita-
tions of various screening tests
and the role of any supplemen-
tal screening. Then, practitioners
could base patient care on exist-
ing evidence and each woman’s
individual risk. Such an approach
might well maximize cancer
detection and minimize the
­ US in patients with mammographically
downsides of screening — espe-
cially false positives and the
risks of overdiagnosis and over-
treatment.
ated from Johns Hopkins School
of Medicine between 1948 and
1964, revealed that 70% had not
had a conversation with their own
personal physician about end-of-
life care. But 64% had an advance
directive that they’d discussed
with their spouse or family, and
more than 80% indicated that
they would choose to receive pain
medication but would refuse life-
sustaining medical treatments at
the end of life.1 Similar prefer-
Breast-Density Legislation
are available with the full text of this article
at NEJM.org.
From the Department of Radiology, Beth
Israel Deaconess Medical Center (P.J.S.),
Harvard Medical School (P.J.S., P.E.F., R.L.B.),
the Department of Radiology, Massachu-
setts General Hospital (P.E.F.), and the De-
partment of Radiology, Brigham and Wom-
en’s Hospital (R.L.B.) — all in Boston.
1. Kerlikowske K, Hubbard RA, Miglioretti
DL, et al. Comparative effectiveness of digital
versus film-screen mammography in com-
munity practice in the United States: a cohort
study. Ann Intern Med 2011;155:493-502.
2. Smith RA, Duffy SW, Gabe R, Tabar L,
Yen AM, Chen TH. The randomized trials of
breast cancer screening: what have we
learned? Radiol Clin North Am 2004;42:793-
806, v.
3. Tice JA, Ollendorf DA, Lee JM, Pearson
SD. The comparative clinical effectiveness
and value of supplemental screening tests
following negative mammography in women
with dense breast tissue. Institute for Clinical
and Economic Review, 2013 (http://www
.ctaf.org/sites/default/files/assessments/
ctaf-final-report-dense-breast-imaging-11.04
.2013-b.pdf).
4. Berg WA, Blume JD, Cormack JB, et al.
Combined screening with ultrasound and
mammography vs mammography alone in
women at elevated risk of breast cancer.
JAMA 2008;299:2151-63.
5. Hooley RJ, Greenberg KL, Stackhouse RM,
Geisel JL, Butler RS, Philpotts LE. Screening
dense breasts: initial experience with Con-
necticut Public Act 09-41. Radiology 2012;
265:59-69.
DOI: 10.1056/NEJMp1413728
Copyright © 2015 Massachusetts Medical Society.
ences were expressed in a 2013
survey of 1147 younger academic
physicians (a group that was more
diverse and included more women):
88.3% indicated that they would
forgo high-intensity end-of-life
treatment.2
Although physicians ought not
assume that their views about
dying should apply to others,
public surveys and research stud-
ies have shown that 80% of
Americans, like the large majority
595
PERS PE C T IV E Care near the End of Life

Teno et al. noted that the rate of

Summary of IOM Committee Recommendations.* acute care hospitalization de-
creased from 32.6% in 2000 to
Delivery of care 24.6% in 2009 but that use of in-
Government health insurers and care delivery programs, as well as tensive care in the last month of
private health insurers, should cover comprehensive care, including life increased from 24.3% to
palliative care and hospice care for persons with advanced serious 29.2%.3 Although hospice use in-
illness who are nearing the end of life.
creased during this period,
Clinician–patient communication and advance care planning 28.4% of the decedents studied
Professional societies and other organizations should develop stan- had used hospice for 3 days or
dards for clinician–patient communication and advance care plan- less in 2009.
ning that are measurable, actionable, and evidence-based. Payers Complex social, cultural, eco-
and delivery organizations should adopt these standards and their nomic, geographic, and health
supporting processes and integrate them into assessments, care system factors and impediments
plans, and the reporting of health care quality. contribute to this discordance be-
tween how doctors treat their pa-
Professional education and development tients and how they themselves
Educational institutions, credentialing bodies, accrediting boards, state (and the majority of surveyed
regulatory agencies, and health care delivery organizations should Americans) wish to be cared for
establish appropriate training, certification, and licensure require-
at the end of life. We are experi-
ments to strengthen the palliative care knowledge and skills of all
clinicians who care for patients with advanced serious illness who encing the greatest demographic
are nearing the end of life. shift in U.S. history. According
to current projections, by 2030,
Policies and payment systems 20% of Americans will be more
Federal, state, and private insurance and health care delivery pro- than 65 years old. Cultural diver-
grams should integrate the financing of medical and social services sity is also increasing, as is the
to support the provision of high-quality care consistent with patients’ percentage of people with one or
values, goals, and informed preferences. Insofar as additional legis- more chronic illnesses. It is there-
lation is necessary to implement this recommendation, the admin- fore imperative that the medical
istration should seek and Congress should enact such legislation. The community listen to patients and
federal government should require public reporting on quality mea- recognize that their end-of-life
sures, outcomes, and costs and encourage private payers and deliv-
preferences may change over time,
ery systems to do the same.
especially as longevity increases.
Public education and engagement The goal should be to help peo-
Civic leaders, public health and other governmental agencies, com- ple receive care in keeping with
munity-based organizations, faith-based organizations, con­ sumer their personal preferences as they
groups, health care delivery organizations, payers, employers, and pro- near the end of life.
fessional societies should engage their constituents and provide fact- In Dying in America: Improving
based information to encourage advance care planning and informed Quality and Honoring Individual
choice based on individuals’ needs and values. Preferences near the End of Life, an
Institute of Medicine (IOM) com-
* The full report is available at www.iom.edu/endoflife.
mittee (which we cochaired) con-
cluded that the U.S. health care
system is poorly designed to
of surveyed physicians, say they’d sponding to the wishes and de- meet the needs of patients and
like to die at home and avoid high- mands of patients’ families who their families at the end of life
intensity care and hospitalization. want more medical therapy than and that major changes are need-
Yet their wishes are too frequently medical providers believe is indi- ed. We need to begin by foster-
overridden by the physicians car- cated or beneficial. In a study ex- ing patients’ ability to take con-
ing for them, who undertake amining the care of more than trol of their quality of life
more medical interventions than 848,000 people who had died in throughout their life and to
patients desire. Physicians also 2000, 2005, or 2009 while cov- choose the care they desire near
sometimes find themselves re- ered by fee-for-service Medicare, the end of life.4 The committee

596 n engl j med 372;7 nejm.org february 12, 2015

The New England Journal of Medicine

Downloaded from nejm.org at UNB on March 10, 2015. For personal use only. No other uses without permission.
Copyright © 2015 Massachusetts Medical Society. All rights reserved.
PE R S PE C T IV E
recognized that these goals
could be achieved only by mak-
ing major changes to the educa-
tion, training, and practice of
health care professionals, as well
as changes in health care policy
and payment systems, Simulta­
neously, individual and public
education would have to be radi-
cally reformed to reshape expec-
tations and allow patients and
clinicians to have meaningful dis-
cussions about end-of-life plan-
ning (see box).
The IOM committee concluded
that “federal, state, and private in-
surance and health care delivery
programs should integrate the fi-
nancing of medical and social
services to support the provision
of quality care consistent with the
values, goals, and informed pref-
erences of people with advanced
serious illness nearing the end of
life.” More specifically, the com-
mittee recommended that insofar
as additional legislation is required
to allow for such financing, rele-
vant laws should be enacted (e.g.,
authorization of payments for ser-
vices delivered in ambulatory or
home settings rather than only
in inpatient settings) and that the
federal government should “re-
quire public reporting on quality
measures, outcomes, and costs
regarding care near the end of
life . . . for programs it funds
or administers (e.g., Medicare,
Medicaid, the Department of
Veterans Affairs)” and encourage
other U.S. payment and delivery
systems to follow suit. We believe
that physicians can and should
work with their professional or-
ganizations to advocate for these
changes — but rather than wait-
ing for new legislation, they can
take action now, in part by set-
ting aside time to encourage pa-
tients to express their preferenc-
es regarding end-of-life care.
Physician practices can also
n engl j med 372;7 nejm.org february 12, 2015
The New England Journal of Medicine
Downloaded from nejm.org at UNB on March 10, 2015. For personal use only. No other uses without permission.
Copyright © 2015 Massachusetts Medical Society. All rights reserved.
organize their clinical services
so as to provide seamless, high-
quality, patient- and family-cen-
tric care that is consistently avail-
able to their patients, especially
those who have advanced serious
illness or are nearing the end of
life. Even by simply providing pa-
tients with a consistent and ac-
cessible place to call when they
need help, physicians can avert un-
necessary trips to the emergency
department or another acute care
setting where patients’ individual
preferences may not be known or
honored. Becoming more acces-
sible in this way improves the
quality of care and should reduce
unnecessary utilization of expen-
sive medical treatments.
Physicians can also work to
ensure that their patients have ac-
cess — in all care settings — to
skilled palliative care or, when
appropriate, hospice care. We be-
lieve that basic palliative care
skills should be part of the knowl-
edge base of all physicians car-
ing for people with advanced se-
rious illness or near the end of
life. Physicians can also seek out
collaboration, whenever possible,
with skilled palliative care spe-
cialists, whether doctors, nurses,
social workers, or clergypersons,
to ensure the best possible care
of their patients. It has been
demonstrated that when pallia-
tive care is combined with active
treatment for patients with ad-
vanced cancer, the quality and
duration of life are enhanced.5
All this care should be coordi-
nated, and handoffs should be
avoided at critical junctures for
patients, such as when they first
encounter a chronic illness or a
life-threatening disease.
Ideally, physicians would initi-
ate discussions about advance di-
rectives with their patients at key
milestones throughout their lives
— perhaps when they get a driv-
Care near the End of Life
er’s license, get married, begin a
new job, relocate, or become eli-
gible for Medicare — not just
when advanced illness or death is
imminent. Many physicians need
to learn how to conduct these
conversations respectfully and
successfully. Physicians can then
make their patients’ preferences
known to all members of the
health care team. Physicians
should be compensated for the
time required to have these dis-
cussions — a change they can
prod the government and other
payers to make.
Changing the culture in these
ways will require intervention at
all stages of physicians’ education.
Physician educators can develop
new models of teaching (includ-
ing the use of simulation) for
students, residents, and fellows.
But physicians can also learn and
teach about compassionate patient
care in their practice settings
and communities. They can then
contribute to public dialogues
about end-of-life issues in their
communities and religious groups
— working especially to help to
dispel misinformation.
Physicians’ experiences with
medical care and dying patients
have helped crystallize their de-
sires for their own end-of-life
experiences. As Dying in America
makes clear, physicians should
now practice what they profess,
to ensure that their patients have
the same options that they them-
selves, and a majority of Ameri­
cans, would choose and that they
honor patients’ preferences at the
end of life.
Disclosure forms provided by the authors
are available with the full text of this article
at NEJM.org.
Dr. Pizzo is a professor of pediatrics and of
microbiology and immunology and former
dean at Stanford University School of Medi-
cine, Stanford, CA; and Mr. Walker is a for-
mer comptroller general of the United
States.
597
PERS PE C T IV E
1. Gallo JJ, Straton JB, Klag MJ, et al. Life-
sustaining treatments: what do physicians
want and do they express their wishes to
others? J Am Geriatr Soc 2003;51:961-9.
2. Periyakoil VS, Neri E, Fong A, Kraemer H.
Do unto others: doctors’ personal end-of-life
resuscitation preferences and their attitudes
toward advance directives. PLoS One 2014;
9(5):e98246.
Finding the Right Words at the Right Time — High-Value
Advance Care Planning
Justin Sanders, M.D.
W hen Ms. C. died, I was sad
but not surprised. I had
met her 4 years earlier, when I
was an intern and she was the
first patient who identified me
as “my doctor.” She did so en-
thusiastically, asking the inpatient
medical teams who frequently
cared for her to run every deci-
sion by me. As a trainee, and
given her complex needs, I found
those requests both absurd and
overwhelming. By 65 years of age,
Ms. C., a lifelong smoker, had
coronary artery disease, atrial
fibrillation, diabetes, and chron-
ic obstructive pulmonary disease
(COPD) complicated by pulmo-
nary hypertension. During the
time I knew her, she was hospi-
talized at least every 3 months
for complications of one or an-
other of her chronic conditions.
The only thing she hated more
than the hospital was the panic
induced by uncontrolled dyspnea,
chest pain, or palpitations —
the panic that led her to dial
911. When she came into the
clinic for follow-up, she’d tell
me that she never wanted to go
back. I would check to make sure
she understood her medication
changes. She would ask about
my family. I would plead with
her to quit smoking. A gregarious
Latina, she always shouted “I love
you” as I walked out of the room.
598
The New England Journal of Medicine
Downloaded from nejm.org at UNB on March 10, 2015. For personal use only. No other uses without permission.
Copyright © 2015 Massachusetts Medical Society. All rights reserved.
3. Teno JM, Gozalo PL, Bynum JPW, et al.
Change in end-of-life care for Medicare ben-
eficiaries: site of death, place of care, and
health care transitions in 2000, 2005, and
2009. JAMA 2013;309:470-7.
4. Institute of Medicine. Dying in America:
improving quality and honoring individual
preferences near the end of life. Washington,
DC: National Academies Press, 2014.
I remember when Ms. C.’s car-
diologist, who was as close to
her as I was, told me he was con-
cerned that she might not live
another year: her arrhythmia had
become more difficult to control.
She had no advance directive. He
suggested that I speak with her
about it, and I said I’d try to find
a good time. I never did.
Some months later, as a newly
appointed attending, I returned
from a vacation to an e-mail in-
forming me of Ms. C.’s death. She
had died in the intensive care unit
after an unsuccessful attempt at
cardiopulmonary resuscitation. I’m
pretty sure that’s not what she
would have wanted, but I couldn’t
say for certain. I stared at my
computer screen, feeling the lead-
en weight of a missed opportu-
nity and a sense of profound dis-
appointment in myself. I felt that
I had failed one of my first and
favorite patients.
Since becoming a palliative
care doctor some years later, I’ve
thought many times about Ms. C.
and the consequences of my own
and others’ inaction. And these
missed opportunities have become
a topic of national conversation.
Last September, the Institute of
Medicine (IOM) released a report
entitled Dying in America, in which
it recommends measures to im-
prove end-of-life care through,
n engl j med 372;7 nejm.org february 12, 2015
Care near the End of Life
5. Smith TJ, Temin S, Alesi ER, et al.
American Society of Clinical Oncology pro­
visional clinical opinion: the integration of
palliative care into standard oncology care.
J Clin Oncol 2012;30:880-7.
DOI: 10.1056/NEJMp1413167
Copyright © 2015 Massachusetts Medical Society.
Related article, p. 595
among other strategies, better
advance care planning (ACP).1
Specifically, it recommends the
development of “standards for
clinician–patient communication
and advance care planning that
are measurable, actionable, and
evidence based” and that these
standards be tied by payers and
professional societies to “reim-
bursement, licensing, and creden-
tialing” (see Perspective article by
Pizzo and Walker, pages 595–598).
If promoting ACP discussions
were as simple as asking or pay-
ing physicians to have them, Dy-
ing in America might not have been
necessary. These discussions are
difficult, and for multiple reasons:
perceived difficulty of prognosti-
cation, uncertainty about how
best to communicate with pa-
tients and families with diverse
communication needs, and inad-
equate time to have them — not
to mention the troubling emo-
tions that talk of death raises for
both patients and physicians.2
During our medical education,
discussions of end-of-life care re-
ceive minimal, if any, attention.
In response to deficiencies in
physician communication about
end-of-life care preferences, policy-
makers, patient advocates, and
payers have endeavored to move
ACP out of physicians’ hands,
from the clinic to the telephone