Honoring the Informed Choices of People Coming to the End of Life

For the Foundation for Informed Medical Decision Making,Washington, DC, January 26, 2011

Joanne Lynn, MD, MA, MSClinical Improvement Expert

Colorado Foundation for Medical CareDrJoanneLynn@gmail.com

Framing Our Issues

• Travelling the Valley of the Shadow of Death…

• Trajectories and categories

• The moral mandate for patient choice and the effectiveness of advance directives

And what we could do…

The Opportunity –US Hospitalists on Sedation for Dyspnea

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Want Sedation for self

Offer Sedation to

Patient

Lynn, Goldstein, Annals Int Med, May 20,2003

How Americans Die: A Century of Change

1900 2000

Age at death 46 years 78 years

Top Causes Infection

Cancer

Accident Organ system

failure

Childbirth

Stroke/Dementia

Disability Not much 2-4 yrs

before death

Financing Private, Public,

substantial- modest

© Copyright 2003, Onion, Inc., All rights reserved.

By permission of Johnny Hart and Creators Syndicate, Inc.

TimeOnset of incurable illness

-- Often a few years from onset,

But decline usually < 2 months

“Cancer” Trajectory, Diagnosis to DeathFu

nct

ion

Death

High

Low

Possible hospice enrollment

“Organ System Failure” Trajectory

Time

Fun

ctio

n

Death

High

Low

Begin to use hospital often, self-care becomes difficult

~ 2-5 years, but death seems “sudden”

TimeOnset could be deficits in

ADL, speech, ambulationQuite variable -

often 6-8 years

Death

High

Low

“Frailty/Dementia” Trajectory

Fu

nc

tio

n

Who is in the Category “End of Life?”

NOT “reliably short prognosis” (e.g., < 6 months) because

– most of us will be stable

– with serious illness and disability

– within a week or two of our deaths

Can we tell who will die soon?

The days just before death

Me

dia

n 2

-mo

nth

Su

rviv

al

Es

tim

ate

0.0

0.2

0.4

0.6

0.8

1.0

7 6 5 4 3 2 1

Prognosis Stays Uncertain Through Most of the Last Part of Life

Days before Death

Med

ian

2-m

on

th S

urv

ival

Esti

mate

0.0

0.2

0.4

0.6

0.8

1.0

7 6 5 4 3 2 1

Lung cancer

Congestive heart

failure

* From SUPPORT, 1988-93

Who Should we Categorize as “End of Life?”

Better answer -

Seriously ill and disabled

With condition(s) that will not substantially improve

Will worsen

And will cause death.(No particular survival time is part of the definition)

The “No Surprise” Population

Would it be a surprise for this person to die within six months? (or a year – doesn’t matter)

If “no surprise” – then “end of life” care– Priorities: planning ahead, comfort, family

– Optimal medical care

– Can continue for a few years

– Includes the short time when dying soon

Gold Standards Framework, Britain www.goldstandardsframework.nhs.uk

Optimal services for the person living with (eventually) fatal illnesses

• Honesty, emotional support, and profound respect

(not the commonplace manipulation and avoidance)

• Negotiating a care plan (ongoing), including planning in advance, and tailored to the patient and family

• Support at home

• Caregiver assessment, training, and support

• Financial and legal help

• Attention to transfers

Advance Care Planning Works

• POLST model www.POLST.org – 22 states

• LaCrosse model: www.Respecting Choices.org

• VA Health System

Serves patient interests

Reduces inappropriate utilization

Alleviates family tensions and uncertainty

Advance Care Planning – Key Elements

• Honesty about likely course (from knowledgeable clinician)

• Normalizing discussion, including uncertainties and trajectories

• Patient and family centered goals, constrained by the situation

• Name a surrogate decision-maker if possible

• Address CPR, perhaps artificial nutrition, and other elements customized to the situation

• Document and assure availability

“Drivers” of Dodging the Discussions

• Patients/families don’t want it

• Providers presume that patients/families will be pained (perhaps unnecessarily)

• Everyone colludes in dodging

• Providers’ financial and regulatory incentives make it easier to keep treating

• “Death panels” makes it hard to provide political/public leadership

• Real dying and disability is not shown on TV or other information media

What works to start the conversation?

• “At this time in your life, what makes you truly happy?”

• Triggers that are expected – in flyers, in early visits, etc – “By the third visit after a hospitalization for heart failure, we will have a discussion of what this means for your life”

• “Tell me about the experience of being so short of breath”

Does a good dying have to be

“losing the valiant battle?”

When hospice is available, many choose it(absolute % change in inpatient deaths by venue nationally)

ICU AcuteNursing

HomeHospice

Change 4 % 3 % 16% 23%

1st Q

FY1023% 35% 7 % 35%

FY04 27% 38% 23% 12%

~5060 veterans affected

Evolving Beyond the “Either-Or”

Traditional Approach

Integrated Approach

“Curative” care

Hospice

“Comfort”

Care

Curative or restorative

Palliative Support

Adapted from Daniel Johnson, MD

How does Pall. Care Reduce EOL Costs?

• Fewer hospitalizations and deaths in hospital– Better care coordination, more hospice, advance

planning

• More admissions directly to PCU from ED

• Shorter LOS, especially in ICUs– Reduce severe symptoms– More transfers out of, fewer into, ICUs

• Less use of labs, radiology, pharmacy, blood– Change, clarify goals of care

An especially sensitive issue….

How can anyone know that the patient’s dying was actually timely?

(appropriate diagnosis and treatment – and not death from inattention, denial of treatment, or deliberate cause)

Possible Answers• Standards about diagnosis and severity

• Standards about choice and planning

• Autopsies

• Reporting concerns, threats to safety

A Particular Opportunity

• Medicare quality measures have been insensitive to preference

• Serious illness near the end of life requires substantial customization – that’s the standard

• So – we need ways to measure the quality of a customized care plan, its continuity over settings and time, its implementation

• FMDIM could help create the demand for this, then the tools for measurement

Negotiated Values

Plan Implement

Outcomes

Goals Integration

Feedback Feedback

Evaluation of

Quality

Defining Quality in Unique Situations

Negotiated Values

Plan Implement Outcomes at T1

Negotiated Values

Plan Implement Outcomes at T2

TIME

Defining Quality for Serious Chronic Illness

Why Bother?

• Suffering (unnecessarily severe)

• Costs (unnecessarily high)

• Track record of successful improvements

• Unpopularity of status quo

So – we might have the political will to reduce suffering, improve care and reduce costs

Why Bother?

It was my father this time,

but next time it will be your father, and then you, and then your child.

I have heard it said by cynics that the quality of medical care would be far better and the hazards far less if physicians, like pilots, were passengers in their own airplanes.

We are.

Berwick, Quality comes home. Ann Int Med 1996; 125:839-832