Assuring Safe Passage: Moving Conversations about the End of Life Upstream

1
Assuring Safe Passage Moving Conversations about
the End of Life Upstream

• Susan D. Block, MD
• Chair, Department of Psychosocial Oncology and
  Palliative Care
• Professor of Psychiatry and Medicine
• Dana-Farber Cancer Institute
• Brigham and Womens Hospital
• Harvard Medical School Center for Palliative Care

2
End-of-Life Conversations

• Its too early until its too late.

3
Objectives

• Provide overview of communication about
  end-of-life care and propose evidence-based
  approach to improve process
• Describe rationale for early values and goals
  discussions as a high-value clinical intervention
• Review rationale for use of checklists in
  medicine
• Describe a systems approach to serious illness
  communication for patients with serious and
  life-threatening illness

4
What are the biggest challenges you face in
communicating with patients about end-of-life
care?
5
The end-of-life experience of cancer patients

• Patients who receive hospice care report better
  quality of life at the end of life
• Aggressive care is associated with reduced
  quality of life, more physical and emotional
  distress at the EOL
• In last month of life, cancer patients
  experience
• Chemotherapy (40)
• Emergency department visits (48)
• Hospital admissions (50)
• Death in ICU (8)
• 55-75 have serious pain, anxiety, discomfort
• No hospice referral (35)
• Late hospice referral (median 16 days)

6
Impact on family/caregivers

• Death in hospital /ICU is associated with 9-11
  times higher rate of PTSD, generalized anxiety
  disorder and prolonged grief disorder among
  bereaved caregivers
• Family members of patients who die report higher
  satisfaction levels with care in hospice than in
  hospital

7
End-of-Life Experiences
Cancer COPD CHF
Hospitalization gt50 last month gt60 last 6 months gt80 last 6 months
ICU admission 8 33 50
Hospice care 65 40
ED visits 48
Invasive procedures in last month 40 chemo 20 CABG, dialysis, AICD, PM, Cath
8
Quality of end-of-life care (Teno 2004)
Families Reporting Hospital Hospice
Inadequate contact or communication with MD 78 34
Inadequate help with emotions 52 35
Poor family support 38 21
Lack of respect 20 4
Excellent care 47 71
9
Advance Care Planning

• About 30 of Americans have an Advance Directive
  (Keeter 2006)
• No evidence that discussion of ACP increases
  anxiety or hopelessness (Wright 2008)
• Discussion in health care setting about ACP is
  associated with increased discussion by patient
  with family, friends (Jones 2011)
• RCTs In older patients, ACP is associated with
  higher level of goal-consistent care (Detering
  2010 Silveira 2010)

10
Cancer Patients and EOL Discussions

• Only 37 of cancer patients with average survival
  of 4 months reported having had discussion about
  EOL issues with physician (Wright 2008)
• Although 95 of hospitalized oncology patients
  believe it is important to have discussions about
  advance directives/EOL care, only 41 have them
• Major barrier Oncologists dont bring up issue
  (Dow 2010)
• 37 of cancer patients admitted to ICU had no HCP
  or living will (Halpern 2011)

11
Prognostication

• 90 of patients on dialysis (20-25 annual
  mortality rate) report they have not discussed
  prognosis (Davison et al. 2010)
• 75-90 of patients and caregivers (multiple
  diseases) say they want all the information about
  their disease , including prognosis (Jenkins
  2001, Mack 2006 Janssen et al. 2012)

12
Butwe dont do it routinely and well

• Discussions happen late, patients are unprepared,
  and are often at their worst
• Large prospective cohort study, lung and
  colorectal cancer
• 87 of patients who died had EOL discussion
  reported or documented
• 55 of first conversations took place in hospital
• First conversation took place a median of 33 days
  before death
• Only 27 were conducted by oncologists

(Mack 2012)
13
Even when we do ACP, there are problems with EHR
documentation

• Advance care planning information found in
• Progress notes 69
• Scanned documents 43
• Problem lists 34 Seale et al. Clin Med
  Research 2012
• Many had documentation in multiple locations, not
  all consistent
• No standardization in type of information
  collected and location of information

14
Does documentation improve care?

• Increased documentation is associated with more
  concordance of patient values with care received
  (Morrison 2005)

15
End-of-life discussions are a high-value clinical
intervention
16
Improving quality

• Early discussion of end-of-life care issues among
  cancer patients is associated with improved
  outcomes
• Patients more likely to have wishes followed
• Increases quality of life
• Reduces rate of hospitalization and ICU
  admission
• Increases use of hospice
• Reduces stress, anxiety, depression, PTSD and
  bereavement morbidity in survivors
• Improves family satisfaction
• Strengthens clinician-patient relationship

(Wright 2008)
17
Reducing costs

• 25 of Medicare costs occur in last year of life
  50 of costs are on hospitalization
• Early conversations associated with 36 cost
  reduction (Zhang 2009)
• MGH early palliative care RCT in patients with
  NSCLC showed 29 reduction in hospital days
  major focus of visits was communication (Temel
  2010, Jacobsen 2011)

18
Analyze failures in the processes of care
A Systematic Approach to Designing Better
Performance
Identify Pause Points
Identify best practices to avert key failures for
each Pause Point
Distill into practical intervention (e.g.,
checklist)
Redraft
Try and retry the intervention
Problem
19
Failures in End-of-Life Planning

• Patients are not prepared to talk about the end
  of life
• No system for identifying appropriate patients
• Unclear who should conduct the conversation
• Inadequate clinician competency/comfort in EOL
  discussions
• Conversations happen too late
• Clinicians focus on procedures, rather than
  values and goals
• Patients often dont know how to discuss with
  family family members dont understand patients
  wishes
• Information about patient wishes is not clearly
  and predictably recorded in EHR

20
Prepare Patients Health Care Agent Designation
for all Newly Diagnosed Patients

• Systems-level standard
• All patients lt 65 or with serious or chronic
  illness should designate HCP (or decline in
  writing) within first three visits
• Normalizes discussion for patients, clinicians
• Cancer is a serious illness unexpected things
  happen we need to be prepared helps assure your
  wishes will be followed

21
DFCI Health Care Proxy Initiative
22
Identify Triggers COPD

• Initiation of new or different treatments
• Lack of further treatment options
• Functional decline
• Symptom exacerbation
• Ongoing oxygen requirement
• Hospitalizations
• Janssen DJA et al. Patient Education and
  Counseling 2012

23
Identify Triggers CHF

• Annual CHF review
• Increased symptoms
• Reduced function
• Hospitalization
• Progressive increase in diuretic need
• Hypotension
• Azotemia
• Initiation of inotrope therapy
• First or recurrent shock (Allen et al.
  Circulation 2012)

24
Identify Triggers Cancer

• Prognosis-related triggers
• Would you be surprised if this patient died in
  the next year?(Moss AH, et al. 2010)
• Disease-based/condition-based criteria, e.g.
• All patients with NSCLC, pancreatic cancer, GBM
• Patients over 70 with AML
• Treatment-based identification
• Prostate cancer 4 visits cabazitaxel
• Renal cancer 4 visits everolimus

25
Train Clinicians

• RCTs Communication skills training programs work
  (Jenkins 2002, Fallowfield 2006, Szmuilowicz
  2010)
• Other studies demonstrate impact of training
  (Sullivan 2005, Back et al. 2007)
• Key elements
• cognitive input
• modeling
• skills practice with feedback
• Improvements are maintained for at least 6 months
  (Maguire 2002, Sullivan 2006)

26
Prompt Clinicians to Discuss End-of-Life
Preferences

• Initiate discussion early
• Outpatient setting, no crisis
• Clinician initiates discussion
• Ongoing clinician-patient relationship
• Focus initially on discussion, exploration, NOT
  decisions
• Allows patient time to adjust to reality of poor
  prognosis without pressure to make major
  decisions
• Allows patient time to talk with family about
  wishes

27
What do checklists do?

• Bridge gap between evidence and real world
  implementation
• Assure adherence to key processes
• Achieve higher level of baseline performance
• Ensure completion of necessary tasks during
  complex, stressful situations

28
A tested structure for communication 7 questions
to guide EOL care
29
Prompt patients to discuss EOL planning with
family members
30
(No Transcript)
31
Document in EHR

• Goal A single source of truth in electronic
  health record
• Structured documentation format supports
  acquisition and communication of key information

32
Partners Health System Advance Care Planning
Module
33
(No Transcript)
34
(No Transcript)
35
Outcomes?
36
Is the checklist discussion harmful?Anxiety
before after Checklist discussion
0 Not at all 1 Several days in the last 2 weeks 2 More than half the days 3 Nearly every day
Mean GAD score Baseline 4.24 Post-SICC 3.09
37
Is the checklist discussion harmful? Depression
before after Checklist discussion
0 Not at all 1 Several days in the last 2 weeks 2 More than half the days 3 Nearly every day
Mean PHQ score Baseline 5.34 Post-SICC 4.00
38
How did clinicians think the Checklist discussion
affected their patients?
39
Do patients find the Checklist conversation
acceptable?
40
Does the Checklist help the clinicians?
Clinicians satisfaction with role in patient's care increased or greatly increased 64
Understand patient's values and goals about end-of-life care quite a bit or a great deal 79
Plan continued use of checklist format for discussing these issues 71
41
Checklist Acceptability
The Checklist-guided discussion Agree or strongly agree (N14)
Allows for discussion about end-of-life issues in a timely manner. 86
Format is simple. 79
Is easy to use. 78
42
How do patients rank their priorities after
discussion?
43
Most patients say they received the right amount
of information
44
Effectiveness
Ability/Comfort in implementing the following practices in next conversation with patient quite a bit or a lot
Acknowledging difficult emotions during conversation 94
Responding to patient/family emotion 100
Eliciting patient concerns 95
Speaking less than 50 of time in this discussion 58
Knowing what to do in challenging situations 73
Overall, how effective was this session in enhancing your confidence in talking with patients in the format described? 89
Overall, how effective did you find this program to be in improving your skills in conducting a discussion about end-of-life care? 95
45
Summary of preliminary data

• Checklist discussions are feasible
• Checklist discussions do not appear to be harmful
• No increase in anxiety or depression
• Patients and clinicians find discussions helpful
• Apparent gap in communication of prognosis to
  patients