Improving the Quality of Palliative Care for Patients in the ICU

1
Improving the Quality of Palliative Care for
Patients in the ICU

• J. Randall Curtis, MD, MPH
• Pulmonary and Critical Care Medicine
• University of Washington

2
Outline

• Epidemiology of death in the ICU
• Communication with patients and families
• Practice of withdrawing life support
• System-level interventions

3
Changing Nature of Death in the ICU
plt0.001
Prendergast Luce, AJRCCM, 1997
4
Variability in Clinician Approach to Withdrawing
Life Support
Cook, JAMA, 1995
5
Physician Biases Around Withdrawing Life Support

• Physicians prefer to withdraw
• Natural rather than iatrogenic cases
• Most recently instituted therapies
• Therapies that will result in delayed death
• Physician factors affect decisions
• Younger, academic, specialist, and less religious
  were more willing to withdraw

Christakis and Asch, Lancet, 1993
6
Pain in Seriously Ill Hospitalized Patients
SUPPORT

• Of 3624 patients able to be interviewed
• 50 reported pain
• 15 reported severe pain
• 15 were dissatisfied with pain control
• Patients with MOF or sepsis report similar pain
  to other patients
• Intensivists had lowest satisfaction for pain
  control

Desbiens, Crit Care Med, 1996
7

Curtis, J Pain Sx Manage, 2002 2417
8
Quality of Dying and Death and ICU Treatments

• QODD score p value
• ICU in last month 0.4
• Yes (n25) 64.6
• No (n179) 67.7
• of invasive treatments 0.02
• 0 (n135) 68.2
• 1 (n16) 66.8
• gt2 (n33) 60.8

Curtis, J Pain Sympt Management, 2002 2417
9
Summary Epidemiology of Death in the ICU

• Majority of deaths in the ICU involve withholding
  or withdrawing life support
• Some evidence suggests we dont do this as well
  as we should
• ICU death need not necessarily be a bad death

10
Outline

• Epidemiology of death in the ICU
• Communication with patients and families
• How are we doing?
• How can we improve?
• Practice of withdrawing life support
• System-level interventions

11
What Do We Know About End-of-life Communication
in the ICU?

• lt5 of patients can participate in ICU decisions
  about withholding treatments
• Communication is primarily with family
• Families rate communication as more important
  than clinical skill
• Families under immense burdens
• Emotional, financial, personal health

Prendergast, AJRCCM, 1997 Covinsky, JAMA 1994
12
Survey of 920 Family Members of Patients in 43
French ICUs

• Psychological symptoms common
• Anxiety 70 Depression 35
• Caregiver factors associated with family anxiety
  or depression
• Absence of regular meetings with MD, RN
• Absence of a family meeting room
• Perceived contradictions in info provided
• Absence of a waiting room

Prochard, Crit Care Med 2001 291893
13
Before-after Study of Communication Intervention
in Medical ICU

• Intervention family conference held within 72
  hours if attending predicts
• ICU stay gt5 days or mortality gt25
• Conference conducted by attending
• Review medical facts and options
• Discuss patients perspective on EOL
• Agree on care plan
• Agree on criteria of success or failure

Lilly, Am J Med 2000 109469
14
Results of Communication Intervention in Medical
ICU

• Before After
• (n134) (n396) p value
• ICU LOS (d) 4 3 0.004
• Worst APACHE quartile
• Survivors 5 4.5 0.8
• Died 5 3 0.02
• Overall mortality odds ratio
• Afterbefore 0.61 (0.38-0.98)

Lilly, Am J Med 2000 109469
adjusted for APACHE III
15
Study of Effectiveness of Clinician-Family
Communication

• 102 consecutive patients in ICU gt2 days 76
  patients visited by family
• Interviewed family after meeting with physician
  to assess comprehension
• Failure to understand basics of
• dx (20), px (43), tx (43)

Azoulay, Crit Care Med 2000 283044
16
Study of ICU Family Conferences

• Daily screen of all ICUs in 4 hospitals
• If conference planned, contact attending
• Is discussion of withholding or withdrawing life
  support likely?
• Willing to have conference recorded?
• Consent all conference participants
• Qualitative and quantitative analyses

Curtis, J Crit Care, 2002 17147
17
Family Conferences and Participants

• 51 family conferences
• 51 unique families
• 36 unique physicians leading conference
• (26 MDs did 1 conference 7 MDs did 2
• 3 MD did 3 or 4)
• 214 family members

Curtis, J Crit Care, 2002 17147
18
Content of the Discussions

• Openings and introductions
• Two-way information exchange
• Discussion of the future
• Prognosis survival quality of life
• Decisions to be made
• Discussions of dying and death
• Closings

Curtis, J Crit Care, 2002 17147
19
Communication Style and Emotional Support

• Support
• Acknowledging the difficulty and emotions
• Discuss patients life and values
• Support for families decisions
• Accessibility

• Style
• Directness with caring
• Active listening
• Addressing conflicts

Curtis, J Crit Care, 2002 17147
20
Support Personalize the Patient

• MD Tell us a little bit about (patient name).
  None of us really know her. What is she like?
  What does she value about life? What kinds of
  things does she like to do?

Curtis, J Crit Care, 2002 17147
21
Support Easing Burden

• MD You can get that sick very quickly just from
  pancreatitis. Bringing her to the hospital any
  sooner wouldnt have made a difference its not
  like we would have been able to give her a
  medication that would have prevented all this
  from happening. So you dont need to worry that
  if youd only brought her in a sooner things
  would be different.

Curtis, J Crit Care, 2002 17147
22
Support Support for Family Decision-Making

• MD With medical science we could prop him up
  and try to get him through this, but it seems
  clear that his wishes would be to not have that
  done. I think its a reasonable decision youve
  all made and a brave one also in that you have to
  put aside your own personal feelings of wanting
  to have him around. Letting go is difficult, but
  I think youre doing him a great service by
  honoring his wishes at this time.

Curtis, J Crit Care, 2002 17147
23
Duration of Family Conferences and Proportion of
Family Speech

• Mean SD
• Duration of conference 32 min 17-45 min
• Proportion family speech 29 14-44

McDonaugh, in progress
24
Proportion Family Speech Correlates with Family
Satisfaction

• Family Speech Duration
• How well did r (p value) r (p value)
• MD communicate 0.37 (0.01) -0.07 (NS)
• Conf. meet needs 0.31 (0.04) 0.08 (NS)
• How much conflict -0.31 (0.04) 0.28 (0.07)

McDonaugh, submitted
25
Summary Discussing Dying and Death in the ICU

• Communication with critical care clinicians
  important to families
• Important to ascertain goals and values directly
  from patients/families
• Develop a protocol for family discussions
• Include agenda and emotional support techniques
• Listen to families
• Critical care clinicians could use help

26
Outline

• Epidemiology of death in the ICU
• Communication with patients and families
• Practice of withdrawing life support
• System-level interventions

27
Withdrawal of Life Support Is a Medical Procedure

• Education of patient and family
• Proper setting and monitoring
• Adequate sedation and analgesia
• Active role for physician
• Prepare for complications
• Documentation
• Quality improvement

28
Setting and Monitoring

• Quiet and privacy
• Remove all monitoring and disable alarms
• Physical exam sufficient to assess pain and
  diagnose death
• Liberalize visitation
• Stop laboratory and radiographic tests

29
Sedation Drugs and Doses

• Drugs used to treat pain and agitation usually
  sufficient
• Narcotic, benzodiazepine, neuroleptic
• Doses
• Explicit dosing guidelines difficult
• No dose is too high if lower doses fail
• Difficult to kill patients with narcotics
• Document reasons for increasing

30
Once the Decision Is Made to Withdraw, Just Turn
It Off

• Only justification for weaning life support is
  when its abrupt removal will cause discomfort
• All life support except ventilator can just be
  turned off
• Stuttering withdrawal can and should be avoided

31
Terminal Discontinuation of the Ventilator
Full Ventilatory Support
Remove supplemental O2 and PEEP
Reduce set rate or PS gradually

• Titrate sedation to ensure comfort
• Takes 5 minutes

• Titrate sedation to ensure comfort
• Takes 5 minutes

• Titrate sedation to ensure comfort
• Takes 5-20 min

32
Should Patients Be Extubated After Withdrawing
Mechanical Ventilation?

• Little evidence to guide decisions
• Clinicians frequently have strong opinions
• Case-based judgment based on
• Family preferences
• Level of support, amount of secretions, level of
  consciousness

33
Outline

• Epidemiology of death in the ICU
• Communication with patients and families
• Practice of withdrawing life support
• System-level interventions

34
Before-after Study of Proactive Palliative Care
Consult in a Medical ICU

• Intervention automatic palliative care consult
  for patients with
• Anoxic encephalopathy after cardiac arrest
• MODS gt3 organs for gt3 days
• Goals of the consult
• Communicate prognosis to family
• Identify patient preferences
• Discuss treatment options with family
• Implement palliative care strategies

Campbell, Chest 2003 123266
35
Results of Palliative Care Consult Intervention
in Medical ICU

• Before After
• (n22) (n21) p value
• ICU LOS (days)
• Anoxic enceph 7.1 3.7 0.01
• MODS 10.7 10.4 0.74
• ICU LOS from diagnosis (days)
• Anoxic enceph 7.1 3.7 0.01
• MODS 5.8 2.1 0.05

Campbell, Chest 2003 123266
36
Before-After Study of a Withdrawal of Life
Support Order Form

• Intervention Implementation of a standardized
  order form for withdrawing life support
• Developed by a multi-disciplinary team
• Presented at institutional multi-disciplinary
  forum for feedback from staff
• In-service education conducted in each ICU

Treece, Am J Resp Crit Care Med, 2003 167A582
37
Components of the Withdrawal of Life Support Form

• Preparation
• DNAR order document discussion with family
  discontinue prior orders
• Analgesia and sedation
• Infusion with broad range no maximum dose
  document reason for increase
• Ventilator withdrawal protocol
• Principles of withdrawing life support

38
Clinician Satisfaction Ratings Generally High

• MD (n61) RN (n73)
• YES YES
• Orders Helpful? 98 84
• Sections helpful
• Preparation 70 36
• Sedate/Analgesia 93 70
• Ventilation 79 44
• Principles 98 6

Treece, Am J Resp Crit Care Med, 2003 167A582
39
Narcotic Dosing and Time to Death After
Ventilator Withdrawal

• Pre Post p value
• (n42) (n57)
• Mean dose (mg)
• 1 hr prior vent w/d 4.0 7.0 0.07
• 1 hr post vent w/d 7.5 9.5 0.31
• Time to death (hrs)
• After vent w/d 6.2 5.3 0.60

Treece, Am J Resp Crit Care Med, 2003 167A582
40
Benzodiazepine Dosing and Time to Death After
Ventilator Withdrawal

• Pre Post p value
• (n42) (n57)
• Mean dose (mg)
• 1 hr prior vent w/d 0.1 4.3 0.001
• 1 hr post vent w/d 0.2 5.5 0.001
• Time to death (hrs)
• After vent w/d 6.2 5.3 0.60

Treece, Am J Resp Crit Care Med, 2003 167A582
41
Resources Recent Reviews

• Prendergast, Puntillo. Withdrawal of life
  support. JAMA 2002 2882732.
• Truog et al. Recommendations for end-of-life care
  in the intensive care unit. Crit Care Med 2001
  292332-46.
• Way et al. Withdrawing life support and
  resolution of conflict with families. Brit Med J
  2003 3251342.

42

• Available from Oxford University Press
• www.oup.com