Prognostication

Question 1

Importance of prognostication

Answer 1

1. Provides patients and families with important information to set meaningful goals, priorities, and expectations for care
2. Key technical prerequisite for many clinical decisions
3. Determines eligibility (prognosis <6 months) for the EI Compassionate Care Benefit and admission to inpatient units in other countries
4. Important for design and analysis of clinical trials

Question 2

Dysfunctional norms of prognostication

Answer 2

1. Avoidance
2. Wait to be asked, rather than volunteering a prognosis (especially if the clinical situation is atypical)
3. Be optimistic (especially if the patient is optimistic)
4. Avoid being specific
5. Do not use prognostication for survival in treatment decision making

Question 3

Three components of prognostication

Answer 3

1. Formulating the prognosis
2. Communicating the prognosis
3. Using the prognosis in clinical decision making

Question 4

How to formulate the prognosis (step 1 of prognostication)

Answer 4

1. Clinical prediction of survival
   - Use of subjective judgment and formulation of the prognosis by the clinician
   - May use a semi-structured approach (median survival, then adjusted for individual factors)
2. Actuarial judgment
   - Uses median survival and hazard ratios
   - Performance status is a great predictor (especially palliative performance scale)
   - Anorexia-cachexia, dyspnea, and confusion associated with shorter prognosis

Unclear which method is most accurate with prognosis

May give a continuous variable (estimation of number of days, weeks, etc.) or the probability of surviving to a certain point (e.g. % chance of being alive at 6 months)

Question 5

Prognostic value of PPS

Answer 5

• Average PPS score declines slowly over the 6 months prior to death (starting at approx 70 to 40%)
• More rapid decline in the last month of life

Question 6

Symptoms associated (and not associated) with prognosis

Answer 6

Predictive of survival

• Self-rated health measures
• Poor mood
• Dyspnea
• Confusion
• Anorexia/cachexia (strongest association)
• Comorbidities (especially in critically ill cancer patients and cancers with a long natural history)
• BNP in CHF

Not predictive

• Pain
• QOL scores

Question 7

SUPPORT study

Answer 7

• SUPPORT model uses a complex algorithm to give a probability of a patient being alive in 2-6 months time
• In the study used for the basis of the data, only some patients had cancer (limiting utility to palliative care)
• First study to show the potential of using actuarial judgment for prognostication

Question 8

Palliative Prognostic Index

Answer 8

• Score based on PS, PO intake, dyspnea, delirium, and edema

PPI > 4 predicts death within 6 weeks (PPV 83%, NPV 71%)

Question 9

Palliative Prognostic Score (PaP)

Answer 9

Score based on:

• Clinical estimation of survival
• KPS
• Dyspnea
• Lymphocyte percentage
• Anorexia

Assigns patients to risk croups with differing probabilities of being alive at 30 days

• Most widely validated, but criticised for exclusion of cognitive function and weight of subjective clinician estimates

Question 10

Prognostic tools for less seriously ill cancer patients

Answer 10

• No prognostic model for predicting survival from cancer that has been validated in the setting of an outpatient palliative are clinic
• Patients often don’t have many of the cardinal ‘end stage’ disease symptoms and may survive for several years

Question 11

Prognostic tools for CHF

Answer 11

• Prognosis is actually worse than many cancers
• NYHA classification category is best prognostic marker
• NYHA IV has 1 year mortality of 30-40%
• More accurate predictions of 6-12 month survival more difficult, due to unpredictable disease trajectory and high incidence of sudden death

Question 12

Prognosis for survival to discharge after CPR

Answer 12

• General hospitalised patient with cardiac arrest, ROSC achieved in 50%, but survival to discharge is less than 20%

Question 13

NYHA classes

Answer 13

1 - cardiac disease but asymptomatic, no limitations

2 - mild symptoms, slight limitation during ordinary activity

3 - Significant limitation, comfortable only at rest

4 - Severe limitations, symptoms at rest

Question 14

Prognostication of COPD

Answer 14

Important factors:

• Age
• FEV1
• BMI <21
• Chronic hypercapnia
• Comorbid CV disease
• Declining PS
• History of recent hospitalizations
• Dyspnea (MRC)

Short term prognosis: Severity of acute illness

Long term prognosis: Stage of COPD and comorbidities

Significant variability in prognosis is incomplete understood.

• Can also use the BODE index for prognostication and prediction of risk of hospitalization
  On BODE, Points assigned for each of:
• BMI
• Obstruction (FEV1 %)
• Dyspnea (MRC dyspnea scale)
• Exercise capacity (6 minute walk distance)

Question 15

BODE index

Answer 15

• COPD prognostication

Points assigned for each of:

• BMI
• Obstruction (FEV1 %)
• Dyspnea (MRC dyspnea scale)
• Exercise capacity (6 minute walk distance)

Question 16

Prognostication of Alzheimer’s

Answer 16

• Generally predictable decline in functional/cognitive status
• Onset of inability to walk unaided indicates final phase of illness (may be protracted)

Current criteria for admission to hospice in US:

• Dementia with impaired ADLs
• Incontinence
• Loss of ambulation
• Onset of additional major medication complication in previous 12 months (e.g. aspiration pna, UTI, decubitus ulcers)

. . . still problematic, 30% of patients who met criteria in one study were alive 3 years later!

Question 17

Communication of prognosis

Answer 17

• Clinicians tend to overestimate patient understanding re: illness and outcome, yet underestimate patient desire for information about prognosis
• Information about prognosis impacts outcomes (aggressiveness of EOL care, earlier hospice referrals, improved bereavement for family members)
• Patients are diverse in what they want from clinicians re: prognosis - important to ask

Important to still facilitate hope

• Be honest, yet avoid being blunt or giving more information than desired by the patient
• Pace information sharing
• Respect patients’ need to follow alternative paths/treatments
• Explore and facilitate realistic goals and wishes where appropriate

Question 18

Approach to sharing prognosis

Answer 18

1. Clarify what information patient desires (e.g. estimation of life expectancy, who info should be shared with, if they want to be told of changes in estimation)
2. Use a prognostic tool to estimate median survival of a group with similar characteristics
3. Explain that median survival of X months means that 50% of patients will live longer, and 50% of patients will live shorter.
4. Explain the typical survival (IQR), best case (what 10% of patients could expect - about 3 - 4x expected median), and worst case (what 10% of patients can expect, 1/6 of predicted median)