Transcript file - UK Council of Clinical Communication in

Communicating Risk
Lecture for Week 4 of Endocrine
Systems module
Peter Washer
Lecturer in Communication Skills
Academic Centre for Medical Education
[email protected]
Aim
This lecture introduces the issues of risk communication. It
examines different ways of communicating risk and their
relative effectiveness in changing attitudes and behavior
and will raise questions around the ethics of risk
communication.
Learning Objectives
By the end of this lecture and SGW session that follows
you should be able to:
• Describe different models of risk
• Have an awareness of the ethical dimension of risk
communication
• Be able to describe the relative effectiveness of different
methods of communicating risk
Risk is…
Noun:
a. Hazard, danger; exposure to mischance or peril.
b. Freq. in phrase to run a or the (also one's) risk.
Verb
1. trans. To hazard, endanger; to expose to the chance of injury
or loss.
2. To venture upon, take the chances of.
Or…
…“the chances that a hazard will give rise to harm”
What are we trying to achieve in
communicating risk?
• At a population level – messages which aim to
reduce risk and improve the population’s health
• At individual level - inform people about risks,
enabling them to make their own choices
What are we trying to achieve in
communicating risk?
Are we trying to simply reduce risk and improve the
population’s health? Or are we trying to inform people,
enabling them to make their own choices, regardless of
whether this reduces risk?
Shared decision making: a middle ground between the
‘doctor / nurse / midwife knows best’ approach and
rampant health consumerism
Ethics of communicating risk
• Doctors can manipulate patients (and populations) by
framing a risk positively or negatively
• Even if doctors have explained risks to patients clearly,
the patient’s attitudes and perception of that risk will
differ
• Uncertainty changes – e.g. the risks associated with
taking hormone replacement therapy
So we need to…
• Take account of the patient’s attitudes and perception of
the risk
• Respect patient autonomy i.e. their right to make an
informed choice
• ‘Do no harm’ – beware manipulating patients by framing
a risk positively or negatively
What GMC says about information patients may want
or ought to know, before consent:
• details of the diagnosis / prognosis & likely prognosis if
the condition is left untreated;
• uncertainties about the diagnosis
• options for treatment (or non-treatment) or management
of the condition
• the purpose of a proposed investigation or treatment
including common & serious side effects;
• explanations of the likely benefits & the probabilities of
success; & discussion of any serious or frequently
occurring risks
Patient Charter
‘You have the right to have any proposed
treatment, including the risks involved in that
treatment and any alternatives, clearly explained
to you before you decide what to do’
Models of risk
There is a large social-scientific literature on the
topic of risk, with different models of risk:
• The Realist Model
• The Social Constructionist Model
Models of risk: Realist Model
Where risk is seen as an objective hazard, threat or danger
that exists 'out there' and can be measured
independently of social or cultural forces.
In the techno-scientific literature, there is often a thinly
disguised contempt for lay people's unscientific, 'correct'
knowledge about risk. The calculations the 'expert'
provides about risk tend to be treated as if they were
value-free, unbiased 'objective' facts.
Models of risk: Risk as a social construction
Social constructionists argue:
• ideas about risk are constructed both through individual
experience and by the mass media, as well as by
experts.
• why are some risks are highlighted and politicised more
than others?
• risk provides a rational, scientific, calculable explanation
for misfortune (illness, accidents etc)
• the concept of risk is used politically to attribute blame
for danger threatening a particular social group.
Giving information about risks
• What information?
• How should it be presented?
• Conducting the consultation
Giving information about risks
• What information?
• How should it be presented?
• Conducting the consultation
Giving information
•
•
•
•
•
•
How much does the patient know?
How much information do they want?
Present information appropriately
Explore patient’s views on information
Share decision making
Check understanding
Giving information about risks
• What information?
• How should it be presented?
• Conducting the consultation
Ways of representing numbers
• Positive or negative framing
• Single event probabilities
• Conditional probabilities
• Relative risks
Ways of representing numbers
• Positive or negative framing
• Single event probabilities
• Conditional probabilities
• Relative risks
Numerical representations
Positive or negative framing:
• There is a 5% chance of this surgery being fatal
• There is a 95% chance of surviving this surgery
Remember all treatments inevitably carry some degree of
risk. Also there are always at least two options (one
being non-treatment). There is no such thing as risk free
medicine.
Ways of representing numbers
• Positive or negative framing
• Single event probabilities
• Conditional probabilities
• Relative risks
Single event probabilities
Doctor says:
“There is a 30% chance of developing impotence
with this drug”
• Patient might understand 30% of people taking
the drug become impotent
or
• In 30% of my sexual encounters I may be
impotent
How might we re-phrase the doctor’s statement
to avoid this confusion?
Ways of representing numbers
• Positive or negative framing
• Single event probabilities
• Conditional probabilities
• Relative risks
Conditional Probabilities
Many diseases, e.g. cancers, have screening programmes
where individuals at risk are encouraged to be tested to
see if they have the early stages of the disease.
The chance of a particular screening test (mammograms,
prostate specific antigen [PSA] etc) actually detecting the
disease is known as the sensitivity of the test.
This is typically communicated as a conditional probability.
Conditional Probabilities
A patient has a positive mammogram. Her doctor has the
following data:
“The probability that a woman has breast cancer is 0.8%. If
she does have breast cancer, the probability that a
mammogram will show a positive result is 90%. If a
woman does not have breast cancer, the probability of a
positive result is 7%. ”
What is the probability that she actually has breast cancer?
(In other words, what is the positive predictive value, the
PPV, of the test?)
Translating Conditional Probabilities into Natural
frequencies
Translate the information on the previous slide into natural
frequencies:
“ [ ] out of every 1000 women have breast cancer. Of
these [
], [
] will have a positive result on
mammography. Of the [ ] who do not have breast
cancer some [ ] will still have positive mammograms.”
Again take for example the woman who has a positive
result. What is the probability that she actually has
breast cancer?
Using natural frequencies
• 8 out of every 1000 women have breast cancer
• Of these 8 , 7 will have a positive result on
mammography.
• Of the 992 who do not have breast cancer some 64.9 will
still have positive mammograms.”
Ways of representing numbers
• Positive or negative framing
• Single event probabilities
• Conditional probabilities
• Relative risks
Relative risk reduction
How would you interpret the following statement?
“Mammography screening reduces a woman’s risk of dying
from breast cancer by 25%”.
Relative risk reduction
“Mammography screening reduces a woman’s risk of dying
from breast cancer by 25%”.
Treatment
Deaths per 1,000 women
No mammography screening
4
Mammography screening
3
Relative risk reduction
“Mammography screening reduces a woman’s risk of dying
from breast cancer by 25%”.
But this 25% figure represents an absolute risk reduction of
only 1 in 1000 or 0.1 % because:
– Of 1000 women who do not undergo mammography,
about 4 will die from breast cancer in 10 years
– Whereas out of those who do undergo mammography,
about 3 in 1000 die
Relative Risks Exercise 1:
Mr X is a patient who has high cholesterol. He has read an
article in the newspaper that says that “People with high
cholesterol can reduce their risk of death by 22 per cent
by taking a cholesterol reducing drug”.
His GP has the following information table from the clinical
trials of the drug.
Treatment
Deaths per 1000 people with
high cholesterol over 5 years
Cholesterol reducing drug
32
Placebo
41
Try and answer the following yourselves:
What does “reduce their risk of death by 22%” mean?
What is the absolute risk reduction (as a %)?
What is the number of people who would need to treated to
save one life?
How might his GP put this information into natural
frequencies?
Answers:
“Reduce their risk of death by 22%” means the mortality
reduction was from 41 to 32 in every 1000 (9/41 =
22%)
The absolute risk reduction is 9 in 1000 or 0.9%
The number of people who would need to treated to save
one life is 111 (roughly 9 in 1000 deaths are prevented
using the drug)
Translating data into information
How might his GP put this information into natural
frequencies?...
“If you imagine 1000 people with high cholesterol like
yourself. Without any medication, you would expect 41 of
them to die from related causes like a heart attack over a
five year period. However, if these 1000 people were
taking a cholesterol lowering drug for the five year
period, then only 32 of them would be expected to die
from related causes.”
Relative Risk Exercise 2: change into natural
frequencies
The standard test for colorectal cancer is the faecal occult
blood test (FOTB). For symptom-free people over 50
screened using this test, the probability that one of these
people has colorectal cancer is 0.3%. If they do have
cancer, there is a 50% probability that they will have a
positive FOTB. If they don’t, the probability of a positive
test is 3%. Imagine a person over 50 with no symptoms
who has a positive test. What is the probability that they
have cancer?
Exercise 2: Natural frequencies format
[
] out of every [
] people have colorectal cancer. Of these [
]
people with colorectal cancer, [ ] will have a positive FOTB test
result. Of the remaining [
] people without cancer, [
] will still
have a positive FOTB. So of [
] people who have a positive test,
only [ ] have cancer, which is a probability of [
]% or 1 in [
].
[
[
[
] people
] colorectal cancer
] positive
[
[
] negative
[
] no colorectal cancer
] positive
[
] negative
Exercise 2: Natural frequencies format
Thirty out of every 10,000 people have colorectal cancer. Of these 30
people with colorectal cancer, 15 will have a positive FOTB test
result. Of the remaining 9,970 people without cancer, 300 will still
have a positive FOTB. So of 315 people who have a positive test,
only 15 have cancer, which is a probability of 4.8% or 1 in 20.
10,000 people
30 colorectal cancer
15 positive
15 negative
9,970 no colorectal cancer
300 positive
9670 negative
Giving information about risks
• What information?
• How should it be presented?
• Conducting the consultation
Conducting the consultation:
• Avoid using descriptive terms only (e.g. ‘low risk’)
• Offer both positive and negative outcomes
• Use absolute numbers or actual numbers wherever
possible, rather than relative risks
• Use natural frequency format rather than percentages
i.e. one in 5 people rather than 20%
• Use consistent denominators e.g 40 out of 1000 versus
5 out of 1000 rather than1 in 25 and 1 in 200.
• Use visual aids where available (e.g. pie charts)
Some general rules:
• Patients should be able to trust the information
they are being given
• Show a competent and caring approach,
• Explore the significance of the risk to the
individual.
• Be honest and acknowledge and share your
uncertainty
Learning Objectives
By the end of this lecture and SGW session that follows
you should be able to:
• Describe different models of risk
• Have an awareness of the ethical dimension of risk
communication
• Be able to describe the relative effectiveness of different
methods of communicating risk
References and Further reading:
Key reading:
Gigerenzer and Edwards (2003) Simple tools for understanding risks.
British Medical Journal 327 (7417) p741-744
Further reading:
Gigerenzer G (2002) Reckoning with Risk. London, Penguin.
(Recommended but only one copy in library)
The British Medical Journal 327 (7417) at
http://bmj.bmjjournals.com/content/vol327/issue7417/
Berry, D. (2004). Risk, communication and health psychology.
Maidenhead: Open University Press.
Department of Health Risk Research Web pages:
http://www.dh.gov.uk/PolicyAndGuidance/HealthAndSocialCareTopi
cs/RiskResearch/
HIV information taken from:
http://www.thebody.com/Forums/AIDS/SafeSex/Archive/PreventionS
exual/Q93372.html