Transcript Simple.

THE SECTION ON
Evidence Based Health Care
OF THE NEW YORK ACADEMY OF
MEDICINE
2014 TEACH Workshop
Simple.
Scientifically Informed Medical Practice and
Learning
Suzana Alves da Silva, MD, MSc, PhD, FNYAM
“Things must be as simple as
possible but not any simpler”
Albert Einstein
2
Clinical
P
I Problems
C
O
Therapy
Diagnosis
Prognosis
Harm
What we know?
Chatterji, M; Graham, M; Wyer, P. Mapping Cognitive Overlaps Between Practice-Based Learning
and Improvement and Evidence-Based Medicine: An Operational Definition for Assessing Resident
Physician Competence. 2009
3
Harm
on outcome
Utility
Impact
of outcome
Prognosis
Performance
Prediction
Diagnosis
of outcome
Therapy
Frequency
Likelihood
What is new?
Silva, S and Wyer P. The Roadmap: a blueprint for evidence literacy within a Scientifically
Informed Medical Practice and Learning Model. 2012
4
Why this is so important?
5
• An elderly man, with dementia, was brought to the
emergency room from a nursing home with ST elevation
acute myocardial infarction. He was submitted to PCI,
but evolved with cardiogenic shock and acute renal
failure. The daughter wants to know “So, if you give him
the most intense treatment, how likely is it that he will
survive? It doesn’t make sense if he is likely to die
anyway!”
Likelihood issue
This scenario converges on an issue of the absolute probability of survival given
maximal, evidence-based, interventions.
6
Elderly Cardiogenic
Shock and
STEAMI
Renal
PCI
Failure
How
many
died?
What is the death rate?
7
• A 40 year old woman is seeing her primary care
physician. Her mother has recently been diagnosed with
Alzheimer’s disease. She is concerned about the
possibility that it could happen to her. The physician
advises that there is a genetic test that could eliminate
that likelihood. However, the woman expresses
ambivalence, saying, “I am not sure I want to know.”
Utility issue
This scenario involves a trade-off between risks versus benefits of sharing
the results of a prognostic information.
8
Testing
Actions
Consequences
Not
Testing
Actions
Consequences
Utility issue
9
• A medical director of a health care organization is
considering inclusion of CYP2C19 within a panel of
molecular biomarkers for optimization of individualized
care of patients submitted to stent implantation that will
be submitted to antiplatelet therapy with clopidogrel.
Performance issue
This scenario involves predicting the likelihood of benefit from a treatment if the predictor is
present. It calls for information regarding the predictive performance of criteria such as a
biomarker in modifying the probability of benefit from a therapeutic option.
10
BMJ 2011
11
Utility
Does it worth
to use?
Performance
Does it work?
Frequency
Is it relevant?
12
Levels of efficacy of Dx
imaging tests
• diagnostic accuracy, sensitivity, and
specificity associated with
interpretation of the test.
Perfor• whether the information produces
mance change in the referring physician’s
diagnostic thinking.
Utility• effect of the information on the
patient management plan and on
patient outcomes.
• societal costs and benefits of a
diagnostic imaging technology.
Fryback and Thornbury. 1991
Phases of novel cardiac
markers
• 1. Proof of concept—Do novel
marker levels differ between
subjects with and without outcome?
• 2. Prospective validation—Does the
novel marker predict development
of future outcomes?
• 4. Clinical utility—Does the novel
risk marker change predicted risk
sufficiently to change
recommended therapy and to
improve clinical outcomes?
• 6. Cost-effectiveness—Does use of
the marker justify additional costs
of testing and treatment?
Hlatky. AHA 2009
13
Likeli
hood
Perfor
mance
Utility
“My doctors estimated that I had an 87
percent risk of breast cancer and a 50
percent risk of ovarian cancer… Once I
knew that this was my reality, I decided to
be proactive and to minimize the risk as
much as I could.”
14
HMOs being pressured for the inclusion of BRCA mutation testing
for screening of women with high risk of breast cancer.
Chief Medical Executive
Healthy Patient
Many patients with family
history of cancer want to be
tested for BRCA mutation.
Many wonder about preventive
surgery or chemotherapy on
top of surveillance.
Scenario
Physician
Many primary care
physicians are ordering
BRCA mutation testing to
their patients or referring
them for genetic
counseling without
knowing exactly what to
do with it.
15
Executive
Physician
Patient
BRCA1 for screening
16
Executive
Physician
Patient
What is the uncertainty of
the information on the risk of
breast cancer? Does it apply
to this patient?
What is my risk of
breast cancer?
P
How accurate is BRCA1
mutation to predict breast
cancer in such groups of
women?
For how many times the
presence of BRCA1
increases my risk of
breast cancer?
If we tested them all for BRCA1
U
mutation, would it decrease
costs?
Preventive mastectomy
would decrease the risk? If
so, for how much?
Should I do preventive
mastectomy to minimize
my risk?
How many women with high
risk of breast cancer? How
L
likely is the development of
breast cancer if BRCA1(+) and
(-)? How many litigations?
BRCA1 for screening
17
Breast Cancer || Ovarian Cancer
Women in general
Preventive surgery will drop the risk of
cancer by ~ 90%
Little info about the effect of
chemoprevention or close surveillance
Little info about impact on death
Source: 1. Senkus. European Society of Medical Oncology. 2014. 2. Risk Assessment, Genetic Counseling,
and Genetic Testing for BRCA-Related Cancer: U.S. Preventive Services Task Force Recommendation. 2013.
3. Berliner. National Society of Genetic Counselors 2013
18
• Recognition of questions of importance to patients,
managers and policy makers in a fashion that facilitates
the use of clinical research to inform the corresponding
decisions.
• Provide guidance regarding how to explore and interpret
patient and stakeholder expressions of the concerns that
will maximally inform decisions.
• Categorization of clinical research designs in relationship
to different types of clinical questions in a way that
embraces the full spectrum of clinically relevant research.
Why this is so important?
19
Why this is so important?
20
Thank you!
21
Therapy Utility
Therapy Frequency
P
In high risk patients,
BRCA1+, for how much
In BRCA1+ patients
I
preventive surgery
who are submitted to
preventive surgery
C
compared to surveillance
O
decreases patient important
outcomes?
what is the expected
likelihood of cancer and
death?
22
Prognosis Utility
Prognosis Frequency
P
In high risk patients,
In high risk patients,
I
does BRCA mutation
testing
who are BRCA1+ carriers
C
compared to usual care
O
decrease patient important
outcomes?
what is the expected
likelihood of cancer and
death?
23
24
= Randomized Trials
25