Transcript June 2011 Power Point Slide Presentation

Communicating Diagnostic and
Therapeutic Information to Patients
Accounting for Lower Health Numeracy
Christopher R. Carpenter, MD, MSc, Richard T. Griffey, MD, MPH, Dan
Theodoro MD,
June 2011 Journal Club
Division of Emergency Medicine
Special Thanks to our guests
• Mary Politi
• Kim Kaphingst
Probability and Bayesian logic is confusing to everyone – not just those
with low health numeracy
Lay persons are not used to thinking beyond positive or negative test
results or therapies recommended by their doctors.
When clinicians consider test characteristics they often don’t consider
these within the context of disease prevalence
Classic Question:
A test correctly detects disease 95% of the time
(sensitivity) in people with the disease and if negative
effectively clears 90% of the patients in whom the
disease is absent (specificity)
If disease is present in 1 out of 1000 people
(prevalence) what is the probability that a randomly
chosen person who tests positive really has the
disease?
Symptom Present
Disease Status
YES
NO
TOTAL
YES
95
9990
10085
NO
5
89910
89915
TOTAL
100
99900
100000
Test characteristics:
Validity (Accuracy) & Reliability
Reliable,
Valid,
Valid and
Not Valid
Not Reliable
Reliable
Patients with
possible disease
Low Risk
Patients
Those testing
positive
Those with
actual disease
How Would You Communicate
Risks & Benefits in the Emergency
Department?
• Need to find and appraise the evidence
summary BEFORE the patient encounter
• Here’s an example using tPA for acute
ischemic stroke and the Wash U Journal Club
archives @
http://emed.wustl.edu/em_journal_club.html
http://emed.wustl.edu/emjclub_July2009_TheEvidenceSupportsThrombolyticsStroke4.5Hours.html
PGY IV Critical Appraisal
Calculating Benefit Pictographs
http://www.nntonline.net/visualrx/
From the Wash U PGY IV Critical Appraisal
What About Harm?
http://www.nntonline.net/visualrx/
PGY I
Needle Aspiration of PTX
• Meta-analysis in 2007 identified only one high
quality RCT upon which to base conclusions
(Noppen 2002 – the PGY I article at that JC)
• Noppen et al. demonstrated
– No difference in immediate success rate
• RR = 0.93 (95% CI 0.62-0.41)
– No difference in early failure rate
• RR = 1.12 (95% CI 0.59-2.13)
– Lower hospitalization rates in aspiration
• RR 0.52 (95% CI 0.36-0.75)
Case 1: PSP
Immediate Success Rate
Immediate Success Rate
Hospitalization Rate
PGY II
CT for PE
• PIOPED II provided the following test
characteristics for PE protocol CT
Case 2:
How accurate is CTA for PE?
First risk stratification by Wells (prevalence):
56% low probability
38% intermediate probability
6% high probability
So, we cannot PERC our patient but her Wells score is 0, so she is low
probability for PE….
(sidestepping d-dimer testing…)
Test characteristics:
In general, when PE is present CTA detects it 83% of the time (Sensitivity)
In general when CTA was negative it was correct 95% of the time (specificity)
Among low prob patients though, a positive CTA indicates true disease only
58% of the time (PPV) and a negative CTA is correct 96% of the time (NPV).
Diagnostic Communication
Two Concepts
• Test Accuracy
• Disease Probability
Concept #1
Test Accuracy
Sensitivity
= Has PE and CT shows it
= Has PE and CT missed it
Specificity
= Does not have PE and CT did not show a PE
= Does not have PE but CT showed a PE
If we knew you had
a PE before we
tested you
If we knew you
did not have a
PE before we
tested you
Concept #2
Disease Probability
?
Fagan Nomogram
No PE on CT
Low Risk ~3.6% mean
probability of PE
3.6% pre-CT and CT no PE = 0.67%
In other words, if 1000 patients with
these odds of having a PE had a CT
that did not demonstrate a PE, about 7
of them would still have a PE
35
Fagan Nomogram
PE found on CT
Low Risk ~3.6% mean
probability of PE
3.6% pre-CT and CT with a PE = 42.3%
In other words, if 1000 patients with
these odds of having a PE had a CT
that demonstrated a PE, about 423 of
them would actually have a PE
36
Fagan Nomogram
PE found on CT
High Risk ~66.7% mean
probability of PE
66.7% pre-CT and CT with a PE = 97.5%
In other words, if 1000 patients with
these odds of having a PE had a CT
that demonstrated a PE, about 975 of
them would actually have a PE
38
PGY III
Steroids to Prevent Recurrent Migraine
• Critical appraisal provides the RR and 95% CI
but not the control event rate so need to pull
the original paper
Control Event Rate = [22 + 10 + 8 + 18 + 20 + 43 + 20] / 353
Control Event Rate = 141/353
Control Event Rate = 0.399
PGY IV
Head CT After Blunt Trauma
• 30 year old male in tornado-related building
collapse without objective signs/symptoms of
injury: Canadian Head CT Rule
Sensitivity
= Has clinically important
brain injury and Canadian Rule shows
it
= Has clinically important
brain injury but Canadian Rule does
not show it
Specificity
= Does not have a clinically important brain
injury and Canadian Rule does not show one
= Does not have a clinically important brain
injury but Canadian Rule suggests one
Fagan Nomogram
Low-risk by
Canadian Head CT
Rule
~9% mean probability
of significant injury
before testing
9% pre-CT and Canadian Rule LowRisk = 0.3%
In other words, if 1000 patients with
these odds of having a significant
intracranial injury were low risk on the
Canadian Head CT rules, about 3of
them would still have a significant
intracranial injury
48
Fagan Nomogram
High Risk by
Canadian Head CT
Rule
~9% mean probability
of significant injury
before testing
9% pre-CT and Canadian Rule HighRisk = 16%
In other words, if 1000 patients with
these odds of having a significant
intracranial injury were non-low risk
on the Canadian Head CT rules, about
160 of them would actually have a
significant intracranial injury
49
What is EBM?
Clinical
Expertise
Research
Evidence
Patient
Preferences
51
References
• Fagerlin A, et al. Making numbers matter: Present and future research in
risk communication, Am J Health Behav 2007; 31: S47-S56.
• Houts PS, et al. The role of pictures in improving health communication: A
review of research on attention, comprehension, recall, and adherence,
Patient Educ Couns 2006: 61: 173-190.
• Barry MJ et al. Reactions of potential jurors to a hypothetical malpractice
suit alleging failure to perform a prostate-specific antigen test, J Law Med
Ethics 2008; 36: 396-402.
• Moulton B, King JS; Aligning ethics with medical decision-making: The
quest for informed patient choice, J Law Med Ethics 2010; 38: 85-97.
• Epstein RM, et al. Communicating evidence for participatory decision
making, JAMA 2004; 291: 2359-2366.
• Lipkus IM; Numeric, verbal, and visual formats of conveying health risks:
Suggested best practices and future recommendations, Med Dec Making
2007; 27: 696-713.
More References
• Fagerlin A, et al. Reducing the influence of anecdotal reasoning on
people’s health care decisions: Is a picture worth a thousand statistics?
Med Decis Making 2005; 25: 398-405.
• Fagerlin A, et al. Measuring numeracy without a math test: Development
of the subjective numeracy scale, Med Dec Making 2007; 27: 672-680.
• Lipkus IM, et al. General performance on a numeracy scale among highly
educated samples, Med Decis Making 2001; 21: 37-44.
• Woloshin S, et al. Assessing values for health: Numeracy matters, Med
Decis Making 2001; 21: 382-390.
• Politi MC, et al. Communicating the uncertainties of harms and benefits of
medical interventions, Med Decis Making 2007; 27: 681-695.
• Hoffrage U, et al. Representation facilitates reasoning: what natural
frequencies are and what they are not, Cognition 2002; 84: 343-352.
And More References
• Hoffrage U, Gigerenzer G; Using natural frequencies to improve diagnostic
inferences, Acad Med 1998; 73: 538-540.
• Loong TW; Understanding sensitivity and specificity with the right side of
the brain, BMJ 2003; 327: 716-719.
• Windish DM et al. Medicine residents’ understanding of the biostatistics
and results in the medical literature, JAMA 2007; 298: 1010-1022.
• Horowitz HW; The interpreter of facts, JAMA 2008; 299: 497-498.
• Halvorsen PA, et al. Different ways to describe the benefits of risk-reducing
treatments: a randomized trial, Ann Intern Med 2007; 146: 848-856.
• Woloshin S, et al. The effectiveness of a primer to help people understand
risk: two randomized trials in distinct populations, Ann Intern Med 2007;
146: 256-265.
• Shalowitz DI, Wolf MS; Shared decision-making and the lower literate
patient, J Law Med Ethics 2004; 32: 759-764.
Textbook References
• Kassirer JP, Kopelman RI; Learning Clinical Reasoning, Williams and
Wilkins 1991.
• Newman TB, Kohn MA; Evidence-Based Diagnosis, Cambridge University
Press 2009.
• Gigerenzer G, Calculated Risks: How to Know When Numbers Deceive
You, Simon and Shuster 2002.