Real Time EBM FINAL.ppt - Duke...
Transcript of Real Time EBM FINAL.ppt - Duke...
EBM in Real Time
Jane Gagliardi, MDReal Inpatient Doctor
Lawrence Greenblatt, MDReal Outpatient Doctor
Objectives
• To see / use EBM as a clinical tool• To see some practical ways to keep up with
evidence in your field• To see how EBM can help even if you are
not a content expert• To recognize EBM as patient-centric• To use the best available evidence in the
everyday care of patients
EBM Strategy
• Take Care of Patients• Identify Area of Uncertainty (Rx, Dx,
Prognosis, Etiology/Harm)• Formulate a Clinical Question (PICOTT)• Search Medline and find article• Critically Appraise the Article• Apply what you learned back to the patient
What Barriers Exist?
• EBM Knowledge and Skills• Time• Access to Resources• Time• Conflicting Information• Time• Inadequate Incentives• Time
EBM in Residency
• Hospital-based questions• Citations for H & P’s• Impress your team• CAT’s• Heavy focus on therapy• Supplements broad “background
learning”• Less time available than desired
EBM in Practice
• Range of settings-Inpatient, Outpatient, ER, Hospital Policy, Health Insurance, Public Policy
• Questions across domains other than therapy
• More focus on latest data• Increased emphasis on foreground• Time available is further compressed
What EBM Resources Are You Using?
• Medline• BMJ Evidence Updates/ACP Journal Wise• Cochrane Library• Guidelines.gov
Global Warming (on a small scale)
• 52 year old woman reports poor sleep for the last 3 months. She has nightly “power surges” which disrupt her sleep. No menses for 4 months. FHx: Mother had CABG at 62 (smoker). She died at 68 of AMI.
• Wants to consider HRT but is concerned about increased risk of CAD events and CAD death.
• PICOTT
Patient: PAG
• ID: 66 year old WM active, generally healthy• CC: Crushing substernal chest pain, requesting
ambulance to ER• Wife believes this may be a panic attack
Soh KC, Lee C. Panic attack and its correlation with acute coronary syndrome - more than just a diagnosis of exclusion. Ann Acad Med Singapore. 2010 Mar;39(3):197-202.
How likely do you think it is that PAG is having unstable angina or an acute coronary
event? (write down/remember your number!)
1 2 3 4 5 6 7
0% 0% 0% 0%0%0%
1. A) 0-10%2. B) 11-20%3. C) 21-30%4. D) 31-40%5. E) 41-50%6. F) >50%
Pre-test Probability
• Estimate made prior to testing of how likely it is a patient has a disease
• Where does this come from?– Clinical judgment after H&P (and other tests)– Prevalence of disorder in your population
• Epidemiologic data– Clinical manifestations of disease articles– Differential diagnosis articles– Clinical Prediction Rules
Patient: PAG• ID: 66 year old WM active, generally healthy• CC: Crushing substernal chest pain, requesting
911. Never had chest pain before this episode• PMH: HTN, Paraesophageal hiatal hernia,
GERD, Barrett’s esophagus• FH:
– Sudden cardiac death (father, MI, age 61)– DM (both brothers)– Depression, dementia (mother, AD, age 86)
• SH: Nonsmoker, 1-2 etoh drinks per month, no drugs, increased stress at work over last year
How likely do you think it is that PAG is having unstable angina or an acute
coronary event? (write down/remember your number!)
1 2 3 4 5 6 7
0% 0% 0% 0%0%0%
1. A) 0-10%2. B) 11-20%3. C) 21-30%4. D) 31-40%5. E) 41-50%6. F) >50%
What likelihood of acute coronary syndrome would you be able to tolerate and
release PAG with outpatient follow-up? (write down/remember your
number!)
1 2 3 4 5 6 7
0% 0% 0% 0%0%0%
1. 40-50%2. 20-40%3. 15-20%4. 10-15%5. 5-10%6. 0-5%
What likelihood of ACS would cause you to send PAG to the CEU for r/o MI and do stress test in the a.m.? (write down/
remember your number!)
1 2 3 4 5 6 7
0% 0% 0% 0%0%0%
1. 40-50%2. 20-40%3. 15-20%4. 10-15%5. 5-10%6. 0-5%
What likelihood of ACS would cause you to send PAG directly to the interventional
cardiac catheterization lab? (write down/ remember your number!)
1 2 3 4 5 6 7
0% 0% 0% 0%0%0%
1. 40-50%2. 20-40%3. 15-20%4. 10-15%5. 5-10%6. 0-5%
Making a Diagnosis
Zone of Action Zone of ActionZone of Uncertainty
Patient: PAG• Medications:
– Lisinopril 10 mg PO daily– HCTZ 25 mg PO daily– Aspirin 81 mg PO daily– Nexium 40 mg PO BID
• ROS: Sudden onset crushing substernal chest pain, radiating to shoulders; shortness of breath, dizziness; slight nausea, no vomiting
• PE: 95/62, 96, 24, afebrile. Uncomfortable appearing WM, breathing OK but complaining of pain. Slight tachycardia. Clear lungs. No edema.
Diagnostic Tests
Where to find Likelihood Ratios?• Go to primary sources of literature
– Make 2x2 tables– Calculate sensitivity/specificity, LR+ and LR-– Many journal articles contain LR’s!
• Go to the Rational Clinical Examination
http://www.jamaevidence.com/content/3484335
http://www.jamaevidence.com/content/3484335
66 year old male with sudden onset crushing substernal chest pain radiating to shoulders; shortness of breath, dizziness; slight nausea, no vomiting
http://www.cebm.net/index.aspx?o=1161
EXERCISE: Use Interactive NomogramUse the nomogram at CEBM.net to arrive at a post-test probability of disease
How likely do you think it is that PAG is having unstable angina or an acute
coronary event?
1 2 3 4 5 6 7
0% 0% 0% 0%0%0%
1. A) 0-10%2. B) 11-20%3. C) 21-30%4. D) 31-40%5. E) 41-50%6. F) >50%
The Rational Clinical Examination > Myocardial InfarctionSection Authors: David L. Simel, Stephen W. Goodacre, L. Kristin Newby
Goldman Chest Pain Decision Rule
Patient: PAG
How likely do you think it is that PAG is having unstable angina or an acute
coronary event?
1 2 3 4 5 6 7
0% 0% 0% 0%0%0%
1. A) 0-10%2. B) 11-20%3. C) 21-30%4. D) 31-40%5. E) 41-50%6. F) He is having an MI
Should I bother?
• Your 76 year old patient, a retired piano teacher presents for a comprehensive visit. She is fairly thin (BMI 23), controlled type 2 DM, and good functional status with no cognitive or physical limitations.
• She objects when you recommend she undergo mammography. “I’ll probably be dead by the time I’m 80 anyway.”
Heart Failure and Potato Chips• 81 year old man presents with weakness and poor
exercise tolerance for 6 months. Recently he notes modest ankle swelling and some orthopnea. Exam has some suggestion of HF (JVP, ankle edema) and echocardiogram shows new, severe systolic failure.
• At his 1st follow up visit you are reviewing treatment and he asks you if he should follow a low sodium diet to minimize symptoms and risk of dying.
• PICOTT
BREAK
Patient: PAG
• 600-mg clopidogrel load• Directly to the ICC lab• Findings:
– Slow flow, no acute coronary obstruction
Patient: PAG
• Findings:– Massive aortic dissection,
root aneurysm– Acute aortic insufficiency
• Recommendation:– Dacron aortic graft– Hypothermic circulatory
arrest
Patient: PAG
• Seeking Reputable Background Information
Patient: PAG
66 year old WM s/p St Jude aortic valve
Dacron aortic arch SVG to the RCA
Intraoperative VF LidocaineAmiodarone loadDaily amiodarone
Patient: PAG
• Postoperative day 4: bad taste in mouth– Amiodarone discontinued
• Postoperative day 6: atrial fibrillation– Digoxin started
Ask/Acquire
“Neither digoxin nor verapamil reduced the likelihood of SVAs after CABG (digoxin: OR = 0.97, 95% confidence interval [CI] = 0.62-1.49; verapamil: OR = 0.91, CI = 0.57-1.46). The likelihood of developing an SVA in patients treated with beta-blockers was markedly decreased compared with controls (OR = 0.28, CI = 0.21-0.36).”
Patient: PAG
• Does prevention of postoperative atrial fibrillation provide benefit?– Mortality– Stroke/embolic phenomena– Hospital length of stay– Incidence of atrial fibrillation
The odds of postoperative atrial fibrillation with pharmacological intervention are 4/10 the odds of postoperative atrial fibrillation
without pharmacological intervention.No real differentiation between types of intervention is possible from these data.
Patient: PAG
• Attempting to resume usual activities, PAG is told by the PA at the cardiology practice never, ever to eat anything with cranberries while taking warfarin.
• Is cranberry consumption bad with warfarin?
Ask
• P• I• C• O• T• T
Patients on warfarin anticoagulationCranberry / Cranberry JuiceNo Cranberry / Cranberry Juice???
Acquire
Appraise
• FRISBE– Follow-up– Randomization / Concealment– Intention to Treat– Similar at Baseline– Blinding– Equal Treatment outside of Intervention
How would you advise PAG regarding cranberry juice consumption?
1 2 3 4 5
0% 0% 0%0%0%
1. It’s too dangerous: never consume it
2. With weekly INR checks, it’s OK.3. Stop taking warfarin and eat/drink
what you want.4. Data for harm are not convincing
despite anecdotal reports.
Evidence-based expert in medical-
decision-making
Dr. Larry Greenblatt
Don’t Ask Me Questions!
• 66 year old professor in for annual exam. Treated for asthma, HL, prediabetes and htn.
• Your 3rd year student wants to know why you choose to treat him with amlodipine and losartan. Why not a beta blocker and diuretic as recommended in the JNC VII guidelines?
• You mock the JNC VII as being ancient and look for a more up to date evidence-based guideline.
My Elbow Hurts!• 49 year old man, very unathletic, complains
of pain over the lateral aspect of the elbow. Pain is made worse by lifting a jug of milk or carrying his briefcase. No pain with elbow movement. There is focal tenderness the area of pain. Provocative testing by resisting wrist extension reproduces the pain.
• His orthopedics PA has recommended a steroid injection. Do you concur?
• PICOTT
Patient: PAG
• Two months post-operatively (2/26):– Fanatically monitoring all Vitamin K intake– Mildly irritated at differing recommendations
from differing providers– INR finally therapeutic
• Sudden onset right flank pain– arterial embolism to R kidney
Patient: PAG
• Renal function is stable (Cr 0.9-1.3)• Aortic dissection is stable, with nearly
equal-sized false and true lumens• Vascular surgeon considers the merits of
endovascular fenestration
(Circulation. 2006;114:II_362.)
BACKGROUND INFORMATION (what are they talking about, anyway?)
• Assess – 66 year old man with impaired renal blood
flow, recent type A dissection all the way to iliacs
• Ask– P (aortic dissection, impaired renal blood flow)– I (fenestration)– C (no fenestration)– O (end-organ damage, mortality, stroke)
• Acquire• Appraise• Apply
A Multitude of Questions:Should PAG get a home automated defibrillator?
Will taking warfarin increase PAG’s risk of prostate cancer?
What about inconsistent Vitamin K in PAG’s diet?Anything to make warfarin safer?
Is a target INR 2.5-3.5 superior to a target INR of 2-3 in a patient with a St. Jude aortic valve?
What is the likelihood of a future embolic event?
What preventive measures can one take after extensive aortic dissection to reduce probability of future bad outcomes?
Is there a benefit of unfractionated heparin over LMWH when PAG’s INR drifts to 1.4?
Should first-degree relativesbe screened for aortic aneurysm?
Should PAG go back to work?
Wrap-Up
• Evidence-Based Medicine is a clinical tool• You can use the process even (especially!)
if you are not a content expert• Knowing how to use the best available
evidence in the care of patients will be useful during your entire career (and life)
• OK to let someone else do the appraisal• EBM starts and ends with the PATIENT
ASSESS
ASK
ACQUIRE
APPRAISE
APPLY
The 5 A’s
EBMCycle
MUST CONSIDER: - Patient preference- Access to care- Quality of life- Goals of care
WHAT’S GOING ON?- History and Physical- Initial Formulation
PICOTT- Patient /
Population- Intervention- Control- Outcome- Type of Question- Type of Study
LITERATURE SEARCH
VALIDITY CRITERIA- Methods- Results- Sources of Bias- Strength of evidence
Questions?Complaints?Applause?