At the Bedside
Evidence Based Medicine
Stephen R. Hayden, MDDepartment of Emergency
MedicineUCSD Medical Center, San
Diego
Teaching
The best teaching is taught by patients themselves
Sir William Osler
EBM at the Bedside
At the bedside, use history and physical exam elements for teaching EBM
Take an item of history or physical exam and think of it as a “diagnostic test”
Presence or absence of a clinical finding changes the probability of disease
EBM at the Bedside Opportunity to discuss many EBM concepts Test properties of clinical exam parameters
Precision (kappa) of clinical examination Accuracy (likelihood ratios, PPV, NPV)
Moving from pretest to post test probability Quantifies the utility of diagnostic tests
Example
How often do you see a case of chest pain in the emergency department?
How precise are clinical findings in chest pain patients?
Interrater reliability (Kappa)
Precision of Clinical Features
Clinical Feature Kappa
Chest pain radiates to L arm
0.89
Pain in substernal location 0.74
Pain described as pressure 0.57
Pain described as sharp 0.30
Pain with movement 0.27
Hickan DH, et al. J Chronic Dis. 1985;38:91-100
Precision of Physical Findings
Physical signs of heart failure in MI
Gadsboll N. European Heart J. 1989;10:1017-1028
Clinical Feature Kappa
Dyspnea 0.62 - 0.75
Neck vein distension 0.31 – 0.51
Dependent edema 0.27 – 0.64
Third heart sound 0.14 – 0.37
Rales 0.12 – 0.31
How accurate are clinical findings in chest pain patients?
Accuracy of Clinical Features
Clinical Feature Positive LR (CI)
Radiation to left armRadiation to right shoulderRadiation to both L and R arm
2.3 (1.7-3.1)2.9 (1.4-6.0)7.1 (3.6-14.2)
Third heart sound 3.2 (1.6-6.5)
Hypotension 3.1(1.8-5.2)
Diaphoresis 2.0 (1.9-2.2)
Nausea or vomiting 1.9 (1.7-2.3)
Past history of MI 1.5-3.0Panju AA, et al. JAMA. 1998;280:1256-1263
Accuracy of Clinical Features
Clinical Feature Negative LR (CI)
Pleuritic chest pain 0.2 (0.2-0.3)
Chest pain sharp or stabbing
0.3 (0.2-0.5)
Positional chest pain 0.3 (0.2-0.4)
Chest pain reproduced by palpation
0.2 - 0.4
Panju AA, et al. JAMA. 1998;280:1256-1263
Can this really be done in a busy ED?
EBM at the Bedside
Don’t attempt to answer all possible questions for every patient
Pick one clinical finding relevant to a individual patient
Choose cases you see frequently in ED
EBM at the Bedside
Requires advance preparation Have Kappa’s, likelihood ratios
with you on index cards, palm pilot, workstation
Need rapid access to high quality evidence
Medcalc3000.com
http://pbrain.hypermart.net/medrules.html
(Freeware)
BestBets.org
Analgesia and Abdominal Pain
http://nhscrd.york.ac.uk/darehp.htm
ACP Journal Club
How do you find articles relating to the precision and accuracy of the H&P?
Search Tips
Add specific terms to search strategy
“Physical examination” “Medical history taking” “Sensitivity” or “specificity” “Clinical assessment” “Observer variation” “Interrater reliability”
MI Reference
Panju AA, et al. Is this patient having a myocardial infarction? JAMA. 1998;280:1256-1263
Summary
Reviewed how to take EBM to the bedside Identifying elements of the H&P as “tests” Describing precision and accuracy of H&P Preparing in advance / bedside tools Rapid access to pre-appraised resources
“I desire no other epitaph than the statement that I taught medical
students in the wards, as I regard this by far the most useful and
important work I have been called upon to do.” Sir William Osler
Farewell Address, 1905