Diagnostics in EM; test thresholds and likelihood ratios Published courtesy of the CEM FOAMed Network

Diagnostics in EM; test thresholds and likelihood ratios

Published courtesy of the CEM FOAMed Networkhttp://www.cemfoamed.co.uk/portfolio/diagnostics-in-

em/

Everything we do in a patient assessment is a test

Including questions we ask

Test thresholds

Within the test thresholds;

appropriate testing

Below a test threshold;

more likely to harm than help a patient by testing

Above a test threshold;

more likely to harm than help a patient by testing

Within a test threshold;

appropriate testing

How do we use all of this information??

100 %

0%

Likelihood of diagnosis being correct

Below test threshold

Diagnosis considered

How do we use all of this information??

0%

Likelihood of diagnosis being correct

Above test threshold

Diagnosis considered

100 %

Above test threshold

How do we use all of this information??

Below test threshold

Appropriate testing

Diagnosis considered

0%

Likelihood of diagnosis being correct

100 %

Likelihood ratios

Positive Likelihood ratio;

a high number is powerful

>>1

Negative Likelihood ratio;

a small number is powerful

<< 1

Likelihood ratio Shift in probability

>10 Large (often conclusive)

5-10 Moderate

2-5 Small

1-2 Minimal

1 Nil

1-0.5 Minimal

0.5-0.2 Small

0.2-0.1 Moderate

<0.1 Large (often conclusive)

Fagan’s nomogram

Application of serial tests100 %

0%

Finding 1

Finding 2

Finding 3

Likelihood of diagnosis being correct

Above test threshold

Below test threshold

Diagnosis considered

Correct testing

Correct testCorrect circumstance

Correct patient

Perry et al. Sensitivity of coputed tomography performed within six hours of onset of headache for diagnosis of subarachnoid haemorrhage: prospective cohort study.

BMJ 2011; 343:d4277 doi. 10.1136/bmj.d4277

Han et al., The Role of Cardiac Risk Factor Burden in Diagnosing Acute Coronary Syndromes in the Emergency Department Setting. Ann Emerg Med. 2007; 49

Who not to test100 %

0%

Likelihood of diagnosis being correct

Below test threshold

Diagnosis considered

Who not to test but to treat

0%

Likelihood of diagnosis being correct

Above test threshold

Diagnosis considered

100 %

Who should be tested

Appropriate testing

Diagnosis considered

0%

Likelihood of diagnosis being correct

100 %

Correct testing

Correct testCorrect circumstance

Correct patient