Protocol for Evaluating Patients with Chest Pain in the Emergency Department

 

Data requirements:

(1) history and physical examination

(2) any prior electrocardiograms, for comparison

(3) an electrocardiogram in the Emergency Department

Diagnostic Group	Management
MI possible (probability of MI >= 7%)	admit to CCU
non-MI (probability of MI < 7%)	do not admit to CCU

 

Question A1:

Are there ST elevations or Q waves in 2 or more leads, not known to be old?

• If Yes, then MI possible (patient group N)

• If No, then proceed to Question 2

 

Question A2:

Did chest pain begin 48 or more (>= 48) hours ago?

• If Yes: Are there ST-T changes of ischemia or strain, not known to be old?

• If Yes, then MI possible (patient group M)

• If No, then non-MI (patient group L)

• If No, then proceed to Question 3.

 

Question A3:

Does the patient have a prior history of myocardial infarction and/or angina?

• If Yes, then proceed to Question B1

• If No, then proceed to Question C1

 

Question B1:

Are there ST-T changes of ischemia or strain, not known to be old?

• If Yes, then MI possible (patient group K)

• If No, then proceed to Question B2.

 

Question B2:

Is the longest episode of pain 1 hour or more?

• If Yes, then proceed to Question B3.

• If No, then non-MI (patient group H)

 

Question B3:

Is the pain worse than prior angina or the same as a prior MI?

• If Yes, then MI possible (patient group J)

• If No, then non-MI (patient group I)

 

Question C1:

Does pain radiate to neck, or left should, or left arm?

• If Yes, then proceed to Question C2.

• If No, Are ST-T changes of ischemia or strain present, not known to be old?

• If Yes, then MI possible (patient group B)

• If No, then non-MI (patient group A)

 

Question C2:

Is the patient 40 years of age or older?

• If Yes, If yes, then proceed to Question C3.

• If No, then non-MI (patient group C)

 

Question C3:

Can the chest pain be reproduced by palpation?

• If Yes, then non-MI (patient group D)

• If No, proceed to Question C4.

 

Question C4:

Does the pain radiate to back, abdomen or legs?

• If Yes, then non-MI (patient group E)

• If No, then proceed to Question C5.

 

Question C5:

Is the chest pain "stabbing" in nature?

• If Yes, then non-MI (patient group F)

• If No, then MI possible (patient group G)

 

Patient Group	Diagnosis	percent MI in retrospective and prospective protocol use
A	non-MI	2%
B	possible MI	26%
C	non-MI	2%
D	non-MI	1%
E	non-MI	8%
F	non-MI	2%
G	possible MI	17%
H	non-MI	4%
I	non-MI	1%
J	possible MI	11%
K	possible MI	26%
L	non-MI	2%
M	possible MI	21%
N	possible MI	75%

 

Performance by the protocol was similar to that of physicians presented with the cases:

• sensitivity (with MI and admitted to CCU): 88%

• specificity (without MI and not admitted to CCU): 74%

• positive predictive value: 32%

• overall accuracy: 76%

 

Limitations:

• The protocol was developed prior to ready availability of cardiac enzyme markers and other newer diagnostic modalities. However, it still may be of use when such resources are not available.

 

Implementation Note: Due to the complexity of expressing a decision tree completely in Boolean terms, key branch points were identified and assigned roman numerals (0, I, II, III, IV, V, VI) that would serve as summaries of decisions made up to that point.