Myocardial infarction: ECG and biochemical markers can help diagnose and exclude it

Clinical bottom line (level 2a)

  1. One in five patients presenting with clinical features suggesting acute cardiac ischaemia have a myocardial infarction.
  2. The following tests make a myocardial infarction more likely
    • abnormal pre-hospital ECG
    • elevated CK-MB
    • elevated myoglobin
    • elevated troponin T or I
  3. The following tests make a myocardial infarction less likely
    • negative Goldman chest pain protocol
    • normal serial CK-MB
    • normal serial myoglobin
    • normal serial troponin T or I
    • normal rest echocardiogram
    • normal technetium-99m sestamibi imaging
Lau et al: AHRQ Publication No. 01-E006, Rockville, MD 2201; : -
Lau et al: Ann Emerg Med 2001; 37 : 453-460
Balk et al: Ann Emerg Med 2001; 37 : 478-494
Expires March 2004

The study

Systematic review of all diagnostic studies of
  • Patients: suspected acute coronary ischaemia attending emergency departments
  • Intervention: ECG, blood tests, stress testing
  • Outcome: acute coronary ischaemia

Articles found in English using Medlnie, 1966 to January 1999 (search terms: Detailed in text, but included "chest pain," "myocardial ischemia" or "infarction," "emergency," and "emergency service." ) and and retrieved studies used in a previous report, searched references of published meta-analyses and selected review articles, and consulted technical experts

Selection criteria: see above and below
Appraisal criteria: Prospective or retrospective cohorts, blinding and reference standard used. . Data were abstracted either independently by two members or by one member and then verified by a second member.
Articles excluded if:
  • no clear reference to emergency department settings
  • populations with special comorbidities (e.g., patients with renal disease)
  • aged < 18


407 articles found
  • 10 on ECGs involving 4481 patients
  • 3 on rest echocardiography involving 397 patients
  • 5 on techetium-99m sestamibi imaging involving 1571 patients mostly with normal/non-diagnostic ECGs and normal serial cardiac markers
  • 12 on creatine kinase at presentation involving 3195 patients
  • 2 on serial creatine kinase involving 786 patients
  • 19 on CK-MB at presentation involving 6424 patients
  • 14 on serial CK-MB involving 11625 patients
  • 18 on myoglobin at presentation involving 4172 patients
  • 10 on serial myoglobin involving 1277 patients
  • 4 on troponin I at presentation involving 1149 patients
  • 2 on serial troponin I involving 1393 patients
  • 6 on troponin T at presentation involving 1348 patients
  • 3 on serial troponin T involving 904 patients
  • 3 on CK-MB and myoglobin at presentation involving 2283 patients
  • 2 on serial CK-MB and myoglobin involving 291 patients
  • 3 on the Goldman chest pain protocol involving 5359 patients

Where possible data was combined - using a random-effects model.

The evidence


diagnostic test number of patients sensitivity for
myocardial infarction
(95% CI)		 specificity for
myocardial infarction
(95% CI)		 LR+ LR-
prehospital ECG 76%
(54% to 89%) 		88%
(67% to 96%) 		6.3 0.27
non-standard ECG 59-83%
(% to %) 		76-93%
(% to %) 		2.5-12 0.18-0.54
rest echocardiography 93%
(81% to 97%) 		66%
(43% to 83%) 		2.7 0.11
technetium-99m sestamibi imaging 92-100%
(% to %) 		49-84%
(% to %) 		1.8-3.8 0.0-0.1
creatine kinase (presentation) 37%
(31% to 44%) 		87%
(80% to 91%) 		2.8 0.72
creatine kinase (serial) 69-99%
(% to %) 		68-84%
(% to %) 		2.2-6.2 0.012-0.46
CK-MB (presentation) 42%
(36% to 48%) 		97%
(95% to 98%) 		24 0.59
CK-MB (serial) 79%
(71% to 86%) 		96%
(95% to 97%) 		20 0.21
myoglobin (presentation) 49%
(43% to 55%) 		91%
(87% to 94%) 		5.4 0.52
myoglobin (serial) 89%
(80% to 94%) 		87%
(80% to 92%) 		6.8 0.13
troponin I (presentation) 39%
(10% to 78%) 		93%
(88% to 97%) 		5.6 0.66
troponin I (serial) 90-100%
(% to %) 		83-96%
(% to %) 		5.3-25 0.0-0.12
troponin T (presentation) 39%
(26% to 53%) 		93%
(90% to 96%) 		5.6 0.66
troponin T (serial) 93%
(85% to 97%) 		85%
(76% to 91%) 		6.2 0.082
CK-MB and myoglobin combination (presentation) 83%
(51% to 96%) 		82%
(68% to 90%) 		4.6 0.20
Goldman chest pain protocol 88-91%
(% to %) 		70-74%
(% to %) 		2.9-3.5 0.12-0.17
total

  • mean pre-test probability of myocardial infarction with
    • any clinical feature suggestive of acute cardiac ischemia: 20% (95% CI: 15% to 25%)
    • chest pain: 28% (95% CI: 24% to 31%)

Comments

  1. By limiting the search strategy to English and Medline, important articles may have been missed.
  2. Few studies blinded both test and reference standard.
  3. Note CK is used as part of the WHO acute myocardial infarction reference standard. Consequently the diagnostic test results for CK appear better than they are in reality.

Citation

  1. Lau J, Ioannidis J, Balk E, et al: Evaluation of Technologies for Identifying Acute Cardiac Ischemia in Emergency Departments. Evidence Report/Technology Assessment Number 26 (Prepared by The New England Medical Center Evidence-based Practice Center under Contract No. 290-97-0019). AHRQ Publication No. 01-E006, Rockville, MD 2201; : -
  2. Lau J, Ioannadis JP, Balk EM, et al: diagnosing acute cardiac ischemia in the ermegency department: a systematic review of the accuracy and clinical effect of current technologies. Ann Emerg Med 2001; 37 : 453-460
  3. Balk EM, Ioannadis JP, Salem D, et al: accuracy of biomarkers to diagnose acute cardiac ischemia in the emergency department: a meta-analysis. Ann Emerg Med 2001; 37 : 478-494
Search Terms: from AHRQ website
Contributor: Chris Ball, March 2002
Reviewer:

Clinical Question.
Patient suspected myocardial infarction
Intervention or Exposure ECG, cardiac enzymes, Goldman chest pain protocol
Outcome myocardial infarction