Current issues of ACP Journal Club are published in Annals of Internal Medicine


Leukocyte differential and creatine kinase-MB levels were accurate predictors of MI

ACP J Club. 1995 July-Aug;123:20. doi:10.7326/ACPJC-1995-123-1-020

Source Citation

Thomson SP, Gibbons RJ, Smars PA, et al. Incremental value of the leukocyte differential and the rapid creatine kinase-MB isoenzyme for the early diagnosis of myocardial infarction. Ann Intern Med. 1995 Mar 1;122:335-41.



To determine the value of independent tests to diagnose myocardial infarction (MI).


A blinded comparison of 6 tests.


U.S. emergency department (ED).


511 consecutive patients ≥ 20 years of age who presented to the ED with anterior or left lateral chest pain. Exclusion criteria included recent infection; recent use of exogenous glucocorticoid; major trauma, bleeding, surgery, or dialysis; and missing data. 384 patients were eligible.

Description of Tests and Diagnostic Standard

Peripheral blood leukocyte counts were estimated to give an automated 3-part leukocyte differential (granulocytes, lymphocytes, and monocytes). Total creatine kinase (CK) activity, rapid creatine kinase-MB (CK-MB) levels, cortisol levels, and ST-segment elevations (> 0.1 mV in 2 contiguous leads) were measured. Diagnosis of MI was based on increased standard CK-MB isoenzyme levels over 48 hours.

Main Outcome Measures

Sensitivity, specificity, and predictive value of tests.

Main Results

69 patients (18%) had an MI. ST-segment elevation had a sensitivity of 39%, a specificity of 99%, and a positive predictive value of 93%, the highest of any of the individual markers. Elevated cortisol had the highest sensitivity (68%) and the lowest specificity (86%) and positive predictive value (52%). Sensitivity, specificity, and positive predictive value were 56%, 93%, and 66% for the initial rapid CK-MB levels and 42%, 93%, and 57% for the initial total creatine kinase levels. The performances of the relative lymphocytopenia (< 20.3%) and granulocytosis (> 73.7%) were similar to that of the rapid CK-MB levels, with sensitivities, specificities, and positive predictive values of 58%, 91%, 58% and 59%, 89%, and 54%. The combination of rapid CK-MB level and decreased relative lymphocyte percentage gave a sensitivity of 44%, a specificity of 99.7%, and a positive predictive value of 97%; the presence of both of these markers increased the sensitivity of MI detection from only 39% for ST-segment changes alone to 65%, while maintaining a specificity of 99%.


Although < 50% sensitive, the combination of elevated rapid CK-MB level and decreased relative lymphocyte percentage was an accurate early predictor of MI in patients presenting to the ED with chest pain, including those without ST-segment elevation.

Source of funding: Not stated.

For article reprint: Dr. R.J. Gibbons, Division of Cardiovascular Diseases and Internal Medicine, Mayo Clinic, 200 First Street SW, Rochester, MN 55905. FAX 507-284-5470.


Early diagnosis of chest pain is important for 2 reasons: It can enhance rapid treatment of MI and can streamline evaluation of and avoid hospitalization of patients without MI.

The findings of Thomson and colleagues are unlikely to directly influence treatment decisions because the current criteria for administering thrombolytic therapy or immediate angioplasty are firm: ST-segment elevation or bundle-branch block (BBB) with a high index of suspicion are the only acceptably specific noninvasive markers for total coronary occlusion. Patients without ST-segment elevation or BBB were not shown to benefit from thrombolysis; this was true even of subgroups at high risk for death, including patients with ST-segment depression (1, 2). Increased ability to do early detection of MI among patients without ST-segment elevation may aid in identifying important subgroups of patients for future trials.

Hospitalization and aggressive medical treatment, such as heparin and intravenous or oral β-blockers and nitrates, are appropriate for both MI without ST-segment elevation and high-risk unstable angina, and still should be done primarily on the basis of history, physical examination, and electrocardiogram.

To identify low-risk patients for discharge from the ED, the current standard for "chest pain units" goes beyond the initial value of CK-MB, which was analyzed by Thomson and colleagues, and extends serial evaluation to at least 9 hours. Many units continue with noninvasive testing, such as that using the exercise treadmill. The negative predictive value of the leukocyte differential is too low to be useful for this second purpose.

New markers with biologic specificity, such as troponins, which are found in myocardial cells only, appear promising in the effort to avoid the hospitalization of low-risk patients, but studies are incomplete.

Steven Borzak, MD
Henry Ford Hospital Detroit, Michigan


1. Fibrinolytic Therapy Trialists' Collaborative Group. Lancet. 1994;343:311-22.

2. TIMI-IIIB Investigators. Circulation. 1994;89:1545-56.

Author's Response

Although thrombolytics are currently not indicated in patients without ST-segment elevation or BBB, relatively few patients with ST-segment depression have been studied (1). The lack of evidence of benefit should not be misinterpreted as evidence of lack of benefit. Further studies clearly are needed and possibly should focus on patients identified by our approach.

Raymond J. Gibbons, MD