Prediction of Early Complications in Patients With Acute Myocardial Infarction by Calculation of the ST Score

Prediction of Early Complications in Patients With Acute Myocardial Infarction by Calculation of the ST Score

Marianne Gwechenberger MD
Wolfgang Schreiber MD
Harald Kittler MD
Michael Binder MD
Bernhard Hohenberger
Anton N Laggner MD
Michael M Hirschl MD


Clinic of Internal Medicine II, Department of Cardiology Department of Emergency Medicine, University of Vienna, Vienna, Austria.


47/1/85428

Study objective:

To assess the relationship between the sum of ST-segment elevations (ST score) in the admission ECG and the occurrence of early complications in patients with acute myocardial infarction (MI).
Methods:

We conducted an observational study of patients who presented with acute anterior or inferior MI to the ED of a 2,000-bed inner-city hospital. Age, sex, time from onset of pain and the start of thrombolysis, and ST score were evaluated by the emergency physician. "Early complications" were defined as acute congestive heart failure or severe rhythm disturbances in the 24 hours after the start of thrombolysis. The outcome measures were the relationship between ST score and the occurrence of early complications; the influence of age, sex, or time between onset of pain and thrombolysis; and identification of a cutoff value with the highest sensitivity and specificity for prediction of complications.
Results:

We included 243 patients (194 men, 49 women; mean age, 56.6 years) with acute MI (anterior, 119; inferior, 124) who underwent thrombolysis in our analysis. ST score was significantly greater in patients with early complications, compared with patients without complications (anterior, 10.3 versus 19.4 mm [ P<.001]; inferior, 6.9 versus 10.4 mm [ P<.001]). Receiver-operator curve analysis revealed an ST score of 13 mm in patients with anterior MI and 9 mm in patients with inferior MI as the cutoff value with the greatest sensitivity and specificity for predicting early complications of MI. (For anterior MI, sensitivity was .79, specificity .73; for inferior MI, sensitivity was .64 and specificity .68.). On multivariate regression analysis, ST score was an independent predictor of the occurrence of at least one complication. (For anterior MI, the odds ratio [OR] was 9.7 and the 95% confidence interval [CI] 3.9 to 25.1; for inferior MI the OR was 5.0 and the 95% CI 2.0 to 12.8). Age, sex, and interval from onset of pain to treatment had no significant effect on the occurrence of early complications.
Conclusion:

The absolute ST score is useful in estimating the probability of early complications in patients with acute MI receiving
564
thrombolytic therapy. A cutoff value of 13 mm for anterior MI and 9 mm for inferior MI stratifies patients into high- and low-risk subgroups for the development of acute congestive heart failure and severe rhythm disturbances during the first 24 hours of hospitalization.

[Gwechenberger M, Schreiber W, Kittler H, Binder M, Hohenberger B, Laggner A, Hirschl MM: Prediction of early complications in patients with acute myocardial infarction by calculation of the ST score. Ann Emerg Med November 1997;30:563-570.]


Received for publication February 28, 1997.
Revision received June 20, 1997.
Accepted for publication July 3, 1997.

Copyright © by the American College of Emergency Physicians.

Address for reprints:
Michael M Hirschl, MD
Department of Emergency Medicine
Wahringer Gurtel 18-20
A-1090 Vienna
Austria
43-1-40400-1964
Fax 43-1-40400-1965
E-mail michael.hirschl@akh-wien.ac.at
INTRODUCTION

The initial ECGs of patients admitted to the ED or CCU have been used to stratify patients into groups at low and high risk for the development of in-hospital life-threatening complications and death. [1] [2] [3] [4] [5] Whereas signs of myocardial infarction (MI), left bundle-branch block, and left ventricular hypertrophy on the initial ECG are associated with a high frequency of adverse events or the need for procedures, [6] normal ECG findings or an ECG with minimal changes indicate a low risk of further complications in patients with suspected MI. [7] In addition, it has been demonstrated that the presence of ST-segment elevation in lead V4R in patients with acute inferior MI is a strong and independent predictor of major complications. [8] [9]

In the last 10 years, thrombolytic therapy has become an important therapeutic tool in patients with acute MI, as survival of patients with acute MI has markedly improved as a result of thrombolysis. [10] [11] However, especially in the early phase after acute MI (within 24 hours of the onset of pain), acute complications including congestive heart failure or rhythm disturbances are commonly observed. Easily available markers for the prediction of these complications would be helpful in stratifying patients with acute MI undergoing thrombolysis into high- and low-risk groups at the time of admission to the ED or CCU. Birnbaum et al [12] demonstrated that different patterns of ST-segment changes in the initial ECG are associated with different prognoses in patients who have sustained acute MI. In contrast to the assessment of the initial ECG pattern, the amount of ST-segment elevation (ST score) seems a more easily available parameter for the treating physician in the ED or CCU. [13] The predictive value of the ST score for early complications in patients with acute MI undergoing thrombolytic therapy has not been investigated until now. We therefore designed a study to assess (1) the relationship between ST score on the admission ECG and occurrence of early complications and (2) the influence of other factors (eg, interval from onset of pain to treatment, age, sex) on this relationship in patients with acute MI undergoing thrombolytic therapy.
MATERIALS AND METHODS

We conducted this study in the ED of the General Hospital (Vienna, Austria) between January 1, 1994, and December 31, 1996. Data from all patients who were admitted to the ED with evidence of an acute MI and given thrombolytic therapy were prospectively collected.

The study protocol was renewed by the local institutional review board and found to be in accordance with the ethical standards of the review board and with the Helsinki Declaration of 1975, as revised in 1983.

In this study we sought to evaluate the relationship between ST score on the admission ECG and the occurrence of early complications and the influence of other factors (eg, interval between the onset of pain and treatment, age, and sex) on this relationship in patients with acute MI who were undergoing thrombolytic therapy.

All patients with suspected MI (eg, chest pain of >30 minutes' duration, persistent ST-segment elevation in two or more anterior or inferior leads on the admission ECG) who were undergoing thrombolytic therapy were eligible for study. Consecutive ECGs and serial creatine kinase-MB (CK-MB) determinations were performed to confirm the diagnosis of acute MI (ie, evidence of new Q waves in serial ECGs, a typical increase and decrease (or both) of the CK-MB concentration in the first 24 hours after ED admission).

Patients in whom Q waves did not develop (.04 seconds or longer) and those in whom the typical increase and decrease in CK-MB concentration did not appear were excluded from further analysis. Patients with evidence of early complications on admission (eg, acute congestive heart failure on the initial chest radiograph or severe rhythm disturbances on the initial ECG) were also dismissed from analysis.

In summary, the inclusion criteria for the study were evidence of acute MI, treatment with a thrombolytic agent, and no signs of early complications at the time of ED admission.

After admission, each patient received thrombolytic treatment with 100 mg recombinant tissue plasminogen activator (rtPA) in accordance with an accelerated front-loading scheme published by Neuhaus et al. [11] Treatment was preceded with an intravenous bolus of 5,000 IU conventional heparin and an oral 100-mg dose of aspirin. beta-Blockers and nitrates were given as deemed clinically appropriate.

Standard 12-lead electrocardiographic tracings were performed immediately after admission. The records were made at a paper speed of 25 mm/second and standardization at 1.0 mV to 1.0 cm. Infarct location was determined
565
on admission, and patients were classified a priori as having anterior MI (ST-segment elevation in leads V1 through V6) or inferior MI (ST segment elevation in leads II, III, and aVF). A small proportion of patients exhibited ST-segment elevation in both anterior (V1 through V6, I, aVL) and inferior leads (II, III, aVF, V5, V6). These patients were classified as having anterior or inferior MI depending on whether the anterior or the inferior site had the greatest ST score. [14] ST-segment elevation was measured to the nearest .5 mm at the J point and summed for all leads but aVR. [14] [15] ST-segment elevation was considered significant if it was more than .1 mm from the baseline. ST-segment depression as observed in patients with posterior AMI as well as elevations in the right ventricular leads were not included in the analysis. Evaluation of the ECG also included the assessment of the number of leads involved and the extent of ST-segment elevation divided by the number of leads involved (ST score/ number of leads with ST-segment elevation).

The ECGs were generally analyzed by the emergency physician on duty. Reevaluation was conducted by two experienced emergency physicians (MMH, WS), blinded to all clinical data. The ST scores measured by all three physicians were averaged and used for analysis. In the case of a marked difference between the physician (ie, >2 mm difference), the ECG was reanalyzed by all three emergency physicians together in an attempt to reach a consensus. The interobserver variability was 96%.

Patients with evidence of acute congestive heart failure on admission or severe rhythm disturbances at the time of the initial ECG were excluded from analysis.

Acute congestive heart failure was defined as evidence of bibasilar rales over the lungs that did not clear with coughing, a third-sound gallop over the heart and signs of congestion in one of the subsequent chest radiographs routinely performed 12 and 24 hours after ED admission. Radiographic signs of acute congestive heart failure were central or perihilar infiltrates, increased size of vessels serving the upper portions of the lungs in the upright position, and increased prominence of interlobular septa (usually bilateral and symmetric).

Severe rhythm disturbances included bradyarrhythmias (asystole, type 2 second-degree or third-degree atrioventricular block, bradycardia associated with hypotension [systolic blood pressure <90 mm Hg] requiring atropine or insertion of a temporary pacemaker), and tachyarrhythmias (ventricular fibrillation, ventricular tachycardia requiring electrical or chemical cardioversion, rapid atrial fibrillation associated with hypotension [systolic blood pressure <90 mm Hg] requiring digitalis or amiodarone).

All patients who died during the 24 hours after admission to the ED were noted.

We retrieved the following data from the ED charts: age, sex, duration of chest pain, peak creatine kinase (CK) and CK-MB, and complications noted during the 24 hours after ED admission. ST score was calculated according to the previously noted criteria. On the basis of these data, we established a model to predict the appearance of early complications according to the extent of the ST score in the admission ECG.

The following parameters were classified as independent variables in the model: age, sex, interval between onset of pain and the start of thrombolytic therapy, and the calculated sum of ST-segment elevation in the admission ECG.

Evidence of acute congestive heart failure or severe rhythm disturbances in the 24 hours after ED admission were the dependent or outcome variables in this model. Cutoff points were identified for the greatest sensitivity or specificity for the prediction of early complications. The influence of time from onset of pain to treatment, age, and sex was evaluated with the use of a multivariate regression analysis.

Diagnostic performance was described in terms of areas under the receiver-operator curves (ROCs) in accordance with the methods described previously. [16] [17] [18] The ROC analysis software of Centor and Keightley (Blue Ridge Express) was used to obtain area calculations.

We described continuous data with the mean and SD or median and interquartile range (IQR) as appropriate. We used the unpaired Student t test or the nonparametric Mann-Whitney U test for the comparison of groups and the chi2 test for comparisons of proportions. Logistic-regression analysis was used for the estimation of odds ratios (ORs). Numeric values were rounded to the nearest integer. Two-tailed P values less than .05 were considered statistically significant.
TABLE 1 -- General data. Features Anterior MI Inferior MI P
No. of patients 119 124
Age (yr) [mean±SD] 57.1±12.2 56.2±11.5 NS
Sex (M/F) 98/21 96/28 NS
Median peak CK
value (U/L) [IQR] 1,221 (502-1,980) 724 (399-1,115) .006
Median peak CK-MB
value (U/L) [IQR] 120 (72-194) 85 (45-130) .004
ST score (mm)
[mean±SD] 14.4±9.9 8.3±4.6 <.001
Median time to therapy
(min) [IQR] 110 (75-175) 110 (85-218) NS

566

RESULTS

We included 243 patients (194 men, 49 women) with a mean±SD age of 56.6±11.8 years in the study (Table 1) . On the basis of electrocardiographic data, MI was located in the anterior wall in 119 patients and in the inferior wall in 124. Peak CK and CK-MB values were significantly greater in patients with anterior-wall infarction compared with those who had sustained inferior-wall infarctions (CK: 1,221 versus 724 U/L, P=.006; CK-MB: 120 versus 85 U/L, P=.004). Overall, in 103 patients (anterior MI: n=53; inferior MI: n=50; NS) either acute congestive heart failure (anterior MI: n=30; inferior MI: n=16; P=.03) or a rhythm disturbance (anterior MI: n=23; inferior MI: n=34; NS) was observed. The median interval from onset of pain to start of thrombolytic therapy was equal in patients with anterior and inferior MI (anterior MI: 110 minutes, IQR=75 to 175; inferior MI: 110 minutes, IQR=85 to 218; NS). Three patients died in the 24 hours after ED admission (anterior MI: n=2; inferior MI: n=1). Causes of death were cardiogenic shock (n=2) and left ventricular rupture (n=1).

In patients with evidence of complications, the sum of ST-segment elevation was significantly higher compared with


Figure 1. ROC curve showing the performance of ST score with regard to the prediction of early complications in patients with anterior MI.
that of patients without complications. This difference was observed in patients with anterior and inferior myocardial infarction (anterior MI: 10.3 versus 19.4 mm, P<.001; inferior MI: 6.9 versus 10.4 mm; P<.001). No significant differences were observed between patients with congestive heart failure and those with rhythm disturbances (anterior MI: 18.7 versus 15.9 mm, P=.42; inferior MI: 10.1 versus 9.5 mm; NS). The mean±SD ST score for the patients who died during the first 24 hours was 21.3±4.6 mm.

The number of leads with ST-segment elevation was significantly higher in patients with anterior MI and early complications than in those without serious events (4.2±1.2 versus 5.7±1.8, P<.05). In patients with inferior MI no significant difference was observed (3.3±.9 versus 3.4±.9, NS). The ST score divided by the number of leads with ST-segment elevation revealed a significantly higher score in patients with early complications compared with those without complications (anterior MI: 2.3±1.3 versus 3.6±1.9 mm/lead, P<.001; inferior MI: 1.9±.9 versus 3.0±1.4 mm/lead, P<.001). ROC curves show the performance of admission ST score in predicting early complications in anterior MI (Figure 1) . The mean±SD area under the curve is .80±.04. The optimal cutoff point maximizing sensitivity and specificity was found


Figure 2. ROC curve showing the performance of ST score with regard to the prediction of early complications in patients with inferior MI.
567
at 13 mm, for a sensitivity of 79% and a specificity of 73%. The relative risk (RR) was significantly greater in patients above the cutoff value than in those below the cutoff (RR, 2.98; 95% confidence interval [CI], 1.9 to 4.8).

ROC curves show the performance of ST score on admission by early complications in inferior MI (Figure 2) . The mean±SD area under the curve was .72±.05. The optimal cutoff point was found at 9 mm, for a sensitivity of 64% and specificity of 68%. The RR is significantly greater in patients above the cutoff value compared with those below the cutoff value (RR, 2.1; IQR, 1.4 to 3.2).

With the cutoff values derived from ROC analysis, logistic-regression analysis showed that the admission ST score on admission was an independent predictor of the occurrence of early complications in patients with anterior MI (OR, 9.7; 95% CI, 3.9 to 25.1; P<.0001) and in patients with inferior MI (OR, 5.0; 95% CI, 2.0 to 12.8; P=.0002) (Table 2 , Figure 3) . An increase of 1.0 mm in ST score increases the odds of complications by 1.1 (95% CI, 1.1 to 1.2) for anterior MI and by 1.2 (95% CI, 1.1 to 1.4) for inferior MI. No other clinical variables available at the time of ED presentation were significantly associated with the occurrence of early complications in patients with anterior or inferior MI.
DISCUSSION

The initial ECG is frequently used to predict acute MI, serious complications and in-hospital mortality. [1] [2] [3] These studies were designed mainly to identify patients at high risk for a subsequent MI. [1] [2] [3] [4] [5] [6] Therefore patients with suspected acute MI were classified into high- and low-risk groups on the
TABLE 2 -- Multivariate analysis for the occurrence of early complications.
Anterior MI (n=119) Inferior MI (n=124)
Features OR (95% CI) P OR (95% CI) P
Age (yr)



70 1.0 NS 1.0 NS
>70 1.36 (.24-7.45)
.53 (.14-1.8)
Sex .84 (.22-3.29) NS .52 (.18-1.45) NS
Time to therapy
(min)



120 1.0 NS 1.0 NS
>120 1.46 (.55-3.99)
1.45 (.6-3.55)
ST score (mm)



13 1.0 <.0001

>13 9.73 (3.94-25.1)


9

1.0 .0002
>9

4.98 (2.04-12.84)

basis of evidence or absence of specific electrocardiographic patterns (eg, ST-segment elevation, left bundle-branch block, left ventricular hypertrophy, ST-segment depression). [1] Brush et al [1] demonstrated that patients with such "positive" admission-ECG findings had a significantly greater risk of a subsequent acute MI, serious complications or death than those with "negative" ECG findings. Also, patients with minimally abnormal admission ECG findings had a low risk of subsequent MI. [7]

In contrast, our study included patients with evidence of acute MI and was initiated to separate MI patients into those at high and low risk for subsequent complications in the 24 hours after admission on the basis of the ST score. In our series of 243 patients, the ST score was a good predictor of early complications during the 24 hours after acute MI; the ST score was significantly higher in patients with evidence of early complications compared with those without any events. This difference was observed in patients with anterior and inferior Mis, indicating that the ST score can be used independent of the infarct location. We also identified cutoff values of the ST score to calculate the risk for the occurrence of early complications. If the sum of the ST-segment elevations was greater than these cutoff values, the odds of early complications were 10 times greater in patients with anterior MI and five times greater in patients with inferior MI (Figure 3) . Other factors--age, sex, and time between onset of pain and thrombolysis--had no effect on the occurrence of early complications. Therefore ST


Figure 3. Schematic representation of the frequency of early complications in patients with anterior or inferior myocardial infarction according to the magnitude of the ST score on the admission ECG.
568
score remains the only predictor of the occurrence of early complications in patients with acute MI.

The extent of the ST score is determined by the total number of leads and by the extent of ST-segment elevation in each lead with ST-segment changes. A higher ST score, therefore, can be due to an increase in the total number of leads with ST-segment elevation, an increase in the extent of ST-segment elevation in each lead, or both.

In patients with anterior MI and early complications increased ST score is caused not only by greater ST-segment elevation in each lead but by a higher total number of leads with ST-segment elevations. Therefore ST score may reflect the area of endangered myocardium in patients with anterior MI. Our data are in line with the findings of previously published reports demonstrating a relationship between ST score and infarct size. [19] [20] These reports also showed a correlation between the amount of ST-segment elevation and serious complications or death. [4] [21] [22] Figueras [4] et al reported a significant relationship between high ST-segment elevation and left ventricular rupture in patients with acute inferior MI. [4] Nielsen et al [21] demonstrated that "major ST elevations" were associated with a higher risk of fatal events compared with the patients with "minor ST elevations." These data are confirmed by our findings; the three patients who died in the first 24 hours were characterized by above-average ST scores (mean, 21.3 mm).

In contrast, the higher ST score in patients with inferior MI and early complications is caused by a greater ST-segment elevation in each affected lead; the total number of leads involved was similar in patients with and without early complications. Therefore, in patients with inferior MI, ST score only partially reflects the area of endangered myocardium.

Because ST-segment depressions are frequently observed in patients with inferior MI, [23] the exclusion of ST-segment depressions from ST-score calculation seems to be responsible for the lack of correlation between the area of endangered myocardium and ST score in patients with inferior MI. The calculation of the ST score without ST-segment depressions also contributes to the significant difference in the cutoff values between patients with anterior and inferior MI (9 mm versus 13 mm). However, ST-segment depression may be due to many nonischemic and chronic conditions (eg, left ventricular hypertrophy, treatment with digoxin or digitoxin, bundle-branch block) that may interfere with the acute ST-segment depressions caused by myocardial ischemia. Therefore inclusion of ST-segment depressions in such patients may lead to an increased ST score, resulting in decreased specificity. Currently the ST-scoring system is not suitable for patients with posterior infarction; these patients mainly exhibit ST-segment depressions. Further studies are required to evaluate the effect of the inclusion of ST-segment depressions in the ST score on sensitivity and specificity of our cutoff values in patients with inferior MI.

It must be emphasized that the results of our study are restricted to patients undergoing thrombolysis. The interval between onset of pain and treatment ranged from a minimum of 30 minutes to a maximum of 6 hours. It must also be emphasized that this interval may be too small to have any influence on the relationship between ST score and early complications. Whether longer intervals (eg, 6 to 12 hours between the onset of pain and the start of treatment) have any influence on the occurrence of early complications cannot be answered with our data. Also, the association between the area of endangered myocardium and the extent of ST score may be weak, especially in patients with inferior MI. Other methods such as echocardiography, scintigraphy or even biochemical markers may provide more accurate data about the area of endangered myocardium or infarct size. [24] [25] [26] [27] [28] Therefore these methods may be used to predict early complications more precisely than may the ST score. However, these methods require technical or laboratory skills, which may not be available in all EDs on a 24-hour basis. Additionally, none of these parameters is as easily available at the time of admission as the ST score.

Despite these limitations, our data have important clinical implications in the treatment of patients with acute MI. First, evaluation of the ST score on the initial ECG provides a simple method with which to identify patients at high risk for early complications. It follows that the patient at risk for complications may be transferred to an intermediate care unit with telemetric monitoring and trained nursing personnel. This procedure may save costs without compromising patient care. Second, patients at high risk of complications may be considered for a more aggressive therapeutic approach (eg, angioplasty or coronary artery bypass grafting) to reduce the risk of severe early complications.

In conclusion, the absolute ST score is useful in estimating the probability of early complications in patients with acute MI receiving thrombolytic therapy. A cutoff value of 13 mm for anterior MI and 9 mm for inferior MI risk-stratifies patients into high- and low-risk subgroups for development of acute congestive heart failure and severe rhythm disturbances during the first 24 hours of hospitalization. We therefore assume that the calculation of ST-segment elevation score on the admission ECG is clinically relevant because patients with acute MI may be classified as being at high or low risk shortly after ED admission. This stratification may have an important effect on the further treatment of the patient.
569
REFERENCES

1. Brush JE, Brand DA, Acampora D, et al.: Use of the initial electrocardiogram to predict inhospital complications of acute myocardial infarction. N Engl J Med 1985;312:1137-1141.

2. Fesmire FM, Percy RF, Wean RL, et al: Risk stratification according to the initial electrocardiogram in patients with suspected acute myocardial infarction. Arch Intern Med 1989;149:1294-1297.

3. Young MJ, McMahon LF, Stross JK: Prediction rules for patients with suspected myocardial infarction: Applying guidelines in community hospitals. Arch Intern Med 1989;147:1219-1222.

4. Figueras J, Curos A, Cortadellas J, et al: Relevance of electrocardiographic findings, heart failure, and infarct site in assessing risk and timing of left ventricular free wall rupture during acute myocardial infarction. Am J Cardiol 1995;76:543-547.

5. Zalenski RJ, Sloan EP, Chen EH, et al: The ED ECG and immediately life-threatening complications in initially uncomplicated suspected myocardial ischemia. Ann Emerg Med 1988;17:221-226.

6. Stark ME, Vacek JL: The initial electrocardiogram during admission for myocardial infarction: Use as a predictor of clinical course and facility utilization. Arch Intern Med 1987;147:843-846.

7. Slater DK, Hlatky MA, Mark DB, et al: Outcome in suspected acute myocardial infarction with normal or minimally abnormal admission electrocardiographic findings. Am J Cardiol 1987;60: 766-770.

8. Zehender M, Kasper W, Kauder E, et al: Right ventricular infarction as an independent predictor of prognosis after inferior myocardial infarction. N Engl J Med 1993;328:981-988.

9. Zehender M, Kasper W, Kauder E, et al: Eligibility for and benefit of thrombolytic therapy in inferior myocardial infarction: Focus on the prognostic importance of right ventricular infarction. J Am Coll Cardiol 1994;24:362-369.

10. ISIS-4 (Fourth International Study of Infarct Survival) Collaborative Group: ISIS-4: A randomized factorial trial assessing early captopril, oral mononitrate, and intravenous magnesium sulphate in 58.050 patients with suspected acute myocardial infarction. Lancet 1995;345:669- 685.

11. Neuhaus KL, von Essen R, Tebbe U, et al: Improved thrombolysis in acute myocardial infarction with front-loaded administration of alteplase: Results of the rt-Pa-APSAC patency study (TAPS). J Am Coll Cardiol 1992;19:892-893.

12. Birnbaum Y, Scamvsky S, Blum A, et al: Prognostic significance of the initial electrocardiographic pattern in a first acute anterior wall myocardial infarction. Chest 1993;103:1681-1687.

13. Wellens HJJ, Conover MB: The ECG in Emergency Decision Making, ed 1. Philadelphia: WB Saunders, 1992:1-26.

14. Clemmensen P, Grande P, Saunamaki K, et al: Effect of intravenous streptokinase on the relation between initial ST-predicted size and final QRS-estimated size of acute myocardial infarcts. J Am Coll Cardiol 1990;5:1252-1257.

15. Aldrich HR, Wagner NB, Boswick J, et al: Use of initial ST segment deviation for prediction of final electrocardiographic size of acute myocardial infarcts. Am J Cardiol 1988;61:749-753.

16. Swets JA: Measuring the accuracy of diagnostic systems. Science 1988;240:1285-1293.

17. Centor RM: Signal detectability: The use of ROC curves and their analyses. Med Decis Making 1991;11:102-106.

18. Metz CE: ROC methodology in radiologic imaging. Invest Radiol 1986;21:720-733.

19. Mamko PR, Libby P, Covell JW, et al: Precordial ST segment elevation mapping: An atraumatic method for assessing alterations in the extent of myocardial injury: The effect of pharmacologic and hemodynamic interventions. Am J Cardiol 1972;29:223-230.

20. Muller JE, Maroko PR, Braunwald E: Evaluation of precordial mapping as a means of assessing changes in myocardial ischemic injury. Circulation 1975:52:16-27.

21. Nielsen BL: ST segment elevation in acute myocardial infarction: Prognostic importance. Circulation 1973;48:338-345.

22. Mauri F, Gasparini M, Barbonaglia L, et al: Prognostic significance of the extent of myocardial injury in acute myocardial infarction treated by streptokinase (the GISSI trial). Am J Cardiol 1989; 63:1291-1295.

23. Bates ER, Clemmensen PM, Califf RM, et al: Precordial ST-segment depression predicts a worse prognosis in inferior infarction despite reperfusion therapy. J Am Coll Cardiol 1990;16: 1538-1544.

24. Bonow RO, Dilsizian V, Cuocolo A, et al: Identification of viable myocardium in patients with chronic coronary artery disease and left ventricular dysfunction: Comparison of thallium scintigraphy with reinjection and PET imaging with 18 F-fluorodeoxyglucose. Circulation 1991;83:26-37.

25. Schiller N, Shah P, Crawford M, et al: American Society of Echocardiography Committee on Standards, Subcommittee on Quantification of Two-Dimensional Echocardiograms. J Am Soc Echo 1989;2:358-367.

26. Iliceto S, Marangelli V, Marchese A, et al: Myocardial contrast echocardiography in acute myocardial infarction: Pathophysiological background and clinical applications. Eur Heart J 1996;17:344-353.

27. Hamm CW, Katus HA: New biochemical markers for myocardial cell injury. Curr Opin Cardiol 1995;10:355-360.

28. Hirschl MM, Gwechenberger M, Binder T, et al: Assessment of myocardial injury by serum tumour necrosis factor alpha measurements in acute myocardial infarction. Eur Heart J 1996;17:1852-1859.
APPENDIX

TABLE 3 -- Sensitivity and specificity of different cutoff values in patients with anterior MI Cutoff Value
(ST Score [mm]) Sensitivity Specificity
51 .00 .98
50 .02 .98
42 .04 .97
38 .06 .97
31 .08 .97
30 .13 .97
29 .17 .97
28 .19 .97
27 .23 .97
26 .25 .97
25 .26 .95
24 .32 .95
23 .38 .95
22 .40 .95
21 .45 .95
20 .45 .92
19 .51 .91
18 .53 .89
17 .55 .86
16 .55 .83
15 .60 .82
14 .72 .79
13 .79 .73
12 .79 .68
11 .85 .65
10 .87 .56
9 .87 .50
8 .89 .35
7 .94 .32
6 .96 .27
5 1.00 .23
4 1.00 .15
3 1.00 .08
2 1.00 .02
0 1.00 .00


TABLE 4 -- Sensitivity and specificity of different cutoff values in patients with inferior MI Cutoff Value
(ST Score [mm]) Sensitivity Specificity
30 .02 1.00
21 .02 .99
18 .06 .97
17 .06 .96
16 .08 .96
15 .16 .95
14 .24 .93
13 .30 .93
12 .40 .93
11 .54 .86
10 .56 .80
9 .64 .68
8 .70 .58
7 .78 .53
6 .86 .41
5 .90 .32
4 .96 .22
3 .96 .07
2 .98 .00
1 1.00 .00

Em