18
Chapter 5 Normal and Nonspecific ECG’s
Chapter 5: NORMAL AND NONSPECIFIC ECG’S
KEY POINTS
· AMI and ACS may present with a nonspecific (Type 3) or normal (Type 4) ECG.
· Computer algorithms may misread ECG’s as normal or nonspecific.
· Serial ECG’s are necessary to identify high-risk patients who may ultimately need reperfusion therapy.
· If serial ECG’s remain Type 3 or Type 4, further workup may be necessary to identify patients at risk for further coronary events.
GENERAL BACKGROUND
Clinical studies have shown that a normal initial ECG occurs in 6.4% of all AMI’s and that a nonspecific initial ECG occurs in approximately 22% of all AMI’s; these studies, however, were performed when the definition of AMI was based on CK-MB, not on troponin, as it is today (21,22,18). Further, these studies did not strictly define the term “normal ECG.” Thus, in many studies, “nonspecific/nondiagnostic ECG’s” included both: 1) ECG’s nondiagnostic of AMI, but diagnostic of UA/NSTEMI; and 2) ECG’s nondiagnostic of any kind of ACS. Management of patients with these two types of “nonspecific” ECG types may be different. Therefore, as described in Chapter 2, we propose a new classification of ECG types based on therapeutic implications. We distinguish between Type 2 ECG’s, which are diagnostic of UA/NSTEMI, and Type 3 ECG’s, which are nondiagnostic of any kind of ACS. Clinical experience and judgment are often necessary to distinguish between ST-T abnormalities of these two ECG types. Here we focus on normal (Type 4) and nonspecific (Type 3) ECG’s, both of which are nondiagnostic of any kind of ACS (see Chapter 8 for discussion of Type 2 ECG’s). Table 5-1 lists percentages of patients in two studies who were found to have normal or nonspecific ECG’s and AMI as diagnosed by CK-MB (21,18). See Cases 5-1 through 5-5.
NORMAL ECG (TYPE 4 ECG)
Definition of a normal ECG (Type 4 ECG)
· Sinus rhythm with normal p-waves
· ST elevation/depression < 0.5 mm relative to corresponding PR segments.
· No LVH, abnormal Q-waves, or conduction abnormalities (QRS must be < 100 ms)
· Size of T-waves is proportional to R-waves and T-wave axis is close to QRS axis
· Normal R-wave progression
· ST segment is neither up- nor downsloping
Treatment precautions in patients with Type 4 ECG’s
· A normal ECG is not unusual in very early AMI.
· A series of normal ECG’s is common in ACS with minimal or no injury, especially if symptoms have resolved.
· Time since pain onset may not affect the negative predictive value (NPV) of the ECG (67).
· A series of normal ECG’s over 15 to 60 minutes is unusual in complete LAD or RCA occlusion, but it is not unusual in circumflex or branch vessel occlusion (68,69,70).
· Posterior leads may detect a posterior AMI that would otherwise show no ST deviation (71,72).
· A Type 4 ECG may develop into a diagnostic (Type 1 or 2) ECG with serial ECG’s.
· Record a repeat ECG every 15 minutes for >/= 60 minutes or perform continuous ST segment monitoring on any patient with a high clinical suspicion of AMI, even if the initial ECG is totally normal. Record a repeat ECG on any patient with a reasonable clinical suspicion of AMI if the initial ECG is totally normal.
NONSPECIFIC ECG (TYPE 3 ECG)
We define a nonspecific (Type 3) ECG as abnormal but nondiagnostic of any ACS, neither AMI nor UA/NSTEMI. Although angiography +/- PCI may be clinically indicated, thrombolytics are NOT indicated. A Type 3 ECG may be nonspecific due to the presence of Q-waves, LVH, or minor, non-dynamic ST or T-wave abnormalities that are not suggestive of ischemia. These include ST depression and T-wave inversion < 1 mm and unchanging. These abnormalities are still defined as nonspecific even if changed from a previous ECG (18), although the probability of AMI is double if there is such a change (73). A Type 3 ECG is not unusual: 1) early after thrombotic coronary occlusion; 2) in thrombotic occlusion with excellent collateral circulation; 3) with nonocclusive, or intermittently occlusive thrombus; 4) after thrombotic occlusion with spontaneous reperfusion; 5) after thrombotic occlusion with injury of an electrocardiographically silent area of myocardium; or 6) with injury of a small area of myocardium.
Treatment precautions in patients with Type 3 ECG’s
· All Type 3 ECG’s are abnormal
· A Type 3 ECG is much more likely to occur in AMI than a normal ECG (22% vs. 6.4%) (21,18).
· In the reperfusion era, a Type 3 ECG in the presence of AMI does not necessarily reflect lower mortality than a Type 1 ECG. AMI patients with Type 3 ECG’s are not eligible for thrombolysis and may have a higher mortality than AMI patients whose ECG’s meet eligibility requirements for thrombolytics and receive them (74). In the prethrombolytic era, AMI patients with a “negative” ECG (approximately Type 3 or 4) had a much better prognosis than those with a positive ECG (Type 1 or 2) (75). Consider immediate angiography +/- PCI in cases of high clinical suspicion, especially if there is hemodynamic instability or pulmonary edema (see Case 26-2).
· A series of Type 3 ECG’s is common in ACS with minimal or no injury.
· An initial Type 3 ECG may be followed by a diagnostic ECG (Type 1 or Type 2) on serial ECG’s.
· Record serial ECG’s every 15 minutes for >/= 60 minutes or perform continuous ST segment monitoring on any patient with an initial Type 3 ECG but high clinical suspicion of AMI.
· Record a serial ECG on any patient with an initial Type 3 ECG with continued symptoms and reasonable clinical suspicion of AMI.
ECG COMPUTER ALGORITHMS
ECG computer algorithms have inadequate sensitivity for STEMI (41-47). These “normal ECG” misreads may occur with ST elevation that is less than standard “thrombolytic criteria” (< 1 to 2 mm in 2 consecutive anterior leads and < 1 mm in 2 consecutive inferior leads). The computer may also misinterpret ST elevation of AMI as another condition, especially early repolarization. You must learn to read the ECG yourself. Look for ST deviation and dynamic and/or resolved ST and T-wave changes.
MANAGEMENT
Specific details of managing patients with persistent Type 3 and Type 4 ECG’s are beyond the scope of this book. Fundamentally, however, with Type 3 and Type 4 ECG’s and some clinical suspicion:
· Examine prehospital rhythm strips (see Cases 12-3 and 31-1 to 31-3)
· Compare with a previous ECG. Obtain earlier ECG’s from other institutions by fax.
· Record serial ECG’s. Record a repeat ECG every 15 minutes for >/= 60 minutes or perform continuous ST segment monitoring on any patient with a high clinical suspicion of AMI, even if the initial ECG is totally normal or nondiagnostic
· Consider posterior leads and echocardiography.
· Patients at moderate risk for ACS should undergo a standard CP unit evaluation. Consider treatment for ACS. Obtain serial biochemical markers, preferably cTnI or cTnT. Observe. Further work-up may include stress electrocardiography, stress echocardiography, sestamibi scanning, or stress nuclear testing, as dictated by institutional protocols. Electron Beam Computed Tomography (CT) may prove valuable.
ANNOTATED BIBLIOGRAPHY
Unless otherwise noted, the following studies used the presence of elevated blood levels of CK-MB for the definitive diagnosis of AMI.
Christian TF et al., Exercise tomographic thallium-201 imaging in patients with severe coronary artery disease and normal electrocardiograms, 1994.
Methods: Christian et al. (76) studied 411 patients with thallium exercise tests and subsequent angiography; all had normal resting ECG’s.
Findings: Of 298 patients with significant CAD, 77 (26%) had left main or 3-vessel disease. Concurrent thallium scintigraphy demonstrated that 25% of these 77 patients had normal ECG’s while they were having ischemia. This indicates that ECG’s are relatively insensitive, at least for non-occlusive ischemia.
Studies of normal and nonspecific ECG’s and AMI
Rouan GW et al., Clinical characteristics and outcome of acute myocardial infarction in patients with initially normal or nonspecific electrocardiograms, 1989 and Karlson BW et al., Early prediction of acute myocardial infarction from clinical history, examination and electrocardiogram in the emergency room, 1991.
Methods: Rouan et al. (18) analyzed data from 7,115 consecutive CP patients (admitted or discharged) in the Multicenter Chest Pain Study. Patients were included if they had CP unexplained by chest wall or chest radiographic findings. Investigators defined nonspecific ST or T-wave changes to be minor ST or T-wave abnormalities not suggestive of ischemia or strain, even if changed from a previous tracing. Karlson et al. (21) studied 4690 CP patients admitted to the hospital. Pain was not subcategorized into ongoing or resolved in either study. Although the ECG’s in both studies were presumably initial ECG’s, this is not explicitly stated, nor is it specified whether serial ECG’s were recorded. Incidence of unstable angina and occurrence of long-term cardiac endpoints were not measured.
Findings: See Table 5-1, above. Karlson et al. (21) also found that, of the 466 patients (10%) for whom the physician had "no suspicion of MI,” 6 patients (1.5%) had an AMI. Rouan et al. (18) found that 461 (45%) of 1024 AMI patients had classic ECG findings (Type 1a). Of the 17% with nonspecific (Type 3) ECG’s, 7% had nonspecific ST-T abnormalities, 5% had old infarction, and 5% had "other" abnormalities including LVH and atrial abnormalities. “Normal” ECG’s (Type 4) were found in 3.5%. In comparison with AMI patients with a diagnostic ECG (Type 1), if the ECG was Type 3 or Type 4 and the patient was admitted, the total CK was lower (643 IU/L vs.1032 IU/L), mortality was lower (6% vs.12%), there was less cardiogenic shock, dysrhythmia, AV Block, CHF, and the circumflex artery was most often involved. Confidence intervals (CI’s) of the risk ratio (RR) for AMI for patient characteristics were as follows: age > 60 years (1.5-2.3); male (1.2-1.6); "pressure" (1.4-1.9), radiation (1.2-1.8); diaphoresis (1.1-1.9); and history of prior MI or angina and normal ECG (1.0-1.8). Of 42 patients with normal ECG’s and all characteristics, 8 (17%) ruled in for AMI. Of 551 patients with normal ECG’s and none of the characteristics, 0 ruled in for AMI. Of 285 patients with nonspecific ECG’s and none of the characteristics, 1 ruled in for AMI. Use of reperfusion therapy is not mentioned; data was apparently gathered in the pre-reperfusion era.
Slater DK et al., Outcome in suspected acute myocardial infarction with normal or minimally abnormal admission electrocardiographic findings, 1987.
Methods: Slater et al. (22) studied 775 patients admitted to the hospital with symptoms "suggestive of acute MI.” Pain was not subcategorized into ongoing or resolved. The ECG was presumably the first and only ECG obtained, although this is not explicitly stated.
Findings: Of 775 patients, 262 (34%) ruled in for AMI. Of 107 patients (14%) with normal ECG's, 11 (10%) ruled in for AMI (4.2% of total AMI), one with complications. Of 73 patients (9%) with nonspecific ECG's, 6 (8%) ruled in for AMI (2.3% of all AMI), 4 with complications. Thus, 6.5% of AMI patients had either Type 3 or 4 ECG’s, which is far less than the 28.9% in Table 5-1, above. This, together with the high rate of rule-ins, suggests hidden methodologic differences from studies by Rouan et al. (18) and Karlson et al. (21) described above.
Patients with cardiac ischemia mistakenly sent home from the ED
Pope JH et al., Missed diagnosis of acute cardiac ischemia in the emergency department, 2000.
Methods: From May to November 1993, Pope et al. (77) conducted a multicenter prospective study of all patients presenting with CP or other symptoms suggesting acute cardiac ischemia. Patients who were discharged returned at 24 to 72 hours for repeat evaluation, ECG, and CK-MB measurement.
Findings: There was 30-day follow-up in 99% of patients. Of 10,689 patients, 8% ruled in for AMI by CK-MB levels, 7% met criteria for UA, 21% had other cardiac problems and 55% had noncardiac problems. Of 889 AMI patients, 19 (2.1%) were discharged home; when retrospectively examined, 17 of these ECG’s showed “no evidence of ischemia” and 2 were “normal.” Of 966 patients with UA, 22 (2.3%) were mistakenly discharged; retrospective examination of these ECG’s revealed that none showed evidence of ischemia and 2 were normal ECG’s. Non-white ethnicity, female sex, chief complaint of dyspnea, and a normal ECG all correlated with mistaken discharge.
Lee TH et al., Clinical characteristics and natural history of patients with acute myocardial infarction sent home from the emergency room, 1987.
Methods: Lee et al. (78) conducted a prospective study of 3077 patients with precordial or left-sided CP (including pleuritic and nontraumatic chest wall pain) unexplained by obvious local trauma or radiographic abnormalities. Pain was not subcategorized into ongoing or resolved.
Findings: Of 1794 admitted patients, 459 (26%) had AMI, 55 (12%) of whom died. Of 1283 patients sent home, 35 (2.5%) had AMI (7% of total AMI’s). Of the 35 AMI patients sent home, 13 (37%) had ECG evidence of acute ischemia, 5 (14%) were of age < 42 years, and 9 (26%) died (0.7% of all CP patients). Of the 9 sent home who died, 6 (67%) had ECG’s that were misread.