Interpretation of the results

Interpretation of the results – Exercise Stress Testing for the Family Physician, part 2

Corey H. Evans

This article discusses analysis of the information obtained from graded exercise testing and interpretation of the results. Much valuable information is obtained from patients undergoing exercise stress testing. Describing the test results as simply “positive” or “negative” is usually inadequate. The written report should specifically mention the five parameters listed in Table 1.

Heart Rate and Blood Pressure

The increase in heart rate that occurs during aerobic exercise correlates linearly with workload and oxygen uptake. However, the maximal heart rate decreases with age. The exercise report should state the maximal heart rate achieved as a percentage of the predicted maximal heart rate. The best measurement of peak heart rate is made during a 10-second monitor strip just before the termination of testing.

As treadmill work increases, systolic blood pressure increases and reaches a maximum at peak exercise levels. The diastolic pressure usually remains at baseline levels or decreases. An elevation of more than 10 mm Hg in the diastolic pressure during exercise is abnormal and should be considered a hypertensive response to exercise. A hypertensive response to stress might contribute to the clinician’s decision to treat a patient with labile hypertension, or to increase the medication in a patient who is already receiving treatment.

The double product (the systolic blood pressure multiplied by the heart rate) correlates closely with measured myocardial oxygen consumption during exercise. The heart rate and blood pressure response are best analyzed by comparing them with normal values, as shown in Figure 1. [1] Ellestad [2pp471-92] provides a series of graph listing age-specific normal ranges for heart rate and blood pressure responses to exercise.

A drop in the systolic blood pressure during exercise is suggestive of acute myocardial impairment due to ischemia. Many authors have found that hypotension associated


Components of the Exercise Test Report

1. Presence or absence of evidence for myocardial


2. Heart rate and blood pressure response

3. Presence or absence of symptoms

4. Dysrhythmias

5. Functional aerobic capacity

with chest pain during exercise is indicative of sever coronary artery disease. [1]

Patients who do not have an appropriate elevation of heart rate also have a poorer prognosis. Several studies have demonstrated that patients whose heart rate does not rise into the normal range with exercise (above 120 beats per minute), so called chronotropic incompetence, have severe myocardial or coronary artery disease. [3,4]

Signs and Symptoms

The presence of symptoms such as chest pain, claudication or exercise-induced wheezing during the test should be mentioned in the written report. Some investigators have shown that patients who have chest pain with electrocardiographic evidence of myocardial ischemia have a worse prognosis than patients who have ECG evidence of ischemia but no chest pain. [2p328] However, other recent studies have suggested


Cardiorespiratory Fitness Levels (*)

mL of oxygen

Fitness level per kg per minute METS

Poor 3.5-13.9 1.0-3.9

Low 14.0-24.9 4.0-6.9

Average 25.0-38.9 7.0-10.9

Good 39.0-48.9 11.0-13.9

High 49.0-56.0 14.0-16.0

METS = metabolic equivalents.

(*)–For 40-year-old men. Adjustments are needed

to apply these standards to others.

From American College of Sports Medicine. Guidelines

for exercise testing and prescription. 4th ed. Philadelphia:

Lea & Febiger, 1991:43. Used with


that patients with asymptomatic myocardial ischemia (silent ischemia) on exercise stress testing have the same prognosis as patients with symptomatic myocardial ischemia. [5] Patients who develop typical anginal chest pain while undergoing testing, even in the absence of ECG evidence of ischemia, should be regarded as having myocardial ischemia.


The occurrence and type of dysrhythmias during exercise or recovery should be carefully noted. Premature ventricular contractions during exercise are frequently seen in both healthy patients and those with heart disease. In healthy persons, the catecholamine released as exercise is first started may induce premature ventricular contractions. Unifocal premature ventricular contractions during exercise are not specifically related to myocardial ischemia. However, patients with “high-grade” ectopy (couplets, multifocal premature ventricular contractions or ventricular tachycardia) are more likely to have severe ischemic heart disease and higher mortality, compared with patients who do not have ectopy. [2p282]

Functional Aerobic Capacity

The treadmill can be used to measure the maximal aerobic capacity if maximal effort is achieved. Although the maximal oxygen uptake is not directly measured, peak performance can be expressed in units of metabolic equivalents (METS, measured as mL of oxygen per kg per minute). The MET level at each stage of exercise on the treadmill (but not the bicycle ergometer) is the same for everyone.

Once the maximal MET level is known, available tables can be consulted to determine a fitness level based on age and sex. Table 2 [6] is an example of a table providing cardiorespiratory fitness levels. The peak MET level can also be compared with the known energy requirements of daily household and recreational activities. [7] This information can be useful in writing an exercise prescription based on objective assessment of the patient’s exercise capacity.

ECG Response

Clinical signs and symptoms are helpful during stress testing, but the most common manifestations of myocardial ischemia during exercise testing are ST-segment changes. Although changes in the P,T and R waves suggestive of ischemia have been described, ST-segment changes correlate best with the presence of myocardial ischemia.

The normal response of the ECG complex to exercise consists of J-point depression and an upsloping ST segment that returns to baseline by 80 msec beyond the J-point. The J-point is the junction between the end of the QRS complex and the ST segment. A normal resting ECG complex and a normal complex during exercise are shown in Figure 2. [2pp471-92] During exercise, the J-point is depressed 1 mm below the baseline. The ST segment slopes up and returns to the baseline by 80 msec (two small boxes on standard ECG paper). When the PQ segment is not horizontal, the baseline, by convention, is considered to be the junction of the PQ segment and the Q wave.

Several abnormal ST-segment responses that indicate myocardial ischemia are described


(Figure 3) [2pp471-92]: (1) upsloping ST-segment depression–when the ST segment does not return to baseline by 80 msec, (2) horizontal ST-segment depression–J-point depression followed by a horizontal ST segment depressed below the baseline, (3) downsloping ST-segment depression and (4) J-point and ST-segment elevation (a rare response).

On the basis of these patterns, most authors consider the ECG response positive for myocardial ischemia when one of the criteria in Table 3 is met (each ST segment is measured at 80 msec past the J-point). Table 4 lists findings that are considered to be suggestive of myocardial ischemia. It is important to remember that the sensitivity and specificity of the test will be affected by the diagnostic cut-off point chosen. Some authors propose


using a more stringent criterion (2 mm of ST-segment depression) for asymptomatic individuals. [8]

The test is considered negative for myocardial ischemia when the above criteria are not present and the patient achieves at least 85 percent of the predicted maximal heart rate. If the patient does not reach 85 percent of the predicted maximum, the test is considered inconclusive.

Clinical Interpretation of Positive

and Negative Tests


Patients with a negative treadmill test have a low likelihood of future myocardial ischemic events. In the Seattle Heart Watch Study, Bruce [9] reported on 2,365 clinically healthy men who were followed after stress testing. At five years, only 1 percent of those with normal treadmill tests had an infarction or sudden death. Ellestad [2p318] reported follow-up data on 1,994 patients with negative treadmill tests. This group included patients either with or without cardiac symptoms who were referred for stress testing. During five years of follow-up, only 5 percent of the patients with a negative test had an infarction or died, compared with 45 percent of the positive responders.

How often does the negative test fail to identify coronary artery disease? The failure rate depends on the population studied, but when a mixture of asymptomatic and symptomatic patients is tested, approximately 20 to 30 percent of tests are false negative for significant coronary artery disease, for a sensitivity of 75 percent. [10]

The false-negative rate also depends on the amount of ST-segment depression required for a positive test. Some patients have mild ST-segment depression that is less than what is required to diagnose myocardial ischemia. By the given criteria, such a finding is not positive for ischemia. This intermediate response has a prognosis that is different from the truly negative response. The risk of future cardiac events is somewhere between the negative and positive response. Therefore, these responses are most appropriately considered possibly suggestive of myocardial ischemia.


Most studies have shown that in patients with a test diagnostic of myocardial ischema, the test is 80 to 90 percent specific for coronary artery disease. In other words, the false-positive rate is 10 to 20 percent in such cases. Detrano and Froelicher [10] recently summarized available studies with the fewest methodologic problems and found an overall specificity of 75 percent (false-positive rate of 25 percent). However, combining the positive results obscures some important concepts. The type of ST-segment depression is an important consideration in correlating the severity of coronary artery disease.

Goldschlager [11] illustrated this by testing 269 patients with angiographically documented coronary artery disease and 141 normal subjects. Each of the four types of ST-segment responses reflected a variable percentage of normal patients and patients with single-, double- or triple-vessel disease (Figure 4). Upsloping ST-segment depression, horizontal ST-segment depression and downsloping ST-segment depression are progressively more suggestive of severe disease. Downsloping ST-segment depression is 99 percent predictive of coronary artery disease and 90 percent predictive of two- or three-vessel disease.

In the same study, the time of onset and duration of ST-segment depression were also predictive of coronary artery disease. ST-segment depression occurring before three minutes of execise on the treadmill and depression lasting beyond eight minutes into recovery were 88 percent and 90 percent predictive of two- or three-vessel coronary artery disease, respectively.

The Stress Test as a Predictive Test

Stress testing is about 70 percent sensitive and 80 percent specific for coronary artery disease. However, a stress test should not be considered this simplistically, because sensitivity and specificity are dependent on the prevalence of coronary artery disease in the population tested. A much better use of stress testing is to determine the patient’s pretest likelihood of coronary artery disease and then use the results of the treadmill test to determine a new post-test likelihood.

Diamond and Forrester [12] predicted the pretest likelihood of coronary artery disease depending on the patient’s age, sex and clinical symptoms. The data are displayed in Figure 5a. The pretest likelihood of coronary artery disease can be modified by the exercise stress test results to develop a post-test likelihood of disease (Figure 5b). [13] The second graph also considers the degree of ST-segment depression.

Some examples illustrate these concepts. Consider a 50-year-old man being tested for atypical chest pain. Before testing, the pretest likelihood of significant coronary artery disease is 20 percent. If the test is negative, the probability drops to 8 percent. If the test shows 2-mm ST-segment depression, the probability increases to 75 percent. In this situation, the test is a good, but not perfect, discriminator between patients with coronary artery disease and those without disease.

In a 60-year-old man with typical angina, the pretest likelihood of coronary artery disease is 92 percent. If a 1-mm ST-segment depression is oberved during the test, the post-test likelihood of significant coronary artery disease is about 97 percent. The positive treadmill test added little additional diagnostic information. If the test had shown no ST-segment depression (a negative test), the post-test likelihood of disease would still be 60 percent. For most clinicians, this would represent an unacceptably high false-negative rate. The patient might thus be better served by proceeding to cardiac catheterization without a treadmill test.

Exercise stress testing is not helpful in evaluating patients with a low pretest probability of coronary artery disease. In this group, most positive tests are actually false positive. For example, testing is performed in a 40-year-old asymptomatic man whose pretest likelihood of coronary artery disease is about 4 percent. If a 1.0-to 1.5-mm ST-segment depression is observed during the test, the post-test likelihood of significant disease is still only 11 percent. If ST-segment depression of more than 2.5 mm is observed, the probability of disease increases to 70 percent, but this


degree of ST-segment depression is unusual in this setting.

In summary, exercise stress testing is best used to evaluate patients whose pretest probability of disease is between 20 and 80 percent. This includes most men and women with atypical chest pain, women with typical angina and asymptomatic men in their 50s or 60s who have multiple risk factors.

Clinical Decision Making

Exercise stress testing is a tool to identify patients with coronary artery disease. Management of patients with the disease depends primarily on the location and severity of the lesions identified on cardiac catherization. Exercise stress testing helps the family physician decide which patients need referral for catherization.

Patients who demonstrate responses correlating with severe two- ot three-vessel disease should definitely be referred. These responses are listed in Table 5.

Another group of patients can be followed with repeat exercise testing despite some degree of ST-segment depression. These patients have ST-segment depression at high workloads. Specifically, when ST-segment depression develops in these patients only after stage 4 (Bruce protocol: 12 minutes) or at a heart rate above 160 beats per minute, they generally have a low mortality risk and a good prognosis. In fact, any patient who exercises for longer than 12 minutes on the Bruce protocol (or the equivalent of 13 METS on other protocols) has a good prognosis. Many cardiologists simply recommend repeating the exercise stress test in six months in such patients.

This approach was applied to patients with atypical chest pain by Goldschlager, [14] who developed an algorithm for evaluating patients with atypical chest pain. Patients with ischemic changes on the ECG who achieved a heart rate greater than 160 beats per minute or exercised for 12 minutes on the Bruce protocol needed periodic retesting without further evaluation (Figure 6).

Mark and colleagues [15] developed a different tool to identify this same subset of patients. They developed the exercise treadmill score, calculated by the following equation: exercise time (minutes) – (5 X ST deviation [mm]) – (4 X angina index). The exercise time is expressed in terms of minutes completed on the Bruce protocol. The ST deviation represents the maximum net ST-segment depression or elevation from baseline in any lead. The angina index is 0 for no angina, 1 for typical angina during exercise and 2 for a test terminated because of angina.

In a group of 2,842 men who underwent exercise stress testing and cardiac catheterization, test scores greater than + 5 were associated with a good prognosis. The five-year survival rate was 97 percent in this low-risk group. Thus, a patient who exercises for 13 minutes, without chest pain, but with 1-mm ST-segment depression, would have a score of 8 and a good prognosis. In the highest-risk group in the study, with a score of less than – 11, the five-year mortality was 72 percent. The report indicates that treadmill testing adds additional prognostic information to the findings of cardiac catheterization. In a follow-up prospective study, Mark and associates [16] showed that the exercise treadmill score is a useful tool to decide which patients need referral for cardiac catheterization.

Some symptomatic patients may have normal test results, but still need further diagnostic studies. These are patients whose post-test likelihood of coronary artery disease is still too high (above 10 to 20 percent). If the physician anticipates this, a radionuclide stress test is a better first choice than exercise stress testing.

Finally, patients with mild to moderate ischemic changes at moderate levels of exercise intensity need further evaluation, usually a radionuclide stress test. Making these clinical decisions is difficult. Family physicians often can help the patient make a decision after getting input from a consulting cardiologist.

These concepts are summarized in the algorithm shown in Figure 7.

Final Comment

Family physicians can gain valuable information from exercise stress testing. Each parameter–the heart rate and blood pressure response, symptoms, dysrhythmias, functional aerobic capacity and the presence or absence of ischemia–should be mentioned in the report. A suggested format for the report is outlined in Table 6.

Family physicians can perform exercise stress testing in their offices. Mastery of the basic concepts of exercise stress testing, careful patient selection and a supporting relationship with consulting cardiologists are mandatory to its success. Many good courses are available. Introductory and advanced courses are offered at the Annual Scientific Assembly of the American Academy of Family Physicians. The annual meeting of the American College of Cardiology is also an excellent source of in-depth information. No specific training guidelines are available for family physicians. After the basic concepts are mastered through study or attending a course and approximately 15 supervised tests are performed, most family physicians feel comfortable performing exercise tests independently.

Standard textbooks on the subject, such as Ellestad’s Stress Testing [2] or Froelicher’s Exercise and the Heart, [1] are invaluable. The American Heart Association’s monograph [8] on exercise testing and exercise training is also recommended.


[1] Froelicher VF. Exercise and the heart: clinical concepts. 2d ed. Chicago: Year Book Medical Publishers, Inc., 1987:99-102.

[2] Ellestad MH. Stress testing: principles and practice. 3d ed. Philadelphia: F.A. Davis Co., 1986:282,308-9,318,328,471-92.

[3] Chin CF, Messenger JC, Greenberg PS, Ellestad MH. Chronotropic incompetence in exercise testing. Clin Cardiol 1979;2:12-8.

[4] McNeer JF, Margolis JR, Lee KL, et al. The role of the exercise test in the evaluation of patients for ischemic heart disease. Circulation 1978;57:64-70.

[5] Weiner DA, Ryan TJ, McCabe CH, et al. Significance of silent myocardial ischemia during exercise testing in patients with coronary artery disease. Am J Cardiol 1987;59:725-9.

[6] American College of Sports Medicine. Guidelines for exercise testing and prescription. 4th ed. Philadelphia: Lea & Febiger, 1991:43.

[7] Chung EK. Manual of exercise ECG testing. New York: Yorke Medical Books, 1986:24-5.

[8] American Heart Association. Exercise standards: a statement for health professionals from the American Heart Association. Circulation 1990;82:2286-322.

[9] Bruce RA, DeRouen TA, Hossack KF. Value of maximal exercise tests in risk assessment of primary coronary heart disease events in healthy men. Five years’ experience of the Seattle Heart Watch Study. Am J Cardiol 1980;46:371-8.

[10] Detrano R, Froelicher VF. Exercise testing: uses and limitations considering recent studies. Prog Cardiovasc Dis 1988;31:173-204.

[11] Goldschlager N, Selzer A, Cohn K. Treadmill stress tests as indicators of presence and severity of coronary artery disease. Ann Intern Med 1976;85:277-86.

[12] Diamond GA, Forrester JS. Analysis of probability as an aid in the clinical diagnosis of coronary artery disease. N Engl J Med 1979;300:1350-8.

[13] Epstein SE. Implications of probability analysis on the strategy used for noninvasive detection of coronary artery disease. Am J Cardiol 1980;46:491-9.

[14] Goldschlager N. Use of the treadmill test in the diagnosis of coronary artery disease in patients with chest pain. Ann Intern Med 1982;97:383-8.

[15] Mark DB, Hlatky MA, Harrell FE Jr, Lee KL, Califf RM, Pryor DB. Exercise treadmill score for predicting prognosis in coronary artery disease. Ann Intern Med 1987;106:793-800.

[16] Mark DB, Shaw L, Harrell FE Jr, et al. Prognostic value of a treadmill exercise score in outpatients with suspected coronary artery disease. N Engl J Med 1991; 325:849-53.

COPYRIGHT 1992 American Academy of Family Physicians

COPYRIGHT 2004 Gale Group