Patients with ST-T abnormalities on electrocardiogram (ECG) have a higher prevalence of many cardiac conditions, including coronary artery disease (CAD), left ventricular dysfunction and cardiac mortality, than those without ST-T abnormalities. Use of a treadmill ECG for prognostic purposes in patients with ST-T abnormalities is, however, difficult, because the ECG findings in these patients are less specific than in those without ST-T changes. Management may be affected because results from an exercise thallium test are often used to determine whether early revascularization is needed.

Kwok and colleagues performed a cohort study to determine prognostic accuracy of treadmill scores in patients with and without ST-T abnormalities. The Duke treadmill exercise score provides a simple method for determining prognoses and future diagnostic evaluations in patients. For example, stress imaging should be done in patients who have ST-segment depression of at least 1 mm. Patients with a history of ST-T abnormalities on resting ECG and chest pain or dyspnea as well as a treadmill thallium test were included. ST-T changes caused by left ventricular hypertrophy, bundle branch block or digoxin therapy were excluded.

Treadmill testing was performed after recording resting ECG, pulse and blood pressure. The exercise test continued until the patient had severe fatigue, moderate angina or at least a 2-mm ST-segment depression. The magnitude and slope of the ST-T deviation were recorded.

The Duke treadmill score was calculated for each patient in this fashion: exercise time (in minutes) −(5 × maximum ST deviation) −(4 × angina index). Low-risk patients were those with Duke treadmill scores of five or higher; those with scores between five and −10 were considered intermediate risk, and those with scores lower than −10 were considered high risk. Outcomes measured during follow-up were death, nonfatal myocardial infarction, coronary artery bypass grafting and percutaneous transluminal coronary angioplasty.

A total of 939 patients had ST-T abnormalities, and 1,466 had normal resting ECGs. Patients with ECG abnormalities were, on average, unable to exercise as long as the control group, or to achieve as high an exercise heart rate or treadmill score. During the seven-year follow-up period, outcomes were significantly worse in the study population. The Duke treadmill score was significantly associated with the outcome in question.

Using a risk-stratification system in which patients with ST-T abnormalities were assigned to a risk group, clear differences in outcomes were seen between the low- and high-risk patients. Cardiac survival after seven years was 97 percent in the low-risk group but dropped to 76 percent in the high-risk group. Similar figures were seen for the other outcomes.

The authors conclude that symptomatic patients with nonspecific ST-T abnormalities can be successfully stratified according to their risk, based on results from exercise treadmill testing. Those deemed to be at high risk can then have early cardiac catheterization and possibly revascularization, rather than waiting for worsening symptoms and late revascularization.

editor's note: Typically, a routine treadmill test has been ordered in patients whose baseline (resting) ECG is normal and who have good exercise capacity. A patient with an abnormal ECG would often be sent for a thallium treadmill test, which is much more expensive. This study offers some guidelines for using a routine treadmill test in patients with nonspecific ST-T changes and then using the result to determine prognosis.—g.b.h.