Dobutamine stress echocardiography in patients with diabetes mellitus: enhanced prognostic prediction using a simple...

We sought to determine the prognostic value of dobutamine stress echocardiography (DSE) for predicting long-term outcomes in a large cohort with diabetes mellitus and to develop a simple risk score using clinical and echocardiographic data.

Neither risk scores nor long-term prognostic value of DSE has been described in a large diabetic population.

We studied 2,349 patients with diabetes mellitus (1,338 men, 67 +/- 11 years of age) during a follow-up of 5.4 +/- 2.2 years.

Mortality and morbidity (myocardial infarction and late coronary revascularization) occurred in 1,044 (44%) and 309 (13%) patients, respectively. Addition of stress echocardiographic variables to the clinical and rest echocardiographic model provided incremental prognostic information for predicting mortality (chi-square = 243 to 270, p < 0.0001) and morbidity (chi-square = 38 to 78, p < 0.0001). For each end point, a simple risk score was derived according to the estimated values of beta coefficients of multivariate predictors (insulin therapy, smoking, failure to achieve target heart rate, percentage of ischemic segments, and impaired left ventricular systolic function) and resulted in an assessment of risk among all age groups. The C-statistic values were 0.60 to 0.64, indicating modest discrimination. The estimated five-year event-free survivals of patients in three risk categories were 94%, 86%, and 80% for morbidity (p < 0.00001) and 69%, 60%, and 47% for mortality (p < 0.0001).

In patients with diabetes mellitus, a simple and practical risk score using clinical variables and results of DSE stratified patients into three risk groups for mortality and cardiovascular morbidity.