OBJECTIVES:This study sought to define the value of dobutamine echocardiography (DbE) for cardiac risk stratification in patients with chronic renal failure (CRF). DESIGN:Outcome study correlating results of DbE with late cardiac events in patients with CRF. SETTING:Academic medical centre. SUBJECTS:All patients with CRF (serum creatinine > 2.5 mg dL-1) undergoing DbE were studied; we analysed 193 consecutive patients (aged 63 +/- 13 years, 73 men). INTERVENTIONS:A standard dobutamine-atropine stress was administered until attainment of peak dose, or the development of severe ischaemia or side-effects. The electrocardiogram (ECG) and echocardiogram were obtained before, during and after stress. Ischaemia was identified by new or worsening wall-motion abnormalities with stress. OUTCOME MEASURES:Patients were followed up after 38 +/- 14 months for cardiac death, myocardial infarction or coronary disease progression requiring revascularization. RESULTS:DbE demonstrated ischaemia in 36 patients (19%), scar in 36 (19%) and a normal study in 121 patients. The heart-rate response to dobutamine was submaximal (< 85% age-predicted heart rate) in the absence of wall-motion abnormalities in 69 patients (36%), 54 of whom completed the protocol. Follow-up data were complete in 191 patients (99%); cardiac events occurred in 33 patients (17%), including 17 with cardiac death, 7 with infarction, and 9 requiring late revascularization. Spontaneous events occurred in 7 patients with ischaemia, 3 with scar (8%), 11 with a nondiagnostic study (16%) and 3 patients with a normal study (6%). Over the entire follow-up, the event-free survival in patients with ischaemia (66%) was markedly lower than those without ischaemia (84%, P = 0.006). However, the event rate in patients with nonischaemic responses increased from 8% to 16% between 24 and 40 months, and whilst ischaemia was an independent predictor of outcome at 24 months, it was not at 40 months. CONCLUSIONS:In patients with CRF, the identification of ischaemia at DbE is associated with a significant risk of adverse cardiac events. Patients with nonischaemic scans have a low frequency of events over short-term follow-up, but this increases at later follow-up. These later events may reflect progressive coronary disease, attributable to the atherogenic milieu of these patients, and imply that repeated testing may be required to maintain cardiac risk stratification in patients with CRF.