Preoperative assessment of heart function has typically focused on evaluating left ventricular ejection fraction (LVEF). Recent evidence suggests that diastolic heart failure is common and may cause substantial morbidity and mortality. We designed this study to examine the prevalence and potential clinical correlates of diastolic filling abnormalities as measured by echocardiography in geriatric surgical patients. Patients >=65 yr of age undergoing coronary artery surgery without concomitant valvular surgery or those with one or more risk factors for cardiovascular disease undergoing noncardiac surgery were prospectively studied. Preoperative precordial echocardiography was performed for patients undergoing noncardiac surgery, and intraoperative transesophageal echocardiography was performed for those undergoing cardiac surgery. LVEF and diastolic filling properties including E/A ratio and deceleration time were measured. Overall, 251 patients were enrolled. The mean age was 72 +/- 7 yr. Multiple linear regression analyses showed that patients with a history of myocardial infarction P = 0.021), angina pectoris (beta = -6.09, 95% CI: -9.66, -2.52; P = 0.01), and valvular heart disease (beta = -5.05, 95% CI: -9.56, -0.55; P = 0.028) had lower LVEF than those without such conditions. Of the patients with normal LVEF, 61.5% had diastolic filling abnormalities. Diastolic filling indices including E/A ratio (beta = -1.11, 95% CI -6.02, 3.78; P = 0.65) and deceleration times (beta = -3.42, 95% CI -31.28, 24.45; P = 0.81) contributed no additional predictive value for LVEF. No clinical predictors could be identified to predict diastolic filling abnormalities. For patients undergoing noncardiac surgery, analysis of variance demonstrates that the clinical assessment of LVEF using history and physical examination data was able to grossly discriminate the different levels of LVEF as compared with echocardiography (P = 0.0004). However, under-estimation of LVEF occurred more frequently than over-estimation. Although physicians' clinical assessment of systolic ejection fraction was generally accurate, geriatric patients with normal LVEF often had isolated diastolic filling abnormalities that could not be predicted by clinical factors. These results suggest that evaluation of LV systolic function alone is not discriminatory in comprehensively characterizing LV function in geriatric surgical patients.
Implications: Although physicians' clinical assessment of systolic ejection fraction was generally accurate, geriatric patients with normal left ventricular (LV) ejection fraction often had isolated diastolic filling abnormalities that could not be predicted by clinical factors. These results suggest that evaluation of LV systolic function alone is not discriminatory in comprehensively characterizing LV function in geriatric surgical patients.