Tissue Doppler in critical illness: a retrospective cohort study

BACKGROUND There is a paucity of published data on tissue Doppler imaging (TDI) in the critically ill. In a critically ill cohort, we studied the distribution of TDI and its correlation with other echocardiographic indices of preload. To aid hypothesis generation and sample size calculation, associations between echocardiographic variables, including the ratio of peak early diastolic transmitral velocity (E) to peak early diastolic mitral annular velocity (E'), and mortality were also explored. METHODS This retrospective study was performed in a combined medical/surgical, tertiary referral intensive care unit. Over a 2-year period, 94 consecutive patients who underwent transthoracic echocardiography with E/E' measurement were studied. RESULTS Mean Acute Physiology and Chronic Health Evaluation III score was 72 +/- 25. Echocardiography was performed 5 +/- 6 days after intensive care unit admission. TDI variables exhibited a wide range (E' 4.7-18.2 cm/s and E/E' 3.3 to 27.2). E' below 9.6 cm/s was observed in 63 patients (rate of myocardial relaxation below lower 95% confidence limit of normal individuals). Fourteen patients had E/E' above 15 (evidence of raised left ventricular filling pressure). E/E' correlated with left atrial area (r = 0.27, P = 0.01) but not inferior vena cava diameter (r = 0.16, P = 0.21) or left ventricular end-diastolic volume (r = 0.16, P = 0.14). In this cohort, increased left ventricular end-systolic volume, but not E/E', appeared to be an independent predictor (odds ratio 2.1, P = 0.007) of 28-day mortality (31%; n = 29). CONCLUSION There was a wide range of TDI values. TDI evidence of diastolic dysfunction was common. E/E' did not correlate strongly with other echocardiographic indices of preload. Further evaluation of echocardiographic variables, particularly left ventricular end-systolic volume, for risk stratification in the critically ill appears warranted.