Left-ventricular global longitudinal systolic strain and strain rate can predict sepsis outcome: comparison between speckle-tracking echocardiography and tissue-Doppler imaging

Background Strain imaging, by either tissue-Doppler imaging (TDI) velocity converted to strain or strain rate or by two-dimensional speckle-tracking echocardiography (STE) analysis, is used to evaluate abnormal left-ventricular (LV) mechanical activation patterns in sepsis.Objective The aim of this study was to predict sepsis outcomes using LV strain and strain-rate measurements as well as to establish a comparison between STE and TDI.Patients and methods This study included 32 patients (43.7±13.7 years, 21 males) [13 patients with sepsis (group 1) and 19 patients with severe sepsis/septic shock (group 2)] and a subset of 10 controls (36.5±8.7 years, eight males). In the first 24 h, color-TDI was performed for LV 16 segments, and Doppler flow profiles were reanalyzed using STE to retrieve LV peak global longitudinal systolic strain (GLSS) and global longitudinal systolic strain rate (GLSSR), which were averaged for the whole segment.Results Compared with the controls, ejection fraction (%EF) of both groups were comparable, but GLSS showed increased values (−17.5±2.9 vs. −20.2±1.6%, P<0.05 by STE; and −14.9±2.6 vs. −19.7±1.8%, P<0.001 by TDI) and for GLSSR values (−1.3±0.2 vs. −1.6±0.1 s−1, P<0.001 by STE, and −1.1±0.4 vs. −1.6±0.1 s−1, P<0.001 by TDI). Compared with group 1, GLSS of group 2 showed increased values (−15.4±1.5 vs. −20.2±2.4%, P<0.05 by STE; and −12.7±6.8 vs. −18.1±2.4%, P<0.05 by TDI). A good correlation was detected between Acute Physiology and Chronic Health Evaluation II score and either GLSS-STE or GLSSR-STE (r=0.88, P<0.001; and r=0.54, P<0.05) and a moderate correlation was detected between %EF and either GLSS-STE or GLSSR-STE (r=0.47, P<0.05; and r=0.45, P<0.05). The area under the curve of GLSS-STE to predict mortality was 0.9 (95% confidence interval: 0.32–0.48), with best cutoff value at −16.8% (sensitivity: 100%, specificity: 86%), and the area under the curve for GLSS-TDI was 0.76% (95% confidence interval: 0.1–0.44), with best cutoff value at −14.9 (sensitivity: 100%, specificity: 82%).Conclusion LV GLSS and GLSSR obtained using STE were more specific and showed a better correlation with both Acute Physiology and Chronic Health Evaluation II and %EF rather than TDI in predicting mortality.