Left ventricular longitudinal wall fractional shortening accurately predicts longitudinal strain in critically ill patients with septic shock

Background

Left ventricular longitudinal strain (LVLS) may be a sensitive indicator of left ventricular (LV) systolic function in patients with sepsis, but is dependent on high image quality and analysis software. Mitral annular plane systolic excursion (MAPSE) and the novel left ventricular longitudinal wall fractional shortening (LV-LWFS) are bedside echocardiographic indicators of LV systolic function that are less dependent on image quality. Both are sparsely investigated in the critically ill population, and may potentially be used as surrogates for LVLS. We assessed if LVLS may be predicted by LV-LWFS and MAPSE in patients with septic shock. We also assessed the repeatability and inter-rater agreement of LVLS, LV-LWFS and MAPSE measurements

Results

122 TTE studies from 3 echocardiographic data repositories of patients admitted to ICU with septic shock were retrospectively assessed, of which 73 were suitable for LVLS analysis using speckle tracking. The correlations between LVLS vs. LV-LWFS and LVLS vs. MAPSE were 0.89 ( p  < 0.001) and 0.81 ( p  < 0.001) with mean squared errors of 5.8% and 9.1%, respectively. Using the generated regression equation, LV-LWFS predicted LVLS with a high degree of accuracy and precision, with bias and limits of agreement of -0.044 ± 4.7% and mean squared prediction error of 5.8%. Interobserver repeatability was good, with high intraclass correlation coefficients (0.96–0.97), small bias and tight limits of agreement (≤ 4.1% for all analyses) between observers for all measurements

Conclusions

LV-LWFS may be used to estimate LVLS in patients with septic shock. MAPSE also performed well, but was slightly inferior compared to LV-LWFS in estimating LVLS. Feasibility of MAPSE and LV-LWFS was excellent, as was interobserver repeatability.