An Increase In Velocity Time Index After A Low Volume Fluid Challenge Can Predict Fluid Responsiveness In Severe Septic Patients

Object To evaluate velocity time index(VTI)variation after a low volume of fluid infusion(100 ml colloid)as a predictor of fluid responsiveness in mechanically ventilated patients with severe sepsis or septic shock.Methods Twenty one patients, diagnosed as septic shock or serious infection with SIRS, treated by invasive ventilation were enrolled from Jul 2015 to Mar 2016 in the department of Intensive Care Unit. 100 ml colloid were infused at a constant speed in 1min, and the last 400 ml colloid were infused in 14 min. T0, T100, T500, treat the three point as before fluid infusion, after 100 ml colloid infusion and after an additional infusion of 400 ml fluid. Fluid responsiveness is defined as the stroke volume(SV) and cardiac output(CO) increased more than 15% after 500 ml colloid infusion, and patients should be divided into two group: responsiveness and no responsiveness. patients,general situation,vital signs, parameters of ventilation were recorded, and ultrasound is used to measure the VTI, SV, CO. Receiver operating characteristic curves were generated for variation of VTI after 100 ml fluid infusion((35)VTI100).Results Among the 21 patients, 10 were found responsive to fluid loading, 12 were not. The results demonstrated that age, height, weight, APACHE â…¡, PEEP, Pplat, VT,VT/IBM, have no different between the two groups. The basis of two groups of patients with heart rate(HR), mean arterial pressure(MAP), central venous pressure(CVP) and pulse pressure variation rate(PPV) did not show obvious difference(P(27) 0.05), T0 reaction time group VTI, SV, CO was lower than that in group without reaction, difference was statistically significant(P > 0.05). Group of patients with responses VTI, SV and CO on T0,T100, T500 three point comparison difference was statistically significant(P < 0.05), while no reaction group VTI, SV and CO on T0, T100, T500 three point was no significantdifference(P > 0.05). There were positive correlation between(35) VTI100 and(35)VTI500(r=0.73;P(27)0.05),(35)SV100 and(35)SV500(r=0.75;P(27)0.05),(35)CO100 and(35)CO500(r=0.74;P(27)0.05).(35)VTI100, CVP, PPV in the area under the ROC curve(AUC) were 0.96 ?0.04(P= 0.001), 0.41?0.13(P = 0.473), 0.66?0.12(P=0.124). VTI0, SV0, CO0 area under the ROC curve(AUC) were 0.19 ? 0.10(P = 0.017), 0.017(P = 0.003), 0.11?0.08(P=0.003),0.21?0.11(P=0.028), the accuracy of predicting fluid responsiveness is poor.(35)VTI100,(35)SV100,(35)CO100 area under the ROC curve(AUC) were 0.96?0.04(P=0.001), 0.92?0.06(P=0.002), 0.91? 0.07(P=0.004), they can accurately predicts fluid responsiveness after100 ml colloid infusion. Using(35) VTI100 ? 10% as the threshold to predict fluid responsiveness, the sensitivity was 90% and specificity was 80%.Conclusion A low volume of fluid infusion(100 ml colloid)can predict the fluid responsiveness in severe sepsis or septic shock patients. For the mechanically ventilated patients with tidal volume less than 8 ml/kg,(35)VTI100 can more accurately predict the fluid responsiveness than PPV and CVP.