According to our study, patients had increased COs and VTIs both after PLRTs and BITs compared with those before these two tests, which indicated that the changes of COs and VTIs during BITs could identify fluid responsiveness in circulatory shock patients with RRT.
The evaluation of fluid responsiveness happens all the time in intensive care units which play a vital role in fluid management. Fluid bolus, which could be treated as preload challenge, was classically used to test if it could induce hemodynamic improvement. But it was possible to be overloaded if no attempt was made to evaluate fluid responsiveness with volume expansion. So static parameters like central venous pressure and dynamic markers like pulse pressure variation and stroke volume variation were used based on heart-lung interaction(10). And PLRT, taken as a reversible preload challenge, could be repeated frequently without any fluid dripping into the body(11) and was accurate even in patients spontaneously complicated with cardiac arrhythmias, and low respiratory system compliance(12). Additionally, an increase in CO after the mini-fluid challenge could also define fluid responsiveness (13). In our opinion, PLRT was proved to be the most useful and convenient maneuver reported to be reliable consistent with studies, and reversible in preload challenges (2). Therefore, we adopted the changes of preload during PLRTs, expressed by CO and velocity-time integral, as the ‘golden standard’ in our study to test if BIT could have the same power to identify the fluid responders from all patients with shock and RRTs. The reliability of CO measurements by pulmonary thermodilution in RRT was challenged because it was found that the thermodilution curve forms were modified resulting in inaccurate calculation of related hemodynamic parameters.(14) But we noticed that the thermal indicator was injected through a dialysis catheter in this study which was not a normal way stipulated by factory settings. Additionally, Dr. Dufour and his colleagues confirmed that hemodynamic measurements derived from transpulmonary thermodilution were not affected by RRTs.(15) So we could take the changes of CO derived from PiCCO2 as the golden standard to recognize fluid responders.
Transthoracic echocardiography was performed excellently in estimating cardiac out based on LVOT-VTI compared with pulmonary artery catheter (16, 17). Additionally, LVOT-VTI combined with PLRT could screen volume responsiveness from end-stage renal disease patients after hemodialysis with the mean VTI increasing from 30.31cm to 34.91cm and the mean ΔVTI between 12.64% and 16.84% (18). And LVOT-VTI is reliable and repeatable in distinguishing fluid responders from all shock patients. As reported in the study of Lill Bergenzaun and his colleagues (19), LVOT-VTI was the best repeatable echocardiographic parameter in the evaluation of left ventricular systolic function.
Little similar studies were found according to our findings except the one published by our team last year which focused on volume changes during BPT at the early stage of continued blood purification(5). From our collected data, general hemodynamic parameters (heart rate, mean arterial pressure, and central venous pressure) had no significant difference before and after PLRTs or BITs. Just as reported in Andreas Umgelter’s study, the above hemodynamic parameters had no change after infusion of 200ml of 20% albumin. Furthermore, no difference was detected between responders and non-responders(20). Additionally, even stroke volume from PiCCO and inferior vena cava-collapse index from echo examination had little change before and after PLRTs or BITs from this study due to various interfering factors in the calculation of stroke volume from PiCCO, like zero setting of arterial pressure, and VTI-CI from echocardiography, like respiratory rate. However, COs and LVOT VTIs increased significantly both in PLRTs and BITs. There were nearly 60% of patients companied by remarkable increased COs with relatively high COs during PLRTs. That was to say, the cause of shock in this population remained partly due to insufficient volume indicating that more fluids should be given to these patients. Such facts implied that, firstly, if the patient had received sufficient fluid management; secondly, the following goals of fluid treatment should be made on account of the results after BIT.
Our study was limited by the following facts. Firstly, it should be a ‘round-trip’ trial if we’re intended to demonstrate BITs do have an impact on fluid responsiveness, which means there should be a BPT just after BIT. However, it couldn’t be carried out in practical daily work due to standards of medical ethics. Secondly, we took PLRT as ‘the golden standard’ in our study instead of fluid-challenge with a certain number of crystalloids. But such a volume as 500ml was not necessary and might be a heavy burden to the patient that could worsen the unstable hemodynamics.