Simulation for Continuous Veno-venous Hemofiltration in Intensive caRe

Overview

Continuous Veno-Veinous Hemofiltration (CVVH) is a frequently used renal replacement therapy in intensive care units for patients with acute renal failure. In theory, this therapy should be continuous, 24 hours/day. However, it is wellknown that actual duration of CVVH sessions is much lower than scheduled, with frequent "circuit down-times" (between 1 to 6 hours/day). There are many reasons for premature ending of CVVH sessions (i.e. catheter dysfunction, wrong settings, low anticoagulation…), but early detection of any dysfunction may prevent (at least in part) the CVVH circuit coagulation. ICU Nurses are on the first line to manage CVVH sessions, particularly in case of alarms. High-fidelity simulation has been recently proposed for health care provider education. It has been consistently associated with large effects for outcomes of knowledge, skills, and behaviors but with moderate effects for patient related outcomes. Indeed, few studies have assess the impact of simulation on patient outcome following complexe procedures, that implies team work in addition to individual skills. In our 12-bed surgical ICU, investigators have decided to implement CVVH technics (in addition to the dialysis that we already used). Before implementing this new technic in the ICU, investigators designed this study in order assess wether a high-fidelity based ICU-nurses education program would improve the outcome of CVVH sessions, compared to the conventional education program, proposed by the CVVH manufacturer.

All nurses of our 12-beds surgical ICU were randomized to one of the two arms for CVVH training (ie "conventional", the control arm, or "conventional + high-fidelity simulation", the intervention arm). Randomization was stratified according to nurse's experience with dialysis (i.e. defined as novice, standard or expert). Then, all CVVH sessions performed in the ICU, were randomized to be supervised by nurses of one of the two arms only. The nurse in charge of the patient could thus be changed, depending on his/her group (note that the usual ratio is 1 nurse for 2.5 patients). Investigators use a minimization algorithm for this randomization, that take into account the presence of shock, counter-indication for heparin use, patient's agitation, site of dialysis catheter insertion (ie. Internal jugular or femoral) and previous early ending of the session. Prescription of CVVH was standardized (ie. Objective of 35 ml/kg/h of ultrafiltrate, with 1/3 in predilution and 2/3 post-dilution), and used unfractionated heparin. All CVVH parameters (i.e. pressures, flow, alarms…) were recorded continuously, using a computer. In addition, hemodynamic parameters and biological parameters were prospectively recorded every 3 hours. Prescription of CVVH ending should be timely written by the doctors. Nurses were asked to clearly record the time of CVVH session ending. An ending appearing less than 4 hours before the scheduled arrest of the session is considered as a "prematurely ending". Total and effective duration of each session were also recorded , as well as the total volume of ultrafiltrate.