The hemodynamic effects during sustained low-efficiency dialysis versus continuous veno-venous hemofiltration for uremic patients with brain hemorrhage: a crossover study Clinical article

Object. Hemodynamic instability occurs frequently during dialysis treatment and remains a significant cause of patient morbidity and mortality, especially in patients with brain hemorrhage. This study aims to compare the effects of hemodynamic parameters and intracranial pressure (ICP) between sustained low-efficiency dialysis (SLED) and continuous veno-venous hemofiltration (CVVH) in dialysis patients with brain hemorrhage. Methods. End-stage renal disease (ESRD) patients with brain hemorrhage undergoing ICP monitoring were enrolled. Patients were randomized to receive CVVH or SLED on the 1st day and were changed to the other modality on the 2nd day. The ultrafiltration rate was set at between 1.0 kg/8 hrs and 1.5 kg/8 hrs according to the patient's fluid status. The primary study end point was the change in hemodynamics and ICP during the dialytic periods. The secondary end point was the difference between cardiovascular peptides and oxidative and inflammatory assays. Results. Ten patients (6 women; mean age 59.9 +/- 3.6 years) were analyzed. The stroke volume variation was higher with SLED than CVVH (generalized estimating equations method, p = 0.031). The ICP level increased after both SLED and CVVH (time effect, p = 0.003) without significant difference between modalities. The dialysis dose quantification after 8-hour dialysis was higher in SLED than CVVH (equivalent urea clearance by convection, 62.7 +/- 4.4 vs 50.2 +/- 3.9 ml/min; p = 0.002). Additionally, the endothelin-1 level increased after CVVH treatment (p = 0.019) but not SLED therapy. Conclusions. With this controlled crossover study, the authors provide the pilot evidence that both SLED and CVVH display identical acute hemodynamic effects and increased ICP after dialysis in brain hemorrhage patients. Clinical trial registration no.: NCT01781585 (ClinicalTrials.gov).