Study protocol: The DOse REsponse Multicentre International collaborative initiative (DO-RE-MI)

BACKGROUND Current practices for renal replacement therapy (RRT) in ICU remain poorly defined. The observational DOse REsponse Multicentre International collaborative initiative (DO-RE-MI) survey addresses the issue of how the different modes of RRT are currently chosen and performed. The primary endpoint of DO-RE-MI will be the delivered dose versus in ICU, 28-day, and hospital mortality, and the secondary endpoint, the hemodynamic response to RRT. Here, we report the first preliminary descriptive analysis after 1-year recruitment. METHODS Data from 431 patients in need of RRT with or without acute renal failure (mean age 61.2+15.9) from 25 centers in 5 countries (Spain, Italy, Germany, Portugal, France) were entered in electronic case report forms (CRFs) available via the website acutevision.net. RESULTS On admission, 51% patients came from surgery, 36% from the emergency department, and 16% from internal medicine. On admission, mean SOFA and SAPS II were 13 and 50, respectively. The first criteria to initiate RRT was the RIFLE in 38% (failure: 70%, injury: 25%, risk: 22%), the second the high urea/creatinine, and the third immunomodulation. A total of 3,010 cumulative CRF were reported: continuous venovenous hemodiafiltration (CVVHDF) 60%, continuous venovenous hemofiltration (CVVH) 15%, intermittent hemodialysis (IHD) 15%, high-volume hemofiltration (HVHF) 7%, continuous venovenous hemodialysis (CVVHD) 1%, and coupled plasma filtration adsorption/CVVD 2%. In 15% of cases, the patient was shifted to another modality. Mean blood flow rates (ml/min) in the different modalities were: 145 (CVVHDF), 200 (CVVH), 215 (IHD), 283 (HVHF), and 150 (CVVHD). Downtime ranged from 8 to 28% of the total treatment time. Clotting of the circuit accounted for 74% of treatment interruptions. CONCLUSIONS Despite a large variability in the criteria of choice of RRT, CVVHDF remains the most used (49%). Clotting and clinical reasons were the most common causes for RRT downtime. In continuous RRT, a large variability in the delivered dose is observed in the majority of patients and often in the same patient from one day to another. Preliminary analysis suggests that in a large number of cases the delivered dose is far from the 'adequate' 35 ml/h/kg.