The use of extracorporeal blood purification therapies (EBPT) is becoming increasingly widespread worldwide in everyday clinical practice, particularly in the critical care setting. Nonetheless, most of the clinical trials aimed at exploring the effect of EBPT on patients' long-term outcomes have failed to demonstrate consistent results regarding 28 day- or hospital- mortality rates. The aim of this observational prospective registry is to evaluate if there is a cluster of critically ill patients that mostly benefits from extracorporeal blood purification therapies with oXiris membrane.

The use of extracorporeal blood purification therapies (EBPT) is becoming increasingly widespread worldwide in everyday clinical practice, particularly in the critical care setting. For most of these treatments, the extracorporeal removal of inflammatory mediators (via enhanced transmembrane clearance and/or selective/un-selective adsorption of bacterial toxins or inflammatory mediators) is already well established in the literature. Unselective cytokine removal is the main rationale for the use of EBPT, and it might explain the effect of these treatments on multiorgan dysfunction in critically ill patients. Indeed, an increment in hemodynamic stability and an improvement in the multiorgan dysfunction scoring system of treated patients are consistently observed in most of the studies available in the literature. Nonetheless, most of the clinical trials aimed at exploring the effect of EBPT on patients' long-term outcomes have failed to demonstrate consistent results regarding 28 day- or hospital- mortality rates.

Interestingly, the post-hoc analysis of large databases suggests that a specific subpopulation of patients, who are likely to benefit more than others from EBPT, could be identified in clinical practice. As an example, press releases from the Euphrates trial suggest that a specific range of baseline endotoxin activity assay might correlate with a positive long-term outcome in patients treated with Toraymixin cartridges. Nowadays, a personalized approach to extracorporeal therapy has been advocated as the only strategy capable of maximizing the clinical effect of EBPT and their impact on patient outcomes (i.e. precision medicine).

The identification of clinical characteristics correlated with positive outcomes during EBPT allows the implementation of specifically designed trials, focused on a well-defined cluster of patients.

The implementation of a large database encompassing the clinical/biochemical characteristics of patients undergoing EBPT is thus widely hoped for to enable the identification of a cluster of patients with specific features who most benefit from these treatments.

Although it is well known that only a clinical trial can formally assess the effect of EBPT on a specific clinical outcome, several drawbacks pose practical limitations to the implementation of a clinical trial in the critical care setting for this purpose. First, most European countries (and particularly Italy) do not allow the enrollment and randomization of non-competent patients (as critically ill patients formally are), due to ethical issues. Second, clinical trials are associated with very high costs. Third, the quality of results and conclusions deriving from post-hoc analysis of randomized clinical trials (often underpowered) may be qualitatively similar to those obtained from a very large database of patients observed during treatments performed according to the routine clinical practice.

As an example, COMPACT and COMPACT 2 trials, failed to demonstrate a significant effect of Coupled plasma-filtration adsorption (CPFA) on the outcome of critically ill septic patients, with the notable exception of specific subpopulations (e.g. those where an adequate volume of plasma was filtered and adsorbed). This conclusion was obtained through two highly-costly multicenter randomized clinical trials, prematurely halted due to ethical issues, using post-hoc analysis of a subgroup of an already limited sample size.

Nowadays, the identification of specific clusters of patients during EBPT is obtained by means of the implementation of a large web-based registry where a network of researchers can upload the clinical data of patients undergoing EBPT in line with the routine clinical practice of each single center. Taking into consideration the feasibility of the research (i.e. mainly related with the lack of critical ethical issues and the easy-to-use interface of web-platforms), these observation-based registries are inexpensive and effective tools able to identify specific clusters of patients within a very large sample size with widely heterogeneous clinical characteristics.

As an example, a web-based registry has been already implemented to describe the clinical effects of polymyxin-B based cartridges for endotoxin removal (EUPHAS 2 registry). Another one has been similarly implemented to describe the clinical effects of un-selective cytokine adsorption obtained with CytoSorb (CytoSorb registry).

Taking into consideration the fact that data on CPFA already exist and web-based registries are already available for Toraymixin and for CytoSorb, the aim of this project is to design and nationally promote a web-based registry specifically designed for oXiris membrane. In particular, an easy-to-use web-based platform will be made available for all centers that spontaneously adhere to this project.

Three main characteristics will distinguish this web-based registry of the clinical use of oXiris membrane when compared with the already available Euphas 2 and CytoSorb registries. First, it will also be available for smartphone and/or tablet applications; this characteristic will guarantee easy-to-use access and data uploading and enhanced compliance from clinical researchers. Second, it will not be confined to a simplistic registration of patients' data; it will instead also provide the clinician with clinical tools that might be helpful for patient management (e.g. automatic calculation of clinical scoring systems, ideal body weight, functions of mechanical ventilation, antibiotic adjustment according to renal function, etc.). All these tools will provide the clinician with real-time feedback; this web-based registry will provide a clear example of translational medicine and translational research, where data from clinical practice will feed a database for clinical research and, contemporaneously, the research tool will improve clinical practice, providing useful instruments for routine practice. Finally, this web-based registry will allow every center to instantaneously evaluate its results, providing a real-time basic statistic for every recorded variable (e.g. age at enrollment, main comorbidities, baseline serum creatinine, mortality rate, rate of multidrug resistant bacteria, etc.). This function may allow each center to continuously monitor outcomes and local practices, and will represent a major improvement when compared to already existing platforms in Italy, such as the Prosafe.

With this rationale, the research questions are:

• Is there a cluster of critically ill patients that mostly benefits from extracorporeal blood purification therapies (EBPT) with oXiris membrane?
• Might some baseline variables be used to identify this subpopulation of "responsive" patients? In particular, may these indicators be employed to guide indications for EBPT with oXiris, in order to personalize treatments and improve patients' long-term outcome?

Inclusion Criteria:

• critically ill patients in the ICU
• treatment with oXiris

Exclusion Criteria:

• age < 18 years

All critically ill patients undergoing EBPT with oXiris membrane in the enrolling centers will be prospectively observed.

Although acute kidney injury will be mostly present among the enrolled population, it will be not strictly required as an inclusion criterion. Similarly, although sepsis will be frequently observed, the systemic inflammatory state leading to multiorgan dysfunction and supported by these extracorporeal treatments might have several different etiologies, such as ischemia-reperfusion, severe acute pancreatitis, intoxication, etc. (i.e. "sepsis-like syndromes").