This project is based on a predictive alghorithm (Multifactorial Dynamic Perfusion Index-MDPI) already published and covered by a patent. The MDPI is based on a dynamic collection of 7 different variables during cardiopulmonary bypass (CPB) and provides a probability index for postoperative acute kidney injury. The study design is a multicenter observational prospective trial developed through 3 work packages, addressing (1) external validation of the MDPI in a series of 800 adult cardiac surgery patients collected in 2 Institutions (2) development of a novel MDPI to be applied in infants < 20 kg undergoing cardiac surgery (200 patients) and (3) verification of of other possible outcomes that may be predicted by the MDPI. Since many of the predictive variables are modifiable by the perfusionist/anesthesiologist during CPB, it is a tool that allows therapeutic manouvres. Ultimately, the MDPI will be incorporated in a dedicated monitor to provide an on-line "flight control" during CPB. Work package 1 will be performed at Units 1 and 2; the parameters composing the MDPI will be collected using the existing CPB monitors that routinely measure the hematocrit, the oxygen delivery, the time of exposure to a pre-defined critical oxygen delivery, the mean arterial pressure, and the CPB duration. Blood lactates and transfusions will be manually inputed. The MDPI will be calculated off-line and tested for association and predictivity (discrimination and calibration) with respect to postoperative AKI defined according to the K-DIGO classification. Workpackage 2 is dedicated to infants, with the purpose of developing an MDPI dedicated to low-weight infants (I-MDPI). This will be develop in Unit 1 that performs congenital heart surgery. The same variables of the MDPI will be collected, plus additional variables specific for infants (blood to prime the oxygenator, plasma for the same purpose; venous oxygen saturation, and others). The variables being independently associated with AKI will enter a logistic regression equation that will be the basis for the I-MDPI. Workpackage 3 considers that AKI is associated with a prolonged mechanical ventilation time, prolonged stay in the intensive care unit and in the hospital; and mortality. Therefore, the MDPI may be predictive of other postoperative complications, apart from AKI, and even of mortality. Some of the factors included in the MDPI may directly (low hematocrit) or indirectly (prolonged CPB duration, excessive hemodilution, low mean arterial pressure) affect the hemostatic system and/or trigger packed red cells transfusions. Additionally, CPB itself is a determinant of a coagulophatic state with associated postoperative bleeding which, in turns, increases the mortality. The specific aim 3 is to confirm the hypothesis that the MDPI may be predictive of one or more of this non-AKI postoperative complications and of 30-days mortality. In the same series of work package and aim 1, these complications will be collected and the MDPI tested for predictive ability of each one of these complications and 30-days mortality. At present, the MDPI can only be calculated off-line, and this greatly limits its applicability. The last step of the aim 3 is based on the involvement of software experts and partnership with companies interested in including the MDPI into their existing monitors; as such, the MDPI patent would be given under licence of the existing patent owned by the IRCCS Policlinico San Donato.

This project is based on a predictive alghorithm (Multifactorial Dynamic Perfusion Index-MDPI) already published and covered by a patent. The MDPI is based on a dynamic collection of 7 different variables during cardiopulmonary bypass (CPB) and provides a probability index for postoperative acute kidney injury. Multicenter observational prospective trial developed through 3 work packages, addressing (1) external validation of the MDPI in a series of 800 adult cardiac surgery patients collected in 2 Institutions (2) development of a novel MDPI to be applied in infants < 20 kg undergoing cardiac surgery (200 patients) and (3) verification of of other possible outcomes that may be predicted by the MDPI. Since many of the predictive variables are modifiable by the perfusionist/anesthesiologist during CPB, it is a tool that allows therapeutic manouvres. Ultimately, the MDPI will be incorporated in a dedicated monitor to provide an on-line "flight control" during CPB. Cardiac surgery associated acute kidney injury (CSA-AKI) is one of the most common postoperative complications, associated with an increased mortality risk. Several risk scores for CSA-AKI exist. They are based on preoperative risk factors and severity of the procedure. They define a static risk (SR) based on non-modifiable risk factors. As so, they do not consider intraoperative variables, that include potentially modifiable risk factors (dynamic risk, DR). In a previous study we have developed a new model for prediction of CSA-AKI that is inclusive of the SR and the DR, producing the Multifactorial Dynamic Perfusion Index (MDPI). The MDPI is based on 7 factors collected during cardiopulmonary bypass (CPB): oxygen delivery time spent on a low oxygen delivery, hematocrit, time on CPB, mean arterial pressure, transfusions and lactate values. The MDPI showed a better discrimination (AUC 0.769) than the other existing models, and a good calibration until a risk of 60%. Of notice 5 out of the 7 predictors composing the MDPI are modifiable risk factors and therefore can be considered as a ¿flight control¿, on-line measure of the quality of perfusion, to prompt interventions by the perfusionist and the anesthesiologist. The MDPI is covered by an Italian patent (n. 102022000012893) owned by the IRCCS Policlinico San Donato with Marco Ranucci as inventor. The patent covers the inclusion of the MDPI into a dedicated monitor collecting the variables on CPB and producing the MDPI. The activities are separated into 3 work-packages (WP).

WP 1: The MDPI has been developed in a single Institution. Additionally, its validation was performed on the same development series using a bootstrap technique. To make this algorithm exportable in different Institutions, a different new series of patients is required (internal validation) and a new series collected in an external Institution (external validation). Additionally, it cannot be excluded that additional risk factors may be identified and included in the MDPI algorithm; moreover, there is the hypothesis that other outcome measures of morbidity and even mortality may be predicted by the MDPI. WP 1 includes the operative Units 1 and 2 and is based on the collection of a new series of consecutive adult patients requiring cardiac surgery with CPB. The study protocol has been already submitted to the Ethics Committee (166/int/2022) and Clinical Trial. Gov for the internal validation at the operative unit 1 and includes 400 patients. An additional amount of 400 patients will be collected at the operative unit 2. In this series, the MDPI parameters will be collected and assessed for discrimination and calibration properties in predicting CSA-AKI (defined as AKI of any kind, AKI stage and AKI stage 2 or greater). Appropriate tools will be applied to define discrimination (ROC analysis) and calibration (calibration plot using LOWESS) properties of the MDPI. This WP is essentially a validation of the existing MDPI as patented by IRCCS Policlinico San Donato WP2: The MDPI has been develop in the adult patient population. There is little information available in the literature about the CSA-AKI risk factors in infants and newborns weighing < 20 kgs. However, CSA-AKI in this segment of population is present at a rate that is equal or even higher than in the adults. It can be hypothesized that above 20 kgs, the patient is probably comparable to the adult patient, whereas there is a gap in knowledge in infants and newborns. WP2 is intended to cover this gap in knowledge by addressing a series of 200 patients weighing < 20 kgs and receiving cardiac surgery with CPB for palliation or correction of congenital heart defects, producing an MDPI for infants (I MDPI). This WP will be totally performed at the operative unit 2, that is the largest congenital heart center in Italy. A new patent on the I-MDPI is anticipated.

WP3: This WP is based on the same patient population of WP 1, but has a complementary aim. It is in fact possible that (a) other factors apart from the seven included in the MDPI may be associated with CSA-AKI, therefore deserving to be included in the model (MDPI 2.0) and (b) other outcomes, and namely 30-days mortality may be predicted by the MDPI. Once defined these aims, this WP (that includes the 2 operative units) includes the preliminary steps for the implementation of the MDPI into existing or newly developed monitoring systems for CPB. Dedicated patents are anticipated for MDPI 2.0.

Inclusion Criteria:

• patients undergoing cardiac surgery with cardiopulmonary bypass
• patients >18 years for WP1 and WP3; patients <6 years and weight <20 kg for WP2
• willingness to participate and sign the informed consent

Exclusion Criteria:

• chronic renal failure on dyalisis for WP1 and WP3
• emergency surgery for all WP
• refusal to partecipate

Patients hospitalized in our Institution for a scheduled cardiac surgery with cardiopulmonary bypass

The original dataset supporting the findings of this study will be deposited in the public repository Zenodo after the publication of the work and accessible upon a reasonable request. The requests should be addressed to the Principal Investigator of the study.