Transfusional practices evolved significantly over the last decades, but there are still important controversies regarding triggers that should be adopted in different clinical scenarios. Most international guidelines recommend using a hemoglobin (Hb) level around 7,0-8,0g/dL as the value to prompt a transfusion of red blood cell concentrates (RBC). Critical care patients usually are in a hyperdynamic state, working with an elevated cardiac output and compromised organ function. In these patients, the dependency on the arterial content of oxygen is greater, making lower Hb levels more associated with organ disfunction and compromised homeostasis.

With this study the investigators hope to help clinicians to make decisions regarding transfusion of RBCs in critical surgical patients, establishing a transfusional trigger, without exposing patients to unnecessary additional risks, in the scenario involving patients with cancer, in post-operative care.

This is a prospective, randomized, controlled, interventional trial, with the aim of evaluating the impact of restrictive versus liberal transfusional strategy on mortality and severe clinical complications in post-operative oncologic critically ill patients. The primary outcome is mortality in 30 days. The interventions consist in transfusion of RBCs according to the allocation to a liberal or restrictive transfusional strategy. In the restrictive strategy arm patients will receive transfusion of RBCs if the Hb falls to a level equal to or below 7,0g/dL. In the liberal strategy arm patients will receive transfusions if Hb level is below or equal to 9,0g/dL. In both arms patients should receive only one unit of RBC per time, with measurement of Hb level after three hours to evaluate the need for additional units. The strategy should be maintained during intensive care unit (ICU) stay for a maximum of 90 days. In case of a permanence in the ICU for a period longer than 90 days, or if the patient is discharged from the ICU, the transfusional support will be determined by the assisting physicians, independently of the allocated study arm. If the patient returns to the ICU during the 90 days of randomization, then he should go back to receiving transfusions according to the liberal or restrictive strategy in use previously in the ICU.

Randomization: patients will be randomized in a 1:1 ratio to one of the transfusion strategies, stratified according to age ≤65 years or > 65 years.

Monitoring and follow-up: Study investigators will collect all of the necessary data with the use of specific forms during patient study follow up. In case of transfusional reactions related to the RBC transfusions the ICU team should contact the study investigators to notify the reaction. In cases where the termination of participation in the study occurs before completion of 90 days since randomization, a study investigator will contact the patient for collection of information of the final follow-up form. An independent data and safety monitoring committee (DMSC) will review study data every 6 months to check for the need for suspending or terminating the study.

Blinding: It will not be feasible to mask the assigned transfusion strategy from health care providers. Information regarding frequency of outcome measures will not be available to the study investigators or health care providers, to minimize comparison of outcomes between study groups. Trial statistician will be blinded for the allocation during analysis. The members of the DMSC will remain blinded unless otherwise requested after the interim analysis provides strong indications of one intervention being beneficial or harmful.

Interim analysis: an interim analysis will be performed when a total of 420 patients (half of the expected target sample) has completed 90 days of follow-up. The independent DMSC will recommend interruption of the trial if the difference in the primary outcome measure between groups has a P <0.001 (Haybittle-Peto criterion).

The trial protocol may be temporarily suspended for an individual patient in case of arterial ischemic events (includes stroke, myocardial infarction, unstable angina, mesenteric ischemia, peripheral ischemia) or life-threatening bleeding, at the discretion of the attending physician. The patient may re-enter the trial protocol after stabilization, at the discretion of the attending doctor.

Inclusion Criteria:

• Patient with a solid tumor and age ≥ 18 years
• Admission to the ICU due to postoperative care after being submitted to a major surgical procedure (defined as having a duration of three hours or more)
• Expected ICU stay of more than 24 hours
• Presenting Hb value of 9,0g/dL or less during ICU stay, within the first 7 days post-surgical procedure

Exclusion Criteria:

• Hematological malignancy
• Chronic anemia (defined as presenting Hb ≤9,0g/dL at hospital admission and / or diagnosed hematological disease)
• Presence of active bleeding at the moment of inclusion in the study (defined as visible bleeding with fall of 2,0g/dL in Hb levels in the last 12 hours or need to reoperate)
• Transfusion of RBC during the current ICU stay before the inclusion in the study
• Patient refusal to receive blood transfusions
• Pregnancy
• Postoperative care of cardiac surgery or surgery for correction of hip fracture
• Patients with terminal oncological disease in palliative care
• Arterial ischemic event in the current hospital stay or in the past twelve months (includes stroke, myocardial infarction, unstable angina, mesenteric ischemia, peripheral ischemia)

It will not be feasible to mask the assigned transfusion strategy from health care providers. Information regarding frequency of outcome measures will not be available to the study investigators or health care providers, to minimize comparison of outcomes between study groups. Trial statistician will be blinded for the allocation during analysis. The members of the DMSC will remain blinded unless they request otherwise after the interim analysis provides strong indications of one intervention being beneficial or harmful.