Knee arthroplasty is performed with restriction of blood flow by use of the pressure tourniquet. Ischemia of skeletal musculature is accompanied by the formation of reactive oxygen species and release of vasoactive factors. Inhaled anesthetics were described as possible protective factors, termed "ischemic preconditioning". The objective of this investigation is to observe the effects of preconditioning with inhaled anesthetic on bleeding in the postoperative period of knee arthroplasty surgery. The study will be prospective and randomized. Patients will be allocated by means of a randomized table in two groups: Group Preconditioning (GP) - subarachnoid and general anesthesia with preconditioning with 1 minimal alveolar concentration (MAC) for 15 minutes before the installation of the ischemia by tourniquet; Group control (GC) - subarachnoid and general anesthesia with sevoflurane without preconditioning. Will be evaluated: hemoglobin and hematocrit, platelet count and coagulogram, electrolytes and creatinine, lactate and glycemia. Volume of drainage will be evaluated until drainage is withdrawn and need for transfusion measuring the number of bags of blood concentrate and volume transfused. Pain assessment will be done by means of visual analog scale. Morbidity will be recorded up to the 30th postoperative day by chart analysis. Patients will be allocated to the study groups by random number table. Data will be presented descriptively, frequently, mean and standard deviation, or medians and distribution in 25-75% percentiles. The comparison between continuous measurements will be made by means of repeated measures intra-group analysis of variance (ANOVA) and Student's t-test for evaluation between groups. The need for transfusion will be compared by Fisher's exact test. Pain intensity will be compared by Wilcoxon test for non-parametric data. A commercially available statistical package - SPSS version 15.0 (SPSS Inc.) will be used. A value of P <0.05 will be considered statistically significant.

The use of pneumatic tourniquet allows a clear surgical field by restricting blood flow to the area to be operated, which implies at least intraoperative bleeding, but with a high risk of postoperative bleeding.

The late bleeding associated with tourniquet use appears to be due to the imbalance between coagulation and fibrinolysis. The use of antifibrinolytics, such as tranexamic acid and epsilon aminocaproic acid, has been shown to reduce postoperative bleeding, reducing the need for transfusion of blood components postoperatively, without increasing the risk of thromboembolism. The effects of tourniquet, which can be maintained at intervals of 75 to 100 minutes, include elevation in blood pressure while it is insuflated, cardiac index elevation during ischemia and its subsequent elevation at the moment of release to the circulation, at which time there is also an increase in oxygen consumption as well as the release of CO2. This increase in oxygen consumption is directly related to the time of ischemia.

Skeletal muscles's ischemia is accompanied by anaerobic glycolysis, formation of reactive oxygen species, and release of vasoactive factors. Inhaled anesthetics, such as isoflurane and sevoflurane, have been described as possible protective factors for both cardiac and skeletal muscles, although the literature still suggests the need for additional studies. Its use in the period before the onset of ischemia is called "ischemic preconditioning".

The objective of this investigation will be to observe the effects of an inhalational anesthetic, commonly used in anesthesia on bleeding, in the results of coagulation tests and in the consumption of blood components in the postoperative period of knee arthroplasty surgery.

Goals

To compare the effects of preconditioning with sevoflurane in total knee arthroplasty surgery by:

1. Coagulogram for coagulation times;
2. Volume of blood drainage in the postoperative period;
3. Need and volume of blood transfusion;
4. Dosage of arterial pH, lactate and CPK enzyme;
5. Immediate postoperative morbidity and up to 30 days postoperatively.

Statistic

Data will be presented by descriptive tables and frequency, mean and standard deviation, or medians and distribution in 25-75% percentiles for data that do not find a normal distribution. The Kolmogorov-Smirnov test will determine if there is a normal distribution.

Values for hemoglobin, hematocrit, platelet count, arterial blood gas values, electrolytes, creatinine, CPK, lactate, blood glucose and coagulogram will be compared by analysis of variance (ANOVA).

The intragroup analyzes will use repeated measures ANOVA and the Dunnett test if there is a significant difference between the moments of the same group, since the test may be better considered for comparison with a control (preoperative) time. Between the groups, the values will be compared moment by moment with Student's t-test.

The need for transfusion, considering the number of CC bags, will be compared by Fisher's exact test and its volume by Student's t-test. Pain intensity in the postoperative period will be evaluated by visual analogue scale and compared by Wilcoxon's test for non-parametric data between the time of entry and discharge of RPA. A commercially available statistical package - SPSS version 15.0 (SPSS Inc.) will be used. A value of P <0.05 will be considered statistically significant.

Patients will be allocated to the study groups by random number table. Briefly, generating a list of two-digit numbers determines the order of entry of patients into the study groups. A minimum expected difference between the means of the treatments (groups) established at 500 ml for the required volume of postoperative transfusion, standard deviation at 400 ml, with test power at 80% and significance level in 5% suggesting a distribution of 15 patients in each group, considering losses in 50%.

Inclusion Criteria:

• indication for knee arthroplasty

Exclusion Criteria:

1. Physical state according to classification from American Association of Anesthesiologists (ASA) III or higher,
2. Obesity (body mass index, BMI, greater than 30),
3. Renal insufficiency - patients on a dialysis program or with a serum creatinine level above 1.4 mg / dL,
4. Hematocrit less than 30% or hemoglobin less than 10g / dL,
5. Records of myocardial infarction during the last 6 months, unstable angina pectoris,
6. History of coagulation disorders,
7. Use of oral anticoagulants, heparin or acetylsalicylic acid in the last 5 days.

Statistical analysis will not open label for group allocation