Percent of patients undergoing acute normovolemic hemodilution by weight group and year. 

Context in source publication

Context 1

... prime (RAP), venous antegrade prime (VAP), and prime volume were recorded. Circuit selection was based on both the proce- dure and a target calculated blood flow rate of 2.2 L/min/m 2 . Arterial-venous loop selection was from the following: 1/8 2 × 3/16 2 , 5/32 2 × 1/4 2 , 3/16 2 × 1/4 2 , 1/4 2 × 3/8 2 , 3/8 2 × 3/8 2 , and 3/8 2 × 1/2 2 . Suckers were 1/8, 3/16, or 1/4 2 depending on the par ticular circuit selected. All tubing was coated, where possible, with Terumo X coating TM (Terumo Cardiovascular, Ann Arbor, MI) down to 3/16 2 . Oxygenators were selected from the following: Terumo RX05 ® /FX05 ® , Terumo SX10 ® or RX15 ® , or Terumo SX18 ® or RX25 ® (Terumo Cardiovascular, Ann Arbor, MI) or Maquet Quadrox (Maquet Cardiopulmonary, Hirrlingen, Germany), depending on patient size. No arterial line filter (ALF) was used for the two smallest circuits: 1/8 2 × 3/16 2 and 5/32 2 × 1/4 2 ; otherwise a Terumo Capiox ® CXAF02, 40 mL prime, CXAF200X, 200 mL prime, or CXAF125, 125 mL prime ALF (Terumo Cardiovascular, Ann Arbor, MI) was used. CPB was performed with the Maquet HL-20 (Maquet Cardiopulmonary, Hirrlingen, Germany) using roller pumps for most procedures. Continuous arterial blood gas and venous saturation/ hematocrit monitoring with the CDI 500 (Terumo Cardiovascular, Ann Arbor, MI) was used on all procedures. All patients had cerebral saturation monitoring with the INVOS 5100 B or C monitor (Somanetics Corporation, Troy, MI). Arterial and venous blood gas, electrolyte, glucose, lactate, and hematocrit (Hct) were measured with the I-Stat ® (Abbott Point of Care, Princeton, NJ). Anticoagulation was monitored with the Hemochron Jr ® (ITC, Edison, NJ) and Medtronic Heparin Management System ® (Medtronic, Inc., Minneapolis, MN). Cell salvage with the Fresenius CATS (Terumo Cardiovascular, Ann Arbor, MI) continuous auto- transfusion system was used on all procedures. Perioperative acute normovolemic hemodilution was considered on all patients starting in 2006. After sampling and documenting a baseline Hct in the OR of ≥ 30%, autologous blood was removed through the arterial or central line, as tolerated, and replaced with .9% normal saline, lactated Ringer’s solution, Normosol ® -R (Hospira, Lake Forest, IL), or plasma protein fraction at the discretion of the anesthesiologist. Autologous blood was anticoagulated with 10 mL of Anticoagulant Citrate Dextrose Solution, Solution A (ACD-A), per 100 mL of autologous blood. Autologous blood was labeled per institutional standard, stored at room temperature, and transfused by anesthesia post-protamine administration unless needed during bypass secondary to low Hct and/or venous saturation. After arterial cannula insertion, RAP was initiated by controlled aspiration on the CPB arterial line (retrograde) into a syringe or transfer bag as previously describe by Rousou et al. (14). Crystalloid prime solution is displaced with the patient’s blood via the arterial manifold purge port back as far as the oxygenator in our two smallest circuits, or to the ALF in the larger circuits. VAP began after a venous cannula was inserted and connected to the venous line. Two clamps were placed on the venous line. One clamp was placed across the venous line at approx- imately 75% occlusion. A second, fully occlusive clamp, was intermittently opened and closed which allowed the crystalloid prime to drain from the venous line and was displaced by the patient’s blood. The crystalloid prime was simultaneously pumped out of the reservoir into a syringe or blood transfer bag. This process continued until the patient’s blood had replaced the crystalloid prime in both the oxygenator and ALF, if present, or as long as tolerated. Trasylol (Bayer Healthcare Pharmaceuticals, Wayne, NJ) was used on all patients through April 2008. Starting in May 2008, Tranexamic acid was used at 100 mg/kg load to the patient and 100 mg/kg added to the pump prime, along with a steroid, dexamethasone, 1 mg/100 mL of pump prime to a maximum of 10 mg. All drugs were added to the bypass circuit after completion of RAP and VAP. Target Hct on CPB was ≥ 20% with notification of the surgeon if lower. Blood was only administered after consul- tation with the surgeon. If RBCs were used, they were first mixed with at least 500 mL Normosol ® -R and washed with the cell saver. Over 95% of all patients underwent arterial- venous modified ultrafiltration (MUF) and conventional ultrafiltration or dilutional ultrafiltration during CPB. Upon completion of MUF all circuits were flushed with one liter of Normosol ® -R to the cell saver for processing. From January 2003 through December 2008 there were 1293 procedures requiring CPB, 633 between 2003 and 2005, and 660 between 2006 and 2008. Table 1 shows the number of procedures by weight category. Prior to 2006 ANH was used on less than 1% of all patients. ANH use increased over the 3 year period from 2006–2008, and was used on 42% of all patients and 12% of patients 0–6 kg ( Figure 1 ) increasing from 6% in this weight group in 2006 to 16.7% in 2008. For the patients greater than 20 kg, ANH use averaged over 60%. RAP/VAP was used on less than 1% of patients prior to 2006. For the three year period 2006–2008, RAP/VAP averaged 70% for all patients and 38% of patients 0–6 kg, increasing from 28–55% over the 3 year period ( Figure 2 ). For the patients greater than 6 kg, RAP/VAP averaged over 80% for each of the weight groups. Figure 3 shows the bloodless surgical cases for the 6 year period 2003–2008. Bloodless surgery was performed on 30% of all patients from 2003–2005 and 43% of patients from 2006–2008. Bloodless surgery more than doubled for the 0–6, 6–15, and 15–20 kg groups from 3.5%, 23%, and 23% respectively in 2003–2005 to 9%, 44%, and 58%, respectively in 2006–2008. There is essentially no literature available describ- ing the results of a programmatic approach to bloodless cardiac surgery in a pediatric center. Through our early development stage of a program of bloodless techniques for the Jehovah’s Witness patient, we found this group of patients experienced low morbidity and mortality, which motivated us to explore these techniques for all patients. Ootaki et al., in a prospective, non-continuous study of 75 pediatric patients undergoing cardiac surgery using a criteria-driven transfusion protocol, achieved 70% transfusion free procedures (9). Mean weight of the transfusion free group was 24.6 ± 13.4 kg. In our series of patients from 2006–2008, we were able to achieve 58% of the CPB patients transfusion free in the 15–20 kg weight range and 78% in the 20–40 kg range. Durandy did a retrospective review of transfusion of 259 consecutive patients weighing < 15 kg that underwent open-heart surgery. In the group of patients weighing less than 6 kg, they were unable to accomplish bloodless procedures. However, of those receiving blood, only 4% required greater than one unit of RBCs (15). Our similar weight group of 0–6 kg, with 185 mL prime volume, which included cardioplegia and ultrafilter, achieved bloodless surgery in 9%, and 45% for patients 6–15 kg. While not reported in this study, we would subjectively concur with Durandy, that when blood is administered in these patients, there is a greater Hct increase with a single unit of RBCs. This increase is accomplished by using smaller size circuits and results in overall less blood transfusion when needed. Boettcher et al. and Ging et al., among others, have written single case reports of bloodless surgical techniques on neonates and infants (8,12). Similarly, we have accomplished bloodless surgery in neonates 3–5 kg, including arterial switch procedures, comprehensive stage II for hypoplastic left heart syndrome, and heart transplants. It is within the group of infants £ 5 kg that is the most chal- lenging and difficult to accomplish without blood. Hübler et al. have reported transfusion free CPB procedures in patients of 1.7 and 2.2 kg with the use of mast mounted pumps, which were remarkable accomplishments (10,13). While hardware may help with circuit miniaturization, it is possible to miniaturize with a traditional heart-lung machine and without the use of mast mounted pumps as we have done. Circuit selection, and thus, prime volume, is extremely important and must take into account the “total” circuit prime volume, including the cardioplegia/ultrafilter/ MUF circuit. Tubing diameter and length, including the position of the pump, oxygenator, and circuit, relative to the table, patient, and surgeon, is significant and may con- tribute to overall circuit prime volume more than the oxygenator selection. The two smallest circuits we used (1/8 × 3/16 and 5/32 × 1/4) are without an ALF secondary to the amount of hemodilution added. This is becoming less of an issue with the introduction of the new Terumo FX (Terumo Cardiovascular, Ann Arbor, MI) series oxygenators, which incorporate an arterial screen filter around the oxygenator fiber bundle without adding additional prime volume. The transfusion of allogeneic blood is associated with increased morbidity and evidence implicating the potential of allogeneic blood as an instigator of inflammation (1–3,16,17). Between the potential of inflammation from blood, and that from cardiopulmonary bypass, it is imperative that circuit miniaturization is optimized for the indi- vidual patient to reduce inflammatory response. The use of ultrafiltration during and after CPB to potentially reverse the effects of both of these potential triggers should be considered for all patients (18). The use of ANH in the OR, and RAP/VAP are not easily accomplished, especially when applying the technique to neonates. Murayama et al. have shown that pre-operative and intraoperative blood donation is safe for pediatric cardiac surgical patients, but this was for patients 5–10 years of age (19). It is imperative that the perfusion-anesthesia- surgeon team work closely together, and communicate for ...