Oxygen transfer performance of a new Sarns membrane oxygenator

Differential diagnosis of suboptimal oxygenator performance

During cardiopulmonary bypass (CPB), the perfusionist must be able to differentiate between: (1) a normal oxygenator with oxygen transfer reserve; (2) a normal oxygenator without O2 transfer reserve; and (3) a failing or suboptimal oxygenator. The purpose of this paper is to report on the use of the oxygen transfer slope, as well as other evaluation techniques previously described, which aided in the differential diagnosis of suboptimal oxygenator performance. We were able to determine the presence and extent of the dysfunction, follow the progression over time, and assess the effectiveness of our intervention. As a direct result of our ability to carefully monitor the oxygenator, replacement was not necessary despite severe dysfunction.

Comparing Oxygen Transfer Performance Between Three Membrane Oxygenators: Effect of Temperature Changes During Cardiopulmonary Bypass

Recently, a new oxygenator (Dideco 903 [D903], Dideco, Mirandola, Italy) has been introduced to the perfusion community, and we set about testing its oxygen transfer performance and then comparing it to two other models. This evaluation was based on the comparison between oxygen transfer slope, gas phase arterial oxygen gradients, degree of blood shunting, maximum oxygen transfer, and diffusing capacity calculated for each membrane. Sixty patients were randomized into three groups of oxygenators (Dideco 703 [D703], Dideco; D903; and Quadrox, Jostra Medizintechnik AG, Hirrlingen, Germany) including 40/20 M/F of 68.6 +/- 11.3 years old, with a body weight of 71.5 +/- 12.1 kg, a body surface area (BSA) of 1.84 +/- 0.3 m(2), and a theoretical blood flow rate (index 2.4 times BSA) of 4.4 +/- 0.7 L/min. The maximum oxygen transfer (VO(2)) values were 313 mL O(2)/min (D703), 579 mL O(2)/min (D903), and 400 mL O(2)/min (Quadrox), with the D903 being the most superior (P < 0.05). Oxygen (O(2)) gradients were 320 mm Hg (D703), 235 mm Hg (D903), and 247 mm Hg (Quadrox), meaning D903 and Quadrox are more efficient versus the D703 (P < 0.05). Shunt fraction (Qs/Qt) and diffusing capacity (DmO(2)) were comparable (P = ns). Diffusing capacity values indexed to BSA (DmO(2)/m(2)) were 0.15 mL O(2)/min/mm Hg/m(2) (D703), 0.2 mL O(2)/min/mm Hg/m(2) (D903), and 0.18 mL O(2)/min/mm Hg/m(2) (Quadrox) with D903 outperforming D703 (P < 0.0005). During hypothermia (32.0 +/- 0.3 degrees C), there was a lower absolute and relative VO(2 )for all three oxygenators (P = ns). The O(2) gradients, DmO(2) and DmO(2)/m(2), were significantly lower for all oxygenators (P < 0.01). Also, Qs/Qt significantly rose for all oxygenators (P < 0.01). The oxygen transfer curve is characteristic to each oxygenator type and represents a tool to quantify oxygenator performance. Using this parameter, we demonstrated significant differences among commercially available oxygenators. However, all three oxygenators are considered to meet the oxygen needs of the patients.

The proposed use of a 'screening test' to assess oxygenator performance

This study aimed to assess whether the use of the physiological shunt equation could (within the first five minutes of initiating CPB) serve as a 'screen' to differentiate normal and dysfunctional oxygenator performance. If dysfunction severe enough to require replacement was necessary, the normothermic patient could be weaned from CPB and replacement would be carried out under safe, controlled conditions. This technique would require postponing the induction of hypothermia (if used), aortic cross-clamping, and arresting the heart until after this screen was completed. This study demonstrates that a strong negative correlation exists between the degree of blood shunting and the membrane's 0 2 transfer performance (r = -0.874). This relation enables us to predict 0 2 transfer performance when only the shunt fraction is known. Of the 41 oxygenators used in this study, 40 demonstrated normal, or below-normal, shunt fractions. Oxygen transfer performance at or above predicted levels would be anticipated for these oxygenators. One of the 41 oxygenators had mildly elevated shunt fractions, which we predicted would be associated with mild 0 2 transfer dysfunction. Based on the performance screen worksheet we created, replacement was not necessary since the oxygenator maintained high levels of 0 2 transfer in reserve despite its marginal performance dysfunction. Assessment of oxygenator performance dysfunction in this earliest phase of CPB would greatly reduce the incidence of emergency oxygenator replacement secondary to actual or perceived oxygenator failure later in the course of the procedure.

Clinical evaluation of a new generation membrane oxygenator: a prospective randomized study

A new generation hollow-fibre membrane oxygenator (Spiral Gold) has been introduced by Baxter Healthcare (Irvine, CA, USA). The purpose of this study was to evaluate the operational performance of this device under clinical conditions and to compare it to the Univox Gold membrane oxygenator. Following institutional review board approval, and the obtainment of informed consent, 26 patients undergoing coronary artery bypass grafting were randomly assigned to either a Spiral Gold (Spiral) (n = 13) or Univox Gold (Univox) (n = 13) group. Study parameters were grouped into the following categories: haematological, haemodynamic, oxygenator performance and perioperative outcomes. All patients received identical surgical, anaesthesia and postoperative care. There were no statistically significant differences in either preoperative or operative parameters between groups. During cardiopulmonary bypass, the Spiral group had a significantly lower pressure drop (26.9 +/- 8.2 vs 46.7 +/- 16.2 mmHg, p < 0.001). The Spiral group had significantly lower plasma free haemoglobin levels during all time periods of CPB compared to the Univox group. Heat exchange coefficients were higher during the rewarming period in the Spiral patients (0.59 +/- 0.28) compared to the Univox group (0.36 +/- 0.19), p = 0.06. There were no differences in oxygen transfer between groups, but ventilation gas sweep rates and FiO2 levels were statistically lower in the Spiral group at two of the three sampling time periods. The ratio of ventilating gas sweep rate to blood flow rate was lower in the Spiral group (0.56 +/- 0.12) compared to the Univox group (0.74 +/- 0.23), p < 0.03. The Spiral Gold oxygenator had superior oxygen transfer efficiency and lower haemolysis rates than the Univox Gold oxygenator.

Clinical evaluation of six hollow-fibre membrane oxygenators

In a clinical evaluation to study the performance of hollow-fibre membrane oxygenators, we compared the gas exchange characteristics and the production of plasma free haemoglobin for different oxygenators. In this study, the data of the Univox, Cobe-Optima, Capiox-SX, Affinity, Safe II and Sarns Turbo 440 oxygenators were evaluated during cardiac surgery in comparable patient groups. Thirteen patients scheduled for elective surgery were enrolled in each group. In all groups, cardiopulmonary bypass (CPB) was conducted using pulsatile blood flow during fibrillation and the period in which the aorta was crossclamped. Arterial and venous blood gases were determined from which oxygen transfer, carbon dioxide transfer, oxygen gradient, shunt fraction and diffusing capacity were calculated. Blood samples were taken 5, 30, 60 and 90 min after beginning CPB. As a result of our measurements we found that the gas exchange capacities of all six oxygenators are within clinically acceptable limits, although the data considering oxygen and carbon dioxide transfer showed a significantly higher capacity for the Sarns Turbo 440 oxygenator (p < 0.05) in comparison with the other oxygenators. Plasma free haemoglobin was, however, significantly higher in the Univox and Sarns Turbo 440 groups (p < 0.005). This difference was already present after 30 min of bypass and increased with time, and is considered as a negative aspect, in relation to optimal patient care.

Clinical evaluation of the oxygenation capacity and controllability of 15 commercially available membrane oxygenators during alpha-stat regulated hypothermic cardiopulmonary bypass

Oxygenation performance was tested in 15 membrane oxygenators by calculating the resistance for oxygenation (R) and the calculation of oxygen transferred versus FiO2. The clinical data reveal that the SARNS Turbo 440 (36.7-51.9%), the AFFINITY (37.2-50.1%) and the HF 5400 (37.5-52.3%) are the oxygenators with the lowest FiO2 settings for comparable amounts of oxygen transferred during hypothermia and during normothermia followed by MAXIMA Plus (39.1-55.8%), MAXIMA Plus PRF (39.1-56.2%), CAPIOX SX 18 (39.7-61.2%), MONOLYTH (43.0-61.3%), OXIM 11-34 (44.1-63.9%), COBE Duo (44.7-64.9%), COBE Optima (47.4-66.4%), COMPACTFLO (48.3-65.3%), SAFE II (49.0-67.6%), UNIVOX (49.8-71.3%), MAXIMA (50.2-70.1%) and the CM 50 (58.6-77.0%). Similar results were found by calculation of R. HF 5400 (2.41-1.87 mmHg/min/ml O2), AFFINITY (2.63-1.87 mmHg/min/ml O2). OXIM II-34 (2.72-2.45 mmHg/min/ml O2), MAXIMA Plus PRF (2.75-2.07 mmHg/min/ ml O2), COBE Optima (2.83-2.13 mmHg/min/ml O2), UNIVOX (2.84-2.17 mmHg/min/ml O2), MONOLYTH (2.89-2.24 mmHg/min/ml O2), SARNS Turbo 440 (3.03-2.12 mmHg/min/ml O2), MAXIMA Plus (3.09-2.28 mmHg/min/ml O2), SAFE 11 (3.19-2.50 mmHg/min/ml O2), CAPIOX SX 18 (3.27-2.44 mmHg/ min/ml O2), COMPACTFLO (3.41-2.50 mmHg/min/ml O2), MAXIMA (3.53-2.72 mmHg/min/ml O2), COBE Duo (3.57-2.71 mmHg/min/ml O2) and the CM 50 (3.53-2.72 mmHg/min/ml O2). As a measure of controllability of oxygenation, the coefficient of variation on the FiO2 ordered on the normothermic blood samples was used, giving as a result CAPIOX SX 18 (2.9-2.5%), AFFINITY (3.6-3.5%), COBE Duo (4.3-2.9%), HF 5400 (5.7-4.7%), MAXIMA Plus (8.1-5.4%), COMPACTFLO (8.2-5.0%), MONOLYTH (8.3-4.0%), MAXIMA (8.7-3.4%), COBE Optima (9.6-6.8%), SARNS Turbo 440 (10.1-7.3%), MAXIMA Plus PRF (10.9-8.7%), CM 50 (11.9-2.4%), UNIVOX (13.3-8.9%), OXIM 11-34 (15.5-17.3%) and the SAFE II (16.1-9.8%). The low FiO2 settings and the lower resistance for oxygenation are an indication of the reserve capacity of the oxygenators whose importance is proven by clinical data of emergency perfusions on patients under full resuscitation.

Clinical evaluation of the Medtronic Maxima Plus membrane oxygenator

The purpose of this study was to clinically evaluate the degree of improvement, if any, in the oxygen transfer performance of the recently released Medtronic Maxima Plus membrane oxygenator. The outside diameter of the hollow fibres was reduced, increasing the membrane surface area from 2.0 m2 to 2.3 m2 without altering the polycarbonate housing. Maximum extrapolated oxygen transfer of the Maxima Plus (444 ml O2/minute) was increased 23.68% when compared to the Maxima (359 ml O2/minute). When expressed per metre squared of membrane surface area, the Maxima Plus had an increase of 13.5 ml O2/m2/minute (7.24%) over the Maxima. Pressure drop across the Maxima Plus was within 3.5 mmHg of the Maxima over the range of clinical blood flows indicating that the fibre bundle packing density was not significantly altered. Oxygen transfer consistency, expressed as a function of the standard deviation of oxygenator performance index values, was not significantly different for the two oxygenators. We concluded that the improvement in total oxygen transfer was due to an increase in membrane surface area as well as enhanced transfer efficiency per metre squared. We believe that the improved oxygen transfer performance was accomplished without impacting significantly upon the other attributes of the oxygenator (e.g., pressure drop, consistency, priming volume).

Clinical oxygen transfer performance of the Sorin Monolyth membrane oxygenator

Oxygen transfer performance of the Sorin Monolyth membrane oxygenator was evaluated. Similar to previous oxygen transfer performance studies conducted at this institution, our purpose was sixfold: (1) to construct an oxygen transfer slope (OTS); (2) to find the maximum extrapolated oxygen transfer; (3) to calculate the oxygenator performance index (OPI); (4) to generate a shunt fraction line; (5) to determine the percentage of predicted shunt (POPS); and (6) to compare the Monolyth's performance to several previously studied membrane oxygenators. From the OTS, the maximum extrapolated oxygen transfer was 346.4 ml O2/min. This absolute value was the lowest of the four oxygenators compared. When maximum oxygen transfer was compared relative to membrane surface area, the Monolyth ranked third (157.5 ml O2/min). The Monolyth produced a relatively narrow range for the OPI (81.64-130.47%) and had the lowest standard deviation (SD) in this group. The Monolyth exhibited higher shunt fractions over the range of clinical blood flows when compared to our three previously studied oxygenators. The range of POPS values (71.65-128.77%) was relatively narrow and the SD was the lowest of the four. We concluded from our evaluation that the Monolyth had relatively low top end oxygen transfer capabilities, but provided very consistent and predictable oxygen transfer performance.

Clinical evaluation of a new hollow fibre membrane oxygenator

Twelve consecutive patients undergoing elective cardiac surgery were perfused with the Cobe Optima hollow fibre oxygenator. Gas transfer characteristics and blood handling were studied. The device had a maximum oxygen transfer of 315 ml/minute. The average shunt fraction was 4.5%, and was not influenced by blood-flow rate. Mean platelet count declined slightly to 91% of the baseline at the end of the study period. Haemolysis was evaluated by monitoring serum-free haemoglobin, serum haptoglobin and serum haemopexin. The evolution was as follows: free haemoglobin increased from 14 +/- 5 mg/100 ml to 85 +/- 0.8/100 ml (p = 0.01) at the end of bypass; haptoglobin decreased from 1.33 +/- 0.90 g/l to 0.89 +/- 0.15 g/l (p = 0.01); and haemopexin decreased from 0.84 +/- 0.13 g/l to 0.74 +/- 0.15 g/l (p = nonsignificant). In all patients the residual capacity of serum haptoglobin to protect against haemolysis was satisfactory. All patients had an uneventful postoperative course.

Use of the oxygen transfer slope and estimated membrane oxygen transfer to predict PaO2

The purpose of this investigation was threefold: (1) could the perfusionist accurately estimate oxygen transfer (VO2/minute) while on CPB; (2) could this estimate, and its position on the oxygen transfer slope (OTS), predict resultant PaO2 values within a specific range; and (3) could previously derived performance 'normals' be used during this study. Fifty-four sets of samples (both arterial and venous) from 27 oxygenators were used in this study. Each oxygenator provided one normothermic and one hypothermic set of samples. In 48 of the 54 samples (88.9%) the predicted VO2/minute was within +/- 10% of the actual VO2/minute. Thirty-nine of these 48 (81.25%) had resultant PaO2 values within our target range of 140 +/- 30 mmHg. The PaO2 for these 39 samples ranged from 110 to 168 mmHg with a mean of 133 mmHg. The percentage of predicted shunt (POPS) ranged from 59.0 to 192.4% with a mean of 109.3% (SD = 23.81%). With this degree of variability, we concluded that the perfusionist must assess VO2/minute as well as POPS in order to predict the resultant PaO2.