In vitro and in vivo assessment of oxygenator blood volume for the prediction of clot formation in an ECMO circuit (theory and validation).

INTRODUCTION: Clotting is one of the major causes of mortality and morbidity during extracorporeal membrane oxygenation (ECMO). A large meta-analysis study suggests that 29% of patients require the oxygenator to be replaced during ECMO. As clots usually form in the oxygenator, the oxygenator blood volume (OXBV) decreases over time. The currently used pressure gradient as a predicator of clot formation is unreliable. OBJECTIVE: The aim of this study was to develop and validate ultrasound dilution technology in a quantitative assessment of clotting, using measurements of OXBV. METHODS: OXBV was measured using the ELSA monitor (Transonic Systems Inc., Ithaca, NY, USA) from the transit time of a saline bolus passing through the oxygenator as recorded by a sensor placed after the oxygenator. The accuracy and reproducibility (coefficient of variation [CV]) of OXBV measurement and its independence from ECMO flow was assessed in vitro in lambs and from a clinical data archive. RESULTS: The in vitro accuracy compared with volumetric measurements of OXBV of 22-134 ml at flows of 300-700 ml/min was -0.8±6.6%. For an OXBV of 355 ml at flows of 1020-7000 ml/min, accuracy was -0.4±1.6%. In 88 animal OXBV measurements, the CV was 1.49±1.12%. For an OXBV of 153 (range 42-387 ml), clinical measurements at flow ranged from 210-5960 ml/min, with a CV of 3.20±2.44 %. CONCLUSION: Dilution technology has the ability to accurately and reproducibly assess the clotting process in the oxygenator. Larger studies are needed to establish guidelines for the prediction of imminent clotting and may help to avoid unnecessary circuit changes.

Pump in Parallel-Mechanical Assistance of Partial Cavopulmonary Circulation Using a Conventional Ventricular Assist Device.

Mechanical assistance of systemic single ventricle is effective in pulling blood through a cavopulmonary circuit. In patients with superior cavopulmonary connection, this strategy can lead to arterial desaturation secondary to increased inferior caval flow. We hypothesized that overall augmentation in cardiac output with mechanical assistance compensates for the drop in oxygen saturation thereby maintaining tissue oxygen delivery (DO2). Bidirectional Glenn (BDG) was established in seven swine (25 kg) after a common atrium had been established by balloon septostomy. Mechanical circulatory assistance of the single ventricle was achieved using an axial flow pump with ventricular inflow and aortic outflow. Cardiac output, mean pulmonary artery pressure (PAP), common atrial pressure (left atrial pressure [LAP]), arterial oxygen saturation (SaO2), partial pressure of arterial oxygen (PaO2), and DO2 were compared between assisted and nonassisted circulation. Significant augmentation of cardiac output was achieved with mechanical assistance in BDG circulation (BDG: median [interquartile range {IQR}], 0.8 [0.9-1.15] L/min versus assisted BDG: median [IQR], 1.5 [1.15-1.7] L/min; p = 0.05). Although oxygen saturations and PaO2 trended to be lower with assistance (SaO2; BDG: median [IQR], 43% [32-57%]; assisted BDG: median [IQR], 32% [24-35%]; p = 0.07) (PaO2; BDG: median [IQR], 24 [20-30] mm Hg; assisted BDG: median [IQR], 20 [17-21] mm Hg; p = 0.08), DO2 was unchanged with mechanical assistance (BDG: median [IQR], 94 [35-99] ml/min; assisted BDG: median [IQR], 79 [63-85] ml/min; p = 0.81). No significant change in the LAP or PAP was observed. In the setting of superior cavopulmonary connection/single ventricle, the systemic ventricular assistance with a ventricular assist device (VAD) leads to increase in cardiac output. Arterial oxygen saturations however may be lower with mechanical assistance, without any change in DO2.

Effect of mechanical assistance of the systemic ventricle in single ventricle circulation with cavopulmonary connection.

BACKGROUND: Previous attempts to support single ventricle circulation mechanically have suggested that a custom-built assist device is needed to push, rather than pull, through the pulmonary circulation. We hypothesized that using a conventional ventricular assist device, with or without conversion of a total cavopulmonary connection to a bidirectional Glenn cavopulmonary connection, would allow assistance by pulling blood through the circuit and improve the cardiac index (CI). METHODS: Cavopulmonary connections were established in each of 5 Yorkshire pigs (25 kg) using ePTFE conduits in a Y configuration with appropriate clamping of the limbs of the Y to achieve a total cavopulmonary Fontan connection (TCPC), superior vena cava cavopulmonary connection (SVC Glenn), and inferior vena cava cavopulmonary connection (IVC Glenn). A common atrium had been established previously by balloon septostomy. Mechanical circulatory assistance of the single systemic ventricle was achieved using a centrifugal pump with common atrial inflow and proximal ascending aortic outflow. The CI was calculated using an ultrasonic flow meter placed on the distal ascending aorta and compared between the assisted and nonassisted circulation for 3 conditions: TCPC, SVC Glenn, and IVC Glenn. The mean pulmonary artery pressure, common atrial pressure, arterial oxygen saturation, partial pressure of arterial oxygen, and oxygen delivery were calculated. RESULTS: The unassisted SVC Glenn CI tended to be greater than the TCPC or IVC Glenn CI. Significant augmentation of total CI was achieved with mechanical assistance for SVC Glenn (109% ± 24%, P = .04) and TCPC (130% ± 109%, P = .01). The assisted CI achieved at least a mean baseline biventricular CI for all 3 support modes. Oxygen delivery was greatest for assisted SVC Glenn (1786 ± 1307 mL/L/min) and lowest for TCPC (1146 ± 386 mL/L/min), with a trend toward lower common atrial and pulmonary artery pressures for SVC Glenn. CONCLUSIONS: SVC bidirectional Glenn circulation might allow optimal augmentation of the CI and oxygen delivery in a failing single ventricle using a conventional pediatric ventricular assist device. The results from our model also suggest that the Fontan circulation itself can be supported with systemic ventricular assistance of the single ventricle.

Comparison of the effect of venovenous versus venoarterial extracorporeal membrane oxygenation on renal blood flow in newborn lambs.

UNLABELLED: Venovenous extracorporeal membrane oxygenation (VV ECMO) using double lumen catheters is an alternative to venoarterial (VA) ECMO and allows for total blood flow using the patient's cardiac output in comparison to partial blood flow provided during VA ECMO. OBJECTIVE: To compare the effects of VV versus VA ECMO on renal blood flow. DESIGN: Prospective study. SETTING: Research laboratory in a hospital. SUBJECT: Newborn lambs 1-7 days of age (n = 15). INTERVENTIONS: In anesthetized, ventilated lambs, femoral artery and vein were cannulated for monitoring and renal venous blood sampling. An ultrasonic flow probe was placed on the left renal artery for continuous renal blood flow measurements. Animals were randomly assigned to control (non-ECMO), VV ECMO and VA ECMO groups. After systemic heparinization, the animals were cannulated and studied at bypass flows of 120 mL/kg/min (partial bypass) for two hours in both ECMO groups and 200 mL/kg/min (full bypass) for an additional 30 min in the VA group. Changes in blood pressure and renal flow on ECMO and during ECMO bridge unclamping were recorded continuously. Plasma renin activity (PRA) levels were sequentially sampled. RESULTS: Systemic blood pressure was not different in VV or VA ECMO at partial bypass flow. However, systemic blood pressure increased significantly at maximal bypass flow in the VA ECMO group. There was no change in renal flow in either VV or VA ECMO groups. PRA levels did not correlate with bypass flow change. During unclamping of the ECMO bridge, blood pressure and renal flow drop significantly in the VA group, but not in the VV group. CONCLUSION: VV and VA ECMO at partial bypass flows had comparable effect on blood pressure, renal blood flow and PRA level in this short-term study. However, unclamping of the ECMO bridges did differentially affect blood pressure and renal blood flow between VV and VA groups. We speculate that this repeated acute change in long-run VA ECMO support may play a role in the persistent hypertension seen in some patients.

Improved oxygenation with reduced recirculation during venovenous ECMO: comparison of two catheters.

OBJECTIVES: To determine whether the new double-lumen catheter made by OriGen Biomedical (Austin, TX) for venovenous (VV) extracorporeal membrane oxygenation (ECMO) would reduce recirculation and improve oxygenation during VV ECMO when compared with the Kendall double-lumen catheter (Kendall Healthcare Products, Mansfield, MA). DESIGN: Prospective intervention study. SETTING: The animal research laboratory at Children's National Medical Center, Washington, DC. SUBJECTS: Nine newborn lambs one to seven days old and weighing 4.4 +/- 0.8 kg. INTERVENTION: Animals were anesthetized, intubated, and ventilated. The ductus arteriosus was ligated. Femoral arterial and venous, cephalic jugular vein, and pulmonary artery catheters were placed. After systemic heparinization, the catheter to be tested, an OriGen catheter, was placed in the right internal jugular vein and advanced into the right atrium. The animal was placed on ECMO and stabilized, with the ventilator settings decreased to a peak inspiratory pressure of 15-20 cmH2O, peak end-expiratory pressure of 5 cmH2O, rate of 15-25 breaths/min, and a fractional inspired oxygen concentration of 0.21-0.30. ECMO flows were increased in 100-ml increments from 200 to 600 ml/min with measurements taken 15 min after each change. The OriGen catheter was removed, the Kendall catheter was placed, and the studies were repeated. MEASUREMENTS AND MAIN RESULTS: Heart rate, mean blood pressure, PaO2, jugular cerebral oxygen saturation, pulmonary artery oxygen saturation, pump venous oxygen saturation, and postmembrane circuit pressures were measured at each study period. The OriGen catheter improved oxygenation, with higher systemic PaO2, higher pulmonary artery and cerebral oxygen saturations, and lower pump venous oxygen saturations (indicating less recirculation). With the OriGen catheter, PaO2 levels ranged from 69 +/- 18 mmHg [9.2 +/- 2.4 kPa] to 114 +/- 45 mmHg [15.2 +/- 6.0 kPa], compared range from 61 +/- 15 mmHg [8.1 +/- 2.0 kPa] to 87 +/- 34 mmHg [11.5+/-4.5 kPa] for the Kendall catheter. These findings indicate that, at all flow rates studied, less recirculation occurred with the OriGen catheter than with the Kendall catheter. The postmembrane pressures were significantly lower for the OriGen catheter at any given flow (from 30 +/- 5 to 122 +/- 18 mmHg) when compared with the Kendall catheter (from 77+/- 16 to 330+/-78 mmHg). CONCLUSIONS: These findings indicate that the OriGen catheter resulted in a reduction of recirculation, thereby resulting in an improvement in oxygenation while on VV ECMO. The lower postmembrane pressure potentially could reduce the risk of ECMO circuit complications such as tubing rupture, bleeding complications, as well as hemolysis. This new catheter makes VV ECMO more effective and represents a design that could be used for neonatal and/or pediatric ECMO.

Continuous blood gas monitoring using an in-dwelling optode method: comparison to intermittent arterial blood gas sampling in ECMO patients.

INTRODUCTION: The ability to measure postmembrane arterial blood gases is essential in the management of critically ill neonates treated with extracorporeal membrane oxygenation (ECMO). A new technology using, the Paratrend 7 system (Diametrics Medical, High Wycombe,UK) allows for continuous measurement of pH, PCO(2) and PO(2), and calculates oxygen saturation, bicarbonate, and base excess. OBJECTIVE: To evaluate and compare the results of continuous blood gas measurement using the Paratrend 7 system with a standard system of blood gas analysis in our intensive care unit. DESIGN: Prospective, controlled, interventional study. SETTING: The neonatal intensive care unit of a tertiary referral center. PATIENTS: Neonates who required extracorporeal life support and were expected to have frequent postmembrane arterial blood sampling during the testing period. RESEARCH DESIGN AND METHOD: To enable Paratrend 7 sensor access to the ECMO circuit, the postmembrane access port extension set that is routinely used for blood drawn for blood gas analysis was used. The study began with the insertion of the Paratrend 7 sensor. Subjects remained on the study until the ECMO was discontinued and/or frequent blood gases were no longer needed. The blood gas results from the Paratrend 7 system were not used in clinical management of the patient. BLOOD GAS MEASUREMENT: During the study period, with each blood sample drawn for laboratory analysis, a printout from the Paratrend 7 monitor was recorded for comparison. RESULTS: A total of 242 pairs of blood gas samples were collected from 10 neonates. The mean bias/precision for pH was -0.02/0.04; for PO(2) 68.35/93.44 mm Hg; and for PCO(2) 1.75/4.23 mm Hg. The correlation (r value) between the sensor reading and the blood gases were 0.89 for pH, 0.96 for PO(2), and 0.73 for PCO(2) (Table 1). CONCLUSION: The blood gases compared in the two methods had a strong correlation for pH, PCO(2) and PO(2). Results of this study indicate that this technology provides an accurate means of monitoring continuous blood gas parameters in neonatal ECMO patients. Use of the Paratrend 7 should allow reduced health-care provider exposure to blood and decreased patient iatrogenic blood loss.

Continuous blood gas monitoring using an in-dwelling optode method: clinical evaluation of the Neotrend sensor using a luer stub adaptor to access the umbilical artery catheter.

INTRODUCTION: Arterial blood gases are essential in the management of critically ill neonates. A new technology using the Neotrend system (Diametrics Medical) allows for continuous measurement of pH, PaCO(2), and PaO(2), and calculates oxygen saturation, bicarbonate, and base excess. OBJECTIVE: To evaluate and compare the results of continuous blood gas measurement using the Neotrend system with a standard system of blood gas analysis in our intensive care unit. DESIGN: Prospective, controlled, interventional study. SETTING: The neonatal intensive care unit of a tertiary referral center. PATIENTS: Neonates with respiratory distress who required respiratory support and frequent arterial blood gas sampling and had a UAC. RESEARCH DESIGN AND METHOD: To enable Neotrend sensor access to an existing Argyle umbilical artery catheter (UAC) the catheter was cut at the 25-cm mark and connected to an 18-gauge blunt needle luer stub adaptor (Vygon 95440). The study began with the insertion of the Neotrend sensor. Subjects remained on the study until the UAC was discontinued and/or frequent blood gases were no longer needed. The blood gas results from the Neotrend system were not used in clinical management of the patient. BLOOD GAS MEASUREMENT: During the study period, with each blood sample drawn for laboratory analysis, a printout from the Neotrend monitor was recorded for comparison. RESULTS: A total of 217 pairs of blood gas samples were collected from seven neonates. The mean bias/precision for pH was 0.01/0.04; for PaO(2) 0.72/18.5 mm Hg; and for PaCO(2) 3.96/2.63 mm Hg. The correlation (r value) between the sensor reading and the blood gases were 0.85 for pH, 0.96 for PaO(2), and 0.92 for PaCO(2). CONCLUSION: The blood gases compared in the two methods had a strong correlation for pH, PaCO(2), PaO(2), and oxygen saturation. Although the bicarbonate and base excess values showed suboptimal statistical correlation, the difference was not clinically relevant. Results of this study indicate that this technology provides an accurate means of monitoring continuous blood gas parameters in neonatal patients. It also allows reduced healthcare provider exposure to blood and decreased patient iatrogenic blood loss.