Should Air Bubble Detectors Be Used to Quantify Microbubble Activity during Cardiopulmonary Bypass?

Transcription

1 The Journal of ExtraCorporeal Technology Should Air Bubble Detectors Be Used to Quantify Microbubble Activity during Cardiopulmonary Bypass? Richard F. Newland, BSc, Dip Perf, CCP (Aust);* Robert A. Baker, PhD, Dip Perf, CCP (Aust);* Annette L. Mazzone, BSc(Hons), Dip Perf, CCP (Aust);* Vijaykumar N. Valiyapurayil, MSc, CCP (Ind)* *Cardiac Surgery Research and Perfusion, Flinders Medical Centre and Flinders University, Bedford Park, South Australia Presented at the 29th ANZCP Annual Scientific Meeting, 2012, Uluru, Australia; the 33rd Annual Cardiothoracic Surgery Symposium, 2013, San Diego, California; and the 50th AmSECT International Meeting, 2013, Las Vegas, Nevada. Abstract: Air bubble detectors (ABDs) are utilized during cardiopulmonary bypass (CPB) to protect against massive air embolism. Stockert (Munich, Germany) ABD quantify microbubbles >300 mm; however, their reliability has not been reported. The aim of this study was to assess the reliability of the microbubble data from the ABD with the SIII and S5 heart lung machines. Microbubble counts from the ABD with the SIII (SIII ABD) and S5 (S5 ABD) were measured simultaneously with the emboli detection and classification (EDAC) quantifier in 12 CPB procedures using two EDAC detectors and two ABDs in series in the arterial line. Reliability was assessed by the Spearman correlation co-efficient (r) between measurements for each detector type, and between each ABD and EDAC detector for counts >300 mm. No correlation was found between the SIII ABD (r =.008,p =.793). A weak negative correlation was found with the S5 ABD (r =.16, p <.001). A strong correlation was found between the EDAC detectors (SIII; r =.958,p <.001),(S5;r =.908,p <.001). With counts >300 mm, the SIII ABDs showed a correlation of small medium effect size between EDAC detectors and ABD1 (r =.286, p <.001 [EDAC1], r =.347,p <.001 [EDAC2]). There was no correlation found between ABD2 and either EDAC detector (r =.003, p =.925 (EDAC1), r =.003,p =.929 [EDAC2]). A correlation between EDAC and the S5 ABD, was not able to be determined due to the low bubble count detected by the EDAC >300 mm. Both SIII ABD and S5 ABD were found to be unreliable for quantification of microbubble activity during CPB in comparison with the EDAC. These results highlight the importance of ensuring that data included in the CPB report is accurate and clinically relevant, and suggests that microbubble counts from devices such as the SIII ABD and S5 ABD should not be reported. Keywords: cardiopulmonary bypass (CPB), equipment, microemboli, embolism. Air bubble detectors (ABDs) are used during cardiopulmonary bypass (CPB) to alert perfusionists to the presence of air in the arterial line and help protect against massive air embolism. Contemporary equipment, technology, and perfusion techniques have reduced the potential for massive air embolism; however, arterial gaseous microemboli (GME) still remain a concern during CPB in the context of postoperative neurological dysfunction (1). GME may arise from a number of sources during cardiac surgery including manipulation of the aorta, cannulation, and cross clamping, and via entrainment of air into Received for publication March 5, 2015; accepted August 8, Address correspondence to: Robert A. Baker, PhD, Dip Perf, CCP (Aust), Cardiac Surgery Research and Perfusion, Flinders Medical Centre and Flinders University, 1 Flinders Drive, Bedford Park, SA rob.baker@health.sa.gov.au The senior author has stated that the authors have reported no material, financial, or other relationship with any healthcare-related business or other entity whose products or services are discussed in this paper. the CPB venous line (2 4) or during drug injection into the sampling manifold (3). In addition, the design of the venous reservoir (5), the use of vacuum-assisted venous drainage (4), high blood flow rates (5), and high temperature gradients may also contribute to GME formation (3). Cognitive decline is a common complication after cardiac surgery, and has been variably reported to occur in up to 50% of patients at discharge, 36% at 6 weeks, 26 33% at 1 year, and 42% at 5 years, postoperatively (6). Although there has been some controversy as to the size and number of GME that are associated with cognitive decline, there is some evidence linking embolization to clinical outcome (2,6,7). Doganci et al. (7) demonstrated significantly greater cognitive decline in patients exposed to >500 GME during CPB compared with those that were exposed to <250 GME at 1 week postoperatively; however, this difference was no longer significant 1 month postoperatively. ABD used with the SIII and S5 heart lung machines (Sorin Group, Munich, Germany) are purported by the 174

2 AIR BUBBLE DETECTOR USAGE TO QUANTIFY MICROBUBBLE ACTIVITY DURING CPB 175 manufacturer to quantify microbubbles greater than 300 mm (SIII (8), S5, Sorin Group, personal communication), from which data may be generated and stored using the Data Management System or CONNECT Software (Sorin Group); however, their ability to generate reliable data has not been reported. Given the implication of GME in adverse patient outcomes, data that are generated for reporting purposes must be reliable and accurate. The aim of this study was to assess the reliability of the microbubble counts quantified by the ABD used with the SIII and S5 heart lung machines. DESCRIPTION Study approval (344.12) was given and the requirement for written informed consent was waived by the Southern Adelaide Clinical Human Research Ethics Committee. Data were collected from 12 patients undergoing routine CPB. Microbubble Detection Microbubble counts from the Sorin Group 3/8 00 ABD with the SIII ABD and S5 ABD were compared with counts from the emboli detection and classification (EDAC; Luna Innovations, Roanoke, VA) quantifier during CPB. Data were generated from the heart lung machine at 20-second intervals and collected using the Data Management System (Sorin Group). Measurements were obtained simultaneously using two EDAC transducers and two ABD in series in the arterial line between the oxygenator and the arterial line filter throughout six CPB procedures for each ABD to allow simultaneous measurement for each detector type (Figure 1). Ultrasonic gel (2 ml) was applied to each EDAC connector before attachment of each transducer. A thin film of ultrasonic gel was appliedtoeachsensorpadoftheabdaccordingtothe manufacturer s recommendations. The SIII ABDs and one S5 ABD had been used clinically and an additional S5 ABD was provided for the study by the distributor (Cellplex Pty Ltd., Dandenong, South, Victoria, Australia). Figure 1. Air bubble detectors and EDAC detectors between the oxygenator and the arterial line filter on the S5 heart lung machine. EDAC, emboli detection and classification. Clinical Management General anesthesia was induced with fentanyl (10 30 g/kg) and supplemented with sevoflurane and/or propofol. All patients underwent cardiac surgery with CPB using either an SIII or S5 roller pump (Sorin Group). The arterial pressure was monitored via a radial artery catheter. CPB was instituted after positioning of either a single 36/51 Fr two-stage atrial cannula (SarnsÔ Terumo Corporation, Tokyo, Japan) or Fr bicaval cannulation (SarnsÔ) and a 22-Fr ascending aortic cannula (DLP; Medtronic, Minneapolis, MN) or a 20 Fr FemFlex (Edwards Lifesciences, Irvine, CA) used in the ascending aortic position. The CPB circuit included a hard-shell membrane oxygenator (Capiox SX25RX, Terumo Corp., Tokyo, Japan), biopassive tubing (SMARxT, Cobe Cardiovascular, Arvada, CO), a 40-mm arterial-line filter (D703; Dideco, Mirandola, Italy), and a.2-mm prebypass filter (Prebypass Plus ; Pall Corporation, Port Washington, NY). The circuit was primed with 375 ml of Plasmalyte solution ((Baxter, Old Toongabbie, NSW, Australia), 500 ml of Gelofusine (B. Braun, Melsungen, Germany; isolated coronary artery bypass grafting procedures), or 4% albumin (other procedures), 50 ml 8.4% sodium bicarbonate solution, 375 ml Hartmann s solution (Baxter, Old Toongabbie, NSW, Australia), and 10,000 IU heparin. The CPB protocol included an arterial, nonpulsatile, target flow rate of L/min/m 2, alpha-stat ph management with a target po 2 of mmhg, gravity venous drainage and tepid systemic temperature management (nasopharyngeal temperature C), with no active cooling. After placement of the aortic cross-clamp, cardioplegic arrest was induced with 4:1 blood cardioplegia (32 34 C), [K+] 30 mmol/l at induction and intermittent boluses ([K+] 16 mmol/l) as required. Mean CPB arterial pressure was controlled using metaraminol, phentolamine, or isoflurane to achieve a target of mmhg. The target nasopharyngeal temperature for separation from bypass was >36 C with a maximum arterial outlet temperature of 37 C and rewarming rate <1 C per min. Hemoglobin during CPB was maintained >70 g/l; none of the patients required red blood cell transfusion either in the priming solution or during CPB. Data Analysis Reliability was assessed by the Spearman correlation co-efficient (r) between simultaneous measurements for each detector type. As the manufacturer reports that the ABD detect microbubbles >300 mm, we evaluated correlation in simultaneous measurements between each ABD and EDAC detector for microbubbles >300 mm using SPSS V20.0 (IBM Corporation, New York, NY). Reliability: Correlation of Simultaneous Air Bubble Detector Microbubble Counts Total Microbubble Counts: A total of 4592 measurements at 20-second intervals were obtained for the SIII

3 176 R.F. NEWLAND ET AL. ABD in six procedures, and 5924 measurements using the S5 ABD. Microbubble counts for each detector are reported in Table 1 for each procedure, and summarized in Figures 2 and 3. No correlation was found between microbubble counts recorded using the two SIII ABD s as shown in Figure 4 (r =.008, p =.793). A weak negative correlation was found with between the two S5 ABD (r =.16,p <.001)(Figure5). A strong correlation was found between microbubble counts recorded using the EDAC detectors in the procedures using the SIII or S5 heart lung machine (r =.958, p <.001;r =.908,p <.001, respectively). Microbubble Counts >300 mm: The SIII ABD, demonstrated a weak correlation with small medium effect size with both EDAC detectors for microbubble counts >300 mm withabd1(r =.286,p <.001 [EDAC1], r =.347, p <.001 [EDAC2]). However, the second detector, SIII ABD2, showed no correlation with either EDAC detector for microbubble counts >300 mm (r =.003, p =.925 [EDAC1], r =.003,p =.929 [EDAC2]). A correlation between EDAC and the S5 ABD was not able to be determined because of the low bubble count detected by the EDAC >300 mm. DISCUSSION This study demonstrated that there was no correlation of microbubble data obtained with simultaneous measurements Table 1. Emboli counts during CPB for each detector and heart lung machine. Emboli counts were recorded in counts/20-second interval. SIII S5 Procedure Device Median Minimum Maximum Procedure Device Median Minimum Maximum 1 ABD ABD ABD ABD EDAC EDAC EDAC1 > EDAC1 > EDAC EDAC EDAC2 > EDAC2 > ABD ABD ABD ABD EDAC EDAC EDAC EDAC ABD ABD ABD ABD EDAC EDAC EDAC EDAC ABD ABD ABD ABD EDAC EDAC EDAC EDAC ABD ABD ABD ABD EDAC EDAC EDAC EDAC EDAC2 > EDAC2 > ABD ABD ABD ABD EDAC EDAC EDAC EDAC Overall ABD Overall ABD ABD ABD EDAC EDAC EDAC EDAC EDAC1 > EDAC1 > EDAC2 > EDAC2 > ABDs, air bubble detectors; CPB, cardiopulmonary bypass; EDAC, emboli detection and classification.

4 AIR BUBBLE DETECTOR USAGE TO QUANTIFY MICROBUBBLE ACTIVITY DURING CPB 177 Figure 2. Boxplot of microbubble counts for the air bubble detectors and for microbubbles >300 mm detected by the EDAC transducers, recorded using the SIII heart lung machine. *Values >3 times the interquartile range. EDAC, emboli detection and classification. using the ABDs on either the SIII or S5 heart lung machines. A strong correlation was observed with simultaneous measurements using the EDAC detectors. In addition, the greater number of microbubbles were detected using the S5 ABD than with the SIII ABD, in contrast to the number of emboli >300 mm detected by the EDAC quantifiers. Furthermore, only one of the ABDs showed a correlation to the EDAC (emboli >300 mm), with a small medium effect size, these results are consistent with our clinical observations that the ABD microbubble data Figure 4. Microbubble counts from each of the SIII ABD plotted at simultaneous time points during cardiopulmonary bypass (n = 4592, Spearman s correlation coefficient, r =.008). are inconsistent in comparison to each other, and to the EDAC quantifier. These results highlight that the data currently being recorded and reported in CPB reports generated using data from the SIII ABD or the S5 ABD is not accurate when compared with either a second ABD in series or when compared with an in-series EDAC quantifier. De Somer et al. (9) reported a comparison of commercially available microbubble detection devices used during CPB including the EDAC and Gampt BC200 (GAMPT mbh, Zappendorf, Germany) against industrial standards Figure 3. Boxplot of microbubble counts for the air bubble detectors and for microbubbles >300 mm detected by the EDAC transducers, recorded using the S5 heart lung machine. O, values between 1.5 and 3 times the interquartile range. EDAC, emboli detection and classification. Figure 5. Microbubble counts from each of the S5 ABD plotted at simultaneous time points during cardiopulmonary bypass (n = 5924, Spearman s correlation coefficient, r =.16).

5 178 R.F. NEWLAND ET AL. of microbubble detection (optical counting using monochromatic laser and backlight shadowgraph technique). Each device uses different algorithms to quantify emboli. The EDAC transducer emits ultrasonic pulses every 1 ms that are reflected by emboli in the blood as they pass through the EDAC connector. The amplitude of the reflected signals is then analyzed to determine bubble size, count, and calculated volume (total embolic load). The manufacturer states that the device is capable of counting emboli with diameters from 10 mm up to the diameter of the connector at a rate of up to 1000 emboli per second, at flow rates between.2 and 6 l per minute (9). The BC200 uses a pulsed ultrasonic Doppler system to generate an amplitude-modulated low-frequency signal depending on the size of the bubble and the time that the bubble is present in the detection signal. Signal transformations allow calculation of the bubble size from the maximum amplitude of the corrected signal. The manufacturer states that the device is capable of counting emboli with diameters between 5 mm and 500 mm at a rate of up to 1000 emboli per second, at flow rates between.5 and 8 L/min(9). Even using commercially available devices designed specifically to quantify microbubbles during CPB, there are some limitations. De Somer et al. s study reported that both devices underestimate the number of microbubbles; at 3 L/min the EDAC counts 38%, the Gampt 18% of total counts and at 6 L/min, both the EDAC and Gampt only count 3% of total counts. The authors concluded that both the EDAC and Gampt can be used in a clinical setting for monitoring basal GME production; however, both devices have limitations when used for studying worst case scenarios. These findings highlight the limitations associated with microbubble measurement. In contrast, ABDs do not provide detail of bubble sizes; therefore, it is impossible to quantify bubble volume to define total embolic load. In our study, we observed a significant correlation between the two EDAC detectors for simultaneous measurements (SIII measurements [r =.958, p <.001], S5 measurements [r =.908, p <.001]) demonstrating reliability, in contrast to the SIII and S5 ABDs. The Sorin Group ABDs transmit an ultrasonic pulse through the circuit tubing that is received by a piezo-element detector, which is able to detect changes in signal amplitude resulting from the difference in speed that sound travels in fluid relative to air. The presence of air bubbles in the blood will result in a drop in the signal amplitude that is received by the detector as the blood travels past the detector. The manufacturer s instructions for operation of the SIII heart lung machine state that the smallest single microbubble that the sensor can detect is 300 mm in diameter. This is not specifically stated in the operating manual for the S5; however, we confirmed with the manufacturer that the same thresholds apply for the S5 (personal communication with Sorin Group). The sensor can detect conglomerates of much smaller microbubbles if the drop in signal amplitude is equal to or greater than that of a 300-mm diameter single microbubble. Since the system detects a total drop in received signal amplitude, the fidelity of the device to count individual microbubbles is limited as is its ability to determine how large the individual bubbles are. Given these limitations, the ABDs should not be used to quantify microbubbles. This is supported by the results of our study in which we not only observed differences in measurements among ABDs on the same heart lung machine, but also among heart lung machines. Although ABDs are interchangeable between Sorin SIII and S5 heart lung machines, the modules used to process the signals from the detectors are not, which may account for the differences in measurements between the SIII and the S5. In our study, ABDs on the S5 routinely detected higher counts of microbubbles than the SIII, by orders of magnitude, which were not demonstrated with the EDAC detectors. From a safety perspective, ABDs have an important role. In a recent U.S. survey (10), 96.8% of respondents routinely used an ABD during CPB, a figure comparable to the 2006 Australian and New Zealand survey that reported 100% use (11), in contrast to the 32% of respondents in the 2006 French survey (12). The reported incidence of air embolism during CPB is relatively infrequent; the French 2006 survey, reported a rate of 1:3757 procedures, and 1:13,426 in an American survey in 2000 (13). It is unclear as to the number of procedures in which the ABD has resulted in the avoidance of air embolus being delivered to the patient although the 2010 Dutch incident survey indicated 13% of respondents had experienced air embolus after the venous reservoir (14). Our study was not designed to assess the use of ABDs as a safety device. Sorin Group ABD s alarm threshold is a bubble detected with a diameter of 5.5 mm or greater in 3/8 00 tubing with the SIII (8) and greater than 4 mm with the S5 (15); therefore, the issues with regard to microbubble quantification are in the context of clinical informatics and medicolegal concerns in the reporting of the microbubble data rather than gross air. Our study did not assess the alarm threshold of the ABD devices as reported by the manufacturer. The limitations of our study include that our data were not limited to open-chamber procedures; hence, we may not have had the opportunity to measure as many emboli as possible. We compared the ABDs against the EDAC rather than an industrial reference; therefore, our comparison of accuracy is limited by the underestimation of microbubble activity as reported by De Somer for the EDAC (9). The correlation between the EDAC and the ABD for emboli >300 mm was limited by the low numbers of emboli detected by the EDAC; however, the data clearly indicate inconsistency in the comparison of each

6 AIR BUBBLE DETECTOR USAGE TO QUANTIFY MICROBUBBLE ACTIVITY DURING CPB 179 ABD against the EDAC, in particular, the large order of magnitude of emboli detected by the ABD using the S5. We performed Bland Altman analysis between the EDAC and ABD measurements; however, due to the unreliability in measurements from the ABD and the limited number of emboli >300 mm, we found that this analysis was not meaningful. An additional potential limitation was the age of the ABDs, with the S5 ABDs both <1 year old, while the SIII ABDs were 2 3 years old. All ABDs had passed routine maintenance testing. Despite these limitations, the data from this study represent what would be detected during routine procedures in our center and therefore our findings are representative of our clinical practice. In conclusion, both SIII ABD and S5 ABD were found to be unreliable for quantification of microbubble activity during CPB in comparison to the EDAC. These results demonstrate that if quantification of microbubbles during CPB is required, then a device primarily designed for this purpose should be used. Furthermore, this study highlights the importance of ensuring that all data that is included in the CPB report is accurate and clinically relevant, and suggests that microbubble counts from devices such as the SIII ABD and S5 ABD should not be reported. REFERENCES 1. Kurusz M, Butler BD. Bubbles and bypass: An update. Perfusion. 2004;19:S Mitchell S, Gorman D. The pathophysiology of cerebral arterial gas embolism. J Extra Corpor Technol. 2002;34: Taylor RL, Borger MA, Weisel RD, Fedorko L, Feindel CM. Cerebral microemboli during cardiopulmonary bypass: Increased emboli during perfusionist interventions. Ann Thorac Surg. 1999;68: Willcox TW, Mitchell SJ, Gorman DF. Venous air in the bypass circuit: A source of arterial line emboli exacerbated by vacuum assisted drainage. Ann Thorac Surg. 1999;68: Liu S, Newland RF, Tully PJ, Tuble SC, Baker RA. In vitro evaluation of gaseous microemboli handling of cardiopulmonary bypass circuits with and without integrated arterial line filters. J Extra Corpor Technol. 2011;43: Martin KK, Wigginton JB, Babikian VL, Pochayd VE, Crittenden MD, Rudolph JL. Intraoperative cerebral high-intensity transient signals and postoperative cognitive function: A systematic review. Am J Surg. 2009;197: Doganci S, Gunaydin S, Murat Kocak O, Yilmaz S, Demirkilic U. Impact of the intensity of microemboli on neurocognitive outcome following cardiopulmonary bypass. Perfusion. 2013;28: SIII System Operators Manual, (version GA E, 5th update: 11/1997). Munich, Germany: Stockert Instrumente; De Somer FM, Vetrano MR, Van Beeck JP, Van Nooten GJ. Extracorporeal bubbles: A word of caution. Interact Cardiovasc Thorac Surg. 2010;10: Kelting T, Searles B, Darling E. A survey on air bubble detector placement in the CPB circuit: A 2011 cross-sectional analysis of the practice of certified clinical perfusionists. Perfusion. 2011;27: Baker RA, Willcox TW. Australian and New Zealand Perfusion Survey: Equipment and monitoring. J Extra Corpor Technol. 2006;38: Charrière JM, Pélissié J, Verd C, et al. Survey: Retrospective survey of monitoring/safety devices and incidents of cardiopulmonary bypass for cardiac surgery in France. J Extra Corpor Technol. 2007; 39: Mejak B, Stammers A, Rauch E, Vang S, Viessman T. A retrospective study on perfusion incidents and safety devices. Perfusion. 2000;15: Groenenberg I, Weerwind PW, Everts PAM, Maessen JG. Dutch perfusion incident survey. Perfusion. 2010;25: S5 System Operating Instructions, (version GA ENG, Version 02/2011). Munich, Germany: Sorin Group; 2011.