The Sensor in the Venous Chamber Does Not Prevent Passage of Air Bubbles During Hemodialysis

Venous chambers in clinical use for hemodialysis have limited capacity to eliminate microbubbles from entering the return bloodline: An in vitro study

BACKGROUND

During hemodialysis (HD), blood passes through an extracorporeal circuit (ECC). To prevent air administration to the patient, a venous chamber (chamber) is located before the blood return. Microbubbles (MBs) may pass through the chamber and end up as microemboli in organs such as the brain and heart. This in vitro study investigated the efficacy of various chambers in MB removal.

MATERIALS AND METHODS

The in vitro recirculated setting of an ECC included an FX10 dialyzer, a dextran-albumin solution to mimic blood viscosity and chambers with different flow characteristics in clinical use (Baxter: AK98 and Artis, Fresenius: 5008 and 6008) and preclinical test (Embody: Emboless®). A Gampt BCC200 device measured the presence and size of MBs (20-500 μm). Percentage change of MBs was calculated: ΔMB% = 100*(outlet-inlet)/inlet for each size of MB. Blood pump speed (Qb) was 200 (Qb200) or 300 (Qb300) ml/minute. Wilcoxon paired test determined differences.

RESULTS

With Qb200 median ΔMB% reduction was: Emboless -58%, AK98 -24%, Fresenius 5008 -23%, Artis -8%, and Fresenius 6008 ± 0%. With Qb300 ΔMB% was: Emboless -36%, AK98 ± 0%, Fresenius 5008 ± 0%, Artis +25%, and Fresenius 6008 + 21%. The Emboless was superior to all other chambers with Qb200 and Qb300 (p < 0.001). Further, the Emboless with Qb300 still eliminated more MBs than all other chambers with Qb200 (p ≤ 0.003).

CONCLUSION

The results from the present study indicate that flow characteristics of the chamber and the Qb are important factors to limiting exposure of MB to the return bloodline. The Emboless chamber reduced MBs more effective than those chambers in clinical use investigated.

Microemboli induced by air bubbles may be deposited in organs as a consequence of contamination during medical care

Background

Larger volumes of accidental air infused during medical care may end up as emboli while microbubbles of air are supposed to be absorbed and cause no harm. The aim of this autopsy study was to investigate if microbubbles of air accidently entering the bloodline may be detected as microemboli (ME) in tissue such as lungs, brain and heart. If so, do differences in prevalence exist between haemodialysis (HD) and amyotrophic lateral sclerosis (ALS) patients.

Methods

Included were data from 44 patients treated by medical healthcare before death. Twenty-five cases had been treated with chronic HD and 19 cases died from ALS. Since air in the bloodline activates coagulation, ME could appear. To discriminate between microbubbles caused by artificial contamination during autopsy versus microbubbles deposited in vivo, tissues were stained with a polyclonal fluorescent antibody against fibrinogen, fibrin and fragments E and D. Fluorescence staining was used to visualize ME counted within 25 microscopic fields (600×) of a tissue preparation. One tissue preparation was used if available from the lung, heart and frontal lobe of the brain and in five cases also the cerebellum.

Results

Microbubbles can be verified at autopsy as ME in the lung, heart and brain in tissue from patients exposed to more extensive medical care. There were significantly more ME in the lungs versus the heart or brain. Women had fewer ME than men. The HD group had a higher median of ME per section than the ALS group (lung: 6 versus 3, P = .007; heart: 2.5 versus 1, P = .013; brain: 7.5 versus 2, P = .001) and had more sections with ME findings than the ALS group (P = .002). A correlation existed between the time on HD (months) and ME in the lungs.

Conclusions

More ME were present in HD patients compared with those who suffered from ALS. Minimizing air contamination from syringes, infusions and bloodlines will decrease ME and subsequent tissue injury.

Non-anticoagulation pediatric continuous renal replacement therapy methods to increase circuit life

INTRODUCTION

Acute kidney injury (AKI) is a clinical condition characterized by an abrupt increase in serum creatinine levels due to functional changes in the kidneys from a newfound insult or injury. For supportive treatment, continuous renal replacement therapy (CRRT) is one of the most widely used modalities due to its precise control of fluid balance over extended periods of time. However, its complications include circuit clotting, the most frequent cause for CRRT interruption. Vascular access and circuit management were found to be major determinants of performance efficiency. Anticoagulation required to prevent clotting has the downside of increasing the risk of bleeding, especially in the setting of overdosage. Hence, a delicate balance needs to be maintained consistently.

METHODS

This study explores the adequacy of non-anticoagulation measures in the prevention of circuit clotting. A comprehensive literature search was conducted using PubMed/Medline and Embase databases to include all relevant studies.

FINDINGS

The most-effective CRRT catheter would be made of nonthrombogenic material, noncuffed and nontunneled with separate lumens for arterial and venous blood. Further, studies show that blood flow during the process is optimized at 200 ml/min, which can be lowered in the pediatric population due to more narrow catheters. Platelet count and hematocrit need to be closely monitored as levels above 450,000 × 10⁶ /L and 0.40, respectively, increase risk of clotting. Predilution is a non-anticoagulation technique to reduce the risk of clotting by returning replacement solution to the blood before it reaches the filter. Also, biocompatible membranes such as polyacrylonitrile or polysulfone activate the coagulation cascade significantly less than the conventional cellulose-based membranes, thereby reducing clotting chances.

DISCUSSIONS

With the advent of such techniques and maneuvers, anticoagulation can be efficiently maintained in patients undergoing CRRT without increasing the risk of bleeding.

Air contamination during medical treatment results in deposits of microemboli in the lungs: An autopsy study

INTRODUCTION

Microbubbles of air may enter into patients during conventional hemodialysis, infusions of fluids, or by injections. The aim of this study was to investigate whether the air that enters the patient during hemodialysis can be detected in the lungs after death, and if so, whether this may be related to tissue damage.

METHODS

The material consisted of lung tissue from five chronic hemodialysis patients who died either during (two) or after hemodialysis (range 10 min from start until 3333 min after the last hemodialysis session); as reference group tissue was taken from seven patients who died due to amyotrophic lateral sclerosis. The lung tissue was investigated by microscopy after autopsy using a fluorescein-marked polyclonal antibody against fibrinogen as a marker for clots preformed before death.

RESULTS

All five hemodialysis patients had microbubbles of air in the lung tissue, whereas two of seven amyotrophic lateral sclerosis patients had such findings (Fisher's test p = 0.0278, relative risk = 3.5, confidence interval: 1.08-11.3). There were more microbubbles of air/10 randomly investigated microscopic fields of tissue in the hemodialysis patients than the amyotrophic lateral sclerosis patients (Student's test, p < 0.05). All hemodialysis patients had a medium graded extent of pulmonary fibrosis that was not found in any of the ALS patients. The microbubbles of air were surrounded by fibrin as a sign of development of clots around the air bubbles while the patients were still alive.

CONCLUSION

Exposure to microbubbles of air during various treatments such as hemodialysis may result in microemboli. Future studies should clarify whether microbubbles of air contribute to tissue scarring. We suggest preventive measures against the exposure to microbubbles of air during especially repeated exposures such as hemodialysis.

Cerebral microembolism in the critically ill with acute kidney injury (COMET-AKI trial): study protocol for a randomized controlled clinical trial

Background

Microembolism is a frequent pathological event during extracorporeal renal replacement therapy (RRT). Some previous data indicate that microemboli are generated in patients who are undergoing RRT and that these may contribute to increased cerebrovascular and neurocognitive morbidity in patients with end-stage renal disease. The current trial aims to quantify the microembolic load and respective qualitative composition that effectively reaches the intracerebral circulation in critically ill patients treated with different RRT modalities for acute kidney injury (AKI).

Methods/design

The COMET-AKI trial is a prospective, randomized controlled clinical trial with a 2-day clinical assessment period and follow-up visits at 6 and 12 months. Consecutive critically ill patients with AKI on continuous renal replacement therapy (CRRT) scheduled for a switch to intermittent renal replacement therapy (IRRT) will be randomized to either switch to IRRT within the next 24 h or continued CRRT for an additional 24 h. Cerebral microembolic load will be determined at baseline, i.e., before switch (on CRRT for both groups) and on IRRT versus CRRT, whichever group they were randomized to. The primary endpoint is defined as the difference in mean total cerebral microemboli count during the measurement period on CRRT versus IRRT following randomization. Microemboli will be assessed within the RRT circuit by a 1.5-MHz ultrasound detector attached to the venous RRT tubing and cerebral microemboli will be measured in the middle cerebral artery using a 1.6-MHz robotic transcranial Doppler system with automatic classification of Doppler signals as solid or gaseous. In addition to Doppler measurements, patients will be examined by magnetic resonance imaging and neurocognitive tests to gain better understanding into the potential morphological and clinical consequences of embolization.

Discussion

The results of COMET-AKI may help to gain a better insight into RRT modality-associated differences regarding microbubble generation and the cerebral microembolic burden endured by RRT recipients. Furthermore, identification of covariates of microbubble formation and distribution may help to encourage the evolution of next-generation RRT circuits including machinery and/or filters.

Trial registration

ClinicalTrials.gov, ID: NCT02621749. Registered on 3 December 2015.

Electronic supplementary material

The online version of this article (10.1186/s13063-018-2561-3) contains supplementary material, which is available to authorized users.

Formation of Blood Foam in the Air Trap During Hemodialysis Due to Insufficient Automatic Priming of Dialyzers

We were encouraged to investigate the reasons for large amounts of foam observed in bloodlines during hemodialysis (HD). Foam was visible in the venous air trap within the Artis Gambro dialysis device. Estimates of the extent of foam were graded (0-no foam, 10-extensive foam) by two persons that were blind to the type of dialyzer used. Thirty-seven patients were involved in the dialysis procedures. Consecutive dialyses were graded using dialyzers from Fresenius Medical Care (CorDiax dialyzers that were used for high flux HD-FX80 and FX100, and for hemodiafiltration-FX1000). The extracorporeal circuit was primed automatically by dialysate using Gambro Artis software 8.15 006 (Gambro, Dasco, Medolla Italy, Baxter, Chicago, IL, USA). The priming volume recommended by the manufacturer was 1100 mL, whereas our center uses 1500 mL. Extensive amounts of blood foam were visual in the air traps. Although the manufacturer recommended extension of priming volume up to 3000 mL, this did not eliminate the foam. Microbubble measurement during HD revealed the air to derive from the dialyzers. When changing to PF210H dialyzers (Baxter) and using a priming volume of 1500 mL, the foam was significantly less (P < 0.01). The extent of foam correlated with the size of the FX-dialyzer surface (P = 0.002). The auto-priming program was updated to version 8.21 by the manufacturer and the extent of foam in the air trap using FX dialyzers was now reduced and there was no longer a difference between FX and PF dialyzers, although less foam was still visible in the venous air trap during several dialyses. In conclusion, this study urgently calls attention to blood foam development in the venous air trap when using Artis devices and priming software 8.15 in combination with Fresenius dialyzers. Updated auto-priming software (version 8.21) of Artis should be requested to reduce the extent of foam for the Fresenius dialyzers. Other interactions may also be present. We recommend further studies to clarify these problems. Meanwhile caution is warranted for the combined use of dialysis devices and dialyzers with incompatible automatic priming.

A microscopic view of gaseous microbubbles passing a filter screen

PURPOSE

The aim of this study was to investigate the filtration efficacy of a 38-µm 1-layer screen filter based on Doppler registrations and video recordings of gaseous microbubbles (GME) observed in a microscope.

METHODS

The relative filtration efficacy (RFE) was calculated from 20 (n = 20) sequential bursts of air introduced into the Plasmodex® primed test circuit.

RESULTS

The main findings indicate that the RFE decreased (p = 0.00), with increasing flow rates (100-300 mL/min) through the filter screen. This reaction was most accentuated for GME below the size of 100 µm, where counts of GME paradoxically increased after filtration, indicating GME fragmentation. For GME sized between 100-250 µm, the RFE was constantly >60%, independently of the flow rate level. The video recording documenting the GME interactions with the screen filter confirmed the experimental findings.

CONCLUSIONS

The 38-µm 1-layer screen filter investigated in this experimental setup was unable to trap gaseous microbubbles effectively, especially for GME below 100 µm in size and in conjunction with high flow rates.

Diagnosis, Treatment, and Prevention of Hemodialysis Emergencies

Given the high comorbidity in patients on hemodialysis and the complexity of the dialysis treatment, it is remarkable how rarely a life-threatening complication occurs during dialysis. The low rate of dialysis emergencies can be attributed to numerous safety features in modern dialysis machines; meticulous treatment and testing of the dialysate solution to prevent exposure to trace elements, toxins, and pathogens; adherence to detailed treatment protocols; and extensive training of dialysis staff to handle medical emergencies. Most hemodialysis emergencies can be attributed to human error. A smaller number are due to rare idiosyncratic reactions. In this review, we highlight major emergencies that may occur during hemodialysis treatments, describe their pathogenesis, offer measures to minimize them, and provide specific interventions to prevent catastrophic consequences on the rare occasions when such emergencies arise. These emergencies include dialysis disequilibrium syndrome, venous air embolism, hemolysis, venous needle dislodgement, vascular access hemorrhage, major allergic reactions to the dialyzer or treatment medications, and disruption or contamination of the dialysis water system. Finally, we describe root cause analysis after a dialysis emergency has occurred to prevent a future recurrence.

Observation of microbubbles during standard dialysis treatments

Background

The infusion of microbubbles as a side effect of haemodialysis was repeatedly demonstrated in recent publications, but the knowledge on the source of microbubbles and on microbubble formation is scarce.

Methods

Microbubbles in the range of 10–500 µm were measured by a non-invasive bubble counter based on a pulsed ultrasonic Doppler system in a non-interventional study of a single centre. Totally, 29 measurements were performed in standard treatments covering a broad range of patient and treatment conditions (types of blood access, treatment modes, blood flow rates and arterial pressures).

Results

Several possible sources of microbubbles could be identified such as an arterial luer lock connector at negative pressure and remnant bubbles from insufficient priming, but some sources of microbubbles remain unknown. Microbubbles were found in all treatments, haemodialysis (HD) and online haemodiafiltration. The lowest average microbubble rates (17 ± 16 microbubbles per minute) were observed in patients treated by online haemodiafiltration at medium blood flow rates and moderate arterial pressures and the highest average microbubble rates (117 ± 63 microbubbles per minute) at high blood flow rates (550 mL/min) and low arterial pressures (−210 mmHg). Generally, the microbubble rate correlated to both blood flow rate (correlation coefficient r = 0.45) and negative arterial pressure (r = 0.67).

Conclusions

Microbubbles are a general side effect of HD; origin and pathophysiologic consequences of this phenomenon are not well understood, and deserve further study.

Pulsatility Produced by the Hemodialysis Roller Pump as Measured by Doppler Ultrasound

Microbubbles have previously been detected in the hemodialysis extracorporeal circuit and can enter the blood vessel leading to potential complications. A potential source of these microbubbles is highly pulsatile flow resulting in cavitation. This study quantified the pulsatility produced by the roller pump throughout the extracorporeal circuit. A Sonosite S-series ultrasound probe (FUJIFILM Sonosite Inc., Tokyo, Japan) was used on a single patient during normal hemodialysis treatment. The Doppler waveform showed highly pulsatile flow throughout the circuit with the greatest pulse occurring after the pump itself. The velocity pulse after the pump ranged from 57.6 ± 1.74 cm/s to -72 ± 4.13 cm/s. Flow reversal occurred when contact between the forward roller and tubing ended. The amplitude of the pulse was reduced from 129.6 cm/s to 16.25 cm/s and 6.87 cm/s following the dialyzer and venous air trap. This resulted in almost nonpulsatile, continuous flow returning to the patient through the venous needle. These results indicate that the roller pump may be a source of microbubble formation from cavitation due to the highly pulsatile blood flow. The venous air trap was identified as the most effective mechanism in reducing the pulsatility. The inclusion of multiple rollers is also recommended to offer an effective solution in dampening the pulse produced by the pump.

NT-proBNP and troponin T levels differ after haemodialysis with a low versus high flux membrane

BACKGROUND

Brain natriuretic peptide (BNP), N-terminal-proBNP (NT-proBNP), and high sensitive cardiac troponin T (TnT) are markers that are elevated in chronic kidney disease and correlate with increased risk of mortality. Data are conflicting on the effect of biomarker levels by hemodialysis (HD).Our aim was to clarify to what extent HD with low-flux (LF) versus high-flux (HF) membranes affects the plasma levels of BNP, NT-proBNP, and TnT.

METHODS AND MATERIALS

31 HD patients were included in a crossover design, randomized to start dialysis with a LF-HD or HF-HD dialyzer. Each patient was his/her own control. The dialyses included in the study were the first treatments of two consecutive weeks with each mode of dialysis. Patients normally on hemodiafiltration (HDF) also performed a HDF the third week. Values after HD were corrected for extent of ultrafiltration.

RESULTS

During LF-HD the biomarkers NT-proBNP and TnT increased (15 versus 6%, P ≤ .001) while there was a slight decrease in BNP (P<.05). During HF-HD the NT-proBNP, BNP and TnT levels decreased (P ≤ .01 for all). During HDF all three markers decreased (P<.01 for all). The rise in TnT during LF-HD correlated with dialysis vintage (months on HD, r = .407, P = .026), Kt/V-urea (r = .383, P = .037), HD time in hours/treatment (r = .447, P = .013) and inversely with residual urinary output (r = -.495, P = .005). The baseline levels of BNP and NT-proBNP correlated with blood pressure.

CONCLUSIONS

Cardiac biomarkers increase slightly during LF-HD. A HF-HD eliminates the biomarkers and can mask increases caused by, e.g., myocardial infarction.

A high blood level in the venous chamber and a wet-stored dialyzer help to reduce exposure for microemboli during hemodialysis

During hemodialysis (HD), microemboli develop in the blood circuit of the apparatus. These microemboli can pass through the venous chamber and enter into the patient's circulation. The aim of this study was to investigate whether it is possible to reduce the risk for exposure of microemboli by altering of the treatment mode. Twenty patients on chronic HD were randomized to a prospective cross-over study of three modes of HD: (a) a dry-stored dialyzer (F8HPS, Fresenius, steam sterilized) with a low blood level in the venous chamber (DL), (b) the same dialyzer as above, but with a high level in the venous chamber (DH), and (c) a wet-stored dialyzer (Rexeed, Asahi Kasei Medical, gamma sterilized) with a high blood level (WH). Microemboli measurements were obtained in a continuous fashion during 180 minutes of HD for all settings. A greater number of microemboli were detected during dialysis with the setting DL vs. WH (odds ratio [OR] 4.07, 95% confidence interval [CI] 4.03-4.11, P<0.0001) and DH vs. WH (OR 1.18, 95% CI 1.17-1.19, P<0.0001) and less for DH vs. DL (OR 0.290, 95% CI 0.288-0.293, P<0.0001). These data indicate that emboli exposure was least when using WH, greater with DH, and most with DL. This study shows that using a high blood level in the venous chamber and wet-stored dialyzers may reduce the number of microemboli.

Microbubbles of air may occur in the organs of hemodialysis patients

During hemodialysis (HD), blood that passes the dialysis device gets loaded with microbubbles (MB) of air that are returned to the patient without inducing an alarm. The aim with this study was to clarify if these signals are due to microembolies of air, clots, or artifacts, by histopathology of autopsy material of HD patients. These first results are from a patient on chronic HD. Due to pulmonary edema he was ultrafiltered. Within 30 minutes after the start, he suffered from a cardiac arrest and died. Autopsy verified the clinical findings. Microscopic investigation verified microembolies of air that were surrounded by fibrin in the lungs, brain, and heart. The study verified that MBs can enter the blood during HD and are trapped in the lungs. In addition, MBs pass the pulmonary capillaries and enter the arterial part of the body and are dispersed throughout the body. This can contribute to organ damage and be part of the poor prognoses seen in HD patients. Data support the importance to reduce MBs in the dialysis circuit.

A high blood level in the air trap reduces microemboli during hemodialysis

Previous studies have demonstrated the presence of air microemboli in the dialysis circuit and in the venous circulation of the patients during hemodialysis. In vitro studies indicate that a high blood level in the venous air trap reduces the extent of microbubble formation. The purpose of this study was to examine whether air microbubbles can be detected in the patient's access and if so, whether the degree of microbubble formation can be altered by changing the blood level in the venous air trap. This was a randomized, double-blinded, interventional study of 20 chronic hemodialysis patients. The patients were assigned to hemodialysis with either an elevated or a low blood level in the air trap. The investigator and the patient were blinded to the settings. The numbers of microbubbles were measured at the site of the arteriovenous (AV) access for 2 min with the aid of an ultrasonic Doppler device. The blood level in the air trap was then altered to the opposite setting and a new measurement was carried out after an equilibration period of 30 min. Median (range) for the number of microbubbles measured with the high air trap level and the low air trap level in AV access was 2.5 (0-80) compared with 17.5 (0-77), respectively (P = 0.044). The degree of microbubble formation in hemodialysis patients with AV access was reduced significantly if the blood level in the air trap was kept high. The exposure of potentially harmful air microbubbles was thereby significantly reduced. This measure can be performed with no additional healthcare cost.

Microemboli, developed during haemodialysis, pass the lung barrier and may cause ischaemic lesions in organs such as the brain

BACKGROUND

Chronic haemodialysis (HD) may relieve some medical problems of terminal uraemia, but the life expectancy of patients is still significantly shortened, and there is a greatly increased morbidity. This includes pulmonary morbidity and chronic central nervous system (CNS) abnormalities. Previous studies have shown that a considerable amount of air microbubbles emanate within the blood lines of the dialysis device and pass the air detector without sounding an alarm. The aim of this study was to investigate whether microemboli can pass to the patient and whether they could be detected in the carotid artery.

METHODS

A total of 54 patients on chronic HD (16 with central dialysis catheter) were investigated with an ultrasound detector (Hatteland, Røyken, Norway) for the presence of microemboli at the arteriovenous (AV) fistula/graft and at the common carotid artery before and during HD. Measurements were taken for 2 and 5 min, respectively. Non-parametric paired statistics were used (Wilcoxon).

RESULTS

The median number (range) and mean +/- SD of microembolic signals detected at the AV access site before commencing dialysis and during HD were 0 (0-3) and 0.2+/- 0.5 versus 4 (0-85) and 13.5 +/- 20 (P = 0.000); at the carotid artery, 1 (0-14) and 1.7 +/- 2.9 versus 2 (0-36) and 3.5 +/- 5.8 (P = 0.008).

CONCLUSIONS

The infused and returning fluid from HD devices contains air microbubbles that enter the patient without triggering any alarms. These small emboli pass the lung and may cause ischaemic lesions in organs supported by the arterial circuit, such as the brain.

Hemodialysis dialyzers contribute to contamination of air microemboli that bypass the alarm system in the air trap

BACKGROUND

Previous studies have shown that micrometer-sized air bubbles are introduced into the patient during hemodialysis. The aim of this study was to investigate, in vitro, the influence of dialysis filters on the generation of air bubbles.

METHODS

Three different kind of dialyzers were tested: one high-flux FX80 dry filter (Fresenius Medical Care AG&Co. KGaA, Bad Homburg, Germany), one low-flux F8HPS dry filter (Fresenius Medical Care AG&Co. KGaA, Bad Homburg, Germany) and a wet-stored APS-18u filter (Asahi Kasei Medical, Tokyo, Japan). The F8HPS was tested with pump flow ranging between 100 to 400 ml/min. The three filters were compared using a constant pump flow of 300 ml/min. Measurements were performed using an ultrasound Doppler instrument.

RESULTS

In 90% of the series, bubbles were measured after the outlet line of the air trap without triggering an alarm. There were significantly more bubbles downstream than upstream of the filters F8HPS and FX80, while there was a significant reduction using the APS-18u. There was no reduction in the number of bubbles after passage through the air trap versus before the air trap (after the dialyzer). Increased priming volume reduced the extent of bubbles in the system.

CONCLUSIONS

Data indicate that the air trap does not prevent air microemboli from entering the venous outlet part of the dialysis tubing (entry to the patient). More extended priming of the dialysis circuit may reduce the extent of microemboli that originate from dialysis filters. A wet filter may be favorable instead of dry-steam sterilized filters.

Air Bubbles Pass the Security System of the Dialysis Device Without Alarming

During hemodialysis microembolic findings have been noted after the venous chamber (subclavian vein). The aim of this study was to evaluate if air could pass the venous chamber and, if so, if it passes the safety-system detector for air-infusion without triggering an alarm. Various in vitro dialysis settings were performed using regular dialysis devices. A dextran fluid was used instead of blood to avoid the risk of development of emboli. Optical visualization as well as recirculation and collection of eventual air into an intermediate bag were investigated. In addition, a specifically designed ultrasound monitor was placed after the venous air trap to measure the presence of eventual microbubbles. Speed of dialysis fluid was changed, as was the level of the fluid in the air trap. Thereby a fluid level was considered "high" if it was close to the top of the air trap and "low" if it was around the mid part of the air trap. By optical vision microbubbles were seen at the bottom of the air trap and could pass the air trap towards the venous line without alarming. During recirculation several mL of air were collected in an intermediate bag after the venous line. Ultrasound monitoring exhibited the presence of microbubbles of the size of approximately 5 microm upwards passing to the venous line in all runs performed. Amount of bubbles differed between devices and in general an increased fluid speed correlated significantly with the increased counts of microbubbles/min. No alarming of the detector occurred. A more concentrated fluid allowed higher counts/min when flow was increased to 600 mL/min. Data revealed that air passes the safety-sensor in the air trap without alarming. The presence of air increased in general with fluid speed and a lower fluid level in the air trap. Differences were present between devices. If this affects the patients has to be elucidated.