Bileaflet Heart Valve Research Articles

Occasional Single Beat Regurgitation Observed with the Medtronic Advantage Bileaflet Heart Valve

The purpose of this clinical study was to obtain further evidence of the underlying mechanism causing the echocardiographically detected phenomenon of single beat regurgitation in a new bileaflet heart valve. As part of a prospective multicenter trial at our institution, 63 patients received the Advantage bileaflet mechanical heart valve (Medtronic, Minneapolis, Minnesota) in aortic position. During routine follow-up performed at discharge and annually after the operation, intermittent moderate transvalvular regurgitation was detected by echocardiography in 5 patients. Fluoroscopy of leaflet motion (n = 4), invasive blood pressure measurements in the ascending aorta (n = 3) and digital phonocardiography (n = 5) was obtained in the patients showing an intermittent regurgitation during echocardiography. Valve thrombosis, sutures, or pannus ingrowth impairing valve closure was not detected. Fluoroscopy of leaflet motion showed intermittent incomplete closure of either one of the two leaflets in the same prosthesis. This could be correlated with a distinct diastolic blood pressure drop in the same cardiac cycle. Digital phonocardiography showed pathologic closure sounds in those cycles in which echocardiographically the intermittent regurgitation was observed. Some patients with the Medtronic Advantage prosthesis in the aortic position show an intermittent inability of complete valve closure that leads to a single beat transvalvular regurgitation. As thrombotic or other material that might cause a disturbance of leaflet motion could not be detected, and the patients seem not to be exposed to any risk except for some chronic regurgitant volume, we decided not to replace the prostheses.

Turbulent stress measurements downstream of three bileaflet heart valve designs in pigs☆

Mechanical heart valves can cause thromboembolic complications, possibly due to abnormal flow patterns that produce turbulence downstream of the valve. The objective of this study was to investigate whether three different bileaflet valve designs would exhibit clinically relevant differences in downstream turbulent stresses. Three bileaflet mechanical heart valves (Medtronic Advantage), CarboMedics Orbis Universal and St. Jude Medical Standard) were implanted into 19 female 90 kg pigs. Blood velocity was measured during open chest conditions in the cross sectional area downstream of the valves with 10 MHz ultrasonic probes connected to a modified Alfred Pulsed Doppler equipment. As a measure of turbulence, Reynolds normal stress (RNS) was calculated at three different cardiac output ranges (3-4, 4.5-5.5, 6-7 L/min). Data from 12 animals were obtained. RNS correlated with increasing cardiac outputs. The highest instantaneous RNS observed in these experiments was 47 N/m2, and the mean RNS taken spatially over the cross sectional area of the aorta during systole was between 3 N/m2 and 15 N/m2. In none of the cardiac output ranges RNS values exceeded the lower critical limit for erythrocyte or thrombocyte damage for any of the valve designs. Reynolds normal stress values were below 100 N/m2 for all three valve designs and the difference in design was not reflected in generation of turbulence. Hence, it is unlikely that any of the valve designs causes flow induced damage to platelets or erythrocytes.

Hemolysis in Mechanical Bileaflet Prostheses: Experience With the Bicarbon Valve

Normal functioning mechanical heart valve prostheses are designed to have a certain degree of intrinsic structural regurgitation as a washout mechanism to avoid prosthetic thrombosis. However, intrinsic regurgitation leads to blood cell trauma and hemolysis. Information on hemolysis associated with mechanical bileaflet prostheses is scarce. This study evaluated factors influencing hemolysis in 197 Bicarbon mechanical bileaflet prostheses implanted in 164 patients. Serial office interviews, laboratory studies, and echocardiography evaluations were done in the surviving patients. An assay for measuring lactate dehydrogenase activity was developed, and the presence and severity of subclinical hemolysis was determined using reported criteria and analyzed at 1 and 2 years. Hospital mortality was 5.5%. Follow-up was 98.1% complete. No patient had clinically significant or severe subclinical hemolysis. Serum lactate dehydrogenase levels were significantly higher when a paravalvular leak was documented (282 +/- 85 U/L versus 242 +/- 64 U/L; p = 0.0026). Subclinical hemolysis was significantly more frequent after mitral valve (p = 0.001) and double valve replacement (p = 0.001) than after aortic valve replacement, and was unrelated to prosthetic size or to geometric area index, even in those cases with effective orifice area index equal to or less than 0.85 cm2/m2 (p = 0.298). Mild subclinical hemolysis is frequently associated with normal functioning Bicarbon heart valves. Subclinical hemolysis was significantly influenced by valve position but not by valve size or effective orifice area index and remained stable through time. The magnitude of hemolysis in Bicarbon prostheses compared favorably with that reported for other bileaflet heart valve prostheses.

11-year single institution clinical experience with the ATS Open Pivot® bileaflet heart valve prosthesis

Aims: At our institution, ATS Open Pivot® (ATS) valves are implanted since 1993. In this 11-year retrospective analysis, mid- to long-term clinical results of this prosthesis are evaluated.

Assessment of hemodynamics properties of a new-type artificial heart valve prosthesis using catheterization and echocardiography

The objectives of this study were to assess the hemodynamic properties of the newly developed artificial heart valve prosthesis experimentally in laboratory simulation, in an animal model, and clinically in a human model and to compare the results measured by catheterization and echocardiography. (1) Laboratory simulation. The prosthesis was tested using a pulsatile flow simulator in the aortic position. Hydrodynamics parameters were automatically analyzed through a custom-designed data processing program. (2) Animal experiment. Six sheep subjected to mitral replacement with 21-mm-valve prosthesis were measured by open cardiac catheterization intraoperatively. Doppler echocardiography and open cardiac catheterization under dobutamine stress were performed in two sheep subjected to implantation 2.5 years ago. (3) Clinical patient observation. Observations were carried out on 14 patients with aortas replacement and 10 patients with bicuspid replacement using both doppler echocardiography and open cardiac catheterization. (1) Laboratory simulation. The results showed that the value of the transvalvular gradient (DeltaP) decreased with the increase of heart rate, and the values were not greater that 10 mm Hg at any given tissue annulus diameter. (2) Animal experiment. The mean DeltaP value of the six sheep was 5.2 +/- 1.7 mm Hg intraoperatively, while the corresponding DeltaP value of the two sheep 2.5 years after implantation was 6.1 +/- 0.3 mm Hg measured by open cardiac catheterization. (3) Clinical patient observation. The mean DeltaP values in the aortic position measured by catheterization and echocardiography were 6.26 approximately 4.10 and 9.42 approximately 7.48 mm Hg, respectively. The gradients in the mitral position were 2.10 approximately 1.9 and 5.28 approximately 4.10 mm Hg, respectively. The results demonstrate that the new-type bileaflet heart valve prosthesis only generates a relatively low transvalvar gradient and thus has good hemodynamic properties.

Invited commentary

The ultimate test of a new heart valve prosthesis remains the implantation of the device in humans. This is not without risk, and unexpected catastrophic results can be obtained in spite of extensive pre-clinical testing. Therefore it is a relief to see the report of a first clinical trial demonstrating uneventful early results. To justify such clinical implantation of newly developed heart valves, one expects at least significant advantages on the existing and generally accepted prostheses. In case of the newly designed low-profile Medtronic Advantage valve (Medtronic, Inc, Minneapolis, MN), there could be superior hemodynamics and reduced thrombogenicity. The present study [1Guenzinger R. Eichinger W.B. Botzenhardt F. et al.Rest and exercise performance of the Medtronic advantage bileaflet valve in the aortic position.Ann Thorac Surg. 2005; 80: 1319-1326Abstract Full Text Full Text PDF PubMed Scopus (7) Google Scholar] reveals that the Medtronic Advantage valve performs well hemodynamically as demonstrated by low mean pressure gradients, large effective orifice areas, and indices. Furthermore, the authors performed hemodynamic studies during treadmill exercise 1 year after valve replacement, the ultimate test of heart valve performance. They also found a low incidence of patient–prosthesis mismatch and a significant regression of left ventricular (LV) mass 1 year after valve replacement. The authors are to be congratulated not only for their carefully performed study, but also for their realistic interpretation of the results. Indeed, a mean pressure gradient ranging from 10.8 to 20.0 mm Hg across all valve sizes during treadmill exercise at 100 W workload is excellent but not superior to the results of existent bileaflet valves. The same holds true for effective orifice area index (EOAI) values between 0.93 and 1.37 cm2/m2, the low incidence of patient-prosthesis mismatch, and an 18% reduction of LV mass at 1 year. It is unquestionably true that “second generation” bileaflet heart valves perform better than their “first generation” counterparts and that actual heart valves designs did “a good job.” In a comparable series of patients for example, we observed EOAI values between 0.60 to 0.92 (across all valve sizes, including size 19) for the older St. Jude Standard bileaflet valve versus 0.91 to 1.37 cm2/m2 for the actual St. Jude Regent valve (St. Jude Medical, St. Paul, MN). We may confirm the conclusion of the present study that this new bileaflet valve has an excellent hemodynamic profile, but we feel that the advantage of the Medtronic Advantage valve on the existing second generation bileaflet heart valves remains to be elucidated.

Mechanical Valve Replacement in Congenital Heart Defects in the Era of International Normalized Ratio Self-Management

As the number of patients with congenital heart defects requiring heart valve replacement increases, the need for durable valve substitutes with good hemodynamic performance and a low incidence of complications becomes more apparent. The use of porcine xenografts is burdened with early fibrocalcific degeneration, whereas the use of mechanical heart valves led to an increased number of thromboembolic events, especially when implanted in the right side of the heart. We report on our experiences implanting bileaflet heart valves in congenital heart defects since the introduction of international normalized ratio (INR) self-management. The data of 68 long-term survivors (33 males, 35 females) who underwent mechanical heart valve replacement in congenital heart defect were reviewed. Patient age at the time of valve replacement ranged from 5 months to 61 years (mean 21 years). Underlying diagnoses were tetralogy of Fallot (n=33), morbus Ebstein (n=4), atrioventricular canal (n=13), truncus arteriosus communis (n=5), transposition of the great arteries (n=10), and congenitally corrected transposition of the great arteries (n=3). In all patients, bileaflet valves were implanted (St. Jude Medical n=40, Carbomedics n=28). Anticoagulation was performed using dicumarol (Marcumar) and INR self-management in all cases. The mean follow-up was 72 months (range 6-132 months; 409 patient-years). Valve thrombosis developed in 3 of 68 patients (4.4%, all three had tetralogy of Fallot, mean age 9.8 years) after a mean follow-up of 3.5 years. In two of these three patients, re-pulmonary valve replacement was necessary, whereas the third patient was treated by thrombolysis. From our experience, we conclude that mechanical heart valve replacement is a good therapy option with a low complication rate for patients with congenital heart defects requiring valve replacement, especially when INR self-management is performed.

Time-dependent analysis of leaflets in mechanical aortic bileaflet heart valves in closing phase using the finite strip method

Background and aims of the study Mechanical heart valves (MHV) are widely used to replace dysfunctional and failed heart valves. The bileaflet MHV design is very popular due to its superior hemodynamics. Since their introduction in 1977, the hemodynamics of bileaflet prostheses has been extensively studied. In this study the dynamic behaviour during the closing phase of a bileaflet MHV under normal physiological conditions has been investigated. Methods Fluid analysis is based on the control volume with moving boundaries in the vicinity of the occluder. Unsteady continuity equation, unsteady momentum equation on the control volume and unsteady Bernoulli's equation have been used to calculate velocity of blood flow and force on the occluder tip. To solve the governing equations for the calculation of pressure and the related force, the finite strips method has been implemented. Only 32 strips are sufficient to calculate the force due to pressure on the leaflets. The equations of motion have been solved using the Runge–Kutta method in the fourth order. Results The maximum velocity of the leakage flow in the closing phase falls within the range of 3.5–4.4 m/s. The maximum velocity of the occluder tip is in the range of 2.4–3.2 m/s. The backflow also exhibits oscillation similar to that of the occluder with net backflow rate in the range of 9.7–12.3 ml/beat. The impact force between occluder and its housing is in the range of 80–140 N and impact occurs during 33.1–41.0 ms and the leaflets are completely settled at 108–115 ms in the closing phase. Conclusion The finite strip method was implemented to study the closing phase of a bileaflet MHV. Results are consistent with the previous experimental data. This method is of general applicability to study dynamic behaviour of MHVs.

Flow in a Mechanical Bileaflet Heart Valve at Laminar and Near-Peak Systole Flow Rates: CFD Simulations and Experiments

Time-accurate, fully 3D numerical simulations and particle image velocity laboratory experiments are carried out for flow through a fully open bileaflet mechanical heart valve under steady (nonpulsatile) inflow conditions. Flows at two different Reynolds numbers, one in the laminar regime and the other turbulent (near-peak systole flow rate), are investigated. A direct numerical simulation is carried out for the laminar flow case while the turbulent flow is investigated with two different unsteady statistical turbulence modeling approaches, unsteady Reynolds-averaged Navier-Stokes (URANS) and detached-eddy simulation (DES) approach. For both the laminar and turbulent cases the computed mean velocity profiles are in good overall agreement with the measurements. For the turbulent simulations, however, the comparisons with the measurements demonstrate clearly the superiority of the DES approach and underscore its potential as a powerful modeling tool of cardiovascular flows at physiological conditions. The study reveals numerous previously unknown features of the flow.

The On-X prosthetic heart valve at five years

Background Multicenter clinical trials were conducted in Europe and North America to evaluate the performance of the On-X bileaflet heart valve prosthesis (Medical Carbon Research Institute, Austin, TX). Methods A total of 532 patients underwent implantation, 303 for aortic valve replacement (AVR) and 229 for mitral valve replacement (MVR), at 20 centers from September 1996 to July 2001. The study followed the guidelines of the AATS/STS. Mean follow-up was 23 months (total 1024 patient-years; maximum 5 years). Poolability analysis was performed to show the equivalence of the populations. Results Patients and results were found to be similar and poolable. Freedom from adverse events at 2 years in the study were as follows: thromboembolism, 96.0% for AVR patients and 96.3% for MVR; thrombosis, 100% for AVR and 100% for MVR; bleeding events, 96.6% for AVR and 95.7% for MVR; and overall mortality, 95.2% for AVR and 92.4% for MVR. Median lactate dehydrogenate levels were in the normal range for AVR and MVR patients at all intervals. At 1 year, AVR echocardiographic results for the 19 to 25 valves, respectively, ranged from 1.5 to 2.8 cm 2 for effective orifice area and 9.2 to 4.7 mm Hg for mean gradient, and MVR effective orifice area by pressure half-time was 2.8 cm 2 and mean gradient was 4.2 mm Hg. Conclusions The two trials have given similarly excellent results for the On-X valve.

Is low-intensity anticoagulant therapy possible with an open pivot valve?: analysis of its development.

In 1976, the authors became the codevelopers of the first successful bileaflet heart valve which has become the standard in valvular surgery. However, there still remains a significant rate of thromboembolism and thrombosis which can be attributed to the inherent design of cavities in the pivot area. These cavities contribute to stagnant areas that lead to thrombus formation. The literature indicates that this rate is approximately 2% to 2.5% per patient year. To maintain these thromboembolism rates, certain levels of warfarin have had to be used leading to a significant rate of bleeding complications. We present the theory and development of the new generation of heart valves without cavities and with an open semisphere pivot area. Results of 912 years of clinical studies of almost 50,000 valves, as well as additional features that have been incorporated to improve hemodynamics and quality of life for the patient, are discussed. The clinical results of the European Community Certification and U.S. Food and Drug Administration premarket application (PMA) studies also are given. The differences in the pivot area suggest that a reduction in thromboembolism is possible at anticoagulant levels as low as 1.5 international normalization ratio (INR) resulting in virtually no bleeding complications. Midterm clinical results are encouraging, and studies are being conducted presently to eliminate totally anticoagulants in certain patients.

Effect of the mechanical prosthetic mono- and bileaflet heart valve orientation on the flow field inside the simulated ventricle

Two groups of typical contemporary mechanical heart valves, the Advancing the Standard (ATS) and the Carbomedics (CM) valve (of bileaflet design) and the Bjork-Shiley (BS) mono and Bicer-Val (BV) valves (of tilting-disc design), were tested in the mitral position under the pulsatile-flow condition. This study extends a previous report studying the effect of orientation of the St. Jude Medical (SJM) valve, representing bilcaflet valve design, and the Meditronic-Hall (MH) valve, representing mono-leaflet valve design. The test program utilized a flow visualization technique to map the velocity field inside the simulated ventricle. The study was carried out using a sophisticated cardiac simulator in conjunction with a high-speed video system (200 frames·s−1). The continuous monitoring of velocity-vector time histories revealed useful details about the complex flow and helped establish the locations and times of the peak parameter values. Comparison of the velocity profiles at corresponding flow phases reveals the effects of the differences in valve design and orientation. Based on precise examination of the data, the following general conclusions can be made: pulsatile flow creates three distinct flow phases consisting of accelerating, peak, and decelerating flow; the bileaflet CM and ATS valves in the antianatomical orientation generally create a single, large circulatory flow; the ATS valve scems to offer smoother flow patterns, similar to the SJM valve; and the monoleaflet BV valve and the BS monostrut valve seem to affect the flow characteristics more dramatically, with the posterior orientation exhibiting simple and stable circulatory flow.

Thrombolysis is an effective and safe therapy in stuck bileaflet mitral valves in the absence of high-risk thrombi

OBJECTIVESWe sought to evaluate the effectiveness and safety of thrombolytic therapy in stuck mitral bileaflet heart valves in the absence of high-risk thrombi.BACKGROUNDCurrent recommendations for the thrombolytic treatment of stuck prosthetic mitral valves are partially based on older valve models and inclusion of patients in whom high-risk thrombi were either ignored or not sought for. The feasibility and safety of thrombolysis in bileaflet models may be affected by the predilection of thrombi to catch the leaflet hinge.METHODSWe studied 12 consecutive patients (men/women = 5/7, age 58.8 ± 14.9 years) who experienced one or more episodes of stuck bileaflet mitral valve over a 33-month period and received thrombolytic therapy with streptokinase, urokinase or tissue-type plasminogen activator. Transesophageal echocardiography was performed in all patients. Patients with mobile or large (>5 mm) thrombi were excluded. Functional class at initial episode was I–II in 4 patients (33.3%) and III–IV in 8 patients (66.6%).RESULTSPatients receiving thrombolytic therapy achieved an overall 83.3% freedom from a repeat operation or major complications (95% confidence interval 51.6–97.9%). Minor bleeding occurred in three patients (25%) and allergic reaction in one (8.3%). Transient vague neurologic complaints, without subjective findings, occurred in four patients (33.3%). Three patients had one or more relapses within 5.2 ± 3.1 months from the previous episode, and readministration of thrombolytics was successful.CONCLUSIONSIn clinically stable patients with stuck bileaflet mitral valves and no high-risk thrombi, thrombolysis is highly successful and safe, both in the primary episode and in recurrence. The best thrombolytic regimen is yet to be established.

CFD investigation of steady flow in bi-leaflet heart valve hinge.

This paper focuses on local flow patterns inside the hinge socket of a bi-leaflet mechanical heart valve (MHV), where experimental measurements are difficult due to the extremely small flow region of about 40 microm. The overall objective of this study is to simulate the steady flow in this confined micro channel within the hinge region of a partially protruded ball hinge concept. A CFD simulation of flow through a bi-leaflet heart valve hinge was carried out. Steady flow with the valve leaflet in the fully open position during the valve systole phase was studied using FLUENT 4.4.7 running on Silicon Graphics. Body Fitted Coordinates (BFC) grid distribution was applied in the overall flow domain and great care was taken at the mesh distribution within the hinge local area. The flow study focused on local flow patterns inside the hinge socket of the valve where experimental measurements in the actual size valve are not practical. CFD results show evidence of flow in local area of hinge and no evidence of stagnation. Flow migrates across the clearance, and small vortices are formed after the hinge stoppers. The results indicate that flow in the hinge region is complex and critical for the valve to function effectively.

A Three-dimensional Analysis of Flow in the Pivot Regions of an ATS Bileaflet Valve

Bileaflet heart valves are currently the most commonly implanted type of mechanical prosthetic valve, because of their low transvalvular pressure drop, centralised flow and durability. However, in common with all mechanical heart valves, implanted bileaflet valves show an inherent tendency for blood clot formation at the valve site. Fluid dynamical phenomena associated with blood clotting are elevated blood shear stresses and regions of persistent blood recirculation, particularly when both occur together. Using three-dimensional CFD modelling, combined with enlarged scale experimental modelling, we investigated the blood flow through the ATS bileaflet valve during forward flow, with particular attention to the leaflet pivot regions. Recirculating regions were found both within and downstream of the valve housing ring. Qualitative assessment of the entire cardiac cycle suggested that recirculating blood within the housing ring will be washed away whilst the valve is closed, but as with all bileaflet valve designs recirculating blood downstream of the valve may have a residence time much longer than one cardiac cycle.

Haemocompatibility of DLC and TiC–TiN interlayers on titanium

An interlayer approach was adopted to achieve adequate adhesion between a DLC coating deposited by plasma-assisted CVD and a titanium substrate as part of a study to coat a bi-leaflet heart valve prosthesis. The interlayers (TiC and TiN) were deposited by d.c. ion plating. TiN was also deposited by reactive r.f. sputtering. The substrate, interlayers and DLC were assessed for haemocompatibility and thrombogenicity using a dynamic blood method and interactions with rabbit blood platelets, respectively. The adhesion, activation and morphology of the platelets were determined by stereological techniques using SEM. All surfaces were assessed topographically by stylus profilometry and AFM. Surface energies were compared by liquid drop goniometry. The structure and crystallographic orientation of the TiN interlayers were determined using XRD. The coatings produced no haemolytic effect, except in the case of TiN deposited by reactive r.f. sputtering, which showed a significant effect after longer assessment periods. In contrast to the DLC coating, all of the interlayers showed a slight tendency towards thrombus formation during the later stages of incubation. Platelet spreading correlated with the surface energy of the coatings.

Visualization of the hinge flow in a 5:1 scaled model of the medtronic parallel bileaflet heart valve prosthesis

In this work, a flow visualization experiment was performed to elucidate features of the retrograde hinge flow through a 5 : 1 scaled model of the Medtronic Parallel bileaflet heart valve. It was hypothesized that this model would provide detailed flow information facilitating identification of flow structures associated with thrombus formation in this valve. The experimental protocol was designed to ensure fluid dynamic similarity between the model and prototype heart valves. Flow was visualized using dye injection. The detailed flow structures observed showed the hinge’s inflow channel was the most suspect region for thrombus formation. Here a complex helical structure was observed.

Prevalence of microembolic signals in patients with different types of mono-leaflet and bi-leaflet prosthetic heart valves.

Microembolic signals (MES) have been described in patients with different types of prosthetic valves. The aim of our study was to compare prevalence of MES in patients with mono-leaflet and bi-leaflet valves, and to correlate them with different clinical and echocardiographic parameters. We evaluated 42 patients, 15 with mono-leaflet and 27 with bi-leaflet valves, with 30-minute bilateral simultaneous transcranial Doppler monitoring for the identification of MES. All patients performed blood testing for international normalized ratio (INR), color-coded duplex of supra-aortic vessels, and echocardiography. Eighteen patients (43%) showed MES; 61% of MES were detected bilaterally, 17% only on the left and 22% only on the right middle cerebral artery (MCA), with no interhemispheric difference. MES did not correlate with patient's age, echocardiographic parameters, INR, presence of atrial fibrillation, or site of valve replacement. One (7%) of the 15 patients with mono-leaflet valves had MES, as opposed to 17 (63%) of the 27 with bileaflet valves (p=0.0003). Patients with bileaflet valves and MES had valves of older implantation and, only for aortic valves, larger valve size.

Analysis of 3D transient blood flow passing through an artificial aortic valve by Lattice–Boltzmann methods

The development of flow instabilities due to high Reynolds number flow in artificial heart valve geometries inducing high strain rates and stresses often leads to hemolysis and related highly undesired effects. Geometric and functional optimization of artificial heart valves is therefore mandatory. In addition to experimental work in this field it is meanwhile possible to obtain increasing insight into flow dynamics by computer simulation of refined model problems. After giving an introductory overview we report the results of the simulation of three-dimensional transient physiological flows in fixed geometries similar to a CarboMedics bileaflet heart valve at different opening angles. The visualization of emerging complicated flow patterns gives detailed information about the transient history of the systems dynamical stability. Stress analysis indicates temporal shear stress peaks even far away from walls. The mathematical approach used is the Lattice–Boltzmann method. We obtained reasonable results for velocity and shear stress fields. The code is implemented on parallel hardware in order to decrease computation time. Finally, we discuss problems, shortcomings and possible extensions of our approach.

Pulse duplicator hydrodynamics of four different bileaflet valves in the mitral position

The hydrodynamic performance of four different types of bileaflet heart valve was tested both in steady flow conditions and in a newly designed computerized artificial left heart system. Four 29-mm valves, manufactured by either St. Jude Medical, CarboMedics, ATS Medical or Medtronic Parallel were studied under a steady state and in pulsatile conditions by pressure and flow recording and colour Doppler echocardiography in the `mitral position' at rest and at exercise (cardiac output range 4–30 l/Min). In the steady state, significant differences were measured at peak flows of 30 l/min, with mean (s.d.) increasing pressure drops (mmHg) recorded at 3.48(0.13) (St. Jude Medical), 3.80(0.11) (Medtronic Parallel), 4.31(0.06) (ATS Medical) (P<0.005), and 4.78(0.06) (Carbo Medics), (P<0.005). The mean ratio of the closing volume to the stroke volume was comparable for all valves despite one faulty sample. In the pulsatile state the effective orifice area calculated by the Continuity and Gorlin equations decreased from 2.10(0.04) (ATS Medical), 1.95(0.05) (St. Jude Medical), 1.87(0.03) (Carbo Medics) to 1.67(0.02) (Medtronic Parallel) independently of method used. No significant difference was found in the mean ratio of the regurgitant volume to the stroke volume. The performance index, defined as the ratio of effective orifice area to geometric orifice area ranks the valves in the order ATS Medical (0.44), St. Jude Medical (0.43), Carbo Medics (0.41) and Medtronic Parallel (0.38). Regurgitation visible on colour Doppler echo was a mere consequence of the working principle of the bileaflet valves in the mitral position and was by no means hydrodynamically important nor representatire of severe insufficiencies. Differences between the valve types were notable under severe conditions, such as high peak flow.