Abstract

Although it is known that pressure gradients and calculated valve areas in bioprosthetic valves are highly flow dependent, no studies have compared bioprosthetic valve performances while adjusting for differences in flow rate. We therefore studied 75 patients undergoing mitral valve replacement who were randomized to receive either Hancock (n = 35) or Carpentier-Edwards (n = 40) bioprosthetic valves. Pressure gradients were measured using transducer-tipped catheters to record left ventricular and left atrial pressures and cardiac outputs by thermodilution. Repeated measurements were made in each patient after either pacing, fluid infusion, or pharmacological intervention to vary flow rates for a total of 239 measurements (mean, 3.2 measurements per patient). Using analyses of variance and covariance, mean valve gradients and the calculated Gorlin area were adjusted for flow rate, valve size, valve type, and interpatient differences to compare hemodynamics. Without flow and interpatient adjustment, the univariate analysis suggested higher mean gradients in the Carpentier-Edwards 29-mm valves (p = 0.038), with a trend toward higher gradients and smaller areas in the Hancock 33-mm valves (p = 0.057 and 0.059, respectively). After adjustment for flow rate and interpatient differences, however, there was no difference at any valve size in the mean pressure gradients (p = 0.13-0.89) or Gorlin valve areas (p = 0.34-0.96). Although measurements within a given patient were consistent, marked interpatient variabilities in gradients and areas were observed for identical valve types and sizes, which were as significant as flow-dependent or size-dependent changes. We conclude that comparisons of valve performance should adjust for variations in flow rate and for interpatient differences with the use of repeated-measures designs.(ABSTRACT TRUNCATED AT 250 WORDS)

View details for PubMedID 2225438