Validating HeartSmart^® against the Cardiopulmonary Bypass Machine

Abstract

Purpose: To assess the utility of HeartSmart ^&#174 , a new method that uses empirical physiological formulae to estimate hemodynamic variables, in estimating cardiac index during open heart surgery when compared with the cardiopulmonary bypass machine. Methods: This was a prospective, blinded study of patients undergoing elective cardiac bypass surgery. We monitored and compared the flow rates provided by the cardiopulmonary bypass machine (roller pump) with estimates derived from the empirical physiological formulae embedded in the HeartSmart ^&#174 monitor in 32 patients. Cardiac index values were compared at the start of cardiopulmonary bypass, during re-warming, and at the end of cardiopulmonary bypass. Results: A total of 256 paired sets of measurements were suitable for comparison. The mean of the differences or bias (95% limits of agreement) was 0.09 l/min/m 2 and the limits of agreement –0.86 to 1.05 l/min/m 2 . The mean difference of the sets of 256 measurements was 585.5 l/min/m for the pump and 575.0 l/min/m 2 for the empirical physiological formulae—a difference of 5% l/min/m 2 . The range of flow rates for the pump was 1.2 to 2.85 l/min/m 2 ; for the empirical physiological formulae, the range was 1.2 to 3.0 l/min/m 2 . Conclusion: The cardiac index estimates derived from the empirical physiological formulae in the HeartSmart ^&#174 software are in good agreement with pump cardiac output rates. These results suggest that HeartSmart ^&#174 measurements are sufficiently accurate for assessing hemodynamic variables in many groups of patients.