Background: Accumulation of lung fluid in the critically ill patient is believed to attenuate impedance cardiac output (CO(IC)) measurements. However, this phenomenon has never been shown experimentally.
Methods: In eight anaesthetized and ventilated dogs (weight 15-22 kg) a high-precision flow probe was placed on the ascending aorta via a left thoracotomy incision and the direct cardiac output (CO(FP)) was measured. Simultaneous CO(IC) measurements were made using a RheoCardioMonitor (ACMA, Singapore). Lung oedema was induced by intravenous oleic acid 0.1 mg kg(-1). Lung fluid was assessed by the decrease in basal thoracic impedance (Z(b)). Percentage errors between the two methods (CO(IC)-CO(FP)) were calculated and compared as Z(b) decreased at 1 Omega intervals.
Results: During the experiment mean Z(b) decreased from 35.9 (sd 5.2) to 27.8 (6.5) Omega (P=0.0037). This occurred over a period of 225 (range 112-338) min and Z(b) decreased by 1 Omega every 51 (22-68) min. The presence of excessive lung fluid was confirmed at post-mortem. Before lung oedema was induced, CO(IC) was 1.5 (0.6) litre min(-1) and the corresponding value of CO(FP) was 1.5 (0.7) litre min(-1) (data from eight dogs). As Z(b) decreased, and lung fluid accumulated, the error between CO(IC) and CO(FP) widened (P<0.0001, anova for repeated measures). Eventually, CO(IC) decreased to 0.7 (0.3) litre min(-1) and the corresponding value of CO(FP) was 1.2 (0.3) litre min(-1) (DeltaZ(b)=5 Omega, data from six dogs). Mean arterial pressure, central venous pressure and systemic vascular resistance were kept constant.
Conclusion: The presence of lung fluid attenuates CO(IC) measurements with respect to CO(FP).