3. Clinical investigation of lithium dilution curves produced by right and left atrial injection of lithium chloride. [Band DM, Linton RAF, O'Brien TK, Jonas MM , Linton NWF. The shape of indicator dilution curves used for cardiac output measurement in man. Journal of Physiology 1997; 498:225-229.]


The objectives were first to determine whether there is significant loss of intravenously injected Li+ from pulmonary capillaries, and secondly to determine the transfer function of the right heart and pulmonary circulation. The latter is important from the point of view of the shape of arterial indicator dilution curves which in turn is important because of the need for extrapolation in deriving cardiac output from a curve.


LiDCO curves were measured following injection of lithium chloride into either the left or right atrium.

Study population

The subjects were 6 postoperative cardiac patients in the intensive care unit. The left atrial catheters had been inserted directly at the time of operation because of poor left ventricular function.


The areas inscribed by the primary lithium dilution curves (and hence cardiac outputs) produced by left atrial injection of the lithium were 3.6 % smaller than those following right atrial injection. This difference was not statistically significant . Within each patient the injections had to be given not less than 5 min apart, so some variation in actual cardiac output would be expected.
For each patient, the curve data produced by left atrial injection was transferred into an Excel spreadsheet and passed through two sequential filters of similar time constant to reduce the peak to the same value which had been obtained in that patient following right atrial injection. A time delay was also added to align the peaks on the x-axis. The close agreement (see
Figure) between the delayed, filtered left atrial curves and corresponding right atrial injection curves provides evidence that the transfer function of the right heart and lungs can be modelled by a delay and 2 sequential filters, which in turn provides further evidence for the lognormal nature of the LiDCO curves.


There is bound to be some loss of lithium in its passage through the pulmonary circulation and so on the basis of this study we allow a 3.6% loss from the pulmonary circulation in the calculation of cardiac output.