Theoretical Considerations and Model Experiments on the Validity of Indicator Dilution Methods for Measurements of Variable Flow

Abstract

Conventional dilution methods can be used to measure pulsatile flow only when heart rate and stroke volume are constant and when there is the mixing region between injection and sampling sites. We determined from theory and model studies the errors which arise when constant flow equations (CFE) are used to calculate variable flow. To eliminate some of these errors we proposed a variable flow equation (VFE) which permits recording of instantaneous and computation of mean flow rates. The VFE indicates that flow is proportional to the reciprocal of the concentration when recorded close to the injection site or distal to a mixing region during a constant rate infusion of indicator. When flow is variable the CFE (which uses time-averages of indicator concentration) will underestimate mean flow. The underestimates increase as the variations in flow get larger but are usually less than 10%. It was also noted that an inappropriate distance between injection and sampling sites caused the concentration-time curve to become distorted. The distortion gave rise to overestimates of flow of up to 40% by either CFE or VFE. The theoretical prediction of the errors were tested in models and confirmed. We used a thermal dilution method which was preferable because indicator concentrations could be measured directly in the stream, thus eliminating errors arising from catheter sampling. Under favorable conditions of mixing and sampling it was possible to record accurately mean and instantaneous pulsatile flow by thermal dilution method and the VFE. The experimental results suggested that the thermal dilution method in conjunction with the proposed VFE should lend itself to continuous measurements of cardiac output in non-steady states in animals and man.