Radionuclide estimation of cerebrospinal fluid shunt flow. Evidence supporting an alternative theoretical model.

Flow of cerebrospinal fluid through a surgically implanted valve may be estimated by analyzing the disappearance curve resulting from the injection of a radiotracer into the valve. The standard method for estimating flow assumes an exponential disappearance of the tracer from the valve. This method models the valve as a single well-mixed compartment. Experimental evidence, showing that estimates of flow were dependent upon the site of injection, is at variance with this assumption. An alternative method of analyzing the disappearance curves, based on the area to height ratio (A/H) of the curves, was found to be more consistent with the experimental evidence and resulted in greater precision than the exponential method. It was concluded that optimal results are obtained using the A/H method with a fixed injection technique.