Assessment of different data representations and averaging methods on the Spiegelberg compliance device

Abstract

The Spiegelberg Compliance Device (Spiegelberg KG, Hamburg, Germany) has been available for the automated measurement and calculation of minute by minute intracranial compliance. Widespread practical use has been somewhat limited by the instability of values; especially at low intracranial pressures. We looked at two aspects of a methodology in an attempt to increase the value of the Spiegelberg device in the clinical setting. Firstly, we discussed the difference in representing measured values as elastance (dp/dv) instead of compliance (dv/dp); and secondly we proposed the use of an averaging algorithm called the Exponentially Weighted Moving Average (ewma), which could be applied as a flexible method to follow trends and rapid changes in the elastance (or compliance). Clinical data from sixteen patients were gathered and statistical analysis was focused on three particular aspects, the coefficient of variation which indicates the variability of data values, the correlation between the elastance (or compliance) time series and the underlying ICP signal and the percentage of outliers greater than 2.5 standard deviations from the mean. Our results showed that expressing elastance (dp/dv) instead of compliance (dv/dp) yielded fewer outliers and had a better correlation to ICP, and the ewma method had a better correlation to ICP than the Spiegelberg method. © 2005 Springer-Verlag.