General Discussion coherence. In chapter 6 we reported high coherence values of nearly one using induced blood pressure changes during CPB. According to Giller’s theory these high coherence values would indicate impaired autoregulation, but for different levels of paCO2 different values of phase were found each with high coherence. Coherence itself therefore does not seem to be a good measure for dCA. To identify if gain or phase is the dominant dCA parameter Panerai 20 simulated the effect of varying ARI on TFA parameters using data generated by the Tiecks-model 36. By generating data for all ten ARI values ten different transfer functions may be constructed. By combining each of the phase spectra with the gain spectrum of ARI 0 ten step responses can be generated with varying phase and constant gain. In the same way this can be done for an ARI 9 response. Similarly this procedure can be repeated with constant phase and varying gain. The overall conclusion of this simulation was that phase is dominant and that a high gain value should not be accepted as an indication of impaired autoregulation in the presence of higher values of phase. Reproducibility An important property for clinical applicability of a test is reproducibility of its results. When measurements are less reliable or when wide day-to-day variation occurs, it is difficult to identify statistically significant change e.g. between normal and pathological situation. A measure for reproducibility is the intraclass correlation coefficient (ICC). For a widely used physiological parameter as for example systolic blood pressure measurement using an automated blood pressure unit ICC values above 0.8 have been found 34 which is considered good to excellent reproducibility 9. ICC values between 0.40 and 0.75 are considered to represent fair to good reproducibility whereas values less than 0.40 represent poor reproducibility. Not many studies have been performed investigating reproducibility of dCA parameters. Studies reporting ICC are performed for spontaneous ABP variations using TFA analysis 5, 11, 38 and multimodal pressure flow analysis 12 and for paced breathing 11 and sit-stand maneuvers using TFA analysis 38. Apart from differences in dCA analysis methods also different parameters are reported, which complicates comparison of results between studies. However, a common finding is high reproducibility (ICC > 0.8) for recorded mean finger ABP and mean CBFV values using TCD. Only van Beek et al 38 report an ICC of 0.32 for ABP, which is an odd finding. They also report invalid negative ICC for phase in the very low frequency band, which 127
proefschrift gommer inhoud met kaft.indd
To see the actual publication please follow the link above