Abstract 1376 Poster Session IV, Tuesday, 5/4 (poster 274)

Doppler mean cerebral blood flow velocity (CBFVM) is used to estimate actual cerebral blood flow (CBF) in neonates in whom the major cerebral artery diameters are too small to measure cerebral flow volume accurately. Pourcelot's resistive index [RI=(Vmax-Vdiast)/Vmax] is widely used as a qualitative measure of cerebral vascular resistance (CVR), although studies have not validated its use. In fact, three reports (Hansen, 1983; Rosenberg, 1985; Batton, 1983) suggest that RI does not reliably predict CBF and CVR. Relative vascular resistance (RVR), calculated as mean arterial BP/CBFVM (RVR), under the Doppler velocity curve) is an alternative to RI. RVR has also been used (Van Bel, 1990; Martinussen, 1996) to estimate regional vascular resistance without adequate validation. We hypothesized that RVR is a better representation of cerebral vascular resistance than in Pourcelot's index. We compared RVR to RI for 143 middle cerebral artery Doppler examinations in 46 babies and found a relatively weak correlation (r=0.28, p<0.05). Because it is not possible to measure actual vascular resistance in premature infants, we reanalyzed our previously reported data (Hansen, 1983) in which we had simultaneously measured CBF (microspheres) and anterior cerebral artery Doppler velocities (n=52) in 1-3 wk old piglets with varying arterial CO2 tensions. We found that RVR was highly correlated with CVR (Fig, square; r=0.7, p<0.001), whereas RI was not (Fig, circles; r=0.4, p=0.07). Furthermore, when multiple linear regression was applied to predict CVR, RVR was highly significant (p<0.001) with RI in the equation, whereas RI provided no further contribution. We conclude that RVR predicts measured cerebral vascular resistance more reliably than does Pourcelot's index. Therefore, we suggest that RVR be used to noninvasively follow changes in cerebral vascular resistance of low birth weight infants.

Fig 1
figure 1

No caption

Funded by American Heart Association B98448P.