Geographic differences in the target-controlled infusion estimated concentration of propofol: bispectral index...

Variability in drug responses could result from both genetic and environmental factors. Thus, drug effect could depend on geographic location, although regional variation is not generally acknowledged as a basis for stratification. There is evidence that the pharmacokinetic set developed in a European population for the target-controlled infusion (TCI) of propofol does not apply in Chinese patients; however, we are not aware of previous studies comparing the estimated concentration-bispectral index (BIS) response of Caucasian patients in Europe with that of Chinese patients in China.

The Diprifusor™ TCI pump, incorporating the pharmacokinetic model proposed by Marsh et al., was applied to 30 Caucasian patients in Austria and 30 Chinese patients in China. The estimated plasma concentration (C(p)) of propofol for the two groups was set at 1 μg·mL(-1) and increased by 1 μg·mL(-1) every minute to gradually reach 5 μg·mL(-1) after 5 min. The BIS values were fitted against the estimated C(p) and the predicted effect-site concentration (C(e)) in a sigmoid E(max) model.

The sigmoid E(max) curves were shifted significantly to the left in the Chinese group compared with the Austrian group. After 5 min, the BIS value in the Chinese group was lower than in the Austrian group (mean ± standard deviation [SD], 47.2 ± 3.6 vs 63.6 ± 5.4, respectively; P = 0.0006). The estimated C(p) at loss of consciousness (LOC), predicted C(e) at LOC, and time to LOC, were lower in the Chinese group than in the Austrian group (3.3 ± 0.8 μg·mL(-1), 1.6 ± 0.4 μg·mL(-1), 2.8 ± 0.6 min, respectively, vs 4.6 ± 2.8 μg·mL(-1), 2.4 ± 1.5 μg·mL(-1), 3.9 ± 0.5 min, respectively; P < 0.0001).

When propofol is given using the same TCI protocol, Chinese patients in China lost consciousness faster and at a lower estimated plasma concentration than Caucasians in Austria. Larger studies are needed to map geographically appropriate TCI infusion models.