Asymmetric target-concentration–response relations of propofol and remifentanil on the bispectral index: a negative-control study of 50,918 steps

Aim To estimate the bispectral index response to propofol and remifentanil target-concentration steps and to measure the change expected without intervention. Methods We analysed 3,245 VitalDB anaesthetics. A step was a sustained target-concentration change of ≥0.2 µg/mL (propofol) or ≥0.5 ng/mL (remifentanil), with the index extracted over −10 to +15 min and one randomly timed pseudo-step per real step; slopes were fitted by direction. Results 50,918 real steps (24,498 propofol, 26,420 remifentanil) from 3,245 cases and 34,008 pseudo-steps. In pseudo-steps ΔBIS rose 0.771 (95% CI 0.705–0.837) per unit of deviation. Corrected, +1 µg/mL propofol gave −3.35 (−3.46 to −3.23) and +1 ng/mL remifentanil −1.93 (−2.00 to −1.86). Remifentanil was asymmetric (0.13, 0.06 to 0.19); propofol's was smaller and not robust (0.75, 0.62 to 0.89; 1.40, 1.02 to 1.85 combined). Deepening Emax 3.51 and 1.75; onset 80–102 s versus 182–235 s lightening. Concurrent adjustment was additive pooled (γ=+0.040, −0.028 to +0.107) but not when both targets moved the same way (+0.96 and −0.48), absent from pseudo-steps. Ratios varied. Conclusion With a negative control, routine steps give concentration–response estimates conditional on residual indication confounding. Remifentanil responded far more to reduction than to increase; the nociception-mediated explanation is consistent with these data, not established by them. A propofol asymmetry is suggested, not established.