Paradox of age: older patients receive higher age-adjusted minimum alveolar concentration fractions of volatile anaesthetics yet display higher bispectral index values

Abstract

Background: Minimum alveolar concentration (MAC) and MAC-awake decrease with age. We hypothesised that, in clinical practice, (i) end-tidal MAC fraction in older patients would decline by less than the predicted age-dependent MAC decrease (i.e. older patients would receive relatively excessive anaesthetic concentrations), and (ii) bispectral index (BIS) values would therefore be lower in older patients.

Methods: We examined the relationship between end-tidal MAC fraction, BIS values, and age in 4699 patients > 30 yr in age at a single centre using unadjusted local regression (locally estimated scatterplot smoothing), Spearman's correlation, stratification, and robust univariable and multivariable linear regression.

Results: The end-tidal MAC fraction in older patients declined by 3.01% per decade (95% confidence interval [CI]: 2.56-3.45; P<0.001), less than the 6.47% MAC decrease per decade that we found in a meta-regression analysis of published studies of age-dependent changes in MAC (P<0.001), and less than the age-dependent decrease in MAC-awake. The BIS values correlated positively with age (ρ=0.15; 95% CI: 0.12-0.17; P<0.001), and inversely with the age-adjusted end-tidal MAC (aaMAC) fraction (ρ= -0.13; 95% CI: -0.16, -0.11; P<0.001).

Conclusions: The age-dependent decline in end-tidal MAC fraction delivered in clinical practice at our institution was less than the age-dependent percentage decrease in MAC and MAC-awake determined from published studies. Despite receiving higher aaMAC fractions, older patients paradoxically showed higher BIS values. This most likely suggests that the BIS algorithm is inaccurate in older adults.

Keywords: age factors; anaesthesia; dose–response relationship; electroencephalography; monitoring.

Copyright © 2019 British Journal of Anaesthesia. Published by Elsevier Ltd. All rights reserved.

Figures

Fig 1

Consolidated Standards of Reporting Trials flow diagram of patient selection. BIS, bispectral index; MAC, minimum alveolar concentration; MRN, medical record number.

Fig 2

Age-dependent changes in minimum alveolar concentration (MAC) and bispectral index (BIS) values. (a) End-tidal delivered MAC fraction (mean of the medians) as a function of age. Unadjusted robust linear regression model estimates decline (95% confidence interval [CI]) of 3.41% (2.93%, 3.90%) per decade, which is significantly lower than the expected age-dependent decline (P<0.001). (b) Plot of intraoperative BIS index values (mean of the medians) by age. Spearman's ρ=0.15; 95% CI (0.12, 0.17); P<0.001. (c) Age-adjusted end-tidal delivered MAC (aaMAC) fraction and BIS values, by age.

Fig 3

Subgroup analyses of end-tidal delivered MAC fraction with locally estimated scatterplot smoothing and robust linear regression models. (a) Age-dependent change in end-tidal MAC fraction differed amongst patients who received three differing fentanyl doses (by tertile). Slope estimates from regression models within fentanyl dose subgroups found declines (95% confidence interval [CI]) of 3.50% (2.59%, 4.41%), 3.21% (2.34%, 4.08%), and 3.28% (2.51%, 4.05%) in the lowest, middle, and highest groups, respectively; all declines were significantly lower than the age-dependent decline (P<0.001 for all three groups). (b) Age-dependent change in end-tidal MAC fraction amongst patients who did vs did not receive nitrous oxide. Slope estimates from regression models within nitrous-oxide-given subgroups found declines (95% CI) of 3.00% (2.55%, 3.46%) and 3.53% (2.70%, 4.36%) in patients who did not vs did receive nitrous oxide, respectively; both declines were significantly smaller than the age-dependent decline (P<0.001 for both). (c) Age-dependent change in end-tidal MAC fraction amongst patients who did vs did not receive a regional or neuraxial block. Slope estimates from regression models within these subgroups found declines (95% CI) of 3.15% (2.65%, 3.65%) and 4.11% (2.41%, 5.81%) in patients who did not vs those who did receive blocks, respectively; both declines were significantly lower than the age-dependent decrease (P<0.001 for patients who did not receive blocks group, and P=0.006 for patients who received blocks. (d) Age-dependent change in end-tidal MAC fraction amongst patients who did vs did not receive ketamine during surgery. Slope estimates from regression models within these subgroups found age-related declines of 3.31% (2.81%, 3.81%) and (95% CI) 5.78% (3.79%, 7.77%) in patients who did not vs those who did receive intraoperative ketamine, respectively; the decline amongst those who did not receive ketamine was significantly lower than the age-dependent decrease (P<0.001), but amongst the 360 patients who received ketamine, the rate of decline was no different than expected (P=0.496).

Fig 4

Locally estimated regression and 95% confidence interval for the relationship between age-adjusted end-tidal minimum alveolar concentration (aaMAC) and bispectral index (BIS) by age strata. Region of separation appears between aaMAC of 0.75 and 1.0.