Are we overdosing older patients?

One of the most fundamental tasks of the anaesthesiologist is to ensure hypnosis during general anaesthesia, for patients of all ages. However, we know remarkably little about individual patients' anaesthetic requirements, using instead population-based metrics such as the minimum alveolar concentration (MAC) of volatile anaesthetic required to prevent movement in response to surgical incision for 50% of the population—an index of movement, not consciousness or recall. It became apparent early on that older patients had MAC values lower than those of younger patients,¹ yielding the concept of age-adjusted MAC (aaMAC): the proportion of the MAC of volatile anaesthetic required to prevent movement to surgical incision for 50% of the population of that age. So, 0.7 MAC for a 40-yr-old individual may represent a relative overdose for an 80-yr-old person.

This is a canonical teaching in modern anaesthesia training. However, only within the past decade has the widespread availability of intraoperative EEG and processed ‘depth of anaesthesia’ indices finally made possible the empirical testing of age adjustment. Documentation of how aaMAC is actually used in routine clinical practice with an alternative metric of anaesthetic depth is now possible. Ni and colleagues² present a single-centre study of the latter question, built upon a meta-regression of previous work addressing fundamental tenets underlying the former.

The study ambitiously attempts to answer three questions: (1) Is the accepted age adjustment factor actually supported by a review and meta-regression of published MAC studies in a range of adults? (2) Are older adults receiving what we would consider ‘appropriately’ age-adjusted concentrations of volatile anaesthetics in a largely unselected single-centre academic population? (3) Does a processed EEG-based measure of anaesthetic depth (here, the bispectral index or BIS) behave as we would expect in response to administered concentrations of volatile anaesthetics in an older population?

The answers are, intriguingly, yes (the canonised adjustment factor of ∼6–7% per decade seems to be appropriate), no (older adults are administered more volatile anaesthetic than their aaMAC indicates they should require), and no (older adults paradoxically display higher BIS values even in the presence of a presumed relative anaesthetic overdose). These latter two conclusions require unpacking, as they have important implications for how we do—and how we should—provide care to this unique, and potentially vulnerable, population.

At the academic centre where this work was performed, anaesthesia machines do not routinely generate an aaMAC; rather, an end-tidal concentration is displayed, with the option to display MAC fraction (without age adjustment). The explanation for the authors' finding—that older adults receive a greater aaMAC than younger individuals—may be more attributable to human engineering than any other factor. When attention is divided among many competing tasks, it is cognitively easier to empirically tweak the volatile anaesthetic delivery in relation to the MAC of a 40-yr-old patient rather than properly calculate a true aaMAC (or equivalent end-tidal volatile gas concentration) for each patient. The increasing availability of aaMAC automatically computed by the gas analyser or anaesthesia machine may fix this tidily.

But what if that isn't the reason for administering what we would consider a relative overdose? Roughly 30% of patients—particularly older patients—in this study underwent concurrent BIS monitoring. We might assume that in those patients, anaesthesia providers were attempting to titrate to individual requirements for hypnosis, rather than a population-based MAC. And here, the authors find that sometimes either BIS misleads, or that aaMAC misleads, for older adults.

The authors confirm that the BIS is relatively invariant at MAC fractions of 1.0 MAC and greater, although interestingly they observe a negative correlation between MAC and BIS in the range of 0.75–1.0 MAC (interesting largely because, although this is how it ought to work, BIS has been criticised for being insensitive to clinically relevant changes in aaMAC³). In comparison with prior work, Ni and colleagues² do not restrict to pharmacokinetically censored data (i.e. their findings may reflect a stronger correlation because of the acute changes in volatile anaesthetic administration), so relevance to the maintenance phase of anaesthesia is somewhat unclear. Perhaps appropriately, then, this finding is not emphasised in favour of the hypothesis-driven, and titular, result that older adults show a paradoxically higher BIS value at equivalent aaMAC values.

That older adults show different EEG signatures at adequate levels of hypnosis compared with younger adults is also well known.⁴ Slow waves and alpha waves decrease with age, and high-frequency activity is relatively more predominant, even at high doses of volatile anaesthetics. Burst suppression is also more common in older adults. Although burst suppression will tend to decrease the BIS number, the other age-dependent features—lower alpha power relative to higher-frequency bands—will tend to increase the BIS number at equi-hypnotic doses of volatile anaesthetic compared with younger adults. Importantly, although it is assumed that processed EEG indices are calibrated using loss of responsiveness as a meaningful change in clinical status that should be reflected in the index value, it is not clear how transition to burst suppression is used by manufacturers. The effect-site concentration of different anaesthetics varies so widely at the point of EEG suppression as to be incomparable with traditional ideas of MAC or its correlate for intravenous anaesthetics.⁵

Thus, there are two competing explanations for Ni and colleagues'² finding of higher aaMAC with older age. The most likely one is the ‘convenience hypothesis’ (mentioned earlier), which suggests anaesthesia providers are using a one-size-fits-all approach when, clearly, one size does not fit all. However, another explanation is that, faced with a black box algorithm accusing the provider of delivering a barely adequate anaesthetic depth—the patient is showing an inappropriately high BIS number for the aaMAC being administered, resulting from known EEG characteristics of older adults—the provider attentively and ‘therapeutically’ increases volatile anaesthetic administration. This explanation is less likely because, on average, the age-related decrease in volatile anaesthetic delivery was the same for the BIS-guided and non-BIS groups. Nevertheless it is a salutary warning that we should be aware of, and perhaps tolerate, higher BIS values in individual older patients.

This reflects a greater issue in medicine: that the complex, older, multimorbid, inhomogeneous population that we routinely care for has their care extrapolated from studies of simple, middle-aged, healthy, homogeneous adults. BIS may perform perfectly well in a young population, but—based on this work by Ni and colleagues² and by others—its utility in the complex geriatric population is suspect. Ni and colleagues² have performed a valuable ‘post-marketing surveillance’ task with this publication. Furthermore, their finding also suggests that it may be particularly difficult to justify decreasing anaesthetic administration to older adults on the basis of BIS number or EEG characteristics (in the absence of burst suppression). Because some older adults show a systematically higher BIS value and corresponding EEG characteristics while receiving a relative overdose of anaesthetic (greater than that needed to ensure hypnosis), how can an anaesthetic provider confidently reduce anaesthetic dose in the face of a relatively high-frequency, low-amplitude EEG trace?

It is important to remember that the mathematical model used by Ni and colleagues² accounts for only about a third of the variation in drug delivery. Some patients may be driving these trends, whereas other older adults respond as expected to increasing anaesthetic doses. A quick glance at Figure 2 shows some patients getting 0.25 MAC and some getting 1.5 MAC. What were those management decisions based upon? Obviously, other (undefined) factors are driving this variability in hypnotic drug titration. The key to future individualised patient care will lie in the further understanding of these factors. Far from replacing anaesthesia provider judgement regarding anaesthetic dosing, Ni and colleagues² provide persuasive evidence that when using BIS or raw EEG in older adults, providers must maintain an even more careful consideration of bias and cognitive availability, clinical scenario, vital signs, and—of course—the EEG, as yet another piece of information to refine our understanding of how our care affects body and brain.

Authors' contributions

Writing and revision of editorial: ELW, JS.

Declaration of interest

The authors declare that they have no conflicts of interest.

Funding

National Institute on Aging of the National Institutes of Health grant (R03AG059822 to ELW) and the Foundation for Anesthesia Education and Research (ELW).