This Invited Commentary accompanies the following article:
Citerio G, Pesenti A, Latini R, et al. A multicentre, randomised, open-label, controlled trial evaluating equivalence for inhalational and intravenous anaesthesia during elective craniotomy. Eur J Anaesthesiol 2012; 29:371–379.
The ‘struggle for power’1 in neuroanaesthesia studies continues. The number of well conducted, randomised studies of anaesthetic techniques in this important sub-speciality branch of anaesthesia is very few, yet neuroanaesthesiologists are as sure as any other sub-specialists that what they do does make a difference.
Work published nearly 20 years ago by Todd et al.2 that investigated three anaesthetic techniques (propofol/fentanyl, isoflurane/nitrous oxide and fentanyl/nitrous oxide) in 121 patients is still often quoted and remains relevant, despite the progression in anaesthetic techniques since it was planned. They found that 29% of patients had a new or worsened neurological deficit (although there were no differences among the three groups) and 11 patients (9%) had a ‘serious complication’ in the postoperative period, of which seven were clearly neurological. However, there was a discussion afterwards that their failure to find a difference in the incidence of some of these ‘rare’ events was not the same as stating there was no difference,3,4 which is akin to the problem of trying to predict rare events in general.5 Most European anaesthesiologists are now likely to choose a total intravenous anaesthesia (TIVA) technique or one based on one of the more modern volatile anaesthetics such as isoflurane, desflurane or, particularly, sevoflurane, and the most likely opioid is remifentanil.
The problem in designing and undertaking the studies we need in neuroanaesthesia is the complexity of detecting what may be subtle changes, or changes which only become apparent some months after the procedure. Allied to this is the rare nature of some of these, even after high-risk surgery. One of the most well conducted studies in neuroanaesthesia in the last two decades was the Intraoperative Hypothermia for Aneurysm Surgery Trial (IHAST) investigating the value of mild hypothermia in aneurysm surgery, surely one of the most high-risk procedures in neuroanaesthetic practice.6 One thousand patients were randomised to be subjected to mild hypothermia (or not) during the procedure when temporary arterial clipping might be used. There was no difference in immediate postoperative progress. The Glasgow Outcome Score was evaluated at 3 months and was similarly ‘good’ in both groups. Wong et al.7 reported the incidence of complications after elective neurosurgery with a propofol–remifentanil technique and observed the Glasgow Coma Scale (GCS) in the recovery (PACU) room; the conscious level was ‘normal’ (median GCS 14–15) in all 145 patients. Manninen et al.8 followed 486 patients after a variety of neurosurgical procedures and found that the incidence of neurological complications was 5.7%, including new sensory or motor deficits, confusion or delirium, dysphasia, delayed awakening or postoperative intracranial haematoma requiring re-operation.
If detecting long-term outcomes is so difficult and expensive, what surrogates are available: early recovery (eye opening, extubation); late recovery (discharge from recovery, time to recover to a GCS of 15, psychomotor testing); imaging to detect new lesions; neuropsychological testing for cognitive decline? The anaesthesia literature is full of studies looking at early recovery – including in neuroanaesthesia.9 Are differences in early recovery parameters helpful? It is certainly pragmatic, as neuroanaesthesiologists are still taught to waken the patient as quickly as possible so that a patent airway is established and to allow early assessment of neurological status. Indeed, surgeons often still gauge the quality of the anaesthetic (and, by extension, of the anaesthesiologists with whom they work) by how quickly the patient is awake. However, the slight, and largely clinically irrelevant, differences in early recovery parameters are mostly predictable from the pharmacokinetics of the agents used. Detecting differences in early recovery is more applicable to assessing the suitability of anaesthetic techniques for short procedures in a day case (ambulatory) setting.
The article in this edition of the European Journal of Anaesthesiology by Citerio et al.10 reports a large study of patients receiving one of three anaesthetic techniques: TIVA with propofol and remifentanil, or sevoflurane combined with either remifentanil or fentanyl. Over 400 patients undergoing elective surgery were enrolled and data on the primary outcome (time to reach an Aldrete score of ≥9) was available for 380. An equivalence design was used. There was no difference in early recovery among the three groups. Surgeon-assessed brain relaxation scores were also similar. However, the group receiving the TIVA technique did have attenuated changes in stress biomarkers (plasma cortisol and urinary cortisol and catecholamine concentrations) compared with the two groups receiving sevoflurane. Certainly, the number of patients studied is persuasive that this study will help in gauging the benefit of one of these techniques over the others.
Some aspects of the study design need further consideration. The use by the authors of an equivalence design is clearly explained and justified. They defined plausible limits for equivalence and were able to conclude that there were no significant differences in the treatments within these limits. The study was a randomised controlled trial (RCT), but as the anaesthesiologists were unblinded, the authors went on to justify the use of a prospective randomised open blinded endpoint (PROBE) design to minimise bias. However, while the PROBE design can perform well, its use in a study in which the blinded observers are expecting no difference (the aim of an equivalence design) should, in itself, lead to a bias of reporting no difference! In the typical RCT in which a difference is expected, the use of a PROBE design more robustly tests the reporting by the blinded observers in detecting a true difference, if it occurs. The use, if possible, of a positive control group may help in more robustly concluding equivalence in the treatment groups of interest with PROBE designs, and indeed for equivalence designs in general.
So, does this study move us forward? Of course, by showing that differences in early recovery are mostly insignificant (and probably irrelevant), we can now perhaps move on and stop performing such studies and replace them with ones powered to look at more clinically relevant outcomes. Similarly, by worrying less about the marginal benefits in early recovery, we can look for other, less obvious, advantages of one technique over another. There is some evidence, for example, that differences in anaesthetic technique may influence cancer recurrence.11 Many patients undergoing craniotomy have an intracranial tumour, either benign (e.g. meningiomas which may recur) or more aggressive tumours such as the gliomas. If choice of technique could influence disease progression, we would have a real outcome to investigate.
Both authors are Editors of the European Journal of Anaesthesiology. M.N. has received honoraria from Abbott Laboratories both for speaking and for membership of Advisory Boards in the last 5 years. M.N. is a co-author of a study of two anaesthetic techniques for neuroradiology procedures9 which used a measure of early recovery as the primary outcome.
This article was checked and accepted by the Editors, but was not sent for external peer-review.
1. Yentis SM. The struggle for power in anaesthetic studies. Anaesthesia
2. Todd MM, Warner DS, Sokoll MD, et al. A prospective, comparative trial of three anesthetics for elective supratentorial craniotomy. Anesthesiology
3. Hartung J, Cottrell JE. Negative inferences about rare events require large samples. Anesthesiology
4. Todd MM, Warner DS, Sokoll MD, et al. Negative inferences about rare events require large samples. In reply. Anesthesiology
5. Basaranoglu G, Lyons G, Columb M. Failure to predict difficult tracheal intubation for emergency caesarean section. Eur J Anaesthesiol
6. Todd MM, Hindman BJ, Clarke WR, Torner JC. Mild intraoperative hypothermia during surgery for intracranial aneurysm. N Engl J Med
7. Wong AYC, O’Regan AM, Irwin MG. Total intravenous anaesthesia with propofol and remifentanil for elective neurosurgical procedures: an audit of early postoperative complications. Eur J Anaesthesiol
8. Manninen PH, Raman SK, Boyle K, El-Beheiry H. Early postoperative complications following neurosurgical procedures. Can J Anaesth
9. Castagnini HE, van Eijs F, Salevsky FC, Nathanson MH. Sevoflurane for interventional neuroradiology is associated with more rapid early recovery than propofol. Can J Anaesth
10. Citerio G, Pesenti A, Latini R, et al
. A multicentre, randomised, open-label, controlled trial evaluating equivalence for inhalational and intravenous anaesthesia during elective craniotomy. Eur J Anaesthesiol
11. Curatolo M. Adding regional analgesia to general anaesthesia: increase of risk or improved outcome? Eur J Anaesthesiol