Debate: Conscious sedation: Is it safe?


Brain to Brain_main

Endovascular therapy for the treatment of acute ischaemic stroke has now been established as effective, so far so, that many of the relevant societies and associations have recently issued updated recommendations suggesting this type of therapy as a first-line treatment. Now, however, increasing emphasis has been placed on the time it takes to reperfuse the patient from their arrival at the hospital. One of the steps that has been taken to improve speed and efficiency is the use of conscious sedation in place of general anaesthesia when possible—but is it safe? Here, Bijoy K Menon and Mayank Goyal face off with Christophe Cognard about which approach should be used and why.


Conscious sedation

By Bijoy K Menon and Mayank Goyal

“It is vain to do more what can be done with less.”

William of Occam.

Mayank Goyal & Bijoy Menon_main
Mayank Goyal & Bijoy Menon

Stroke treatment has been revolutionised by recent trials showing benefit of endovascular treatment. One of the primary lessons learnt from these trials is that speed and efficiency in administering this treatment is essential in improving patient outcome. This approach to endovascular treatment relies on collaborative teamwork and eliminating any kind of waste in workflow and health system processes. Stroke treatment has benefited from using the lean six-sigma approach to improving efficiency in administering treatment to patients.

The lean six-sigma approach is data-driven. Over the last many years, we as well as others have looked in detail into our own systems and workflow. Through a data-driven approach that looks at each step from stroke onset to treatment success, this research has identified many steps in the workflow that are either unnecessary or can be executed using a parallel workflow, thus saving time and improving patient outcome. We have previously published on strategies that are necessary to consistently open vessels within 90 minutes of reaching the hospital. The key ones are: fast, efficient, decision oriented imaging; not waiting for a clinical response to t-PA before taking the patient to the angiosuite; preorganisation of interventional suite and team, etc. This data driven approach has also identified the default use of general anaesthesia in patients with acute stroke who are candidates for endovascular therapy as inefficient and in many cases, harmful.1,2

Data from multiple clinical trials including IMS III, MR CLEAN and ESCAPE show that administering general anaesthesia prolongs time to treatment administration without in any way improving patient outcome.2,3 This evidence is supported by data from registries like STAR and numerous single and multicentre studies that show similar results.1,4 Moreover, data also show that by reducing systemic blood pressure and consequently cerebral perfusion pressure, general anaesthesia may increase the rate at which infarct grows before reperfusion is achieved.1 We are yet to identify a systematically done data-driven study that suggests to any benefit from administering general anaesthesia by default to all endovascular treatment eligible stroke patients. All the recent trials showed a consistent and extremely low procedural complication rate; this complication rate was not influenced by the presence or absence of GA. As physicians, we realise that in exceptional patient circumstances like respiratory distress, cardio-respiratory compromise or extreme agitation, general anaesthesia is essential. The lean six-sigma approach towards efficiency in administering treatment, a focus on saving time and brain and all past data shows us that these exceptions should never become the rule.

From our personal experience, one other major advantage of not using GA is the ability to examine patients during the procedure. This can help guide treatment decisions including when to stop; whether to chase a tiny M3 clot. We are able to send most of our patients to a step-down unit rather than an ICU and likely reduce the likelihood of chest infections as well. A detailed economic analysis of data from the recent trials is currently underway and results should be available soon. We suspect that we would find that the use of GA significantly increased the cost of the procedure.

The debate on default use of general anaesthesia should be settled by now. In the best interest of patients, we request stroke physicians and interventionists to only use general anaesthesia when absolutely needed. Yes, we acknowledge that for some interventionists, this would mean getting outside of their comfort zone; however, the fact that we could accomplish this philosophy through training and data across all the centres in ESCAPE trial (only 15 patients in the endovascular arm were done under GA; mostly the ones where intubation was needed for other medical reasons, not for the procedure) suggests that it can be done.3

As Peter Drucker so thoughtfully said, “There is nothing so useless as doing efficiently that which should not be done at all.”

Bijoy K Menon is assistant professor in the Departments of Clinical Neurosciences, Radiology and Community Health Sciences, Cumming School of Medicine, University of Calgary, Canada

Mayank Goyal is professor and director of Research, Department of Diagnostic Imaging, University of Calgary, Canada


  1. Davis MJ, Menon BK, Baghirzada LB, Campos-Herrera CR, Goyal M, Hill MD, et al. Anesthetic management and outcome in patients during endovascular therapy for acute stroke. Anesthesiology. 2012;116(2):396-405.
  2. Abou-Chebl A, Yeatts SD, Yan B, Cockroft K, Goyal M, Jovin T, et al. Impact of General Anesthesia on Safety and Outcomes in the Endovascular Arm of Interventional Management of Stroke (IMS) III Trial. Stroke. 2015;46(8):2142-8.
  3. Goyal M, Demchuk AM, Menon BK, Eesa M, Rempel JL, Thornton J, et al. Randomized assessment of rapid endovascular treatment of ischemic stroke. N Engl J Med. 2015;372(11):1019-30.
  4. Menon BK, Almekhlafi MA, Pereira VM, Gralla J, Bonafe A, Davalos A, et al. Optimal workflow and process-based performance measures for endovascular therapy in acute ischemic stroke: analysis of the Solitaire FR thrombectomy for acute revascularization study. Stroke. 2014;45(7):2024-9




General anaesthesia

By Christophe Cognard

Christophe Cognard
Christophe Cognard

Recent randomised controlled trials have demonstrated the efficacy of mechanical thrombectomy in improving the functional outcome of patients suffering from acute ischaemic stroke due to a large vessel occlusion (LVO). Several national and international scientific societies guidelines were since published establishing thrombectomy as the treatment of choice (with or without IV thrombolysis) in large vessel occlusion anterior strokes.
There is today a wild field of uncertainty considering thrombectomy LVO stroke and lot of pending questions:

  • Which patients should receive thrombectomy depending on clinical situation (patient age, NIHSS score, time from onset, level of occlusion)?
  • Which imaging should be performed for thrombectomy decision-making (NCT, CTA/CT-perfusion, MRA/MR-perfusion, software evaluated target mismatch)?
  • Which technique should be used for mechanical thrombectomy (8F balloon/intermediate catheter aspiration, type and size of stent retriever).
  • But one of the most critical questions concerns the type of anaesthesia that should be done to practice mechanical thrombectomy—general anaesthesia (GA) or conscious sedation (CS)?

Possible advantages of general anaesthesia are:

1) Complete immobilisation of the patient allowing better visualisation of vessels and devices during the procedure to prevent thrombectomy-related complications. Patient immobilisation could as well allow easier, faster recanalisation and higher rate of complete recanalisation.

2) Adequate ventilation and airway protection.

3) Limiting patient discomfort and pain.

Possible advantages of conscious sedation are:

1) It is supposed to reduce time to thrombectomy procedure initiation.

2) It does not induce blood pressure lowering.
The drawback of CS is eventual need to convert acutely to GA accompanied by emergency intubation, which may be associated with a higher rate of aspiration pneumonia and poor outcome.

In the recent RCTs, GA was performed in: 37.8% in MR CLEAN; 9.1% in ESCAPE; 36% in EXTEND-IA; 37.1% in SWIFT PRIME; 6.7% in REVASCAT; and 50% in THRACE. This wide difference of patients treated under GA or CS in these recent RCTs perfectly illustrates the total absence of consensus concerning the type of anaesthesia. It is obvious that some centres have decided since they started doing thrombectomy years ago to perform the procedures under CS when the others decided to use GA. The initial reason to start with or without GA was probably mainly dictated by organisational issues and mainly the anaesthetist’s team availability. Besides, we all started doing stroke procedures years ago and all began by using intra-arterial lytics injections. Then the procedures became more and more sophisticated with nowadays the use of stent retrievers with co-axial or three-axial approach and aspiration making GA more useful than before.

Several studies were published before the recent RCTs about type of anaesthesia in thrombectomy procedures. All of them were reviewed in a recent meta-analysis.1 Nine studies enrolling 1,956 patients (814 with GA and 1142 with CS) were included. Compared with patients treated under CS, patients undergoing GA had higher odds of death (OR2.59; 95% CI, 1.87–3.58) and respiratory complications (OR2.09; 95% CI, 1.36–3.23) and lower odds of good functional outcome (OR0.43; 95% CI, 0.35–0.53) and successful angiographic outcome (OR0.54; 95% CI, 0.37–0.80). No difference in procedure time (p=0.28) was seen between the groups. Pre-intervention NIHSS scores were available from six studies; in those, patients receiving GA had a higher average NIHSS score. Following adjustment for the baseline NIHSS score, GA was associated with lower, but not statistically significant, odds of good functional outcome.

Among all the recent RCTs on thrombectomy, only the MR CLEAN group has already published an evaluation of the impact of the type of anaesthesia on functional outcome.2 Among 348 patients, 70 patients received GA and 278 patients CS. CS was significantly associated with good clinical outcome (odds ratio 2.1, 95% confidence interval 1.02–4.31). After adjusting for pre-specified prognostic factors, the point estimate remained similar; statistical significance, however, was lost (odds ratio 1.9, 95% confidence interval 0.89–4.24). The additional multivariable ordinal regression analysis showed a shift in distribution on the mRs in favour of the non-GA group (adjusted common odds ratio 1.6, 95% confidence interval 0.98–2.54). This also was not statistically significant.

Patients in the GA group had a longer time from onset of symptoms to start of IAT of 00:20 hours (median 04:01; interquartile range 01:53 hours versus 03:40; interquartile range 01:41) and were more frequently treated with mechanical thrombectomy only (32/70 [46%] versus 61/278 [22%]). The distribution of occlusion site was similar in both groups. In contrast to previously published series, in the MR CLEAN study both the GA and CS group had equal scores on baseline NIHSS. The authors concluded that difference in baseline stroke severity is not the reason for improved clinical outcome after CS patients in their cohort.

In the end, the literature before and after the publication of the recent RCTs is extremely confusing. It is very likely that in all these papers with both GA and CS performed, GA was performed in more severe and complex cases, the reason to perform GA being patient agitation, respiratory disturbances or loss of consciousness or more challenging vascular occlusions (cervical or tandem occlusion, foreseen difficult catheterisation). Even if the baseline NIHSS is equal in both groups, the patient global clinical status severity is probably the main reason for improved clinical outcome after CS patients in MR CLEAN cohort. Indeed, in the seven recent RCTs, even if the mean NIHSS score in all series is around 17, the rate of good functional outcome (mRs 0 to 2 at three months) varies from 19.1 to 42.1% in the IV group and from 32.6 to 71% in the thrombectomy group.

Finally, all the published literature concerning the difference in functional outcome in GA versus CS patients does not provide any reliable data and information allowing for the drawing of any conclusion on the best option and no recommendation should be done. The only way to really know is indeed to perform randomised trials comparing for both types of anaesthesia the three months functional outcome and evaluating precisely: 1) time delay, 2) blood pressure level, 3) recanalisation rate, 4) procedure complication. There are already several RCTs ongoing or pending dedicated to the type of anaesthesia and we should wait for their results instead of massaging statistics in non-randomised registries or trials and giving unreliable conclusions.

In our centre, all thrombectomy procedures are done under GA with the following protocol:

  • The neurointerventionist and neuroanaesthesiologist are called by the neurologist when the patient is driven to the MR (for mother-ship patients) and when the patient is on the way to the hospital (for drip and ship patients). Both arrive in the angioroom before the patient.
  • GA is performed immediately after the patient arrival in the angioroom. During this time, the neurointerventionist prepares and flushes all the devices (8F sheet, 8F balloon and or intermediate catheters, microcatheter, stent-triever). If the GA takes too long for any reason, the groin puncture is performed in the same time. That means that as soon the patient is intubated (10 to 15 minutes), the neurointerventionist is ready to do the catheterisation.
  • As soon as the patient is intubated, the anaesthesiologists are focusing on blood pressure control.
  • The patient is extubated in the angioroom at the end of the procedure.
  • Our opinion is that if the team is used to practicing GA in routine:
  • GA does not delay thrombectomy procedure initiation.
  • GA induced hypotension can be easily controlled.
  • The thrombectomy procedure is much more easily done; it is also faster and safer.

It is only if future RCTs show superiority of CS that we will change our practice.


Christophe Cognard is head of the Department of Diagnostic and Therapeutic Neuroradiology, Purpan University Hospital, Toulouse, France


  1. Brinjikji W, Murad MH, Rabinstein AA, Cloft HJ, Lanzino G, Kallmes DF. Conscious Sedation versus General Anesthesia during Endovascular Acute Ischemic Stroke Treatment: A Systematic Review and Meta-Analysis. AJNR Am J Neuroradiol. 2015; 36: 525-9.
  2. Type of Anesthesia and Differences in Clinical Outcome After Intra-Arterial Treatment for Ischemic Stroke, Lucie A. van den Berg, Diederik L.H. Koelman, Olvert A. Berkhemer, Anouk D. Rozeman, Puck S.S. Fransen et al., Stroke, 2015; 46:1257-1262.