Ten false beliefs in neurocritical care

Only neurointensivists should care about the brain.

In acute brain injury, the need for specific expertise on central nervous pathophysiology is evident. However, even when the primary reason for ICU admission is extracranial, the brain may be affected too, through inadequate substrate and oxygen delivery, blood–brain barrier leak, harmful effects of sedatives, and excitotoxicity. The resulting spectrum of brain dysfunction includes delirium, encephalopathy, coma, and non-convulsive seizures. Therefore, all intensive care should integrate neurointensive care, with the primary goal to preserve the brain [1].

Clinical examination of neurocritically ill patients is impossible.

The patient’s clinical state is our most important neuromonitor. Clinical assessment of consciousness, cognition, brainstem, and motor function should be attempted at least upon admission and daily [2]. Sedatives confound neurological examinations, and should be used sparingly in severe brain injuries, except for specific indications, such as intracranial pressure (ICP) control, seizure treatment, or targeted temperature management (TTM). They should be titrated and stopped if no longer indicated.

We should no longer monitor ICP in traumatic brain injury (TBI).

The ICP monitor has been accused of increasing therapeutic intensity, potentially harming patients without improving their outcomes. The Best-TRIP trial [3] is often wrongfully interpreted as evidence against ICP monitoring—a view that inappropriately conflates monitoring use and therapy titration. The controversy is not whether to monitor or treat ICP, but how this signal should be interpreted and responded to.

The threshold to treat ICP is 20 or 22 mmHg.

Changing ICP treatment thresholds from 20 to 22 mmHg in influential guidelines [4] implies that an ICP of 21 mmHg is fine, whereas an ICP of 23 mmHg should be treated aggressively. The absurdity of such a strategy is obvious, because it neglects measurement errors and ignores modern concepts such as the ICP intensity–time burden [5]. Interventional studies where aggressive treatments, decompressive craniectomy [6] or hypothermia [7], were applied early after crossing the 20 mmHg threshold showed harm rather than benefit. A tiered approach is rational, and aggressive measures should probably be reserved for sustained ICP elevations above 25–30 mmHg unresponsive to lower-tier therapy.

Ketamine increases the ICP.

Ketamine-induced ICP elevations were reported in small studies in non-ventilated (and not acutely brain-injured) patients [8]. In fact, ketamine, as an adjunct in sedated mechanically ventilated patients, might decrease the ICP [9, 10], produce neuroprotection through NMDA-receptor antagonism [11], suppress harmful cortical spreading depolarisations [12], and control refractory seizures [13].

Subarachnoid haemorrhage (SAH) patients should get ‘triple H’ therapy.

Aggressive fluid loading (with consequent haemodilution), coupled with vasopressor administration to increase arterial blood pressure (ABP), was used in the past in the hope of preventing delayed cerebral ischemia (DCI) and vasospasm. This strategy is discredited, and might be deleterious in this population, where the median age is 50–60 years, and cardiopulmonary complications are common. Current recommendations [14] advise against haemodilution, and to aim for normovolemia. Clinical DCI should prompt a stepwise trial of ABP augmentation, titrated to neurological assessment. Where diagnosis or response to therapy are uncertain, additional investigations can help confirm or refute the diagnosis of DCI—not all late neurological deterioration in SAH is due to vasospasm.

There is no need to control the temperature after cardiac arrest (CA).

The evidence for TTM at 32–34 °C after out-of-hospital CA is less robust than initially assumed [15]. No difference in outcome was found between TTM at 33 or 36 °C [16]. However, these results do not justify neglecting temperature control after CA. Targeting 32–34 °C may still be defensible because of equipoise between targets, but many sites involved in the TTM trial have adopted the 36 °C target [17], since it avoids potential adverse events of more aggressive cooling.

Hypoglycaemia is harmful for the brain, hyperglycaemia is not.

The optimal glycaemic target for the injured brain is controversial. Tight blood glucose control to the normal fasting range (TGC) increases the risk of (deep) hypoglycaemia, especially in inexperienced hands. In small observational studies, the (lower) normoglycaemic range has been associated with low cerebral microdialysis glucose values [18], raising concerns about substrate delivery. On the other hand, hyperglycaemia is an independent predictor of poor neurological outcome and death after CA [19]. According to a recent meta-analysis, the use of TGC showed a small but significant reduction in the risk of poor neurological outcomes in TBI [20]. So, even when the discussion on the optimal glycaemic target is far from resolved, it is important not to ignore that both hypo-and hyperglycaemia are associated with worse clinical outcomes in neurocritically ill patients.

In acute ischemic stroke (AIS), revascularization should be done within 3 h of symptom onset.

The conventional time window for thrombolysis in AIS is 3 h, extendable to 4.5 h in patients ≤ 80 years of age, without a history of both diabetes mellitus and prior stroke combined, a National Institutes of Health Stroke Scale (NIHSS) score of ≤ 25, not taking any oral anticoagulants, and without imaging evidence of ischemic injury involving more than 1/3 of the middle cerebral artery territory [21]. When mechanical thrombectomy is considered, the recommended timeframe is 6 h post-ictus, but specific penumbra-like conditions on perfusion imaging allow for longer time windows up to 16 h [22]. General anaesthesia during thrombectomy should be avoided [23].

Blood pressure control in intracerebral haemorrhage (ICH): contradictory trials.

Interpreting recent trials [24‐26] on the treatment of hypertension after ICH is complicated by differences in inclusion criteria, intervention timing, outcomes, antihypertensive drugs, and systolic blood pressure (SBP) targets. All these studies were conducted in patients with relatively small ICH volumes, with varying latency to achieving target SBP (ranging from 4.5 to 24 h). Early intensive SBP control to targets above 140 mmHg reduces haematoma expansion, but does not improve neurological outcome or mortality. More aggressive SBP reduction to 110–140 mmHg in the ATACH-2 study [26] found no incremental benefit as compared to the 140–180 mmHg range, but a higher rate of renal complications. In brief, SBP control in ICH may reduce haematoma expansion, but not below 140 mmHg (Table 1).