Climate change


The earth might be warming up and so might our patients in the ICU. It is easy to fall into the Fever ➝ CRP ➝ antibiotics trap, but our goal is to be more responsible.


  • Raised temperature: fever vs. hyperthermia
  • Central fever
  • Stroke and fever
  • Temperature management strategies for stroke
  • Infectious causes of fever
  • Noninfectious causes of fever
  • References

Raised temperature

might be fever or hyperthermia.

Fever is more common and defined by a raised hypothalamic setpoint, due to

  • infectious or
  • other inflammatory reasons, or due to
  • central stimulation

of the hypothalamus (“central fever”, e.g. blood in brain, see below). Note that CRP does not really distinguish between the three causes of fever, while procalcitonin might at least hint at an infectious etiology. Also there is no proper consensus as to what constitutes fever.

Non-fever hyperthermia is failure of heat regulation with intact setpoint, e.g. in exsiccosis. Typically, antipyretics are ineffective in pure hyperthermia.

Fever control in the ICU has been studied, if not extensively, and never been shown to be helpful. Most recently, Acetaminophen was not effectively in improving anything (Young NEJM 2015). It may be harmful, especially in septic patients (see Schulman 2005 and Lee 2012).

Central fever

(see this 2016 review on the subject) is always suspected in neuro patients, but hard to prove.

Pathophysiology, it is due to damage to the hypothalamus or contact of this structure to blood 0r pus (this can be reproduced in animals). In brain injury, diffuse axonal damage and frontal lesions indicate shear stress on the hypothalamus and correlate with central fever.

Clinically, central fever might have less diaphoresis and tachycardia, but this is not very specific. The diagnosis relies on exclusion of other infectious and inflammatory causes. The literature (Predicting central fever in NICU, Hocker 2013) says

The combination of negative cultures; absence of infiltrate on chest radiographs; diagnosis of subarachnoid hemorrhage, intraventricular hemorrhage, or tumor; and onset of fever within 72 hours of admission predicted central fever with a probability of .90.

Therapywise, central fever is harder to treat, so that physical measures and endovascular cooling are often employed. I grew up with the lytic cocktail (blocking every neurotransmitter you know), but there is no proper literature on that.

You should bear in mind that there is another central neurologic complication with fever that complicates severe brain injury, namely paroxysmal sympathetic hyperactivity – this is a chapter on its own.

Stroke and fever

  • Very common: (40-61%) in the first 2d of stroke have elevated temperatures, depending on the definition
  • Very bad: Raised temperature correlates with bad outcome, both in animal experiments (40° x 3h leads to 3 times the stroke volume) and patients (e.g., Greer 2008).
  • Early is worse: Stroke is more vulnerable to fever in the first 24 hours – it accelerates the conversion of penumbra to stroke and all bad pathophysiologic cascades (apoptosis, inflammatory).
  • Can be controlled: Fever control is feasible in principle and regarded as one of the active components of stroke unit care.

Temperature management strategies for stroke

  • Hypothermia: Bi 2011, de Georgia 2004, Ictus-L 2006, Ovesen 2013 show no benefit
    EuroHyp-I (ongoing), Ictus 2/3 (terminated; no results yet) – endovascular methods, no evidence (but very effective)
  • Prophylactic
    • antipyretics: PAIS (den Hertog 2009), PISA (Dippel 2003) – only modest effect on temperature, no benefit (NRO, survival)
    • antibiotics: EPIAS, PANTHERIS – reduced infection rates, no benefit (NRO, survival)
  • Fever treatment: blankets/air cooling probably not better than drugs, endovascular highly effective – no benefit shown
    QASC trial shows that the combination of controlling fever, dysphagia and glucose is beneficial

Infectious causes of fever

  • Head: meningitis/encephalitis, brain abscess, ventriculitis, sinusitis, dental, HEENT (including epiglottitis)
  • Circulatory: CVC, endocarditis/myocarditis, peripheral cannula, aortitis, mediastinitis
  • Respiratory: pneumonia/bronchitis, empyema, VAP
  • GI: esophagitis, pancreatitis, diverticulitis, rectal/anal abscess, C. diff
  • Urogenital: prostatitis, pyelonephritis, cystitis, PID
  • Hematological: malaria, HIV
  • Integument: Osteomyelitis, cellulitis, fasciitis, myositis

Noninfectious causes of fever

  • Vascular: stroke, IVH, ICH, SAH, MI, ischemic bowel, DVT
  • Idiopathic inflammatory: Gout, postoperative, acalculous cholecystitis, pancreatitis, aspiration pneumonitis, GI bleed, ARDS
  • Traumatic: Hematoma, Ulcers
  • Toxic
    • Drug fever: high spiking fever, chills, maybe leucocytosis, eosinophilia; drugs are beta-lactams, PHE, iv contrast
    • Malignant neuroleptic syndrome, Serotonin syndrome (beware: linezolid, MCP, setrons), malignant hyperthermia
  • Autoimmune: vasculitis, hemolysis, transplant rejection, transfusion
  • Psychiatric: Withdrawal
  • Neoplastic: renal cell CA, tumor lysis, lymphoma, leucemia
  • Endocrine: Ovulation, Thyroiditis, Thyreotoxicosis, adrenal insufficiency


Get a grip!

Not a HaNDL but a bavarian HeNDL

When a disease is discovered nowadays, it needs to be assigned a proper acronym – see CLIPPERS, MELAS, CADASIL. If we were living in the good ole eponymic days of, say, Steele-Richardson-Olszewski, the HaNDL syndrome would be named Swanson-Bartleson-Whisnant – the authors of a 1980s Neurology paper on the condition we covered in our rounds today. Or better: by Stigler’s law it should be named Berg-Williams-Syndrome after the 1995 Neurology article that coined the acronym HaNDL.

By all we know about the inflammatory pathophysiology of migraine, an increase in frequency and intensity of migraine auras should be able to produce a pleocytosis, but this is not the explanation for HaNDL, for it should enjoy all the epidemiologic characteristics of migraine then. And it doesn’t.

It is men of middle age (40-70, not young women) with only rarely (slightly more than chance predicts) any migraine in their history, who – sometimes after a preceding viral illness – suffer a series of episodes over days to weeks that resemble aura (except: less visual symptoms, more aphasia, more sensorimotor disturbances) in their development and migratory nature, but take way too long (a few hours rather than 10-60 minutes), accompanied or followed by headaches that resemble migraine with slightly less phobias (osmo-, kineto-, photo-, phonophobia). The usual workup (CT, MRI, labs) is negative, but there is a lymphocytotic pleocytosis of > 15/µl and often a raised CSF opening pressure.

Of course, as in any case of focal symptoms plus pleocytosis (aka encephalitis), we send off the standard microbiology tests (HIV, TPHA, borreliosis, PCR for HSV, VZV), start acyclovir and maybe ceftriaxone at once and wait until everything comes back negative. Then we are left with the hopes of a spontaneously resolving syndrome – by definition it should take weeks or months to clear.

In my experience, the usual migraine therapy (iv high dose NSAIDs = metamizol or ASS, plus antiemetics – MCP or dimenhydrinate) covers each episode but does not prevent the next. Steroids might, but don’t last long enough. I usually treat with spreading depression drugs (valproate, topiramate) and hope for the best.

The disease certainly is underrecognized, being replaced by “a minor encephalitis”. There should be some studies of  autoimmune mechanisms and antibodies, but I know of no proper results yet. So, if you are faced with your next HaNDL why not send of an experimental panel?




First described in 1883, the concept of watershed strokes was further developed pathophysiologically in the 50s and 60s. (There is a proud  article in Stroke this week, discussing the  history of the concept.)


Several terms are used in the literature: border-zone stroke, watershed stroke, misery perfusion, Letzte Wiese. Watershed is probably the best term, as it describes the idea that the land most distal to two supplying rivers suffers from even slight variations in flow in either. I personally apply the name Letzte Wiese only in specific cases – it captures an area  insufficiently supplied by just one vessel and thus can be applied only to internal border zone strokes.


For watershed strokes you need either quite severe hypoperfusion (as in hemorrhagic shock) or only minor hypoperfusion but at least one severely stenosed vessel.

  • The most frequent stenosis affects the proximal ICA, leading to the ACA/MCA and PCA/ACA border zones, as well as (sometimes) strokes in the MCA internal border zone.
  • High grade MCA stenosis can lead to watershed strokes in the internal border zone along and above the lateral ventricles (rosary pattern). In this case the region between the supplied region of the deep penetrating endarteries (mainly basal ganglia) and the small branches of the MCA main branches (which enter the brain from the cortex down) suffers, which amounts to the white matter in the centrum semiovale.
  • Finally the cerebellum knows watershed zones between the 3 feeding vessels, but this is not of practical relevance because treatment is similar to embolic strokes.

ICA stenosis

  • In chronic near occlusion of the ICA, the borderzone region can move if you give it enough time, with the PCA/MCA region moving forward and the ACA/MCA region backward.
  • To complicate matters further, the borderzones are variable depending on the localization of the stenosis and the integrity of the ECA collateral as well as the Circle of Willis (e.g., if A1 or the anterior communicating artery is hypoplastic or the PCA has the fetal variant).
  • In my experience you need at least 2 patent collaterals (out of ECA, ACA, PCA) to ensure hemodynamic stability even in near occlusion.
  • In the last years we learnt more about degenerative distal ICA stenosis just below the carotid T which is not well collateralized via the ECA pathway – this region unfortunately is not well imaged with CTA due to calcification-related artifacts, being  better amenable to duplex ultrasound and MRA.
  • Similar problems can arise with dissections extending to the intracranial sections of the ICA.
  • In very slowly progressive combined distal ICA and MCA stenosis (degenerative, vasculitis, Moya Moya disease) a web of tiny collaterals can form (Moya Moya picture), which has its own intricate pathophysiology.

Pathophysiology of hypoperfusion

Local hypoperfusion can be graded as follows:

  • Grade I: reduced CBF, enhanced CBV, functionally intact or only slightly compromised, near normal oxygen extraction, reduced vasodilatatory reserve
  • Grade II: severely reduced CBF, reduced CBV, reduced function, no or negative local vasodilatatory reserve (the latter is called the reverse Robin Hood phenomenon)

Practical management

  • When faced with an imaging pattern of watershed strokes, duplex ultrasound and angiography (CTA, MRA or even conventional angiography) are urgent to get as much information as possible about the flow patterns and the collateral situation.
  • Perfusion imaging: Severe proximal stenoses leads to difficulties in interpreting perfusion imaging, but this can be used to try to differentiate between grade I and II hemodynamic compromise.
  • To judge how imminent the danger is, functional ultrasound of the MCA (using CO2, apnea or – easier – acetazolamide 1g) is used.
  • If  in doubt, I recommend a trial of therapeutic hypertension (usually with 25-50-100µg Noradrenaline, only with intact coronaries!) to see whether the neurologic deficit fluctuates with blood pressure.
  • In this case, an emergency revascularization is necessary and without alternative. To bridge the time to surgery or stenting, you can use continuous therapeutic hypertension, aiming for an RR of > 180 or 200 mmHg.
  • Sensitivity to blood pressure drops: Quite often the reverse happens – someone accidentally treats an impressively high RR of 220 mmHg (which might actually be due to the brains own reaction to hypoperfusion) with an iv bolus of labetalol or urapidil and the patient deteriorates. If this happens, quickly counteract your medication (Noradrenaline again) and due your vascular studies.
  • Remember that patients with severe misery perfusion, the danger of hyperperfusion syndrome after revascularization is quite real and this is difficult to treat.


The thalamus of secrets

thalmusThe thalamus is a  tightly packed collection of nuclei and fibres in the center of the brain that is involved in everything sensory (except olfactory) and extrapyramidally expressive. It is connected to everyone and his mother. Any proper neuropsychological problem can be caused by thalamic lesions – neglect, aphasia, dementia, delirium, visual, sensory, motor disturbances, ataxia, pain. Thus as the caudate nucleus, the thalamus is always a good answer if you are asked for localization of your lesion. It helps immensely to organize the connections and deficits into a simple system in order to understand thalamic stroke deficits. I follow the literature on thalamic strokes and distinguish 4 vascular regions.

The four arteries

Bildschirmfoto 2016-03-07 um 07.15.12

Polar artery region (anterior thalamoperforating artery). The anterior nucleus is involved in cognition, episodic memory, language and emotion with connections to the limbic system (the corpora mamillaria, the hippocampus, the cingulum and all else) and the frontal lobes (see this cool article). Together with parts of the lateral nuclei the region is supplied by the A. thalamoperforans anterior (also called polar or tuberothalamic artery) which arises from the mid of the posterior communicating artery, although in about a third of people the paramedian artery replaces it. Strokes in this region are etiologically diverse (as e.g. in the anterior choroidal artery) and lead to thalamic aphasia as well as diverse neuropsychological deficits that resemble caudate strokes, such as change in personality, abulia, apathy and – via its mamillary connections – memory deficits.

Bildschirmfoto 2016-03-07 um 07.51.10

Posterior choroidal artery region. This artery arises from  (P1 or) P2 and supplies the pulvinar and the geniculate bodies, all of which connect to the auditory (temporal lobe) and visual system (occipital lobe). Resulting deficits are visual field defects (hemianopia, wedge shaped), hearing deficits and more rarely aphasia and other neuropsychological deficits. Since the artery also sometimes reaches the posterior and even lateral ventral nuclei, its occlusion can also lead to hemihypesthesia.

Bildschirmfoto 2016-03-07 um 07.50.38

Paramedian artery/arteries. Does the name Percheron ring a bell? He is the one discussing the various anomalies in the paramedian (or posterior thalamoperforating artery): it usually arises from P1 but often one side misses and this side  is supplied from the contralateral one, so that a thrombus in one P1 can lead to inferomedial strokes in both thalami (artery of Percheron). As one of the first arteries after the basilar head, basilar artery embolism often leads to paramedian artery  and thus to inferomedial thalamic strokes. The most prominent version is Caplan’s top-of-the-basilar-syndrome (a short aside: have you ever read up on the embryology of the vertebrobasilar system? if not – look here).

The artery reaches the medial nuclei (which somehow interact with all other thalamic nuclei) as well as intralaminary nuclei (these lie inside the internal medullary lamina that crosses the thalamus longitudinally), also parts of the pulvinar and sometimes the ventral lateral nuclei. More importantly, a branch of the paramedian artery goes off to the mesencephalon and pons and leads to disturbances in vertical ocular motor control (damaging the rostral = interstitial nucleus of the MLF aka riMLF). In principle a stroke in the paramedian artery can affect the functions of all other thalamic nuclei, but mainly it causes coma or somnolence, vertical gaze paralysis or skew deviation, and neuropschological deficits (“thalamic dementia”).

The thalamogeniculate artery arises from P2 and irrigates the ventral nuclei, including the anterior, lateral, intermediate, posterolateral, posteromedial nuclei. Functionally this implies sensory deficits (hemihypesthesia or hemianalgesia, but also wedge shaped sensory disturbance) via the vental posterolateral (body below face) and ventral posteromedial (trigeminal) nuclei. Interestingly the sensory disturbance in thalamic strokes can be limited exactly in the median (as a counter example for functional deficits) and tends to give rise to intractable pain syndromes (thalamic pain syndrome) later. Through deafferentation the hand moves in a weird dystonic way (with flexion of the wrist and the MCPs, adduction of the thumb and athetosis of the fingers) when the patient is asked to hold up his hands with eyes closed (thalamic hand). Through the ventral anterior and lateral nuclei as well as the subthalamic nucleus connections to the basal ganglia, the premotor cortex and the cerebellum can be harmed – this leads to atactic hemiparesis and hyper- or hypokinetic movement disorders. As the name of the artery suggests, it can also reach the genu and the posterior limb of the capsulae internae, causing the lacunar syndrome sensorimotor hemiparesis. Strokes in this region are – as those in the capsule – mainly (70%) microangiopathic.

Powell’s cross

The following scheme has been adapted from Powell’s wonderfully concise article:

Bildschirmfoto 2016-03-07 um 07.18.43Bildschirmfoto 2016-03-07 um 07.14.55

Specific thalamic syndromes

Thalamic dementia. Although it occurs frequently after paramedian artery strokes, it can also happen in anterior (through the connection to the Hippocampus and Corpora mamillaria) or even medial thalamic lesions. The neuropsychological profile is different from Alzheimer’s and akin to caudate dementia with reduced initiative, recall, short term memory (resembling Korsakoff’s), spontaneity, vigilance. Patients seem indifferent more than incapable of answering. Social intelligence may be affected more severely than formal intelligence.

Thalamic aphasia may be quite diffferent from cortical aphasia; similar to the dementia syndrome patients seem to be less interested in producing speech rather than severely handicapped. Word finding difficulties (as in any aphasia), reduced spontaneous speech, perseverations dominate, with grossly intact grammar and often preserved reading and writing capabilities as well as repetition.

Thalamic sensory  deficits. From the full blown syndrome of Déjerine-Roussy with complete hemianesthesia giving way to hypersensitivity, paresthesias and a thalamic pain syndrome to more restricted variants of hypesthesia – dissociated, face or body only, pain only, the lateral thalamic strokes can lead to severely disabling pain syndromes. Sensory deficits can be associated with (usually temporary) hemiparesis (sensorimotor stroke) or atactic hemiparesis.

Thalamic hand. This is more or less a deafferentation syndrome. Without proper feedback on the position of the fingers and hand (eyes closed), the wrist and fingers assume a dystonic posture with flexed wrist, flexed MCPs, straight PIPs and DIPS and hyperadducted thumb. When holding out the hand, deafferentation athetosis may occur, the finger wandering up and down.



Imploding head syndrome

What are the contraindications to lumbar puncture? The easy ones are coagulation disorders, infection around the LP site and spinal cord trauma – although to be honest nothing much can be found about the threshold of the three Ps (PT, PTT, platelets) you need for doing an LP; obviously no proper studies have been performed, so there is a project for all those budding research Neurologists or Anesthesists out there; can you do an LP on a patient with Tirofiban? How long after tPA?

Herniation does occur in meningitis

Of course, the most important question is: how much do we have to make sure that herniation will not occur? This has only been studied in bacterial meningits, where LP is felt to be that urgent that the CT cannot be waited for (as an aside: in Germany, our guidelines solved the problem by just requiring ER docs to infuse Ceftriaxone AFTER blood cultures but BEFORE LP) – again without proper evidence base. But meningitis is the most important case, because this is the only disease, where herniation does occur at least rarely (as opposed to never – cf. all other diseases) – the literature estimates about 5-10% of patients. It also can happen without LP.

Intracerebral pressure is not the problem

The first thing to understand is that herniation is usually attributed much more to brain shift than to pressure. Pressure can be very high (see pseudotumor) and still no coning after LP. In fact, most bacterial meningitides do have quite high pressures (30 cm H2O upward). Since papilledema (if at all present) really only signifies raised ICP, a lack of papilledema does not help at all.

Brain shift is the problem

The crux lies in pressure differences, either from left to right (along the falx, from temporal to brainstem), cephalad to caudad or vice versa. These pressure differences stem from space occupying lesions (e.g. abscesses), generalized but locally different edema or even from slight differences in local adhesiveness due to inflamed meninges, leading to reduced mobility of brain tissue. Once the pressure difference is accentuated by LP, brain shift can occur and this then leads to herniation. As an intensive care or emergency neurologist you ought to know your herniation syndromes, if not, you can look them up in any Neurointensive Care Textbook.

Acute brain shifted patients do have neurology

If an acute process such as meningitis or an abscess leads to brain shift, you can expect neurological symptoms. This has been established reasonably well in children and adults (see the NEJM 2001 paper). Here focal neurological symptoms means

  • Severe disorder of consciousness
  • Seizures
  • Pupillary disorders
  • focal deficits

This rule only applies in acute patients and not if your pathology develops slowly (tumors, chronic infections, in particular in immunocompromised patients)!

Normal CT, no neurology – still can herniate

Now this is the downer: not all patients that are going to herniate do have evidence of brain shift in their CT. The Rennick study had 5 of their 14 herniations in children with normal CT. The reasons could be

  • insufficient sensitivity of CT for brain shift (say in comparison to MRI)
  • other pathological processes that lead to herniation without mass effect (see above)

Be prepared for herniation in all cases

Since we cannot rule out herniation even by CT we have to ensure close neurological monitoring in bacterial meningitis after  or even without LP. It is the one reason why meningitis patients do die even nowadays. If herniation does occur – be quick to react: mannitol, tube, CT (is it hydrocephalus, herniation or something else?), OR.

What are CT signs of impending herniation?

  • Evidence of unequal pressure across the falx cerebri (midline shift, one-sided ventricular dilatation, one-sided effaced sulci)
  • Evidence of unequal pressures between supratentorial and infratentorial compartments (loss of suprachiasmatic, pentagon cisterns, cisterna ambiens)
  • Evidence of transforaminal herniation (low standing tonsils, loss of perimedullary space)
  • Evidence of non-communicating hydrocephalus, aqueduct blockade
  • Generalized cerebral edema (effaced sulci anywhere, loss of all cisterns)

What to do if signs of mass effect?

  • Cultivate everything else (blood, sputum, …)
  • In abscess patients try to find the focus or try to get at the content of the abscess surgically, CSF is of little value in this case anyway.
  • Use calculated antibiotics
  • If you really suspect strange bugs (immunocompromised or patient has been to strange places) – get ventricular CSF surgically



So to wrap this all up:

You can forgo CT in neurologically intact acute patients, but they can still herniate



CT versus LP – dawn of the thunderclap headache

As for dogmas to be lysed, no falsity is more resilient than the one that every thunderclap headache (TCH) needs to be tapped to rule out subarachnoid hemorrhage.

Walking through the numbers, our 90,000 pt emergency room probably treats 2,000 headache patients yearly, with 5-10% being thunderclaps, i.e. about 100-200 patients – let’s assume 200. We know from old (Landtbloom 2002) and new (Sayers 2015) series that about 7-15%% of these will be real subarachnoids or about 30 bleeds a year in the worst case. If we tap all of them, we will necessarily induce 10-20% = 40 artificial bleeds, which – according to the best literature we have – is at least hard if not impossible to differentiate from real bleeds:

  •  the red blood cell (RBC) count need not decrease in artificial bleeds
  •  but it can in SAH
  •  Xanthochromia, Ferritin, erythro- and siderophages need time to develop
  • there is no proper threshold for RBCs.

If you look at Sayers 2015 series of LPs in TCH, about 15% were uninterpretable and 13% were inconclusive, while 4% were clearly positive. This amounts to 32% LPs that lead to further workup or 64 patients a year for us. Of these 64 patients between 2% (if you use CTA) or up to 5% (if using DSA) harbour incidental harmless aneurysms, so about 2 patients will have one, which then is coiled or operated on.

On the other hand, at most 1 patient of our 40 real subarachnoids yearly will be CT negative (regardless of the time – see Sayers or Dubosh 2016, but at least in the first 6-24 hours). Unfortunately our LP workup for SAH in Germany (visual inspection for xanthochromia) only has about a third sensitivity. We don’t really know how many SAH do have enough red blood cells in their lumbal CSF and what a proper threshold might be. So it turns out, that we might actually miss the one CT negative patient after all.

In all this discussion I did not even consider other adverse events of lumbar puncture.

Here, then, is my take on the issue:

  •  TCH is way more complex than just ruling out SAHs. Think about the differential (see below) and do a proper history, physical exam and ensure follow-up, if possible with MRI (think of RCVS, SVT and other stuff). A rate of two thirds etiologically unresolved TCHs („primary“) is too much.
  • Always keep in mind that Hunt and Hess I-II may actually have a better prognosis than H&H III-V.
  • Discuss the probabilities with your department and your patient and be reasonable: LP is not the goldstandard test for subarachnoid!
  • If you do an LP to rule out SAH, leave at least 12 hours time (from onset) for xanthrochromia to develop and yes: all this time the patient is in danger, but the probability that he has an SAH is extremely low.


ICU fever

Differential for (new, resistant) fever in the acute care ward aka ICU.

  • Wrong site:
    • The obvious five: lung, lines, abdomen, skin, urine
    • The hidden five: endocarditis, meningitis, translocation, sinusitis, abscess
    • The weird fevers: drug, central
  • Wrong bug – viruses, fungi
  • Wrong antibiotic – wrong spectrum, MDR pathogens