Acute status asthmaticus is a fragile thing which, most ICU specialists would agree, is a good reason to get out of bed and drive to the hospital. These people tend to die rather readily, typically of cardiac arrest due to dynamic hyperinflation and the resulting loss of preload. This is made worse by the fact that the excellent outpatient management of asthma has filtered out the mild and moderate patient groups, and left only the incredibly brittle severe asthmatics. Only these are the people who ever get admitted to ICU for mechanical ventilation. Thankfully, even the ICU management of these people is improving, and in spite of an overall worsening severity of their illness, their ICU mortality has been steadily improving.
Severe asthma is very popular with the CICM examiners. Past paper SAQs on this topic include the following:
In brief summary, the management of severe asthma can be compressed into a short grey box:
- Use the largest tube possible.
- FiO2 : lowest required to achieve SpO2 of 90-92%
- Tidal volume: small, protective 5-7ml/kg
- Respiratory rate: slow, 10-12 breaths per minute (or even less!)
- Use a long expiratory time, with I:E ratio 1:3 or 1:4
- Use a volume-controlled mode, or any other mode with a square flow waveform (i.e. constant flow) - this decreases the peak airway pressure
- Reset the pressure limits (i.e. ignore high peak airway pressures). .
- Use heavy sedation.
- Use neuromuscular blockade.
- Minimise the duration of neuromuscular blockade.
- Use minimal PEEP when the patient is paralysed, and titrate PEEP to work of triggering once the patient is breathing spontaneously.
- Keep the Pplat below 25cmH2o to prevent dynamic hyperinflation.
First-tier therapies with strong supporting evidence
- Humidified oxygen titrated to SpO2 90-92%
- Nebulised beta-agonist bronchodilators
- Nebulised anticholinergic drugs
- Steroids: IV hydrocortisone or oral prednisone
Second-tier therapies with weak supporting evidence
- Intravenous beta-agonist bronchodilators for refractory bronchospasm
- Nebulised adrenaline
- Magnesium sulfate
- Helium-oxygen mixture
Third-tier therapies without any supporting evidence
- Volatile anaesthetics
- ECMO in asthma
There are several excellent resources for this topic, though there does not appear to be any grand unifying resource to combine all the information. Oddo et al (2006) is a good introduction into the ventilator management of this condition, which has not changed much since 2006
There is a fairly well-defined set of features which qualifies asthma as "severe".
Pulmonary function test data:
How do you get yourself admitted to the ICU as a severe asthmatic?
Well. There are several physiological behavioural and environmental features which seem common.
...But this list is not exclusive. Its not as if you would keep a severe asthmatic puffing away in the emergency department, waiting for these features to develop.
Auto-PEEP, the trapping of pressurised gas inside the chest, is repsonsible for most of the problems in acute severe asthma.
Hyperoxia, it turns out, is harmful in asthmatics - for reasons completely divorced from CO2 retention.
Consider the hyperinflated alveolus. It, sitting there full of pressurised gas, is slowly filling up with CO2, and emptying of oxygen. Its ventilation is terrible, and the normal mechanism of hypoxic vasoconstriction has intelligently diverted blood flow from this useless cavity, to collect oxygen from better ventilated regions.
Now, see what happens when you blow extra oxygen into this intermittently ventilated alveolus. Suddenly, the burst of extra oxygen reverses the hypoxic vasoconstriction, and blood is diverted away from well-ventilated regions. The well-ventilated regions will still participate in gas exchange, to be sure - but now, CO2 clearance will be impaired, because the CO2-containing blood is being diverted to regions from which expoiratory gas escape is impossible.
On top of that, there is probably some contribution from the Haldane effect, which describes the affinity of hemoglobin for CO2. In brief, hyperoxic environments degrade this carrying capacity, decreasing the exchange of CO2 at the alveolus.
Is there any evidence for the validity of these fancy theories? Yes there is. Perrin et al published a trial in Thorax which compared liberal untargeted oxygen therapy with conservative targets (for 93-95% saturation). Turns out, the liberally oxygenated patients became hypercapnic.
This phenomenon is analogous to the need for careful O2 control in patients with COPD, and elsewhere in this site there is an exploration of the empirical evidence for the changes in CO2 with excessive oxygen administration. Though data of this sort is not available for asthma, one could surmise that it would be the same sort of relationship, if not worse.
Even though probably less than 10% of the nebulised drug reaches the lung, salbutamol nebs are the mainstay of treatment. Whether to nebulise the drugs with a continuous flow of oxygen, or to administer via spacer - there seems to be no real difference in effectiveness.
It is generally held that IV bronchodilators should be started when it is clear the the patient is not responding to the nebulised administration, and the reason they are held in reserve is the perception that they tend to have a massive excess of adverse effects, for comparatively little clinical benefit.
Lactic acidosis and hypokalemia reward the overvigorous salbutamol user. The general consensus is that these should be avoided unless absolutely necessary. There is some data that (at least in children) it may have some benefit in terms of earlier discharge from the emergency department, but no such data in adults exists. Tobin (2005) gives a perspective of his time, in case the reader wants to have a look at what their fathers thought about this strategy.
What would make them "absolutely necessary"? Well, for one, in order for nebulised bronchodilators to be effective, they would need to have access to the bronchi. A patient in a state of near-total respiratory arrest is unlikely to be moving enough air to get any mist to its site of action. In these people, IV salbutamol may be the right choice. In effect, any nebs administered to them will only be having effect by absorbing directly into the bloodstream via the oral mucosa anyway. One can identify such patients from their ABG results; continuous nebs which are well-absorbed via the bronchi and alveoli will have a substantial systemic effect and will cause a rise in the lactate, whereas pointless nebs will not.
A contrarian would, at this stage, pipe up with the complaint that continuous nebs, landing on the face and oropharyngeal mucosa, will often still do the bronchodilator thing by systemic absorption, which for salbutamol is surprisingly good. When Salmeron et al (1994) compared the blood levels of patients receiving nebulised salbutamol, they found them to be as high as those in their IV infusion group, suggesting the IV route was completely unnecessary.
But let's say you have decided to give IV salbutamol anyway. In general we tend not to give doses higher than 10μcg/min (5-10 µg/min is the usual range, which is about 5-10ml/hr of a 6mg/100ml infusion). One important aspect of managing such patients would be not to overreact to their hypokalemia, which is mostly the result of intracellular redistribution. The potassium level should be maintained by supplementation, as hypokalemia is bad for muscle (cardiac and other), but the target level should be down-adjusted. The reason for this is the post-salbutamol redistribution of potassium back into the circulation, potentially giving rise to lifethreatening rebound hyperkalemia within a couple of hours of the cessation of the infusion.
Ipratropium nowadays gets used for all acute asthmatics, because of its suspected benefit, and absence of harm. The anticholinergics seem to be useless on their own, but potent in combination with beta-agonists.
From internship, one recalls the tendency for these asthmatic people to be put on 100mg qid of hydrocortisone, which equates to about 100mg of oral prednisolone. There are many benfits:
However, it seems oral agents are at least as effective as intravenous agents, and that there is no benefit in giving truly massive doses (eg. over 800mg/day of hydrocortisone).
Aminiphylline is the only one available in Australia. Its use tends to be reserved for patients refractory to the first tier of therapies.
In general, the current opinion of aminophylline is that it ...
There is also an entire population of patients in whom it should never be even considered:
Additonally, aminophylline interacts with a whole host of drugs, making its use a fiddly and complicated affair. It can even decrease the sedating effects of propofol.
Not a lot of aminophylline experience exists in the modern era. Most of the time, when one comes across aminophylline in the literature, some variation of the phrase "still used in spite of". Aralihond et al (2020) describes mostly paediatric experience, which is likely to be also representative of the adult experience. In short, the risks and disadvantages appear to be overstated in the critical care folklore, perhaps because everyone is aware of all the possible side effects and takes active steps to mitigate them.
So, say you decided to use this horrible stuff anyway. The British Thoracic Society recommend a loading dose of 5mg/kg, or about 350-400mg (one and a half of the 250mg-ampules), followed
by an infusion of something like 20-25mg/min. Brave souls have even used 10mg/kg as a loading dose. What would you expect, in terms of positive effects? It appears that some authors have found a trend towards improved airflow when using this drug in combination with other bronchodilators. It may have some kind of synergistic effect with β-agonists, for example.
Some might say that the vasoconstricting effects of adrenaline are beneficial in asthma because they address the inflamed vascular engorgement of the bronchial lining. It may also have some kind of antihistamine effect (Baldwin et al, 1994). In early studies, it had compared favourably. However, well-designed trials which followed had demonstrated that it is no different to salbutamol in efficacy, and that salbutamol has fewer adverse events.
IV or nebulised adrenaline is one of the things you try when you are truly stuck, and have nothing else to add to the management.
The use of magnesium sulfate in asthma is controversial; generally speaking everybody knows that "thats what you give for severe asthma" but when one examines the evidence, there is actually nothing to support its routine use in anything but the most desperate situations. The MAGNETIC trial has demonstrated that nebulised magneisum sulfate may actually have a little bit of benefit in children with severe asthma, and a 2013 meta-analysis has confirmed that IV magnesium has a statistically significant influence on resiratory function among adults.
Oh's Manual only gives it a lukewarm recommendation. Costello & Howes (2004) could not find support for this in their literature review, and
ketamine is generally though to be a bronchodilator by a variety of mechanisms, the chief of which is probably sympathomimetic. This all sounds very nice, but the evidence for its use in adults is not very strong, and a recent meta-analysis has demonstrated that it has no benefit in acutely asthmatic children. In its support, we only have case series of questionable quality.
That said, I think nobody should die of asthma without having had a trial of ketamine.
Heliox, the trade name of the 70% helium - 30% oxygen mixture, is another measure one may consider when faced with a comletely impenetrable bronchospasm. This therapy, in contrast to such baddies as adrenaline and aminophylline, remains on the menu purely because it has so few adverse effects (and who cares about the lack of benefit, right?)
The theory behind it is well summarised in the 2006 paper by Wigmore and Stachowski. Characteristically of these authors, interesting details of physics are offered, including a table of the physical properties of gases at room temperature and pressure. One immediately notices that the density of helium is vastly lower than the density of oxygen, or all the other atmospheric gases for that matter. This leads to an improvement in the flow of gas through narrow cylinders; laminar flow occurs in tight spaces where turbulence would otherwise appen.
Thus, the benefit of this therapy lies in the decreased effort of breathing. This seems to have some benefit in the setting of severe asthma (but, as expected, the mild and moderate asthmatics get nothing out of it). Ther are really very few risks - helium is very well tolerated. The major risk is hypoxia, because the ideal mixture of gases has only 30% FiO2, and not every asthma patient can tolerate this.
Like ketamine, these are only described in small case series. There is an inherent difficulty in fitting a volatile vapouriser to the ICu ventilator. That said, it seems they are potent bronchodilators, and can make all the difference in an intubated severely bronchospastic patient. In essence, this therapy should be brought into the equation as a prelude to ECMO.
NIV has a number of theoretical benefits in asthma:
NIV also has a number of theoretical disadvantages.
The evidence for the use of NIV in asthma is not as strong as the evidence for its use in COPD, but among the experts the general opinion is that NIV in asthma is useful. So what if there haven't been enough trials to confirm the improvement in mortality?
However, the evidence has significant limitations:
Indications for NIV in the acute asthmatic are thought to be:
Full face masks are usually best.
No "routine" pressure settings can be recommended; it is best to stand at the bedside and adjust the pressures until the patient feels their breathing is getting easier. Oh's Manual suggests initial settings of 5cmH2O EPAP, and 8-10cmH2O of pressure support on top of this (i.e. an IPAP of 13-15 cmH2O).
If n=206 was depressing, then you're going to love this area. It seems we have only this conference abstract to work with (Samaria et al, 2011). In this n=54 prospective cohort, NIV failure was defined as the need for intubation. The following features identified patients who would go on to be intubated after a trial of NIV (as defined by an hour of continuous NIV)
A gradual course of onset makes you think the patient will fail NIV. If the asthma attack has been going for days, the patient arrives to the ED when they are already exhausted.
Another way of asking the question about the predictors of NIV failure is to ask which criteria one would use to intubate the asthmatic patient. This is not a consequence-free decision. It is possible to cause worsening bronchospasm, and to inadvertantly cause pneumothorax or worsening hyperinflation.
The following criteria are "consensus indications for intubation" listed in an article by Brenner et al (2009 - see Table 2).
There are several principles to abide by:
The end result of all these measures is a ventilator monitor which looks something like this:
Observe the waveforms. Classically, when one it trying to detect for dynamic hyperinflation, one is taught to lok for it in the flow waveform. However, expiratory flow is very poor, and the flow rate of the gas escaping from the patient is probably somewhere well below 10L/min. As you can see, the scale of the flow graph in the image above is from -200 to +200 L/min. At this scale, the miniscule expiratory flow becomes impossible to detect. If the other waveforms were not available, one would not immediately shout "bronchospasm!" at this flow graphic. Thus, it is much better to use the volume over time waveform. There is clearly an expiratoy flow limitation here; this modest 500ml tidal volume is taking about 4 seconds to escape. In fact, from this one can determine that the mean expiratory flow rate must be something like 7.5L/min. Anyway, in summary: look to the volume/time waveforms for evidence of dynamic hyperinflation.
Speaking of volumes and times; how does one safely select a tidal volume for a severe asthmatic? One needs to ask themselves two questions.
The first is a trick question. Nobody needs a large volume. 6-8ml/kg is going to be enough for the vast majority of the ICU population. As to the second question, the best answer is a diagram, representing a volume over time curve in somebody with severe airflow limitation:
So, in the form of words:
In summary, use lung-protective tidal volumes.
Though various highly reputable sources (Stather & Stewart, 2005; Leatherman et al, 2015; Laher et al, 2018) might recommend "to shorten inspiratory time as much as possible", there is obviously a limit to this recommendation. For instance, it does not mean that in all status asthmaticus scenarios one should reduce the inspiratory time to the shortest possible duration permitted by the ventilator. Thinking along those lines might lead to absurd ventilator settings; for instance, the Siemens SERVO-i can deliver a maximum flow rate of 3.33L/sec, and so could theoretically blow a 500ml tidal volume over an I-time of 0.15 seconds. Obviously no sane person would ever leave the patient on those settings. So, there must be some safe (short) inspiratory time which everybody is comfortable with, and which still maximises expiratory time.
How long does the inspiratory time have to be, and for what scientific reason does it need to be longer than 0.15 seconds? There are two main reasons. For most conventional ventilation scenarios, a prolonged I-time allows the recruitment of lung units with longer time-constants, which favours oxygenation by increasing the available gas exchange surface, and by decreasing shunt by defeating the typical tendency of mechanically ventilated patients to develop atelectasis in their lung bases. These matters are less relevant in the context of severe status asthmaticus. For one, the lung units with the long time constants tend remain recruited throughout the respiratory cycle by autoPEEP. Atelectasis does not tend to be a major feature of severe acute asthma, and oxygenation is rarely the problem. The more important benefit of a longer inspiratory time in asthma is the tendency of short time constant units to fill quickly, taking much of the volume and therefore becoming the most hyperinflated during inspiration. A longer inspiratory time is therefore likely to result in a more even inflation among heterogeneous lung units, and therefore a reduced risk of pneumothorax. Sarnaik et al (2004) used this argument to support the use of a pressure-controlled mode of ventilation for asthamtic children.
Another relevant benefit of a longer inspiratory time is the reduced flow rate. Using a "square" constant flow pattern, one could deliver the 500ml volume over 0.15 seconds with a flow rate of around 200L/min, or over 1.0 seconds with a flow rate of 30L/min. In severely bronchospastic highly resistant airways, flow rate will be directly related to the peak inspiratory pressure; the lower flow rate will produce much lower peak pressures, and fewer ventilator alarms.
The better question to ask is, how much good does it do to shorten inspiratory time to a minimum? Turns out, the benefit from it is probably minor, when compared to decreasing the respiratory rate. Tuxen & Lane (1987) found that changes to the respiratory rate and minute volume were by far the more effective. The ultimate benefit is from the increased expiratory time; removing extra breaths from a minute of ventilation increases the total expiratory time by a much large value than decreasing the inspiratory time of each breath from 1.0 to 0.5 seconds.
In summary, the strategy of shortening the inspiratory time should be viewed as second fiddle to the strategy of decreasing the respiratory rate. It should be resorted to only in those scenarios where one cannot possibly decrease the respiratory rate any further, and but still wishes to have some additional minor advantage. Another minor advantage is passing the CICM fellowship exam, where this topic comes up frequently, and where the official college answers (eg. in Question 27 from the second paper of 2012 and Question 4 from the first paper of 2006 ) occasionally mention "minimise inspiratory time) as one of the strategies. In order to score full marks, one would be expected to join the examiners in being inaccurate.
So. As with virtually everything in mechanical ventilation of asthma, this topic is somewhat controversial. The conventional teaching is to minimise the PEEP for these patients. In fact you're often ventilating them with zero PEEP (ZEEP). The rationale for this is:
There is also the erroneous belief that the added ventilator-generated PEEP might somehow "stack" with the intrinsic PEEP, increasing the total PEEP by some sort of weird additive mechanism. Clearly, this is wrong, as it defies logic.
Anyway. There is also a rationale for having some PEEP:
So, some sort of careful minimalist PEEP is probably either not harmful, or actually helpful. In fact, as long as the ventilator-generated PEEP is not higher than the intrinsic PEEP, there should be no harmful effects.
This is not exactly a highly evolved well-investigated scientific territory. We have case reports and expert opinion to guide us. For a representative example, Golchin et al (2018) had used a small amount of PEEP in their asthmatic, and demonstrated an improvement in the degree of dynamic hyperinflation (in terms of haemodynamics and ventilator measurements) but this was not even a case report, but a poster presentation from the International Conference of the American Thoracic Society, barely meeting the definition of a peer-reviewed publication. That's the level of evidence we are dealing with here. It is probably also possible to extrapolate some of the findings from the COPD literature, because surely bronchospasm is bronchospasm?... Kondili et al (2004) found that some PEEP (5 or 10) produced a more uniform pattern of lung emptying in patients with COPD, whereas ZEEP produced a higher respiratory resistance and time constant. The same was demonstrated by Tuxen et al (1989), who tested asthmatics at PEEP of 5, 10 and 15. The investigators found that, at any given respiratory rate, the increase in PEEP produced an improvement in gas trapping. However (probably because they were using monstrous pre-ARDSNet tidal volumes) the increase in FRC with extra PEEP exceeded the decrease in gas trapping, i.e. the patient was now hyperinflated because of the insane ventilation strategy rather than because of the asthma. "PEEP produced significant but unnecessary improvements in arterial oxygenation and potentially dangerous increases in lung volumes and airway and intrathoracic pressures", Tuxen et al concluded unselfconsciously.
Without the support of published evidence, the author reluctantly finds himself pretending to be a mechanical ventilation expert. So, what recommendation can be made with a straight face, regarding the use of PEEP in severe asthma? Perhaps the following:
This "responsiveness" to extrinsic PEEP is going to change dynamically over the course of time, perhaps over the timeframe of minutes (eg. after every neb). Fortunately, if the intrinsic PEEP increases, the extrinsic PEEP is not going to cause much of a problem, as it may decrease the airway collapse (and unless comically huge tidal volumes are used, you should not see any problems with this). Alternatively, if the intrinsic PEEP decreases, the extrinisic PEEP will not going to cause problems either, because you're probably using a very low level of PEEP (i.e. under 10 cm H2O).
Tenacious mucus plugging is a major problem and historically people have viewed mucolytics +/- bronchoscopy as either a major curative strategy or a end-of-the-line last ditch rescue therapy for intubated severe asthma patients. A representative article from the heyday of this strategy is Henke et al (1996), whose team tended to give 3ml of an N-acetylcysteine/salbutamol mixture into each lobe of the lung via a bronchoscope and then suctioned the resulting mucus.
The options for mucolytic agent are
This practice is far from dead. A more recent survey of British intensivists found that "recombinant human DNase (rhDNase) does get administered by 63% of clinicians, with 54% and 19% that administer hypertonic saline or N-acetylcysteine, respectively" (Snoek et al, 2015). Case reports with miraculous rescues exist (eg. Hull et al, 2010).
The largest part of the literature on this topic comes in the from of observational studies and case reports. The rest is evidence extrapolated from cystic fibrosis patients, burns patients, surgical patients and other non-asthma territories. For these reasons, most authors do not recommend mucolytics. In their chapter (Ch 5, Acute Exacerbations of Asthma) for Pulmonary Emergencies (2016), Bhakta and Lazarus placed mucolytics into the "Therapies without supporting evidence" table. Irwin and Rippe (p 708, section 4) also could not recommend either rhDNAse or N-acetylcysteine. In fact, even in cystic fibrosis the benefits of rhDNAse were not statistically significant (seems antibiotics and chest physiotherapy did all the work). The most recent consensus guidelines (Expert Panel Report 3, 2007) do not support the routine use of mucolytics.
When all else fails, patients who are dying from this reversible condition should be offered ECMO wherever it is available. A publication from 2003 details a series of case reports, which support the use of ECMO in near-fatal asthma. If hypercapnea is what ails you, CO2 removal alone by ECCOR is also an option. Obviously, owing to the extreme potential complications, and the violently invasive nature of this treatment, it should probably be reserved for patients in whom all other management strategies have failed.