Acute, severe asthma describes the serious asthmatic attack that places the patient at risk of developing respiratory failure, a condition referred to as status asthmaticus [13,14]. The time course of the asthmatic crisis as well as the severity of airways obstruction may vary broadly [14]. In some patients who present with asthmatic crisis, repeated PEF measurements when available may document subacute worsening of expiratory flow over several days before the appearance of severe symptoms, the so-called 'slow onset asthma exacerbation'. In others, however, lung function may deteriorate severely in less than one hour, the so-called 'sudden onset asthma exacerbation' [14,15]. Slow onset asthma exacerbations are mainly related to faults in management (inadequate treatment, low compliance, inappropriate control, coexisting psychological factors) that should be investigated and corrected in every patient in advance. On the other hand, massive exposure to common allergens, sensitivity to nonsteroidal anti-inflammatory agents, and sensitivity to food allergens and sulphites are mainly considered the triggers in sudden asthma exacerbations. Without prompt and appropriate treatment, status asthmaticus may result in ventilatory failure and death.

Two different patterns of fatal asthma have been described (Table 1). The greater number of deaths from asthma (80–85%) occurs in patients with severe and poorly controlled disease who gradually deteriorate over days or weeks, the so-called 'slow onset – late arrival' or type 1 scenario of asthma death [2,3,4,16,17,18]. This pattern of asthma death is generally considered preventable. A variation of this pattern is a history of unstable disease, which is partially responsive to treatment, upon which a major attack is superimposed. In both situations, hypercapnic respiratory failure and mixed acidosis ensues and the patient succumbs to asphyxia, or if mechanical ventilation is applied, to complications such as barotrauma and ventilator-associated pneumonia. Pathologic examination in such cases shows extensive airways plugging by dense and tenacious mucous mixed to inflammatory and epithelial cells, epithelial denudation, mucosal edema, and an intense eosinophilic infiltration of the submucosa. In a small proportion of patients, death from asthma can be sudden and unexpected (sudden asphyxic asthma), without obvious antecedent long-term deterioration of asthma control, the so-called 'sudden onset' or type 2, scenario of asthma death [18,19,20,21]. Affected individuals develop rapidly severe hypercapnic respiratory failure with combined metabolic and respiratory acidosis, and succumb to asphyxia. If treated (medically and/or mechanically ventilated), however, they present a faster rate of improvement than patients with slow-onset asthmatic crisis. Pathologic examination in such cases shows 'empty' airways (no mucous plugs) in some patients, and in almost all patients, a greater proportion of neutrophils than eosinophils infiltrating the submucosa is observed [20,21,22].

Intrapulmonary shunt appears to be practically absent in the majority of patients because of the collateral ventilation, the effectiveness of the hypoxic pulmonary vasoconstriction, and the fact that the airway obstruction can never be functionally complete [24]. Hypoxemia is therefore common in every asthmatic crisis of some severity; mild hypoxia is easily corrected with the administration of relatively low concentrations of supplemental oxygen [25]. More severe hypoxemia and the need for higher concentrations of supplemental oxygen may relate to some contribution of shunt physiology.

Analysis of arterial blood gases is important in the management of patients with acute, severe asthma, but it is not predictive of outcome. In the early stages of acute, severe asthma, analysis of arterial blood gases usually reveals mild hypoxemia, hypocapnia and respiratory alkalosis. If the deterioration in the patient's clinical status lasts for a few days there may be some compensatory renal bicarbonate secretion, which manifests as a non-anion-gap metabolic acidosis. As the severity of airflow obstruction increases, arterial carbon dioxide (PaCO₂) first normalizes and subsequently increases because of patient's exhaustion, inadequate alveolar ventilation and/or an increase in physiologic death space.

Hypercapnia is not usually observed for FEV₁ values higher than 25% of predicted normal, but in general, there is no correlation between airflow rates and gas exchange markers. Furthermore, paradoxical deterioration of gas exchange, while flow rates improve after the administration of β-adrenergic agonists is not uncommon.

Respiratory acidosis is always present in hypercapnic patients who rapidly deteriorate and in severe, advanced-stage disease, metabolic (lactic) acidosis may coexist. The pathogenesis of lactic acidosis in the acutely severe asthmatic patient remains to be fully elucidated. There are several mechanisms that are probably involved [13]: the use of high-dose parenteral β-adrenergic agonists; the highly increased work of breathing resulting in anaerobic metabolism of the ventilatory muscles and overproduction of lactic acid; the eventually coexisting profound tissue hypoxia; the presence of intracellular alkalosis; and the decreased lactate clearance by the liver because of hypoperfusion.

During an asthma attack, all indices of expiratory flow, including FEV₁, FEV₁/FVC (forced vital capacity), PEF, maximal expiratory flows at 75%, 50%, and 25% of vital capacity (MEF₇₅, MEF₅₀, and MEF₂₅ respectively) and maximal expiratory flow between 25% and 75% of the FVC (MEF_25–75) are reduced significantly. The abnormally high airway resistance observed (5–15 times normal) is directly related to the shortening of airway smooth muscle, edema, inflammation, and excessive luminal secretions, and leads to a dramatic increase in flow-related resistive work of breathing. Although the increased resistive work significantly contributes to patient functional status, however, the elastic work also increases significantly, and enhances respiratory muscle fatigue and ventilatory failure [26,27].

In asthmatic crisis, remarkably high volumes of functional residual capacity (FRC), total lung capacity and residual volume can be observed, and tidal breathing occurs near predicted total lung capacity. Lung hyperinflation that develops as a result of acute airflow obstruction, however, can also be beneficial since it improves gas exchange. The increase in lung volume tends to increase airway caliber and consequently reduce the resistive work of breathing. This is accomplished, however, at the expense of increased mechanical load and elastic work of breathing.

Lung hyperinflation in acute, severe asthma, is primarily related to the fact that the highly increased airway expiratory resistance, the high ventilatory demands, the short expiratory time, and the increased post-inspiratory activity of the inspiratory muscles (all present at variable degrees in patients in status asthmaticus) do not permit the respiratory system to reach static equilibrium volume at the end of expiration (Fig. 1). Inspiration, therefore, begins at a volume in which the respiratory system exhibits a positive recoil pressure. This pressure is called intrinsic positive-end expiratory pressure (PEEP_I) or auto-PEEP. This phenomenon is called dynamic hyperinflation and is directly proportional to minute ventilation (V_E) and to the degree of airflow obstruction.

Dynamic hyperinflation has significant unfavorable effects on lung mechanics. First, dynamic hyperinflation shifts tidal breathing to a less compliant part of the respiratory system pressure-volume curve leading to an increased pressure-volume work of breathing. Second, it flattens the diaphragm and reduces generation of force since muscle contraction results from a mechanically disadvantageous fiber length. Third, dynamic hyperinflation increases dead space, thus increasing the minute volume required to maintain adequate ventilation. Conceivably, asthma increases all three components of respiratory system load, namely resistance, elastance, and minute volume. Finally, in acute severe asthma, the diaphragmatic blood flow may also be reduced. Under these overwhelming conditions, in the case of persistence of the severe asthma attack, ventilatory muscles cannot sustain adequate tidal volumes and respiratory failure ensues.

Acute, severe asthma alters profoundly the cardiovascular status and function [28,29]. In expiration, because of the effects of dynamic hyperinflation, the systemic venous return decreases significantly, and again rapidly increases in the next respiratory phase. Rapid right ventricular filling in inspiration, by shifting the interventricular septum toward the left ventricle, may lead to left ventricular diastolic dysfunction and incomplete filling. The large negative intrathoracic pressure generated during inspiration increases left ventricular after-load by impairing systolic emptying. Pulmonary artery pressure may also be increased due to lung hyperinflation, thereby resulting in increased right ventricular afterload. These events in acute, severe asthma may accentuate the normal inspiratory reduction in left ventricular stroke volume and systolic pressure, leading to the appearance of pulsus paradoxus (significant reduction of the arterial systolic pressure in inspiration). A variation greater than 12 mmHg in systolic blood pressure between inspiration and expiration represents a sign of severity in asthmatic crisis. In advanced stages, when ventilatory muscle fatigue ensues, pulsus paradoxus will decrease or disappear as force generation declines. Such status harbingers impeding respiratory arrest.

Pneumothorax eventually associated with pneumomediastinum, subcutaneous emphysema (Fig. 2), pneumopericardium and tracheoesofageal fistula (in the mechanically ventilated patient) (Fig. 3) are rare but potentially severe complications of acute, severe asthma. Myocardial ischemia should be considered in older patients with coronary artery disease. Mucus plugging and atelectasis are not rare and usually respond to effective treatment. Other complications to consider include theophylline toxicity, lactic acidosis, electrolyte disturbances (hypokalemia, hypophosphatemia, hypomagnesemia), myopathy and ultimately anoxemic brain injury [13].

Close monitoring by serial measurements of lung function (PEF or FEV₁ at bedside) to quantify the severity of airflow obstruction and its response to treatment are of paramount importance. PEF or FEV₁ <30–50% of predicted or personal best indicates severe attack. Attention should be paid to measuring lung function in the severely ill patient, however, because the deep inspiration maneuver involved in PEF or FEV₁ measurement may precipitate respiratory arrest by worsening bronchospasm [31]. Failure of treatment to improve expiratory flow predicts a more severe course and the need for hospitalization [32].

Continuous or repetitive nebulisation of short-acting β₂-agonists is the most effective means of reversing airflow obstruction and can be given safely. Continuous nebulisation of β₂-agonists may be more effective in children and severely obstructed adults [37,38,39]. Salbutamol (albuterol) is the most frequently used agent because of its potency, duration of action (four to six hours) and β₂-selectivity. Continuous or repetitive nebulisation of salbutamol should be preferred because duration of activity and effectiveness of β₂-agonists are inversely related to the severity of airways obstruction [13]. The usual dose is 2.5 mg of salbutamol (0.5 ml) in 2.5 ml normal saline for each nebulisation (Table 3) [34]. Nebulised β₂-agonists should continue until a significant clinical response is achieved or serious side effects appear (severe tachycardia, or arrhythmias). Prior ineffective use of β₂-agonists does not preclude their use and does not limittheir efficacy [13]. Inhaled therapy with β₂-agonists appears to be equal to, or even better than, their intravenous infusion in treating airways obstruction in patients with severe asthma [13].

Anticholinergics, such as ipratropium bromide (Table 3), may be considered in the emergency treatment of asthma but there is controversy on their ability to offer additional bronchodilation [40,41]. Methylxanthines, such as theophylline (Table 3), in the emergency department are of debated efficacy and not generally recommended [1,42]. Several authors, however, consider that there is enough data demonstrating that theophylline treatment in the emergency department benefits patients after 24 hours. Its non bronchodilating properties, including its action on the diaphragm and its anti-inflammatory effects, may thus warrant the use of theophylline in the emergency care of acute, severe asthma [43].

Systemic corticosteroids (to speed the resolution and reduce relapse) are recommended for most patients in the emergency department, especially those with moderate-to-severe exacerbation and patients who do not respond completely to initial β₂-agonist therapy [1,44,45]. The intensification of a patient's corticosteroids therapy, however, should begin much earlier, at the first sign of loss of asthma control. Corticosteroids in the emergency department may also help to reduce mortality from asthma [3]. Since benefits from corticosteroid treatment are not usually seen before six to twelve hours, early administration is necessary. A recent study reports for the first time that large doses of inhaled corticosteroids (18 mg flunisolide in three hours) administered in the emergency department, in addition to β₂-agonists, speed the resolution of acute bronchoconstriction [46]. Furthermore, another study recently demonstrated that a single dose of inhaled budesonide (2.4 mg) significantly reduced sputum eosinophils and airway hyperesponsiveness in six hours [47]. These studies appear to also support a role for inhaled steroid treatment in asthma exacerbations, but additional studies are necessary to assess steroid behavior in these conditions [48].

Long-term treatment with inhaled corticosteroids has been shown to reduce hospitalization rates in younger patients with asthma [49] and to reduce the risk of rehospitalization and all-cause mortality in elderly asthmatics [50]. The optimal dose and dosing frequency of systemic corticosteroids in the severe hospitalized asthmatics are not clearly established. One common approach is the intravenous administration of 60–125 mg methylprednisolone every six hours during the initial 24–48 hours of treatment (Table 3), followed by 60–80 mg daily in improving patients, with gradual tapering during the next two weeks [13,51]. In addition to common and well-known side effects of corticosteroid administration (hyperglycemia, hypertension, hypokalemia, psychosis, susceptibility to infections), myopathy should be considered seriously in the intubated and mechanically ventilated patient.

Subcutaneous administration of epinephrine or terbutaline should be considered, in patients not responding adequately to continuous nebulised salbutamol, and in those patients unable to cooperate (depression of mental status, apnea, coma) [52] It should also be attempted in intubated patients not responding to inhaled therapy. Epinephrine may also be delivered effectively down the endotracheal tube in extreme situations [13]. Subcutaneously, 0.3–0.4 ml (1:1000) of epinephrine can be administered every 20 min for three doses (Table 3). Terbutaline can be administered subcutaneously (0.25 mg) or as intravenous infusion starting at 0.05–0.10 μg/kg per min (Table 3). When administered subcutaneously, however, terbutaline loses its β-selectivity and offers no advantages over epinephrine [53]. Terbutaline administered subcutaneously should be preferred only in pregnancy because it appears safer [13]. Subcutaneous administration of epinephrine or terbutaline should not be avoided or delayed since it is well tolerated even in patients older than 40–50 years with no history of cardiovascular disease (angina or recent myocardial infarction). Intravenous administration of β-agonists (epinephrine, salbutamol) is also an option in extreme situations and should be considered in the treatment of patients who have not responded to inhaled or subcutaneous treatment, and in whom respiratory arrest is imminent, or in patients not adequately ventilated and severely hyperinflated, despite optimal setting of the ventilator.

Antibiotics are not recommended for the treatment of asthma exacerbations and should be reserved for those with evidence of infection (e.g. pneumonia, sinusitis) [1]. Aggressive hydration is not recommended for adults or older children but may be indicated for infants and young children [1]. Chest physical therapy, mucolytics, and sedation are not recommended [1].

Effective sedation is necessary to prepare for intubation and to allow synchronism between the patient and the ventilator [13,29]. In addition, sedation improves patient comfort, decreases oxygen consumption and dioxide production, decreases the risk of barotrauma, and facilitates procedures. There appears to be no standard sedation protocol for the asthmatic patient (Table 4). One purposed approach is the intravenous administration of midazolam, 1 mg initially, followed by 1 mg every 2–3 min until the patient allows positioning and inspection of the airway [13].

Ketamine is an intravenous general anesthetic with sedative, analgesic, anesthetic, and bronchodilating properties that appears useful in the emergency intubation for asthma [54,55,56]. During intubation the intravenous administration of 1–2 mg ketamine/kg at a rate of 0.5 mg/kg/min results in 10–15 min of general anesthesia without significant respiratory depression. Bronchodilation appears within minutes after intravenous administration and lasts 20–30 min after cessation. Because of its sympaticomimetics effects, ketamine is contraindicated in hypertension, cardiovascular disease, high intracranial pressure, and preeclampsia. Additional side effects of ketamine include lowering of the seizure threshold, altered mood, delirium, laryngospasm, and aspiration. Furthermore, since ketamine is metabolized by the liver to norketamine, which also has anesthetic properties and a half-life of about 120 minutes, drug accumulation may occur and lead to prolonged side effects. Paralysis with the short-acting neuro-muscular blocker, succinylcholine, may offer some additional advantages [57].

Propofol is an excellent sedating agent since it has a rapid onset and a rapid resolution of its action [58]. In addition it has bronchodilating properties and since patients can be titrated to anesthetic-depth sedation, it may avoid the need for paralytic agents. Propofol is administered intravenously during the peri-intubation period at a dose of 60–80 mg/min initial infusion, up to 2 mg/kg, followed by an infusion of 5–10 mg/kg/hour as needed, and for sedation for protracted mechanical ventilation 1–4.5 mg/kg/hour [13]. Prolonged propofol administration may be associated with generalized seizures, increase carbon dioxide production, and hypertriglyceridemia [59,60].

Opioids (e.g. morphine sulfate) are not usually recommended for sedation in asthmatics because of their potential to induce hypotension through a combination of direct vasodilation, histamine release, and vagally-mediated bradycardia. Opioids also induce nausea and vomiting, decrease gut motility, and depress ventilatory drive.

The use of neuromuscular-blocking agents (e.g. vecuronium, atracurium, cis-atracurium, pancuronium) lessens the patient–ventilator asynchronism (thus permitting more effective ventilation), lowers the risk for barotrauma, reduces oxygen consumption and dioxide production, and reduces lactate accumulation. The use of paralytics, however, presents a number of disadvantages such as myopathy, excessive airways secretions, histamine release (atracurium), and tachycardia and hypotension (pancuronium). The concomitant use of corticosteroids may increase the risk for postparalytic myopathy [61]. Paralytic agents, when administered, may be given intermittently by bolus or continuous intravenous infusion. If a continuous infusion is used, either a nerve stimulator should be used or the drug should be withheld every four to six hours to avoid drug accumulation and prolonged paralysis. Currently the use of paralytics is usually recommended only in those patients who cannot adequately be controlled with sedation alone.

Soon after the intubation of the severe asthmatic patient and the application of positive pressure ventilation, systemic hypotension commonly appears. The determinant factors of systemic hypotension are sedation, hypovolemia, and primarily lung hyperinflation. Dynamic hyperinflation, as previously discussed, occurs when the highly increased airway expiratory resistance prolongs expiratory flow in such a way that the next breath interrupts exhalation of the inspired tidal volume, leading to end-expiratory gas trapping. The mechanisms of intrinsic positive end-expiratory pressure (PEEP_I) in acute, severe asthma are as follows:

1. Deceleration of the expiratory flow

High resistance to airflow

Shortened expiratory time

Increased post-inspiratory activity of the inspiratory muscles.

2. Increased expiratory muscle activity.

3. High ventilatory demands.

Additional factors in its pathogenesis are insufficient time during expiration to equilibrate alveolar pressure with atmospheric pressure, persistent inspiratory muscle activity during expiration, and an increased expiratory muscle activity [62,63].

Dynamic hyperinflation is directly proportional to V_E and to the degree of airflow obstruction. In the severely obstructed patient, dynamic hyperinflation may further increase in the immediate post intubation period, as a result of the mechanical ventilation, even after the delivering of normal or reduced V_E. High levels of dynamic hyperinflation may cause hemodynamic compromise in a somewhat similar way to that caused by the tension pneumothorax. The large intrathoracic pressures that develop due to dynamic hyperinflation increase the right atrial pressure, leading to a decrease in venous return to the right heart. An additional mechanism of decrease in venous return may be related to the positive intrathoracic pressures throughout the respiratory cycle during mechanical ventilation. Sedation, eventually muscle relaxation, and hypovolemia (due to increased water loss or decreased water uptake) reduce the mean systemic blood pressure and cause a further decrease in venous return to the heart. The end result may be sudden cardiovascular collapse, with systemic hypotension and tachycardia. Similar cardiovascular effects also may be observed in patients with tension pneumothorax. In such cases a trial of hypoventilation (delivering of 2–3 breaths/min of 100% oxygen for few minutes) may become necessary to exclude this eventuality. With such setting of the ventilator, in the absence of pneumothorax, the mean intrathoracic pressure will fall, systemic blood pressure will rise, pulse pressure will increase, and pulse rate will fall [13].

Controlled modes of ventilation are preferred in the immediate post intubation period for different reasons. Patients in status asthmaticus not only have an excessive work of breathing but in addition most of them are intubated and mechanically ventilated when ventilatory muscles are fatigued. Previous studies have shown that at least 24 hours are needed for complete recovery of fatigue. It appears reasonable, therefore, to rest the muscles for some time in the controlled mode of ventilation before switching the ventilator to the assist mode.

Two different modes of controlled setting of the ventilator are widely used. One mode is pressure-controlled ventilation, where the airway pressure can be maintained at a predetermined level independently of any eventual changes of the mechanical properties of the respiratory system. The other method is volume-controlled ventilation, where the ventilator delivers a predetermined tidal volume and the airways pressure becomes the dependent variable. Pressure-controlled ventilation seems more appropriate to maintain airway pressure, especially in status asthmaticus in which airways resistance varies rapidly.

Correction of the hypoxemia is one of the first priorities in setting the ventilator [64]. In non-complicated status asthmaticus, low V/Q ratio is the predominant mechanism and true shunt does not contribute significantly to the hypoxemia. Correction of the hypoxemia is therefore easily obtained by the administration of relatively low concentrations of supplemental oxygen. In the majority of patients, the administration of a fraction of inspired oxygen at the level of 30–50% is sufficient to raise PaO₂ above 60 mmHg. Failure of this level of fraction of inspired oxygento increase PaO₂ above 60 mmHg indicates some kind of complication (e.g. mucoid impaction atelectasis, pneumonia, or pneumothorax). Complete correction of the respiratory acidosis is not an urgent priority and, since it requires normalization of the PaCO₂, which may cause significant deterioration of dynamic hyperinflation, most authors prefer a carefully used buffered therapy for the correction of patient's pH.

The appropriate setting of the ventilator in mechanically ventilated patients is of paramount importance to manage, and to avoid further significant increase of, lung hyperinflation. Several studies have shown that for patients in status asthmaticus, a large part of the morbidity and mortality are related to the mechanical ventilation itself rather to the disease process [64]. In patients on controlled modes, there are three strategies that can decrease dynamic hyperinflation: decrease of V_E; increase of expiratory time; and decrease of resistance. Controlled hypoventilation (permissive hypercapnia) is the recommended ventilator strategy to reduce the degree of lung hyperinflation and provide adequate oxygenation and ventilation while minimizing systemic hypotension and barotrauma. [1,65,66,67]. Hypoventilation can be obtained by decreasing either the tidal volume or the respiratory frequency, or both. Hypercapnia is generally well tolerated as long as PaCO₂ does not exceed 90 mmHg and acute variations in PaCO₂ are avoided. Low values of arterial pH are well tolerated, and slow infusions of sodium bicarbonate appear to be safe in patients with acidosis.

In the immediate post intubation period, and as a result of the mechanical ventilation, especially when patients are ventilated to eucapnia, dynamic hyperinflation may further increase. In order to avoid dangerous levels of dynamic hyperinflation, the ventilator should be set to allow sufficient exhalation time (increasing the expiratory time) while continuing bronchodilator and corticosteroid treatment (decreasing resistance to expiratory flows). The increase in expiratory time can be achieved by increasing inspiratory flows at the expense of increasing peak dynamic pressures, and by the elimination of end-inspiratory pause. The strategy of increasing expiratory time, although less powerful than controlled hypoventilation, decreases dynamic hyperinflation considerably, whereas it improves cardiovascular function and gas exchange [64]. Decreasing resistance to expiratory flows includes pharmacologic treatment to overcome asthma and an effort to decrease the external resistance related to ventilator tubing.

Several studies have shown that setting the ventilator with a level of V_E of 8–10 L/min (achieved by a tidal volume of 8–10 ml/kg and respiratory rate of 11–14 breaths/min) combined with an inspiratory flow rate of 80–100 L/min, may be appropriate to avoid dangerous levels of dynamic hyperinflation (Table 5).

There are several techniques for the estimation of the level of dynamic hyperinflation. One way is to measure the end-inspired volume above apneic FRC (V_EI) [68]. Apneic FRC is determined by measuring the total exhaled volume of gas during 20–40 seconds of apnea in a paralyzed patient. V_EI is the sum of the tidal volume and the volume at end exhalation above FRC (Fig. 4). Values of V_EI above a threshold value of 20 ml/kg (1.4 L in an average adult) have been shown to predict complications of hypotension and barotrauma [68]. The need for paralysis, however, limits the use of V_EI.

In the mechanically ventilated patient other estimates of the degree of dynamic hyperinflation are the determination of the end-inspiratory plateau pressure (Pplat) and the measure of PEEP_I [69]. To obtain both measurements patient relaxation, but not paralysis, is necessary. Pplat is an estimate of average end-inspiratory alveolar pressures and is easily determined by stopping flow at end-inspiration (Fig. 5). PEEP_I is the lowest average alveolar pressure achieved during the respiratory cycle and is obtained by an end-expiratory hold maneuver (Fig. 6). When evaluating these methods, The American College of Chest Physicians consensus conference on mechanical ventilation concluded that Pplat is the best predictor of hyperinflation in ventilated asthmatic patients and recommended that Pplat must be kept lower than 35 cmH₂O [70].

After successful treatment and when the patient's conditions improve, the ventilator should be switched to assisted modes (pressure or volume), and the process of weaning should begin. In order to improve the patient/ventilator synchronism, two issues should be taken into account, the response of ventilator to patient effort, and the response of patient effort to ventilator-delivered breath.