Airway mucus in healthy subjects usually acts as an essential protective physical barrier by forming a thin layer on top of the airway lumen and then constructing MCC together with cilia (about 200 per cell, 6-7 μm long). Under the cyclic beatings of cilia including a fast effective stroke and a slow recovery stroke (10~15 Hz), the inhaled particles and pathogens can be trapped in the airway mucus and eliminated from the airway system (Fig. 1) (Loreng and Smith, 2017; McAuley et al., 2019; Ridley and Thornton, 2018; Roy et al., 2014; Sanderson and Sleigh, 1981; Widdicombe, 1997; Xu and Jiang, 2019b).

According to its location to cilia, the airway mucus layer can be divided into an apical high viscosity layer (mucus gel, 10-50 μm depth) and a bottom low viscosity layer (periciliary layer, PCL, 7 μm depth) (Button et al., 2012; Thornton and Sheehan, 2004). Since the height of cilia is similar to the depth of PCL layer, most of the movements of cilia occur in the PCL layer. Therefore, during the effective strokes the cilia intermittently sweep the subside of the mucus gel, which leads to the mucus cephalad transport (~100 μm/s) (Silberberg, 1990). This high turnover rate of airway mucus by MCC is a major feature of the respiratory system in healthy subjects. However, in COVID-19 patients, SARS-CoV-2 infection is known to seriously disrupt the function of MCC. For example, it has been reported that SARS-CoV-2 infection inadvertently leads to the shedding of ciliated cells from the airway epithelium (Fang et al., 2020). This is partly because that a subpopulation of the ciliated cells highly express ACE2 and TMPRSS2 and thus mediate the initial infection and cellular entry of SARS-CoV-2 (Chilvers et al., 2001; Lee et al., 2020; Sungnak et al., 2020; Ziegler et al., 2020; Zou et al., 2020). In addition, it has been demonstrated in a reconstituted human bronchial epithelial model that SARS-CoV-2 preferentially replicates in ciliated cells, which causes a rapid loss of the ciliary layer and thus compromises the motile function of the cilia (Brosnahan et al., 2020). On the other hand, the intrinsic epithelial defense mechanisms likely occur too late to prevent such cilia loss due to SARS-CoV-2 infection (Robinot et al., 2020).

Apart from the structural dysfunction of MCC, SARS-CoV-2 also disrupts the cilia beating through interference with the physiologically relevant signaling that modulates the cilia beating frequency, such as ATP, acetylcholine, cyclic adenosine monophosphate, cyclic guanosine monophosphate, nitric oxide, and inositol trisphosphate (Essaidi-Laziosi et al., 2018; Kuek and Lee, 2020). Both the structural and signaling dysfunctions of the airway epithelium and the airway cilia would inevitably reduce the ability of MCC to remove airway mucus from the lung, which partially explains why airway mucus accumulation is so common and severe in critically ill COVID-19 patients.

The principal components of airway mucus both structurally and functionally are mucins, and there are at least nine types of them in human airways (Chatterjee et al., 2020; Lillehoj et al., 2013). Among these mucins, MUC1, MUC4 and MUC16 mainly locate in the PCL layer tethered to the plasma membrane of airway epithelium via a transmembrane domain, which attract water, thereby forming a lubricate fluid layer surrounding and protecting the cilia to promote movement of the mucus gel above. The gel-forming MUC2, MUC5A/C, MUC5B and MUC19, and the non-gel forming MUC7 and MUC8 mainly locate in the mucus gel layer after secretion by goblet cells (Fig. 1) (Denneny et al., 2020; Dhanisha et al., 2018; Fahy and Dickey, 2010; Hattrup and Gendler, 2008; Ridley and Thornton, 2018; Thornton et al., 2008).

Airway mucins always contain highly repeated domains of proline, threonine and serine. Since both threonine and serine mediate glycosylation via their oxygen, airway mucins are thus heavily glycosylated (20% protein core and 80% carbohydrates). In addition, the gel-forming mucins also contain abundant cysteine, which mediates the non-covalent and covalent crosslinks between the ends of different mucins due to their hydrophobic property and the disulfide formation, respectively. Consequently, the gel-forming mucins form polymers up to 5 μm long from only 0.2–0.6 μm long mucin monomers (Bansil et al., 1995; Sheehan et al., 1991). Once secreted, these heavily glycosylated long polymeric mucins interact with ions and absorb water and then increase the volume by several hundred times within seconds, forming a three dimensional lattice hydrostatic mucus gel (Quraishi et al., 1998). Therefore, the airway mucus in the gel layer acquires both viscosity and elasticity, a physical property that is characteristic of complex fluids.

Besides their roles in determining airway mucus physical properties, airway mucins also have vital roles in protection of airways from infection by viruses such as SARS-CoV-2. For example, it has been shown that the terminal sialic acid glycans in gel-forming mucins including MUC5AC and MUC5B can facilitate the interaction with coronavirus (Ganesan et al., 2013). Consequently, the airway mucins may act as the primary site to entrap the viral particles and then remove them from the respiratory tract via MCC. The finding that COVID-19 patients have a higher level of MUC5B than their healthy counterparts seems to corroborate the protective role for MUC5B in this disease (Iyer et al., 2020).

In addition, MUC1, MUC4 and MUC16 in the PCL layer (transmembrane mucins) can bind to coronavirus such as SARS-CoV-2 via the interaction of adhesion and sialic acid, which may prevent the entry of virus into the airway cells (Chatterjee et al., 2020; Zou et al., 2020). The transmembrane mucins are also known to shed their extracellular domains from the cell surface into the airway lumen, which is cleavaged by endogenous protease that is activated due to mucin-virus interactions (McAuley et al., 2017). Therefore, transmembrane mucins such as MUC1, MUC4, and MUC16 located on the surface of cilia can act as releasable decoy receptors and prevent the virus infection (Chatterjee et al., 2020).

In COVID-19 patients, the expression of airway mucins may be enhanced in response to SARS-CoV-2 infection via several signaling pathways, which leads to airway mucus accumulation (Bose et al., 2021; Khan et al., 2021; Plante et al., 2020). For example, SARS-CoV-2 infection may stimulate differentiation and hyperplasia of the goblet cells in the airways, which is followed by hyperproduction and secretion of mucins. Up to date, enhanced expression of various mucins particularly MUC1, MUC2, and MUC4, MUC5AC, MUC5B, and retention of highly viscous mucus in bronchial airways have been observed in COVID-19 patients (Chen et al., 2020; Leng et al., 2020; Lu et al., 2021; Wang et al., 2020; Ye et al., 2020; Zhang et al., 2021).

One of the causes of airway mucin hypersecretion in COVID-19 is the SARS-CoV-2 induced excessive inflammatory response known as cytokine storm, which is summarized by Castelli et al. (2020) in their review article. Briefly, numerous cytokines facilitate mucin expression via various signaling pathways. Among them, IL-2, IL-4, IL-8, IL-9, IL-13 and IL-19 facilitate mucin expression via activation of STATs signaling, IL-1β, IL-5, IL-6, and IL-19 facilitate mucin expression via activation of MAPK signaling, while EGF, ATP and adenosine facilitate mucin expression via activation of NF-κβ signaling. In addition, it has been reported that SARS-CoV-2 increased IFN-β/γ expression, which enhanced expression of airway mucins such as MUC1, MUC2, MUC5AC, MUC5B, MUC13 and MUC16 via activation of IDO1-AhR axis (Hadjadj et al., 2020; Liu et al., 2020b). Since cytokine storm is a general feature in critically ill COVID-19 patients, these patients are also generally associated with airway mucin hypersecretion (Wang et al., 2020).

Airway mucus is a material that behaves like both a viscous fluid and a soft elastic solid, a rheological feature known as viscoelasticity. This airway mucus rheology can be impacted by virus infection and thus differently modulates the cilia beating and mucus transport.

In COVID-19 it has been reported that a proportion of the critically ill patients are often seen with highly viscous airway mucus that even forms colloidal mucus plugs in the bronchioles (Wang et al., 2020). The abnormal rheological properties and mucus accumulation/plugging in the COVID-19 patients may result from various factors such as increased content of solids in the airway mucus (up to 15% as reported by (Button et al., 2016) as well as changes in hydration, pH, and ion composition of the airway mucus (Ridley and Thornton, 2018).

In healthy persons, mucus is continuously secreted into airways with totally about 10–100 mL/day. Under the function of MCC, mucus will move from peripheral airways to the glottis, and then be cleared by swallowing or coughing (Button et al., 2012; Dickey, 2018). For critically ill COVID-19 patients, viral infection and the following inflammatory responses lead to airway mucus accumulation by changing mucus viscoelasticity, stimulating mucus hypersecretion, and disturbing mucus elimination as described above (Button et al., 2016; Ridley and Thornton, 2018; Wang et al., 2020).

Once the sticky airway mucus accumulates in the airways, it provides a favorable growth media for bacteria (Konrad et al., 1994; Li Bassi et al., 2008; Safdar et al., 2005). Bacterial infection will cause airway inflammation, which then induces goblet cells hyperplasia and submucosal gland hypertrophy, and leads to further mucus secretion and thus a vicious cycle of infection, inflammation, and hypersecretion. Additionally, the sticky airway mucus leads to the shedding of ciliated cells and impairs ciliary beating, and thus disrupts the function of MCC. Both the increased secretion of sticky airway mucus and the impaired MCC will further accelerate airway mucus accumulation (Anderson et al., 2015; Zahm et al., 1989).

Airway mucus accumulation can increase airway resistance and thus deteriorate lung function presenting as decreased forced expiratory volume in one second, which will lead to atelectasis, ARDS, ultimately the total obstruction of respiratory tracts and respiratory failure (Cerveri and Brusasco, 2010; Chen et al., 2019b; Curran and Cohn, 2010; Darbee et al., 2005; Dunican et al., 2018; Fahy and Dickey, 2010; Hogg, 2004; Sprung et al., 2010). Therefore, airway mucus accumulation may be one of the causatives for the high death rate of critically ill COVID-19 patients (Chen et al., 2020; Lu et al., 2021; Zhang et al., 2021), and aggressive management of airway mucus accumulation by using appropriate ACTs is essential in clinical practice of treating respiratory conditions including critically ill COVID-19 patients (McKim et al., 2011; Thomas et al., 2020).

Airway mucus rheology determines the efficacy of MCC via modulating the mechanical coupling between the cilia and the mucus (Macchione et al., 1995; Quraishi et al., 1998; Rajendran and Banerjee, 2020; Silberberg, 1983). Therefore, a suitable viscoelasticity is required for the airway mucus to facilitate the conversion of energy from the cilia beating into mucus movement. Indeed, it has been widely reported that the airway mucus with an elastic modulus at 1–2 Pa and a zero shear rate viscosity (yield) 12 Pa∙s (0.01 s⁻¹) is most suitable for the function of MCC (Majima et al., 1991; Puchelle et al., 1987; Rubin et al., 1990; Shih et al., 1977). Studies have also shown that the ratio of elasticity to viscosity and the overall impedance rather than viscosity and elasticity alone are more significantly associated with the function of MCC (Lorenzi et al., 1992).

Interestingly, airway mucus can switch between viscous and elastic features depending on the magnitude of external deformation rates (Chen et al., 2019a; Cone, 2009). For example, when it is sheared airway mucus undergoes a process of thinning due to a temporary realignment of mucins (Volsko, 2013). Because of these, several types of mechanical force including gravitational force, friction force, shearing force, and adhesion force may function together to determine the airway mucus rheological features and thus correlate the tendency of mucus movement (Fig. 2). This confers the ability of external mechanical forces due to tidal breathing (0.3 Hz), ciliary beating (10–15 Hz), forced expiratory maneuvers to regulate mucus rheology and the following mucus transport (Branson, 2007; Lai et al., 2009).

Based on the rheological response of airway mucus to different external forces, it is possible to develop a variety of ACTs for clinical usage. For example, gravity, oscillation, and inspiration/expiration flow have been utilized to modulate airway mucus transport. Indeed, using in vitro and in vivo models, it has been reported that oscillation of mucus can decrease its viscoelasticity in a time, frequency, and magnitude-dependent manner. Interestingly, the oscillation to reduce mucus viscoelasticity is found to be most potent at 12–15 Hz, which is similar to the cilia beating frequency (App et al., 1998). Therefore, oscillation-based working mechanisms have been developed and integrated into different types of ACTs (Huang et al., 2020; Rogers, 2007; Zhu et al., 2020).

If mucus obstructs the airways/alveoli, ventilating behind the obstructed regions is helpful to push the retained airway mucus out, which can be realized via interdependence of the adjacent alveoli and the collateral channels as shown in Fig. 3 (Cetti et al., 2006; Terry and Traystman, 2016).

For obstructed alveoli that are adjacent and physically interconnected, the expansion of alveoli exerts traction forces and opens adjacent collapsed alveoli, thereby assisting ventilation to these regions, which is called interdependence of adjacent alveoli (McIlwaine et al., 2017).

For obstructed airways especially the peripheral small airways, ventilation can be realized through some collateral channels that connect from either bronchioles to bronchioles (channels of Martin), bronchioles to alveoli (channels of Lambert) or alveoli to alveoli (pores of Kohn, ~50/alveolus). Once the airways are obstructed, these collateral channels can assist ventilation behind the obstructed regions to promote clearing the airway mucus since airflow prefers the path of least resistance (Slebos and Shah, 2017).

Ventilation through these accessory pathways can be generated by holding breath, increasing inspiratory flow rate/volume, and increasing positive pressure, which forms the basis of several ACTs such as breathing maneuvers, hyperinflation, and positive expiratory pressure (PEP) (Branson, 2007). In addition to ventilation behind the obstructed airways/alveoli, there are several other ways to physically enhance the fluidity and thus clearance of airway mucus that are discussed respectively in the following.

Expiratory airflow especially peak expiratory flow (PEF) provides shearing forces on airway mucus surface to mobilize its cephalad movement via a two-phase gas-liquid flow mechanism (Fig. 4) (Kim et al., 1986a; Kim et al., 1987; Kim et al., 1986b; Leith, 1968). Basically, once the shearing force exceeds the surface tension and the adhesive force of the mucus gel layer, the airway mucus starts to flow concurrently with the expiratory airflow in a pattern dependent on the magnitude of PEF, i.e., the expiratory air flows as either 1) fine bubbles in the airway mucus (bubble flow) for PEF <18 L/min; 2) large plugs in the airway mucus (plug flow) for PEF at 18-300 L/min; 3) a core inside an annulus of airway mucus (annular flow) for PEF at 300–750 L/min; or 4) fine drops combined with airway mucus (mist flow) for PEF >750 L/min. Among them, both the annular and mist flows can initiate cephalad movement of retained airway mucus (Kim et al., 1986a; Kim et al., 1987; Kim et al., 1986b; Leith, 1968).

It is generally known that PEF is proportional to the intrapulmonary pressure and inversely proportional to the airway resistance. Therefore, increasing the intrapulmonary pressure and decreasing the airway resistance have been used to increase PEF for airway mucus clearance in various ACTs (Maxwell and Ellis, 1998). For example, spontaneous forced expiration maneuvers such as cough and huff, mechanical device assistance such as mechanical insufflation-exsufflation (MI-E) can be used to generate increased intrapulmonary pressure. In addition, according to Poiseuille’s law the airway resistance (R) is proportional to the length (l) and inversely proportional to the 4^th power of radius (r) of the airways tract, as well as proportional to the viscosity (η) of air and the airway mucus (R = 8ηl/πr⁴). Therefore, the airway resistance can be decreased by either increasing airway radius via hyperinflation or decreasing the viscosity of the airway mucus via shearing-thinning by cough, huff, and MI-E (Kim et al., 1987).

On the other hand, the intrapulmonary pressure during forced expiratory will gradually decrease from the alveoli to the mouth because of frictional pressure loss, and there will be a point where the intrapulmonary pressure equals to the surrounding pleural pressure (i.e., equal pressure point, EPP) (Oberwaldner, 2000). From the EPP toward the mouth in the downstream, the pleural pressure exceeds the intrapulmonary pressure and airway compression occurs, which narrows the airway transiently and thus increases PEF and favors a cephalad mucus flow according to Bernoulli’s principle (Fig. 5) (Hollandl and Buttonl, 2006; Kaminsky and Chapman, 2020; van der Schans, 2007).

The location of EPP is mainly determined by the expiratory force and the elastic recoil of the lung, which usually occurs in the central airways. This is why cough and huff are usually effective for clearing airway mucus in the central airways especially above the sixth or seventh generations of airway branching (Button and Button, 2013; Oberwaldner, 2000). However, increasing expiratory force and elastic recoil can shift the site of EPP toward the alveoli, which provides the rationale for ACTs to clear airway mucus in small airways and alveoli by using enhanced force expiratory flow, PEP, and high frequency chest wall oscillation (HFCWO). Since the elastic recoil of the lung is also determined by the lung volume, it is another rationale for ACTs to remove mucus from the distal and central airways by using low and high lung volumes, respectively (Fink, 2007).

In addition to increasing expiratory flow, the flow bias between inspiratory and expiratory flows also influences the cephalad transport of airway mucus (Maxwell and Ellis, 1998; McIlwaine et al., 2017; Volpe et al., 2020). This is due to the fact that although the volume of air moving in and out of the airways during inspiration and expiration is equal, the peak flow during inspiration (PIF) is always different from that during expiration (PEF).

Early studies proposed the flow bias threshold for initiating a cephalad mucus movement in terms of PEF/PIF ratio > 1.1 (Kim et al., 1987). Later study, however, demonstrated that it is the difference between PEF and PIF rather than the PEF/PIF ratio that determines the airway mucus mobilization (Volpe et al., 2008). Although PEF-PIF >17 L/min is the normal threshold for initiating cephalad mucus movement, PEF-PIF >33 L/min is observed in vivo as a threshold to initiate mucus movement in mechanically ventilated patients and this threshold is probably more clinically relevant (Li Bassi et al., 2012). Recently, it has been updated to include three criteria for an expiratory flow bias to mobilize airway mucus cephalad movement, i.e., PEF/PIF ratio > 1.1, PEF-PIF > 17 L/min, PEF/PIF ratio > 4.3 and PEF-PIF > 33 L/min (Oliveira et al., 2019b; Volpe et al., 2008). It is worth emphasizing that this rationale does not depend on a high flow rate and can be applied even under tidal volume ventilation.

Oscillating intrapulmonary airflow can enhance mucus clearance by increasing the rate and bias of expiratory flow, ciliary beat frequency, or mucus hydration as well as by altering the rheological properties of mucus, which can occur in both the central and the peripheral airways (Boucher, 2002; Dasgupta et al., 1998; Hansen et al., 1994; King et al., 1983; Puchelle et al., 1995; Winters et al., 2007; Winters and Yeates, 1997).

The oscillation of intrapulmonary airflow can be induced by either directly oscillating the airflow at the mouth with an oscillatory positive expiratory pressure (Osc-PEP) or vibrating the thorax using HFCWO, both of which have been proved sufficient to overcome the mucus adhesion to the airway wall in the central and peripheral airways and consequently propel the mucus up and out of the airway (Button et al., 2007; Dietl et al., 2001; Tarran et al., 2005).

As mentioned above, shear stress can physically thin airway mucus, which is also an essential pathway to promote mucus clearance (Shen et al., 2018). Apart from shear stress, nanoparticles can also be used to physically thin airway mucus. For example, it has been reported that Fe₂O₃ nanoparticles (rod-shape, 8 nm diameter, 100 nm length) can experimentally induce physical thinning of simulated asthmatic airway mucus (Fig. 6), which may be mediated by the surface interaction between the Fe₂O₃ nanoparticles and the mucin polymers (Wang et al., 2017). Our preliminary data also showed that magnetic field-driven movement of Fe₂O₃ can further decrease the viscoelasticity of the stimulated asthmatic airway mucus (data not shown). Since nanoparticles can be easily delivered into the central and peripheral airways, the phenomenon of mucus thinning induced by Fe₂O₃ nanoparticles provides another rationale for ACTs to physically thin the airway mucus and thus enhance the removal of mucus from the airways. These results suggest that novel nanotechnologies may provide alternative options for modulating airway mucus rheology for the purpose of promoting efficiency of ACTs.

Postural drainage is an ACT that uses different body positions such as upright, side-lying, supine and prone to directly enhance airway mucus movement from the five lobes of the lung (Berney et al., 2012; Button and Button, 2013; Li Bassi et al., 2008; Safdar et al., 2005). These precise postures comprehensively utilize the anatomy of the bronchial tree, and thus facilitate the sliding of airway mucus from the periphery toward the central airways and improve ventilation and perfusion via gravity-dependent strategies. Among them, the prone position is most effective and has been strongly recommended for mechanically ventilated COVID-19 patients, because it promotes drainage of airway secretion, homogenizes lung perfusion, and improves ventilation mismatch (Gattinoni et al., 2020). It is also reported that although machine vibration or manual clapping does not affect severe-to-critical progression of COVID-19 patients, the prone position has been proved effective to avoid this progression via promoting mucus drainage (Wang et al., 2020).

Various breathing maneuvers are known to be used as helpful ACTs to clear sticky mucus out of the obstructed regions, including 3-second breath holdings, thoracic expansion exercises, mobilization by increasing expiratory flow rate or volumes, forced expiration, active cycle of breathing, and autogenic drainage (Lewis et al., 2012).

Each of these maneuvers works differently to promote mucus clearance. Specifically, the 3-second breath holdings can increase the constant time that allows the air to flow to the obstructed regions of the lung via accessory pathways as described before (Maxwell and Ellis, 1998; McIlwaine et al., 2017). The thoracic expansion exercises use a deep, slow inspiration to increase the lung volume and thus promote ventilation to the obstructed airways via accessory pathways. Mobilization increases oxygen demand and thus increases the lung volume, which leads to ventilation behind the obstructed regions via accessory pathways (Amidei, 2012; Hodgson et al., 2014; Thomas, 2013a, b). Forced expiration such as cough and huff relies on an adequate expiratory flow rate and volume to clear the airway mucus. The active cycle of breathing combines breath holding, deep inspiration and forced expiration (Bott et al., 2009). This maneuver consists of one or two forced expirations followed by relaxed breathing. Autogenic drainage uses gently accelerated tidal volume breathing and gradually increased lung volumes to increase the expiratory flow rate and thus improve mucus clearance from peripheral to central airways (Maxwell and Ellis, 1998; McIlwaine et al., 2017). It needs to note that these breathing maneuvers may not be practical for critically ill COVID-19 patients because of their severe infirmity.

Percussion is performed with cupped hands to clap on the thorax of the patients to generate a transient oscillation with a frequency of 3–6 Hz. Such oscillation transmits through the airways to dislodge the airway mucus from the airway wall so that the mucus can be removed out of the airways.

Similarly, the chest wall can be compressed either manually or using mechanical devices such as HFCWO to generate a transient oscillation and compression with a frequency of 3–17 Hz (McCarren et al., 2006b). This vibration is proposed to remove mucus by increasing PEF (about 50%) and the expiratory flow bias or by decreasing the viscosity of mucus (Kim et al., 1987; McCarren and Alison, 2006; McCarren et al., 2006a). In addition, HFCWO utilizes an inflatable vest that is inflated with an oscillatory pressure below 25 cmH₂O at 5–25 Hz. This oscillation pressure can vibrate the chest wall and generate oscillatory airflow to clear airway mucus (Darbee et al., 2005; Kendrick, 2007).

PEP helps to increase functional residual capacity and thus the tidal volume which can open obstructed airways via accessory pathways and support the airways during expiration to avoid small airway collapse (Branson, 2007; McIlwaine et al., 2015). The patients need to exhale against an expiratory resistor for 8-10 cycles with moderately increased tidal breathing to finally achieve a positive pressure of 10–25 cmH₂O (Demchuk and Chatburn, 2021).

Apart from 10–25 cmH₂O, PEP can also function with either 40–140 cmH₂O high pressure (Hi-PEP), oscillation (Osc-PEP), or ≤1 cmH₂O temporary PEP (T-PEP) (Figueiredo et al., 2012). Evidence suggests that Hi-PEP helps to clear airway mucus by ventilating obstructed airways via collateral channels (Clini, 2009; Volsko et al., 2003). It is especially suitable for patients who have an airway instability during forced expiration since it avoids the premature collapse of the airways and thus allows them to exhale a greater volume than the usual forced vital capacity. Osc-PEP is characterized by breathing against intermittent expiratory resistance to induce oscillation of variable frequency and can directly dislodge the sticky mucus and also decrease mucus viscoelasticity in a frequency- and amplitude-dependent manner. T-PEP applies an expiratory pressure at ≤1 cmH₂O only for a fraction of the expiratory phase. This increase in low pressure is created through a pulsatile flow at approximately 42 Hz in frequency. The vibration generated by the pulsatile flow is transmitted throughout the airways to detach the mucus. Analysis suggested that T-PEP provides greater benefit to patients with emphysema or on oxygen therapy, while PEP therapy would be of more benefit for patients on mechanical ventilation (D’Abrosca et al., 2017; Nicolini et al., 2018).

Intrapulmonary percussive ventilation (IPV) provides originally high-frequency mini bursts of air at a rate of 4–10 Hz, which creates an internal vibration in the lung and thus promotes mucus clearance. IPV is recommended for weak and fatigued patients with ineffective cough to loose secretions, which is feasible and safe in spontaneously breathing and non-intubated subjects in critical care (Berlinski, 2019; Hassan et al., 2020). Additionally, IPV can be superimposed on conventional ventilation. The potential benefits of the concomitant use of IPV with ventilation include secretion clearance, improved oxygenation, and intermittent use (Dellamonica et al., 2008).

Lung hyperinflation clears airway mucus by increasing inspiratory flow volume. This will recruit collapsed airways, reduce inspiratory resistance, and thus generate high expiratory flow rate bias and induce annular flow to remove airway mucus, which is predominantly used to move airway mucus from the peripheral airway to the tracheal for mechanically ventilated patients (Savian et al., 2006).

It starts with a slow deep inspiration flow to deliver a high tidal volume up to a peak pressure of 40 cmH₂O, and follows by a 3-second breath holding, and then a fast uninterrupted expiratory flow that mimics a forced expiration (i.e., cough) (Berney and Denehy, 2002; Hodgson et al., 2007). This deep inspiration will increase tidal volume by 50% and increase peak intrabronchial pressure to 40 cmH₂O (Haake et al., 1987).

Such lung hyperinflation can be realized either manually, i.e., manual hyperinflation (MHI) or mechanically by a ventilator, i.e., ventilator hyperinflation (VHI). For mechanically ventilated COVID-19 patients, VHI is more suitable than MHI, since the patients are already connected to the ventilator and maintained with PEEP, and the outcome of VHI is more reproducible than that of MHI (Haake et al., 1987; Volpe et al., 2020). Furthermore, it has been clarified that among different modes of VHI, the volume-controlled ventilation with inspiratory flow of 20 L/min and pressure support achieves the best effect to remove airway mucus (Ribeiro et al., 2019).

Expiratory rib cage compression (ERCC), also known as manual chest compression, is one of the most commonly used ACTs in ventilated patients (Volpe et al., 2020). ERCC is usually applied either to assist secretion clearance from peripheral to central airways, or to remove secretion from central airways. Unlike the previously described percussion and vibration, ERRC is applied with a steady force that gradually increases from gentle to strong intensity to prolong exhalation after the onset of the expiratory phase, which is used to remove the secretions from distal airways. To remove the secretions from proximal airways, ERRC is applied with hard compression to increase PEF and synchronized with the onset of expiration (Borges et al., 2017; Guimaraes et al., 2013; Martí et al., 2013).

Positive end-expiratory pressure-zero end expiratory pressure (PEEP-ZEEP) consists of increasing PEEP to 15 cmH₂O during 5 cycles with peak inspiratory pressure limited to 40 cmH₂O, followed by an abrupt reduction of PEEP to 0 cmH₂O. By increasing the pressure change at the onset of the expiratory phase, this technique increases the PEF and, consequently, the expiratory flow bias, especially in the volume controlled mode (Amaral et al., 2020; Oliveira et al., 2019a; Volpe et al., 2020). Although both ERCC and PEEP-ZEEP are responsible for increasing the expiratory flow bias, the latter is more potent in this aspect (7 L/min vs. 49 L/min) (Oliveira et al., 2019a).

Mechanical insufflation-exsufflation (MI-E) is the most recently developed technique for mechanically ventilated patients, which delivers a positive inspiratory pressure to the patient to generate a deep inspiration (insufflation) and immediately (10 ms) followed by a negative expiratory pressure generating a deep expiration (exsufflation). In this way, a PEF greater than that induced by coughing can be achieved to remove airway mucus more effectively. In addition, the rapid switch from positive to negative pressure stimulates the bias flow. The advantage of MI-E is that it induces negative pressure throughout central and peripheral airways, suggesting that airway mucus in all generations of the airways can be removed (Ferreira de Camillis et al., 2018; Terzi et al., 2018).

Tracheal suctioning can clear sticky mucus from the trachea and the lower airways, which can be performed via an in-line closed system or an open system (Dodek et al., 2004). Although tracheal suctioning should not be used as a routine ACT for non-intubated patients, it is recommended for the mechanically ventilated COVID-19 patients with suction catheters of an in-line closed system, considering the risk of generating virus aerosol with an open system.