European Respiratory Society statement on airway clearance techniques in adults with bronchiectasis

Evidence overview

A total of 22 studies were identified, all meeting the inclusion criteria (supplement 1, figure S1 and table S2). Of these studies, nine studies were primary research [9, 21–28] and 13 were secondary research studies (one systematic review [29] and 12 narrative reviews [3, 30–40]). Data for 11 studies came from in vitro experiments [21, 22, 24–26, 32–34, 36, 38, 40] and three studies were experimental/clinical trials [21–23]. Six studies reported data from people with bronchiectasis [3, 9, 25, 27–29], five studies provided a mix of data from different respiratory diseases including bronchiectasis [24, 30, 35, 37, 39], seven studies outlined data from other respiratory diseases such as chronic obstructive pulmonary disease (COPD) and cystic fibrosis [26, 32–34, 36, 38, 40] and one study involved healthy adults [23].

Impaired mucociliary clearance in people with bronchiectasis

Impaired mucociliary clearance in people with bronchiectasis is often demonstrated through productive cough, abnormally high presence of sputum or difficulty in sputum expectoration. Abnormalities in mucus production, ciliary function and biophysical and surface mucus properties directly contribute to a decreased mucus clearance rate compared to healthy people [34, 37, 40]. There are limited data about the function of the airway surface layer in bronchiectasis and the hypothesis that mucus layer dehydration impairs mucus transport is derived from other chronic respiratory diseases [26, 38, 40]. Still, considering that sputum samples from people with bronchiectasis are abnormally hyper-concentrated (dehydrated), this could be a possible explanation [9].

First, neutrophil elastase activity plays an important role in the pathogenesis and progression of bronchiectasis [27, 29]. Excessive neutrophil elastase activity within the inflamed airway has been reported to decrease ciliary beat frequency and directly stimulates mucin secretion [27, 29]. Mucin 5B appears to be the most predominant mucin in bronchiectasis [9]. Mucin 5AC and higher airway mucin levels are associated with increased disease severity [9, 27]. However, these findings are based on only a few studies in bronchiectasis [9, 27] and further research is needed to confirm them.

An excess of secreted mucins leads to mucus layer dehydration and generates an osmotic imbalance between the mucus layer and the periciliary layer. This phenomenon ultimately compresses the periciliary layer and ciliary system [33]. Consequently, ciliary beating is slowed down and mucus layer adhesion to the airway epithelial surface is facilitated (adhesivity is a surface property of the sputum, defined as the ability to bond to a solid surface [37]); therefore, mucus transport is reduced, which results in mucus accumulation [26, 40]. This perpetuates the pathogenesis of bronchiectasis, which has lately been described as a vicious vortex [3]. In fact, a hyper-concentrated mucus layer can lead to local epithelial hypoxia, which may limit the action of cystic fibrosis transmembrane conductance regulator channels and produce higher levels of dehydration in the airway surface layer [25]. Most ciliary dysfunction in bronchiectasis is due to the effects of chronic inflammation, but the genetic condition primary ciliary dyskinesia, caused by more than 50 recognised gene defects affecting the structure and function of motile cilia, is increasingly recognised [39].

Second, when respiratory muscle strength is preserved as in the case of bronchiectasis, mucus adhesivity appears to be the strongest factor determining cough effectiveness for airway clearance [34, 37]. This property is independent of mucus viscosity (the loss of energy from an object through a substance and thus the resistance to flow) and mucus elasticity (the recoil energy transmitted back to an object) [32, 37]. Greater adhesivity appears when there is high interfacial tension between the mucus layer and the airway epithelium and/or low mucus wettability (the surface energy at a solid-sputum–air interface) [34, 37]. The limited available data from in vitro experiments suggest that mucus transport via coughing is impaired in bronchiectasis [25].

Consequently, the rationale for the use of ACTs in bronchiectasis is based on improving the biophysical and surface properties of the mucus layer to enhance the clearance of inflammatory markers and to help modulate the pathogenic microorganism load in the airways. Thus, ACTs aim to break the pathogenic vortex and slow down disease progression.

Physiological mechanisms to enhance mucociliary clearance

Mechanical stress applied to the airways could stimulate hydration of the mucus layer and enhance airway clearance [24, 36]. During normal breathing, two mechanical stresses are generated during both respiratory phases of inspiration and expiration, and they are essential for the normal regulation of airway surface hydration: the airflow and the trans-airway pressure gradient [24]. Previous studies reported that fluid shear stress, compression or stretch and osmotic shock are the main physical mechanisms that stimulate airway surface hydration [24]. Additionally, an in vitro flow model suggests two conditions that promote airway clearance [21, 22, 30]: the peak expiratory flow rate should be greater than the peak inspiratory flow rate (rate difference >10%) for mucus to move proximally, and a peak expiratory flow rate of 30–60 L·min⁻¹ is required to break the adhesive bonds generated between the mucus layer and the airway epithelial surface. Accordingly, airway clearance strategies are based on generating greater mechanical stress on the airways compared to normal breathing and the achievement of one of the above conditions may play an important role in improving airway clearance for people with bronchiectasis.

Achieving a sufficient volume of air behind the lung regions that are obstructed by mucus accumulation is another mechanism that may enhance mucus clearance [30]. Three different strategies have been described to achieve this: slow, deep inspirations to take advantage of the parenchymal interdependence and generate traction force to keep open or re-expand the smaller airways [31]; end-inspiratory breath-hold to reduce asynchronies in time constants between lung regions with different resistance or compliance constants [23]; and promotion of ventilation via collateral channels using adjacent lung units [30]. These mechanisms use the Pendelluft effect, which allows air to move into the lung units that are most obstructed by mucus accumulation [30].

In summary, evidence suggests that sputum-related symptoms result from increased mucus production, dehydration and impaired biophysical properties of mucus and reduced ciliary function due to primary and secondary ciliary dysfunction. The physiological mechanisms by which ACTs could enhance mucociliary clearance include improving the rate of mucus clearance by stimulating airway surface hydration, increasing the velocity of airflow and thus the air–mucus interaction, and facilitating a homogeneous distribution of ventilation. These mechanisms provide a physiological justification for the role of ACTs in bronchiectasis.

Question 1: Statements

• • Sputum from people with bronchiectasis is abnormally hyper-concentrated (dehydrated) and mucin concentration is related to disease severity. This indicates that the level of mucus layer dehydration plays an important role in the pathophysiology of the disease.

• • The main physiological mechanism that promotes mucus clearance involves mechanical stress, such as fluid shear stresses, compressions or stretching, and osmotic shocks. ACTs which implement these mechanisms of action have the potential to enhance mucociliary clearance in bronchiectasis, because they can potentially achieve a greater expiratory to inspiratory flow rate or direct a volume of air behind lung regions that are obstructed by mucus accumulation.

Question 1: Recommendations for research

• • Investigate how biophysical and surface sputum properties such as viscoelasticity, adhesivity and cohesivity (defined as the tendency for a gel to remain attracted to itself [37]) change across the disease trajectory in relation to underlying aetiologies or with different endotypes or phenotypes in people with bronchiectasis.

• • Investigate how biophysical and surface sputum properties influence the effectiveness of ACTs in people with bronchiectasis. Evaluate whether these sputum biomarkers could support the identification of good candidates/responders for specific ACTs in order to personalise airway clearance management.

• • Explore whether the order or specific combination of the physiological mechanisms described above can improve mucus clearance, specifically in people with bronchiectasis according to their disease aetiology, endotype or phenotype. This may help in selecting the most suitable ACTs or combination of ACTs in clinical practice.

Evidence overview

A total of 30 studies were identified, all meeting the inclusion criteria (supplement 1, figure S2 and table S3). Of these studies, 18 were primary research papers, including 14 clinical trials, comprising one randomised controlled trial [41], nine crossover trials [42–50] and four quasi-experimental trials [51–54]. Five studies provided data from in vitro or animal experiments [21, 22, 36, 51, 53]. Only four studies reported data exclusively from people with bronchiectasis [50, 55–57], nine studies reported mixed data from various respiratory diseases (e.g. bronchiectasis, COPD and cystic fibrosis) [30, 46, 58–64], 10 studies included patients with other respiratory diseases [36, 41–45, 48, 49, 52, 65] and one study reported data from healthy adults [47].

The physiological rationale for each of the following ACTs was considered: forced expiration technique (FET), active cycle of breathing techniques (ACBT), manual percussions, manual vibrations or shaking, autogenic drainage, slow expiration with glottis opened in lateral posture (ELTGOL), gravity-assisted drainage (GAD) technique, positive expiratory pressure (PEP) devices, positive expiratory pressure devices with oscillation (O-PEP), high-frequency chest wall oscillation (HFCWO) and intrapulmonary percussive ventilation (IPV). These techniques appear to achieve one or several of the physiological principles proposed to enhance sputum clearance: improving collateral ventilation and interdependence, increasing expiratory airflow velocity, reducing the total airway cross-sectional ratio, use of gravity, changing airway pressures and producing airway oscillations. Specific data on frequencies and flow rates achieved through ACTs in other patient populations can be found in McIlwaine et al. [30].

The identified ACTs have a range of advantages. For instance, many ACTs can be used independently by the patient, and they are portable and easy to learn. Common disadvantages include the need for some level of concentration when performing the techniques, as well as the need for instructions or training to ensure optimal execution, especially when access to a specialist respiratory physiotherapist is limited. If the performance of ACTs includes the use of a device, the need for cleaning and periodic replacement, noise and/or transport difficulties are the main disadvantages associated with its use. Table S4 in supplement 1 presents the physiological rationale for the ACTs and table 2 their advantages and disadvantages from the respiratory physiotherapists’ and patients’ perspectives. Although some techniques may be combined with others, each technique was reported separately. Further information on the procedures for performing ACTs is available in online resources that include videos and illustrations [66–68].

Question 2: Statements

• • The ACTs enhance sputum clearance by incorporating one or more of the following mechanism of actions: improving collateral ventilation and interdependence, increasing expiratory airflow velocity, reducing the total airway cross-sectional ratio, use of gravity, changing airway pressures and generating airway oscillations. Data specifically evaluating the above physiological principles in people with bronchiectasis are scarce.

• • The main advantages of specific ACTs are that they can be performed independently, they are feasible in different environments or can easily be implemented in a daily routine.

• • The main disadvantages of specific ACTs are the level of concentration and effort that is required to perform them, the need for cleaning and periodic replacement of devices, the noise or size of devices, difficulty of transport, the lack of biofeedback and the cost.

Question 2: Recommendations for research

• • Assess if the physiological mechanisms described for each ACT work specifically in people with bronchiectasis and what are the related physiological actions (pressure, frequencies, flow rate, etc.) in this population.

• • Establish whether the physiological effects of ACTs change depending on clinical status (i.e. clinical stability versus acute exacerbation) or disease severity (i.e. mild versus severe).

• • Conduct studies that involve people with bronchiectasis in their design and incorporate strategies that enhance the advantages and mitigate the disadvantages of ACTs in clinical practice.

Evidence overview

A total of 2934 studies were screened for eligibility and seven papers were included [69–75] (supplement 1, figure S3 and table S5). Five studies assessed the clinical use of ACTs via surveys [69–71, 73, 74], and two studies through audit [75] or registry data [72]. One registry recorded the ACTs during a clinically stable stage or exacerbation of the disease [72], and one survey only during exacerbations [74], while the other studies were conducted during a clinically stable stage or did not report this information. One study assessed ACTs during COVID-19 [73].

All surveys were administered to healthcare professionals [69–74], mainly physiotherapists (n=482) [69–71, 74]. In three studies that reported on survey response rates for healthcare centres and professionals, these ranged from 70% to 88% [69–71], while the response rate was only 0.5% in the survey that had the highest number of invited healthcare professionals (n=26 000) [74].

All studies assessed a variety of ACTs, apart from Santos et al. [71] that specifically assessed the use of different PEP devices. The ACTs were not always defined, but results were mainly presented as frequencies of use (figure 1). Based on six studies that compared a variety of ACTs, the most common routinely used ACT was the ACBT (range 48–91%). PEP or O-PEP (range 7–75%), techniques based on gravity such as GAD (range 8–76%) and modified GAD (range 10–55%), and techniques based on optimal positioning (range 35–84%) were also frequently reported. Other less-frequently used ACTs were manual percussion, deep breathing exercises, positions of ease (possibly for ease of breathlessness), manual and high-frequency vibrations, sustained maximum expiration, FET, autogenic drainage or other (figure 1 and supplement 1, figure S4).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1

Clinical use of airway clearance techniques (ACTs). The graph presents the ACTs that were reported as the most routinely used. The terminology of the graphs follows the terminology of the original studies, i.e. PEP versus PEP mask and Flutter versus O-PEP. GAD was presented as postural drainage in the original studies. The study by Santos et al. [71] was not included in the graphs because it reported on the frequency of a specific type of ACT: PEP devices. Results of this study, presented as percentage of physiotherapists, were PEP mask 2%, PEP mouthpiece 32%, PEP bottle 72%, Flutter 36% and Acapella 46%. Percentages do not add up to 100% because the physiotherapists could choose all ACTs that apply. ACBT: active cycle of breathing techniques; ELTGOL: slow expiration with glottis opened in lateral posture; GAD: gravity-assisted drainage; O-PEP: oscillatory positive expiratory pressure; PEP: positive expiratory pressure.

Evidence for the clinical use of ACTs was mainly available from Australia, and ACBT with FET or directed cough were the most frequently selected choices [70, 73–75]. Owing to the limited data from other countries, it was not possible to identify additional geographical patterns.

Question 3: Statements

• • There is limited evidence about the clinical use of ACTs in specific countries. Based on the available data (i.e. Australia, New Zealand, USA, Japan, UK), the ACBT is the most commonly used ACT in bronchiectasis. PEP, O-PEP and techniques based on the effect of gravity are also commonly used.

• • Studies reporting on the clinical use of ACTs do not always adequately describe the responding population and sample. They also do not always clearly define ACTs.

• • Data on the use of ACTs in clinical practice are scarce and some data are likely to be out of date given the progress in bronchiectasis management in the past decade.

Question 3: Recommendations for research

• • Surveys, audits and clinical registries need to assess the ACTs that are currently used in clinical practice in different geographic areas and investigate potential variations.

• • Surveys need to consistently report on the responder population and sample characteristics.

• • Studies need to clearly and adequately define ACTs in line with the clinical practice of each country, so as to enable future comparisons.

Evidence overview

A total of 1936 studies were screened for eligibility and 30 papers were included (supplement 1, figure S5). All included studies were randomised; 10 had a parallel group design and 20 had a crossover design. They were mainly short-term studies with a range from 1 day to 4 weeks, while two studies had a duration of 3 months and 12 months each [76, 77]. Overall, there were 811 participants (57% female), with mean age of 58 years and mean forced expiratory volume in 1 s (FEV₁) of 59% predicted. Most studies included patients during a clinically stable condition (n=22), three during an acute exacerbation [78–80] and one study during both clinical stages [81], although this was not clearly reported in four studies [82–85]. Results are presented in table 3.

Most included studies were active comparator studies, with one ACT technique compared to another. Most studies investigated the effectiveness of O-PEP; specifically, Flutter (n=13), Acapella (n=7), Aerobika (n=1), RC-Cornet (n=2), Quake (n=1), temporary positive expiratory pressure (n=1), Lung Flute (n=1) and bubble PEP (n=2) were studied in different or the same trials. Studies also assessed GAD (n=13), ACBT (n=11), manual techniques (i.e. percussions/vibrations) (n=9), ELTGOL (n=4), autogenic drainage (n=4), PEP (n=2) and HFCWO (n=1). GAD and manual techniques were used alone or in combination with other ACTs, also referred to as conventional physiotherapy treatment. Most techniques were self-administered by the patients at the hospital, predominantly after training by a respiratory physiotherapist or another experienced healthcare professional. Alternatively, clinical supervision was provided in each session or selected ones.

The effectiveness of the techniques was assessed using function and disability outcomes, while there were limited trials that used activity and participation measures (table 3). Sputum volume or weight during or after treatment (n=30), patient-reported preference and comfort (n=16), dyspnoea (n=11) and HRQoL (n=7) were the most common outcome measures used, and most studies did not report on patients’ adherence to treatment. Three studies included the number of coughs or presence of cough as one of their secondary outcomes [86–88] and only two studies used frequency of exacerbations or time to first exacerbation [76, 77]. There is no evidence about the optimal frequency or the number of sessions needed to ensure correct procedure of the ACTs. The studies that assessed ACT adverse events did not identify serious adverse effects that are related to ACTs.

Question 4: Statements

• • Although data on the effects of performing ACTs for periods over 6 or 12 months are limited, the findings demonstrate a reduction in the impact of cough, improvement in HRQoL and reduction in the risk of exacerbations. These findings support previously published clinical recommendations for the use of ACTs as part of bronchiectasis management in adults [2, 11–14]. However, there is no evidence about the optimal frequency or number of ACT sessions.

• • Randomised controlled trials have assessed a variety of ACTs, with O-PEP devices (mainly via Flutter and Acapella), GAD and ACBT being the most commonly studied techniques. The existing literature does not demonstrate superiority of one technique over another but supports the use of ACTs.

• • Wet sputum weight or volume were the most commonly used outcome measures. The ACTs increase the expectorated sputum during or following a single session of ACTs. Despite being frequently used in clinical practice, the interpretation of sputum changes is ambiguous.

• • To date, there are no studies that have investigated the effect of ACTs on mortality or changes in disease severity using the bronchiectasis severity index or FACED. There are also no studies providing a health economics estimation for ACTs in bronchiectasis.

Question 4: Recommendations for research

• • Investigate the effectiveness of ACTs using large-scale and prospective randomised controlled trials, particularly during acute exacerbations.

• • Assess the effect of ACTs in the long term, particularly in reducing exacerbations, hospitalisations, bronchiectasis disease severity and mortality. A follow-up of at least 6 months needs to be implemented in these studies.

• • Assess the cost-effectiveness of ACTs based on direct and indirect costs, such as savings on medications and hospitalisation compared to therapist time and equipment expenses.

• • Consider including patient adherence and disease-specific HRQoL questionnaires as a primary or secondary outcome in all clinical trials.

• • Include alternative assessment tools and outcomes for the ACTs studies, such as impulse oscillometry for pulmonary function, lung clearance index for ventilation impairment, magnetic resonance or high-resolution computed tomography imaging and airway inflammatory markers or changes in airway microbiota.

• • Identify the optimal frequency of ACTs and factors that enhance accessibility to physiotherapy, such as home techniques and telehealth.

Evidence overview

A total of 22 studies met the inclusion criteria (supplement 1, figure S6); nine studies were included for examining participant experience and impact on symptoms [79, 82, 89–91, 93, 105–107], 17 studies examined preference for techniques or adherence [76, 78, 82, 83, 86, 87, 89, 90, 97, 99, 100, 102, 104, 105, 107–109] and one study explored participant satisfaction with ACTs [110]. Two studies were conducted in people experiencing an acute exacerbation [78, 79], while the remaining studies were in clinically stable individuals [82, 83, 86, 87, 89, 90, 93, 97, 99, 100, 102, 104, 105, 107] or the clinical state was unspecified [106, 108–110]. The instruments used to evaluate patient satisfaction and perceived experience were an adapted questionnaire that was validated in cystic fibrosis [110], visual analogue scales [91, 93, 107], Likert scales [79, 82, 86, 90, 106], tools developed by the authors [79, 82, 86, 90, 106] or the instrument was unspecified [89]. Adherence rates were measured by diary card recording [76, 104], while assessment of patients’ preference was heterogeneous and included Likert scales and standard questionnaires [78, 82, 86, 87, 99, 100, 102, 108], visual analogue scales [97, 107] and asking the subject to indicate the preferred ACT [89, 90], or were not reported [105, 109].

There are mixed reports related to patient satisfaction, preference, experience related to symptoms and perceived impact of ACTs (table 4). Three studies focused on patient satisfaction after a single ACT session or following an unclear duration. A cross-sectional study [110] evaluated patient satisfaction for a mix of ACTs that did not require equipment (manual-assisted or self-administered) and O-PEP. Efficacy, convenience, comfort, satisfaction and cost-effectiveness were rated highly for all techniques. It was proposed that conventional chest physiotherapy (GAD, manual vibrations or percussions) may be easy to learn, cost-effective and convenient for home use, but ACBT and O-PEP were highly rated due to the patient's active participation, independence, convenience and/or device portability.

Ten studies [76, 86, 87, 90, 97, 99, 102, 104, 105, 109] compared O-PEP (Acapella or Flutter) to a mix of ACTs that did not require equipment (ACBT with or without GAD and manual vibration or percussions, autogenic drainage and ELTGOL) or a control condition in individuals in a stable clinical state. A short-term study showed that most participants preferred autogenic drainage (49%) followed by ELTGOL (35%) [99]. In contrast, Semwal et al. [97] identified a lower preference for autogenic drainage over O-PEP (Acapella), which may be linked to the complexity of autogenic drainage. Eaton et al. [90] found that ACBT-GAD was considered more valuable at clearing sputum compared to ACBT, but was less preferable compared to O-PEP (Flutter). This finding was also demonstrated in three other studies, with their treatment ranging from single sessions to 4 weeks of ACBT-GAD or O-PEP (Acapella, Flutter or bottle PEP), with a greater preference for O-PEP [87, 102, 109]. In contrast, a recent study found O-PEP (bottle PEP) more useful for clearing secretions than ACBT (seated position) or no therapy, but still more tiring and with similar levels of discomfort to ACBT [86]. A different study [105] reported no difference in time consumption, tediousness or need for additional training between O-PEP (bottle PEP) and ACBT.

Two studies recruited people experiencing an acute exacerbation of bronchiectasis [78, 79]. GAD with breathing control, O-PEP (Flutter) with breathing control, and breathing control only were all perceived to be similar in ease of application. GAD and O-PEP were reported as equally effective and superior to breathing control for clearing secretions [79]. Patients in the same clinical state demonstrated a greater preference for O-PEP (Acapella) when it was newly introduced versus their usual ACT, with a proportion of patients still using the O-PEP device daily at 1 month follow-up [78].

Seven studies [82, 83, 89, 91, 93, 100, 106] compared a combination of techniques that were administered simultaneously (GAD, manual vibrations or percussions and/or ACBT) or equipment techniques (O-PEP, HFCWO, IPV). A preference for the RC-Cornet over Acapella, based on usefulness for clearing secretions, convenience, comfort and performance [83], and Flutter over the Lung Flute based on usefulness for clearing secretions, convenience, comfort and performance [100], was demonstrated in single studies. The Acapella was perceived to be more helpful in clearing secretions and a preferred technique compared to inspiratory muscle training [82], but ACBT-GAD was patient-reported as more effective compared to inspiratory muscle training, despite an equal patient preference for both techniques [89]. Moreover, both IPV and combination of GAD, FET or manual percussion and vibrations achieved patient-reported improvements in subjective perception of sputum expectoration [93]. Post-treatment discomfort was lower with IPV, which may be related to frequent position changes incorporated into GAD. Following a year of HFCWO use, there were improvements in subjective ratings of respiratory health and ability to clear secretions, compared to not using this treatment [106]. In the only comparison between techniques not requiring any equipment, Syed et al. [91] found greater comfort with ACBT with GAD compared to GAD with deep breathing and manual vibrations and percussions, which may influence compliance for the treatment.

Evidence overview

Five studies addressed barriers and/or enablers together with adherence to ACT in a mix of study designs, including cohort [111, 112] and qualitative studies [113–115] (supplement 1, figure S6). In all studies, participants were in a clinically stable state. A study for predictors of adherence measured the compliance to ACTs over a 12-month period [112]. A total of 41% of patients self-reported adherence to ACT. Those who were adherent to ACTs had a better Physical Function domain score on the Quality of Life-Bronchiectasis questionnaire (QoL-B) compared to those who were nonadherent (mean±sd: 42±28 versus 29±26, respectively) [112]. Higher adherence to ACT was associated with lower Treatment Burden domain score on QoL-B (regression coefficient (95% CI): −15.46 (−26.54 to −4.37)) and lower Respiratory Symptoms domain score on QoL-B (regression coefficient (95% CI): −10.77 (−21.45 to −0.09)) [112]. This cohort who reported using ACBT (53%) or O-PEP (Acapella) (61%) also completed a modified Beliefs about Medicine questionnaire specific to ACTs. In determining independent predictors of adherence, those adherent to ACT (41%) were older (OR (95% CI): 2.94 (2.74–3.18)), based on a 10-year increase in age, and believed their ACT was necessary (OR (95% CI): 1.3 (1.1–1.53)). Those with fewer concerns about treatment were also more likely to be adherent to ACTs [111].

Three qualitative studies with patients and clinicians, including respiratory physicians, respiratory physiotherapists and nurses, described barriers and enablers to ACTs [113–115]. From the patient perspective, identified barriers were late referral to the multidisciplinary team, lack of engagement with a healthcare professional [114], lack of perceived health benefit, lack of motivation and time commitment required [113]. Enablers were working with a multidisciplinary team, which incorporates chronic disease management and support; recognition of the patient role in management and their substantial burden of disease; and a personalised approach to therapy [113, 114]. From the clinicians’ perspective, barriers to management were availability of resources for ACTs, time and space restrictions, and funding. Enablers were working with the multidisciplinary team and using a chronic disease approach, as well as patient engagement [115].

Question 5: Statements

• • Patient experience was generally well rated for ACTs. Preference was mainly based on the independence of technique, patient satisfaction with symptom relief and perceived efficacy or difficulty.

• • Patient adherence to ACTs could be related to older age, good physical function, milder respiratory symptoms, less treatment burden and belief in treatment necessity.

• • Optimal engagement of patient and healthcare professionals, adequate motivation, time and resources were some of the enablers of and barriers to ACTs.

Question 5: Recommendations for research

• • Further investigate the barriers to and enablers of ACTs from the patients’ and healthcare professionals’ perspectives, and the factors that influence patient preference and adherence to treatment, in qualitative studies. Within this topic, examine the patients’ perspectives upon changing techniques, through mixed-methods study designs.

• • Investigate the barriers to and enablers of using ACTs in varying geographical locations and the underlying training and clinical experience of therapists that may be challenges to or facilitators of ACT therapy.

• • Use standardised patient-reported outcome measures (PROMs) and patient-experience outcomes measures (PREMs), including patients’ perceptions and preferences. To allow standardisation and comparison among studies, there are available disease-specific PROMs but there is a need to develop and validate PREMs in bronchiectasis.

• • Conduct pragmatic trials, which consider the patients’ preference, experience and satisfaction of ACT prescription in their design. For instance, trials investigating the effectiveness of two or more interventions could use stratified randomisation based on patient preferences, thus providing a more accurate reflection of real life.

• • To facilitate clinical implementation and patient adherence, standardised tools that assess patient-related factors such as discomfort, fatigue, ease of performing ACTs, perceived impact of treatment effect and preference should be considered in ACT trials. Future studies need to embed patients’, stakeholders’ and public’s perspectives in their design and delivery, through the patient and public involvement or study co-production.

Evidence overview

In total, 34 randomised trials (30 full papers and four abstracts) were included (supplement 1, figure S7). The majority were crossover studies (n=21), conducted at a single centre (n=30) in European countries (n=13). A total of 915 patients with bronchiectasis were included. Gender was reported by 28 studies, with females (n=445; 59%) being more represented than males (n=306; 41%). The mean age range was 39–75 years, while lung function (FEV₁) ranged from 29% to 96% predicted. Most studies included patients in a clinically stable condition (n=30), who had a productive cough or self-reported sputum expectoration; when reported, daily sputum volume ranged from 1.2 mL to 132 mL.

Sample size estimation was reported in 16 of the full-text studies (53%) (range 8–68 participants), with only three full-text studies including possible dropout rates (range 20–25%) in their sample size calculation. Primary end-points were clearly reported in 20 studies (59%), sputum quantity being the most frequent outcome measure used (wet sputum weight, n=8 (24%); dry sputum weight, n=2 (6%); wet sputum volume, n=7 (21%)). Sputum was collected during the ACTs intervention in seven studies (21%) and post-intervention in 10 studies (30%). Lung function, sputum quantity, HRQoL, symptoms (particularly breathlessness) and patients’ feedback were other common outcomes (supplement 1, table S6).

The risk of bias of the included studies is presented in figure 2. Only one study had low risk of bias [76] and when considered across studies, none of the risk assessment domains were free of bias (supplement 1, table S7). The remaining 33 trials did not consistently report sufficient information to adequately assess risk of bias. For the domains of the Cochrane Risk of Bias Tool that could be assessed as high or low risk of bias (i.e. not unclear), there were six assessment domain scores that had a high risk of bias and 78 that had a low risk of bias.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2

The Cochrane Collaboration's tool for assessing the risk of bias, based on reviewer's judgements for each risk of bias item and presented as percentages across all included studies (n=34).

On the risk of bias, allocation concealment, blinding and selective reporting were frequently classified as unclear (figure 2). Most studies failed to provide sufficient information about the method used to conceal the allocation sequence (selection bias) and most trials did not blind participants nor personnel (performance bias), although we need to acknowledge that the Cochrane tool was designed for placebo-controlled drug studies. Most studies combined objective and subjective (e.g. self-reported) outcome measures, which is a strong point. Nevertheless, data collection procedures for the subjective outcomes were often unclear. Selective reporting was classified as unclear risk of bias in most studies owing to insufficient information. Eight trials were included on a clinical trial registry and were classified as low risk because their results reported all primary outcomes and most secondary outcomes [76, 77, 94, 96, 99, 103, 104, 117].

Question 6: Statements

• • The risk of bias amongst the studies that assess ACTs was heterogeneous, but generally unclear.

• • For most studies, reporting was unclear for allocation concealment or there was selective reporting.

• • Blinding of the ACTs was also limited for patients and personnel, although this is often challenging due to the nature of the intervention.

• • Futures studies should be adequately powered, based on sample size estimation of one or two primary outcome measures, which have well-explored psychometrics properties. Blinding of outcome assessment and statistical analysis of the ACTs should be implemented to help minimise bias. Study reporting should be clear and following the CONSORT reporting guidelines [19].

Question 6: Recommendations for research

• • Ensure that study reporting is clear and facilitates risk of bias assessment by following the CONSORT reporting guidelines. To reduce the risk of reporting bias, we recommend registering the studies in clinical trial registries or publishing their protocol in clinical journals, according to SPIRIT reporting guidelines [118].

• • To minimise bias and maximise the validity of the results, trials should blind as many individuals as possible. Blinding of outcome assessment and statistical analysis of the ACTs is usually feasible and needs to be implemented. Where possible, studies should use blinding of the investigators who direct or supervise the treatment and/or the patients who perform ACTs. To achieve this last point, sham interventions, placebo-controlled designs for treatment-naive patients or cluster trials that include settings where ACTs are not part of standard care can be implemented.

• • Studies should be adequately powered, based on sample size estimation of one or two primary outcome measures. Recruitment from multiple centres that can follow standardised procedures may be an optimal strategy.

• • Future trials that use a core set of outcome measures, with well-explored psychometric properties according to COSMIN [119], could simplify future meta-analyses and support stronger conclusions. Current suggestions for core outcomes in bronchiectasis [116] together with exploration of core outcomes that are specific to physiotherapy [120] should be considered in future trials.