Mechanical insufflation-exsufflation for people with neuromuscular disorders

  • Review
  • Intervention

Authors

  • Brenda Morrow,

    Corresponding author
    1. University of Cape Town, Department of Paediatrics, Cape Town, South Africa
    • Brenda Morrow, Department of Paediatrics, University of Cape Town, 5th Floor ICH Building, Red Cross Memorial Children's Hospital, Klipfontein Road, Rondebosch, 7700, Cape Town, South Africa. Brenda.morrow@uct.ac.za.

    Search for more papers by this author
  • Marco Zampoli,

    1. Red Cross War Memorial Children's Hospital and University of Cape Town, Pulmonology, and Paediatric Medicine, Cape Town, South Africa
    Search for more papers by this author
  • Helena van Aswegen,

    1. University of the Witwatersrand, Physiotherapy Department, Faculty of Health Sciences, Johannesburg, South Africa
    Search for more papers by this author
  • Andrew Argent

    1. Red Cross War Memorial Children's Hospital and University of Cape Town, Pediatric Intensive Care, Division of Pediatric Critical Care and Children's Heart Disease, Cape Town, South Africa
    Search for more papers by this author

Abstract

Background

People with neuromuscular disorders (NMDs) may have weak respiratory (breathing) muscles which makes it difficult for them to effectively cough and clear mucus from the lungs. This places them at risk of recurrent chest infections and chronic lung disease. Mechanical insufflation-exsufflation (MI-E) is one of a number of techniques available to improve cough efficacy and mucus clearance.

Objectives

To determine the efficacy and safety of MI-E in people with NMDs.

Search methods

On 7 October 2013, we searched the following databases from inception: the Cochrane Neuromuscular Disease Group Specialized Register, CENTRAL (The Cochrane Library), MEDLINE, and EMBASE. We also searched ClinicalTrials.gov and the WHO International Clinical Trials Registry Platform for ongoing trials. We conducted handsearches of reference lists and conference proceedings.

Selection criteria

We considered randomised or quasi-randomised clinical trials, and randomised cross-over trials of MI-E used to assist airway clearance in people with a NMD and respiratory insufficiency. We considered comparisons of MI-E with no treatment, or alternative cough augmentation techniques.

Data collection and analysis

Two authors independently assessed trial eligibility, extracted data, and assessed risk of bias in included studies according to standard Cochrane methodology. The primary outcome was mortality throughout follow-up or at six months follow-up.

Main results

Five studies with a total of 105 participants were found to be eligible for inclusion in this review. All included trials were short-term studies (two days or less), measuring immediate effects of the interventions. There was insufficient detail in the reports to assess methods of randomisation and allocation concealment. All five studies were at a high risk of bias from lack of blinding. The studies did not report on mortality, morbidity, quality of life, serious adverse events or any of the other prespecified outcomes. One study was a randomised cross-over trial conducted over two days, in which investigators applied two interventions twice daily in randomly assigned order, with a reverse cross-over the following day. Four studies applied multiple interventions for cough augmentation to each participant, in random order. One study reported fatigue as an adverse effect of MI-E, using a visual analogue scale. Peak cough expiratory flow (PCEF) was the most common outcome measure and was reported in four studies. Based on three studies, MI-E may improve PCEF compared to an unassisted cough. All interventions increased PCEF to the critical level necessary for mucus clearance. The included studies did not clearly show that MI-E improves cough expiratory flow more than other cough augmentation techniques. Based on one study, which was at risk of assessor bias, the addition of MI-E may reduce treatment time when added to a standard airway clearance regimen with manually assisted cough. MI-E appeared to be as well tolerated as other cough augmentation techniques, based on three studies which reported comfort visual analogue scores.

Authors' conclusions

The results of this review do not provide sufficient evidence on which to base clinical practice as we were unable to address important short- and long-term outcomes, including adverse effects of MI-E. There is currently insufficient evidence for or against the use of MI-E in people with NMDs. Further randomised controlled clinical trials are needed to test the safety and efficacy of MI-E.

Résumé scientifique

L'insufflation-exsufflation mécanique chez les patients souffrant de troubles neuromusculaires

Contexte

Les patients atteints de troubles neuromusculaires (NMD) possèdent parfois des muscles respiratoires (la respiration) affaiblis, ils rencontrent ainsi des difficultés pour tousser ou dégager le mucus dans les poumons. Cela accroit le risque d'infections pulmonaires répétées et de maladies pulmonaires chroniques. L'insufflation-exsufflation mécanique (IE-M) est l'une des nombreuses techniques disponibles pour améliorer l’efficacité de la toux et l’élimination du mucus.

Objectifs

Déterminer l'efficacité et l'innocuité de l’IE-M chez les patients atteints de NMD.

Stratégie de recherche documentaire

Le 7 octobre 2013, nous avons effectué des recherches dans les bases de données suivantes depuis leur date de création : le registre spécialisé du groupe Cochrane sur les affections neuromusculaires, CENTRAL (The Cochrane Library), MEDLINE et EMBASE. Nous avons également effectué des recherches dans ClinicalTrials.gov et le système d’enregistrement international des essais cliniques de l’organisation mondiale de la santé pour les essais en cours. Nous avons effectué des recherches manuelles dans des références bibliographiques et des actes de conférence.

Critères de sélection

Nous avons pris en compte les essais cliniques randomisés ou quasi-randomisés et les essais croisés randomisés d’IE-M utilisés pour aider le dégagement des voies respiratoires chez les patients souffrant d’insuffisance respiratoire et de NMD. Nous avons pris en compte les comparaisons d’IE-M à l'absence de traitement, ou à d'autres techniques d'augmentation de la toux.

Recueil et analyse des données

Deux auteurs ont indépendamment évalué l'éligibilité, extrait les données et évalué le risque de biais dans les études incluses selon la méthodologie standard Cochrane. Le critère de jugement principal était la mortalité durant tout le suivi ou à six mois de suivi.

Résultats principaux

Cinq études avec un total de 105 participants ont été identifiées comme étant éligibles pour être incluses dans cette revue. Tous les essais inclus étaient des études à court terme (deux jours ou moins), mesurant les effets immédiats des interventions. Les rapports ne fournissaient pas suffisamment de détails pour évaluer les méthodes de randomisation et d'assignation secrète. Les cinq études présentaient un risque de biais élevé dû au manque de masquage. Les études n'ont pas rapporté sur la mortalité, la morbidité, la qualité de vie, les effets indésirables graves, ni sur aucun des autres critères de jugement prédéfinis. Une étude était un essai croisé randomisé réalisé sur deux jours, dans lequel les investigateurs appliquaient deux interventions deux fois par jour de manière aléatoire, avec un inversement croisé le jour suivant. Quatre études appliquaient de multiples interventions afin d’augmenter la toux, ceci pour chaque participant et en ordre aléatoire. Une étude rapportait la fatigue comme un effet indésirable d’IE-M, à l'aide d'une échelle visuelle analogique. Le débit expiratoire de pointe (DEP) à la toux était le critère de jugement le plus couramment rapporté dans quatre études. Sur la base de trois études, l’IE-M pourrait améliorer le DEP à la toux par rapport à une toux sans assistance. Toutes les interventions augmentaient le DEP à la toux au niveau critique nécessaire pour l'élimination du mucus. Les études incluses n'ont pas clairement montré que l’IE-M améliore le débit expiratoire de la toux plus que d'autres techniques d'augmentation de la toux. Sur la base d'une étude, qui présentait un risque de biais, l'ajout de l’IE-M pourrait réduire la durée du traitement lorsqu'elle est ajoutée à un schéma posologique standard de dégagement des voies respiratoires d’une toux assistée manuellement. L’IE-M semble être aussi bien tolérée que d'autres techniques d'augmentation de la toux, sur la base de trois études qui rapportaient l’échelle visuelle analogique.

Conclusions des auteurs

Les résultats de cette revue ne fournissent pas suffisamment de preuves susceptibles d'orienter la pratique clinique car nous n'avons pas été en mesure d’apporter de résultats importants à court et à long terme, y compris les effets indésirables des IE-M. Il n'existe actuellement pas suffisamment de preuves pour favoriser ou non l'utilisation d’IE-M chez les patients atteints de NMD. D'autres essais cliniques contrôlés randomisés sont nécessaires pour tester l'efficacité et l'innocuité d’IE-M.

Plain language summary

Mechanical insufflation-exsufflation to improve mucus clearance in people with neuromuscular disorders

Review question

Our aim in this review was to find out how well mechanical insufflation-exsufflation (MI-E) works and how safe it is to use in people with neuromuscular disorders (diseases of the peripheral nerves or muscles) who have breathing problems.

Background

People with neuromuscular disorders (NMDs) sometimes have weak breathing muscles. This can make it difficult for them to cough and clear mucus from the lungs well, putting them at risk of repeated chest infections and ongoing lung disease. MI-E is one of a number of methods used to improve cough and mucus clearance. MI-E is given through a mask, mouthpiece, or via a tracheostomy (an opening in the neck into the windpipe). MI-E acts like a cough by first pushing air into the lungs when the person breathes in (insufflation), then sucking it out again (exsufflation).

Methods

We carried out a wide database search for trials of MI-E in people with NMDs. We only included trials in which people were assigned to the treatments by chance, as these studies provide the best quality evidence.

Results and quality of the evidence

We found five trials, with 105 people. They all studied the immediate effects of a single treatment with MI-E. The studies compared MI-E to other ways of helping people cough, or normal cough without help. One trial studied MI-E when added to other treatment. Based on three trials, MI-E may improve the outwards flow of air during coughing compared to a normal cough without help. MI-E was not clearly better than other methods of improving cough. None of the studies measured the outcomes that we thought were important for making decisions about the usefulness of MI-E. For example, the studies did not report on survival, length of hospital stay, quality of life, or serious side effects. One study reported extreme tiredness as a side effect of MI-E. There was often not enough information in the reports to tell whether the studies were well run; in some we found design problems that could have affected the results.

The findings of this review do not give enough evidence on which to make decisions. We were unable to find any information from trials on important short- and long-term effects, including side effects of MI-E in NMDs.

There is currently insufficient evidence for or against the use of MI-E to help people with NMDs clear mucus from their lungs. Further studies are needed to better understand the benefits and risks of MI-E in relation to other methods of cough assistance.

The evidence in the review is up to date as of 7 October 2013.

Résumé simplifié

L'insufflation-exsufflation mécanique pour améliorer l'élimination du mucus chez les patients souffrant de troubles neuromusculaires

Question de la revue

L’objectif de cette revue était de déterminer l'efficacité de l’insufflation-exsufflation mécanique (IE-M) et si elle est sûre pour être utilisée chez les patients atteints de troubles neuromusculaires (maladies des nerfs périphériques ou des muscles) qui ont des problèmes respiratoires.

Contexte

Les patients atteints de troubles neuromusculaires (NMD) possèdent parfois des muscles respiratoires affaiblis. Ils peuvent ainsi rencontrer des difficultés pour tousser ou dégager le mucus dans les poumons, ce qui accroit le risque d'infections pulmonaires répétées et de maladies pulmonaires chroniques. L’IE-M fait partie d’une des nombreuses méthodes utilisées pour améliorer la toux et l'élimination du mucus. L’IE-M est administrée au moyen d'un masque, d’un embout buccal, ou via une trachéostomie (une ouverture dans le cou à l’intérieur de la trachée). L’IE-M agit comme une toux en poussant tout d’abord de l'air dans les poumons lorsque la personne respire (insufflation), puis en l'éliminant de nouveau vers l’extérieur (exsufflation).

Méthodes

Nous avons réalisé une vaste recherche dans les bases de données d'essais d’EI-M chez les patients atteints de NMD. Nous avons uniquement inclus les essais dans lesquels les patients ont été assignés aux traitements de façon hasardeuse, car ces études fournissent les meilleures preuves de qualité.

Les résultats et la qualité des preuves

Nous avons trouvé cinq essais, totalisant 105 personnes. Tous les essais étudiaient les effets immédiats du traitement unique avec l’IE-M. Les études ont comparé l’IE-M à d'autres méthodes visant à aider les personnes à tousser sans assistance. Un essai étudiait l’IE-M lorsqu' elle était combinée à un autre traitement. Sur la base de trois essais, l’IE-M peut améliorer le flux d'air dirigé vers l’extérieur lors de la toux par rapport à une toux normale sans assistance. L’IE-M n'était manifestement pas plus efficace que d'autres méthodes pour améliorer la toux. Aucune des études n’a mesuré les critères de jugement que nous considérions être importants pour la prise de décision concernant l'utilité de l’IE-M. Par exemple, les études n'avaient pas rendu compte de la survie, de la durée de séjour à l'hôpital, de la qualité de vie ou des effets secondaires graves. Une étude a rapporté une fatigue extrême comme un effet secondaire de l’IE-M. Les rapports ne fournissaient souvent pas suffisamment d'informations pour savoir si les études étaient bien exécutées; dans certains rapports, nous avons trouvé des problèmes de conception qui pourraient avoir affecté les résultats.

Les résultats de cette revue ne fournissent pas suffisamment de preuves permettant de prendre des décisions. Nous ne sommes pas parvenus à trouver d'informations issues d'essais sur d'importants effets à court et à long terme, y compris les effets secondaires de l’IE-M dans les NMD.

Il n'existe actuellement pas suffisamment de preuves pour savoir si l’IE-M aide ou non les patients atteints de NMD à dégager le mucus dans les poumons. D'autres études sont nécessaires pour mieux comprendre les bénéfices et les risques de l’IE-M par rapport à d'autres méthodes assistant la toux.

Les preuves de la revue sont à jour le 7 octobre 2013.

Notes de traduction

Traduit par: French Cochrane Centre 14th January, 2014
Traduction financée par: Financeurs pour le Canada : Instituts de Recherche en Santé du Canada, Ministère de la Santé et des Services Sociaux du Québec, Fonds de recherche du Québec-Santé et Institut National d'Excellence en Santé et en Services Sociaux; pour la France : Ministère en charge de la Santé

Background

Description of the condition

There are many types of neuromuscular disorder (NMD), both acute and chronic, hereditary and acquired (Gozal 2000). The five major groups into which hereditary NMDs can be classified are: muscular dystrophies, congenital and metabolic myopathies, neuromuscular junction disorders, peripheral neuropathies and anterior horn cell diseases (Gozal 2000).

People with a NMD run the risk of significant morbidity and mortality from acute respiratory tract infections and chronic respiratory insufficiency as a consequence of diaphragmatic or intercostal muscle weakness or both and progressive chest deformities (Boitano 2006; Finder 2010; Gozal 2000; Panitch 2009). Weak cough is an important factor contributing to respiratory morbidity in patients with NMDs (Boitano 2006).

Despite the heterogeneity in respiratory pathophysiology amongst the different NMDs, it is accepted that two main factors influence the progression of respiratory insufficiency: respiratory muscle strength and thoracic cage abnormalities. These factors are also affected developmentally by advancing age. Infants with a NMD have more cartilage in their rib cages than other infants, which results in increased chest wall compliance, more than twice that of controls (Papastamelos 1996). In addition, intercostal muscle weakness contributes to ribcage deformity, which further impacts on respiratory efficiency (Panitch 2009). Progessive pulmonary impairment, in terms of reduction in total lung capacity and forced vital capacity (FVC), occurs with progressive respiratory muscle weakness. Postural deformities, such as kyphosis, scoliosis and spinal rigidity, as well as shortening and fibrosis of the chest wall muscles due to an inability to fully expand the chest, result in a progressive decrease in chest wall compliance leading to a restrictive pattern of disease (Fauroux 2008; Gozal 2000; Panitch 2009). Microatelectasis from breathing at low lung volumes, secretion retention (Fauroux 2008), and the loss of sigh capacity (Bach 2000) cause further loss of lung compliance (Panitch 2009; Sharma 2009). Although total lung volumes are decreased in NMDs, residual volume may be preserved or even increased as a result of preferential expiratory muscle weakness (Gozal 2000).

An effective cough is essential for the clearance of pulmonary secretions, during both respiratory infections and stable periods. Components of the cough which may be missing in people with NMDs include an inability to take a deep inspiration to up to 80% of the vital capacity, failure to close the glottis in certain conditions (Bach 2003), and insufficient expiratory flow rates (Finder 2010). In addition to an impaired cough, other causes of impaired secretion clearance include the presence of infection with altered sputum viscosity, difficulty swallowing (dysphagia) and gastro-oesophageal reflux (Bannister 1985; Finder 2010; Iannaccone 2007). 

Retention of pulmonary secretions leads to airway obstruction, increased work of breathing, hypoxia and ultimately respiratory failure. Long-term retention of secretions may predispose to lower respiratory tract infections, atelectasis and chronic lung disease (Homnick 2007). The vast majority of episodes of respiratory failure in patients with muscular dystrophy are reported to be a result of ineffective coughing during intercurrent chest infections (Bach 2003). Identification of effective, safe measures to optimise cough efficacy is therefore key to improving quality of life and preventing morbidity in people with NMDs.

Description of the intervention

A number of physiotherapy techniques aimed at mobilising secretions and increasing lung volumes are used to assist airway clearance in people with NMDs. Manual techniques to assist airway clearance include postural drainage, chest wall shaking, percussion and vibrations; respiratory muscle training with inspiratory or expiratory resistance or both in early disease. Techniques using different breathing patterns include the active cycle of breathing technique, forced expiratory technique, and autogenic drainage; and positive pressure therapy including the use of flutter valves, positive expiratory pressure (PEP) therapy, intermittent positive pressure breathing (IPPB), and continuous positive airways pressure (CPAP) (Anderson 2005; Bott 2009; Finder 2010). Most PEP devices are effort-dependent and therefore not useful in people with severe respiratory muscle weakness (Finder 2010).

For successful secretion clearance, one needs both secretion mobilisation and effective cough (Finder 2010). Cough augmentation may be achieved by several manual or mechanical methods (Finder 2010). When voluntary deep breathing becomes impossible, breath stacking or glossopharyngeal breathing and mechanical or manual (bagging) hyperinflation provide sufficient inspiratory lung volumes (Bott 2009). Manual chest compressions or abdominal thrusts and mechanical exsufflation achieve cough augmentation by improving expiratory flow rates (Anderson 2005; Finder 2010). Suctioning may be needed in patients unable to clear secretions, but this intervention is generally uncomfortable and poorly tolerated (Bott 2009). If an artifical airway is present, suctioning is associated with significant complications, including hypoxia, changes in blood pressure and cerebral blood flow, cardiac arrhythmia, increased intracranial pressure, mucosal trauma and atelectasis (Anderson 2005; Morrow 2008).

Mechanical insufflation-exsufflation (MI-E) was first described in 1952 and used for patients with poliomyelitis (Barach 1952). Initially, commercially available in-exsufflators used high negative pressures (tank devices) for insufflation followed by rapid assisted exsufflation to atmospheric pressures (Barach 1952). Following the advent of invasive positive pressure ventilation, this form of negative pressure MI-E was discontinued but the technique regained popularity as an adjunct to non-invasive ventilation (NIV) in the late 1980s (Bach 1996). Modern MI-E devices such as the CoughAssist In-Exsufflator (Respironics Corporation, PA), the Pegaso (Dimla Italia, Bologna, Italy), and the Nippy Clearway (B & D Electromedical, Warwickshire, UK), deliver a preset positive pressure into the airways for a set duration during inspiration (insufflation), immediately followed by an abrupt change to a preset negative exsufflation pressure, thus simulating a cough with high expiratory flow rates (Anderson 2005; Boitano 2009; Fauroux 2008). MI-E can be delivered noninvasively via the nose (mask) or mouth (mouthpiece), or via tracheostomy (Boitano 2009). MI-E has been shown to enhance peak cough flow in patients with advanced NMDs, and may assist in maintaining lung volume, both of which are necessary for effective secretion clearance (Anderson 2005; Bach 1993; Chatwin 2003). A study of 11 consecutive older children and adults (age range 11 to 72 years) with a NMD and acute respiratory tract infections treated with chest physiotherapy and MI-E, compared with 16 historically matched controls who had received chest physiotherapy alone, found that treatment failure (i.e. the need for mini-tracheostomy or intubation) was significantly less in the MI-E group and there were no serious complications of MI-E (Vianello 2005). There were no statistically significant differences in other measured outcomes including days requiring mechanical ventilation, duration of hospital stay and number of patients requiring bronchoscopy-assisted aspiration(Vianello 2005). Significant short-term improvements in oxygenation and dyspnoea have been reported in adults treated with 40 cmH2O MI-E settings, but differences were not significant when using lower pressures (Winck 2004). A small cross-over study of eight people with a NMD aged four to 44 years demonstrated that airway clearance treatment time was significantly shortened with the addition of MI-E. Both standard intervention and MI-E resulted in improved auscultation scores and secretion clearance but there were significantly higher levels of fatigue after MI-E (Chatwin 2009).

An observational study with no control group showed a single treatment session of MI-E to be well tolerated and physiologically beneficial in 17 stable children (age five to 18 years) with NMDs (Fauroux 2008). Benefits of a single treatment session included improved end-tidal expired carbon dioxide pressure, and improved peak cough flow and self-assessed respiratory comfort (Fauroux 2008). A retrospective descriptive study of 62 children with NMDs concluded that MI-E was “mostly safe” and effective in preventing and treating pulmonary complications (Miske 2004). Four individuals demonstrated resolution of chronic atelectasis following the implementation of MI-E and five children (eight per cent) experienced a reduction in the frequency of lower respiratory tract infections (Miske 2004). In a recent case series, 13 infants and young children with spinal muscular atrophy type 1 were followed from the time of diagnosis. Seven subjects used MI-E at home in combination with goal-directed NIV (Chatwin 2011).The median age at initiation of MI-E was 13 months (range 10 to 43 months) and the typical MI-E pressures used were +25 to -30 cm H2O, increasing to +35 to -45 cm H2O as clinically indicated. In these subjects, MI-E was reported to be well tolerated and may have prevented the need for intubation and invasive ventilation (Chatwin 2011).

It has been suggested that MI-E may be used to prevent and treat intercurrent infections in both the hospital and home settings (Hanayama 1997; Whitney 2002), in combination with NIV where indicated (Anderson 2005; Tzeng 2000). In a retrospective cohort study of children (over 10 years old) and adults, the trial authors found that participants managed with a respiratory protocol, including prevention or reversal of oxyhaemoglobin desaturation by the use of noninvasive intermittent positive pressure ventilation (IPPV) and assisted coughing (manual and mechanical, including MI-E) as needed, experienced reduced rates of hospitalisation and hospital days compared to non-protocol participants who were on IPPV via tracheostomy (Bach 1997). A similar retrospective cohort study of 94 adults and children demonstrated a reduction in the number of hospitalisations per year and hospital days per year in individuals using a respiratory muscle aid protocol, which included MI-E as needed (Tzeng 2000).

How the intervention might work

The peak cough expiratory flow (PCEF) needed for an effective cough in adults is 162 L/min, with adults having normal peak cough flows of 360 to 720 L/min (Bach 1993; Bach 1995; Bach 1997; Gómez-Merino 2002; Tzeng 2000). The value for infants and children is not known. An initial inspiration to up to 90% of maximum insufflation capacity (vital capacity > 1.5 L), and sufficient thoraco-abdominal pressure (> 100 cmH2O) is needed to achieve an effective peak cough flow (Anderson 2005). MI-E has been shown to improve cough expiratory flow rates in adults and children with NMDs (Bach 1993; Chatwin 2003; Winck 2004). Bach 1993 reported an increase in the mean peak cough flow rate from 108.6 ± 61.8 L/min unassisted to 448.2 ± 61.2 L/min in adults when using the exsufflator, whilst Winck 2004 reported an increase in mean peak cough flow from 180 L/min at baseline to 220 L/min following MI-E of 40 cmH2O. Change in cough flow rates using lower MI-E pressures were not measured (Winck 2004). In a controlled study of adults and older children (over 10 years old) with NMDs, MI-E resulted in a significant increase in mean peak cough flow from 169 L/min unassisted to 235 L/min (Chatwin 2003).

It has been suggested that MI-E may obviate the need for suctioning in people with a NMD, as peak cough flow rates generated are sufficient to clear secretions (Bach 1993; Hanayama 1997).

Why it is important to do this review

MI-E is recommended in a number of international guidelines for the management of people with NMDs (Bott 2009; Finder 2004; McCool 2006; Rosiere 2009; Wang 2007; Wang 2010). Of particular concern, is that the pressures that have been recommended for all age groups reach or exceed -40 cmH2O (exsufflation) to +40 cmH2O (insufflation), with four to five sets of breaths being performed as often as needed to clear secretions (Boitano 2006; Gómez-Merino 2002; Miske 2004; Respironics 2009; Tzeng 2000). Whilst some studies have used these pressures in adults and children for both insufflation and exsufflation (Bach 1993; Fauroux 2008; Sivasothy 2001), another study of MI-E in older children and adults used lower, "comfortable" pressures of 15 cmH2O (Chatwin 2003). The optimum pressures, frequency of use and insufflation-exsufflation times are not currently known (Anderson 2005).

With any mechanical positive-pressure device, there is a risk of complications such as abdominal distention, discomfort, gastro-oesophageal reflux, cardiovascular effects, such as changes in blood pressure and cardiac arrhythmia, and pneumothorax (Homnick 2007). Cases of pneumothorax have been described in adult patients following use of MI-E (Suri 2008) and long-term non-invasive positive pressure ventilation (Vianello 2004). Considering the differences between the paediatric and adult respiratory systems (particularly relating to high chest wall compliance in the infant), there may be a greater risk of baro- or volutrauma in young children with the use of such high pressures, and more so in infants with NMDs where chest wall compliance is further increased relative to lung compliance. It is notable that applied volume is not measured during mechanical insufflation, and high tidal volume has been implicated in ventilator-induced lung injury (Albuali 2007), along with repeated alveolar collapse and re-expansion (i.e. atelectrauma) (Saharan 2010). Current lung-protective ventilation strategies include limiting inspired tidal volumes and preventing derecruitment by loss of PEEP or wide swings in pressure or both (Saharan 2010). Use of MI-E appears to contradict all these strategies.

Whilst MI-E is gaining popularity amongst professionals and consumers alike, this therapy requires expensive equipment which may not be readily available, whilst other, inexpensive, readily available techniques have also been shown to be effective in improving cough flow (Anderson 2005; Finder 2010). Although assisted airway clearance is necessary in people with an advanced NMD, the optimal respiratory management to clear secretions is not evident.

Specific respiratory anatomical and physiological considerations must be made according to different age groups and specific neuromuscular conditions when assessing the effects of MI-E. Application of positive pressure will affect the lungs differently according to the prevailing pathology and physiology, including the presence of heterogenous lung disease, baseline lung volumes, and respiratory system compliance and resistance (Gattinoni 2003; Gattinoni 2010).

Objectives

To determine the efficacy and safety of MI-E in people with NMDs.

Methods

Criteria for considering studies for this review

Types of studies

Prospective randomised controlled trials (RCTs) or quasi-RCTs and randomised cross-over trials.

Types of participants

Participants of all ages diagnosed with a NMD and respiratory insufficiency. We planned to stratify participants according to age and whether the problem being managed was acute, chronic or mixed.

Types of interventions

MI-E to assist airway clearance, both as maintenance therapy and treatment of intercurrent respiratory tract infection.

We planned to compare MI-E to no treatment or other cough augmentation methods.

Types of outcome measures

Outcome measures typically used to determine the clinical efficacy of treatments for people with a NMD include survival, frequency of pulmonary exacerbations and hospitalisation, duration of hospital stay and quality of life indicators. The adverse event of most concern is likely to be pneumothorax (Suri 2008), but we also considered other potential complications.

Primary outcomes

The primary outcomes were the pooled mortality hazard ratio (HR) calculated using deaths throughout follow-up, or the pooled mortality risk ratio (RR) at a single follow-up time (at six months).

Secondary outcomes

In formulating secondary outcomes, we differentiated between MI-E used in acute and chronic conditions.

Secondary outcomes in the acute use setting were as follows.

  1. Measures of gaseous exchange (oxygenation (PaO2) or expired CO2 or both), or pulmonary function measured by forced expiratory volume in one second (FEV1), FVC and peak expiratory flow rate (PEFR), at six months.

  2. Morbidity, measured by duration of hospital stay.

  3. Quality of life measured by any validated measure, for example McGill Quality of Life (MQOL) instrument (Robbins 2001); Life Satisfaction Index for Adolescents (LSIA) (Reid 1994); Nottingham Health Profile (NHP) and the Medical Outcome Study 36-item Short-Form questionnaire (MOS SF-36) (Boyer 2006), at six months.

  4. Proportion of patients intubated and invasively ventilated at six months.

  5. Adverse events occurring as a consequence of the intervention.

Secondary outcomes in the chronic use setting were as follows.

  1. Measures of gaseous exchange (oxygenation (PaO2) and or expired CO2), or pulmonary function measured by FEV1, FVC and PEFR, at six months.

  2. Morbidity, measured by:

    1. number of hospital admissions, up to six months; and

    2. number of antibiotic courses, up to six months.

  3. Quality of life measured by any validated measure, for example MQOL instrument (Robbins 2001); LSIA (Reid 1994); NHP and the MOS SF-36 (Boyer 2006) at six months.

  4. In children under 18 years of age, growth (weight for age or weight for height Z scores); in adults, weight, from baseline to six months.

  5. Proportion of patients intubated and invasively ventilated at six months.

  6. Adverse events occurring as a consequence of the intervention.

We planned to include all of these outcomes in a ‘Summary of findings’ table for studies comparing MI-E to no treatment. We planned to produce separate tables for acute and chronic use of MI-E. We would have presented further tables if there were sufficient data for other comparisons.

Search methods for identification of studies

Electronic searches

On 7 October 2013, we searched the Cochrane Neuromuscular Disease Group Specialized Register, CENTRAL (2013, Issue 9 in The Cochrane Library), MEDLINE (January 1966 to September 2013), EMBASE (January 1980 to October 2013), ClinicalTrials.gov (www.clinicaltrials.gov) and the World Health Organization International Clinical Trials Registry Platform (ICTRP) (www.who.int/ictrp/en/).

The detailed search strategies are reported in the appendices: Cochane Neuromuscular Disease Group Specialized Register (Appendix 1), CENTRAL (Appendix 2), MEDLINE (Appendix 3), EMBASE (Appendix 4), and ClinicalTrials.gov and ICTRP (Appendix 5).

Searching other resources

We scanned conference abstracts for relevant studies. We checked all references in the identified trials and contacted authors where appropriate to identify any additional published or unpublished data.

Data collection and analysis

Selection of studies

Two authors (BM and HvA) independently reviewed abstracts and titles identified from the search and compiled a list of studies that potentially met the inclusion criteria. We compared the lists and resolved any disagreement by discussion and consensus. We retrieved the full text of every study deemed potentially relevant and two authors independently assessed eligibility using standardised study eligibility forms. Following discussion and an attempt to reach consensus, the authors selecting studies planned to consult a third author if there was still any disagreement.

Data extraction and management

Two review authors (BM and MZ) independently extracted data using a pre-structured data extraction form. We collected information on participants (i.e. age, gender, diagnosis, severity of symptoms, recruitment or enrolment method, inclusion and exclusion criteria, comorbid conditions, trial setting, allocation procedure, blinding, number of patients randomised, number lost to follow-up), interventions (i.e. equipment used, pressure settings, duration, frequency, cross-overs), outcome measures and results (i.e. point estimates, precision, measures of variability, frequency counts for dichotomous variables, number of participants). Participants were stratified according to age and diagnosis. The lead reviewer (BM) collated and entered all data into the Cochrane software Review Manager (RevMan5) (RevMan 2012) and a second author checked the data entry. Where we identified incomplete data, we attempted to contact the investigator for clarification or additional data. We resolved any disagreements on data extraction by discussion and consensus, referring to a third author if necessary.

Assessment of risk of bias in included studies

Two review authors (BM and MZ) independently assessed risk of bias in each included study using the Cochrane 'Risk of bias' tool (Higgins 2011a). The authors assessed the following key criteria: selection bias (i.e. adequate sequence generation and allocation concealment); performance bias (i.e. blinding of participants and providers); detection bias (i.e. blinding of outcome assessors); attrition bias (i.e. incomplete outcome data addressed); reporting bias (i.e. free of selective reporting); and other bias. The following judgements were applied: low risk, high risk and unclear risk (i.e. either lack of information or uncertainty over the potential for bias). The review authors resolved conflict by discussion and consensus in consultation with a third author if necessary.

Measures of treatment effect

The extracted data were insufficient to perform a meta-analysis (see Results). We therefore described the individual results in narrative and tabular form using the effect measures described in the original studies.

We will perform the following analyses with available data in subsequent updates of the review:

We will analyse all data for acute and chronic MI-E use separately. We will extract data from included trials and enter them into RevMan 5 for statistical analysis. For time-to-event outcomes (e.g. mortality calculated by deaths throughout follow-up) we will calculate the hazard ratio (HR) and corresponding 95% confidence interval (CI). For dichotomous outcomes (e.g. mortality at six months) we will calculate the risk ratio (RR) and corresponding 95% (CI). We will calculate Peto odds ratio (Peto OR) and corresponding 95% for rare adverse events. In the case of statistically significant results, we will calculate the risk difference (RD) and 95% CI and the number needed to treat for an additional beneficial outcome (NNTB) or for an additional harmful outcome (NNTH) as appropriate. For continuous outcomes (i.e. quality of life indicator scales, pulmonary function measurements, duration of hospital stay, number of hospitalisations and exacerbations) we will calculate the standardised mean difference (SMD) or mean difference (MD) and 95% CI as appropriate. Where standard errors of the means (SEM) are reported, we will convert these to standard deviations (SD) where possible.

Unit of analysis issues

In future updates of the review we will include only first period data from cross-over trials (Higgins 2011b). Long-term studies with multiple repeated measures of outcome may be included, in which case we will define outcomes based on the specified time points (Deeks 2011). If studies have more than one intervention group we will combine groups to create a single pair-wise comparison, or select one pair of interventions and exclude the others, combine groups to create a single pair-wise comparison, or include intervention groups separately in the meta-analysis with the control group divided in half (Higgins 2011b; Ramsay 2003).

Dealing with missing data

We attempted to contact authors of studies to obtain incomplete data. Where statistical data were missing we estimated the results from the available information, where possible, by imputing or calculating missing standard deviations, using methods described by Higgins and Deeks (Higgins 2011c).

In future updates of this review, where we are unable to obtain missing data, we will consider the studies adequate if more than 85% of the participants are included in the outcome analysis or if fewer participants were analysed but sufficient measures were taken to ensure or demonstrate that this did not bias the results. Where this is not clear, an intention-to-treat analysis will be performed from extrapolated data. Where statistical data are missing we will calculate the data from the available information when possible.

Assessment of heterogeneity

In the updates of this review, where data are available, we will assess the degree of heterogeneity using the Chi2 test and I2 statistic (Higgins 2003). A Chi2 significance level of P < 0.1 will indicate statistically significant heterogeneity amongst studies. For this meta-analysis, we will use random-effects models if the Chi2 test is significant and the I2 statistic indicates moderate to severe heterogeneity (i.e. I2 > 50%).

Assessment of reporting biases

If future updates of this review include more than 10 studies of different sizes, we will assess the potential for publication bias examining funnel plot symmetry (Sterne 2011). We will create a funnel plot using RevMan 5, comparing treatment effect against a measure of the size or precision of the study.

Data synthesis

If data are available in updates of this review we will perform the following analyses as appropriate. We will combine data from individual trials for meta-analysis if the interventions, outcomes and patient groups are sufficiently similar (to be determined by consensus). For time-to-event outcomes (e.g. mortality calculated by deaths throughout follow-up) we will calculate the pooled HR and corresponding 95% CI. For dichotomous outcomes (e.g. mortality at six months) we will calculate the pooled RR and corresponding 95% CI. We will calculate the pooled Peto OR and corresponding 95% for rare adverse events. In the case of statistically significant results, we will calculate the pooled risk difference (RD) and 95% CI and the number needed to treat for an additional beneficial outcome (NNTB) or for an additional harmful outcome (NNTH) as appropriate. We will pool data using a fixed-effect model in cases of low heterogeneity (I2 statistic < 50% and non-significant Chi2 test) and a random-effects model in cases of moderate to high heterogeneity (I2 statistic > 50% and significant Chi2 test). We will conduct meta-analyses where there is minimal clinical or methodological heterogeneity. Where we cannot pool data, we will report the results in narrative form.

Subgroup analysis and investigation of heterogeneity

The factors which may lead to heterogeneity across the studies include age of participants (i.e. infancy versus older children versus adults); clinical setting (i.e. acute versus chronic use); MI-E set pressures (i.e. +40 to -40 cm H20 versus other pressures); frequency and duration (i.e. repetitions) of application; ventilation requirements (i.e. NIV, invasive ventilation or spontaneously breathing); and MI-E interface (i.e. mask, tracheostomy or mouthpiece). We plan to investigate these factors with subgroup analyses if updates of this review include a sufficient number of studies.

Sensitivity analysis

We plan to conduct a sensitivity analysis in updates of this review, where appropriate, based on the methodological quality of the trials (i.e. randomised versus quasi-randomised); published versus unpublished data; and different statistical models (i.e. fixed-effect versus random-effects models).

This review has a published protocol (Morrow 2012).

Results

Description of studies

See Characteristics of included studies and Characteristics of excluded studies.

Results of the search

The literature search identified a total of 53 papers (see Figure 1 for study flow): 0 from the Cochrane Neuromuscular Disease Group Specialized Register, 5 from CENTRAL, 22 from MEDLINE and 26 from EMBASE. After removing duplicates, we reviewed the titles and abstracts of 46 papers. We selected eight studies for full-text review (Chatwin 2003; Chatwin 2009; Mustfa 2003; Sancho 2003; Senent 2011; Sivasothy 2001; Trebbia 2005; Vitacca 2010) and excluded three of these studies (Sancho 2003; Trebbia 2005; Vitacca 2010). Five studies met the inclusion criteria for the review (Chatwin 2003; Chatwin 2009; Mustfa 2003; Senent 2011; Sivasothy 2001). Six ongoing clinical trials were identified through the WHO ICTR platform (five from ClinicalTrials.gov) and one of these studies was identified for possible inclusion in future reviews (NCT01518439).

Figure 1.

Study flow diagram.

Included studies

The five included studies were all from Europe (Chatwin 2003; Chatwin 2009; Mustfa 2003; Senent 2011; Sivasothy 2001). Four studies were conducted in the United Kingdom (Chatwin 2009; Sivasothy 2001; Mustfa 2003; Chatwin 2003), and one was conducted in France (Senent 2011). All were short-term studies (i.e. two days or less in duration) of the immediate effects of MI-E in a hospital or clinic setting. None of these studies investigated the effects of a full course of MI-E treatment either for long-term maintenance therapy or for the treatment of an acute exacerbation.

Study participants were adults and children with an age range from 4 to 73 years. A total of 5 children under the age of 13 years were included. No paediatric-specific studies were identified. Three studies included adults only (Mustfa 2003; Senent 2011; Sivasothy 2001). Two studies included both adults and children. Chatwin 2003 included 8 children and adolescents between the ages of 10 and 17 years; and 14 adults aged 18 to 56 years. Chatwin 2009 included 2 children aged 4 and 12 years, and 6 adults older than 20 years. Only one study investigated participants admitted to hospital with acute respiratory tract infections (Chatwin 2009). Three studies investigated stable participants without intercurrent infection (Sivasothy 2001; Chatwin 2003; Senent 2011). One study did not report the infection status of patients and no specific exclusion criteria were described (Mustfa 2003). One study specifically excluded people with bulbar palsy (Chatwin 2003).

The total number of participants with NMDs included in the studies was 105 (range 8 to 47). Two studies only included patients with amyotrophic lateral sclerosis (ALS), who were further divided into bulbar and non-bulbar groups (Mustfa 2003; Senent 2011). Other studies included a range of neuromuscular conditions including spinal muscular atrophy (SMA), Duchenne muscular dystrophy, congenital myopathy, poliomyelitis, and Becker's muscular dystrophy (Chatwin 2003; Chatwin 2009; Sivasothy 2001). One study divided participants into those with and without scoliosis (Sivasothy 2001). In three studies, participants without a NMD were also included as separately analysed groups (Chatwin 2003; Mustfa 2003; Sivasothy 2001), but results from these participants are not eligible for this review.

Mustfa 2003 further subdivided participants into those with cough gastric pressure < 50 cmH2O; vital capacity < 50% and PCEF < 160 L/min.

MI-E for subjects with a NMD was given using the CoughAssist MI-E device (JH Emerson Co, Cambridge, USA) in three studies (Chatwin 2003; Mustfa 2003; Sivasothy 2001) and the CoughAssist (Philips Respironics, Murraysville, Pennsylvania) device in two studies (Chatwin 2009; Senent 2011). All studies used a face-mask interface. Average insufflation pressures ranged from +15 ± 3 cmH2O (Chatwin 2003), through +20 cmH2O (Chatwin 2009; Sivasothy 2001), to +40 cmH2O (Senent 2011). Exsufflation pressures ranged from -40 cmH2O (Chatwin 2009; Senent 2011), through -20 cmH2O (Chatwin 2009; Sivasothy 2001), to -15 cmH2O (Chatwin 2003). In-exsufflation pressures were not reported in one study (Mustfa 2003).

Insufflation and exsufflation times of 2 to 4 and 4 to 5 seconds were used in one study (Chatwin 2009) and in four studies in-exsufflation times were not reported (Senent 2011; Chatwin 2003; Sivasothy 2001; Mustfa 2003).

All studies compared MI-E to one or more alternative cough augmentation techniques (Chatwin 2003; Chatwin 2009; Mustfa 2003; Senent 2011; Sivasothy 2001). These techniques included: manually assisted cough using abdominal or thoraco-abdominal pressure (Chatwin 2003; Mustfa 2003; Sivasothy 2001); spontaneous cough after augmented inspiration (Chatwin 2003); manually assisted cough after augmented inspiration (Senent 2011); exsufflation-assisted cough (Chatwin 2003); insufflation-assisted cough (Chatwin 2003; Mustfa 2003); and a combination of MI-E and manually assisted cough (Sivasothy 2001). One randomised cross-over trial conducted over two days compared standardised airway clearance therapy with and without the addition of MI-E (Chatwin 2009). Three studies compared MI-E assisted cough to unassisted coughing (Chatwin 2003; Mustfa 2003; Sivasothy 2001). The MI-E intervention in one study (Sivasothy 2001) consisted of two cycles of in-exsufflation followed by an insufflation and maximal voluntary cough. Exsufflation was not used during the cough itself.

All of the five included studies reported only short-term outcome parameters of individual interventions. Objective outcomes measured were: PCEF in four studies (Chatwin 2003; Mustfa 2003; Senent 2011; Sivasothy 2001); physiological variables of heart rate, transcutaneous oxygen saturation, transcutaneous carbon dioxide tension in one study (Chatwin 2009); transpulmonary pressure and time integrated cough volume in one study (Mustfa 2003); cough expiratory volume and peak value time in one study (Sivasothy 2001); treatment time (Chatwin 2009) and oesophageal or cough gastric pressures or both in two studies (Mustfa 2003; Sivasothy 2001).

Subjective outcome measures were patient comfort visual analogue score (VAS) in three studies (Chatwin 2003; Chatwin 2009; Senent 2011); VAS scores for breathlessness, mood and secretion production (Chatwin 2009); efficacy VAS (Senent 2011); and auscultation score (Chatwin 2009).

None of the short- or long-term outcome measures outlined in the protocol, including survival or mortality, frequency of exacerbations and hospitalisation, duration of hospital stay and quality of life indicators, were reported in any of the included studies.

Four of the studies reported no adverse events associated with the study interventions (Chatwin 2003; Mustfa 2003; Senent 2011; Sivasothy 2001). One study reported fatigue as an adverse effect of MI-E, using a VAS (Chatwin 2009). It is not clear whether serious adverse events such as pneumothorax were systematically investigated in any of the studies.

In four studies, all participants received every intervention in random order, with a variable washout period between interventions (Chatwin 2003; Mustfa 2003; Senent 2011; Sivasothy 2001). One study (Chatwin 2009) was a randomised cross-over trial conducted over two days in which eight participants were assigned to receive MI-E for one treatment session and no MI-E for a second treatment session, with a reverse cross-over the following day. Data for the first period of the cross-over trial were not supplied, precluding analysis.

Excluded studies

Three trials were excluded: two because of lack of randomisation or quasi-randomisation (Sancho 2003; Vitacca 2010), and the other because the intervention was not actually MI-E (Trebbia 2005).

Risk of bias in included studies

See Figure 2 and Characteristics of included studies.

Figure 2.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study. Red = high risk of bias, yellow = unclear risk of bias, green = low risk of bias.

Allocation

All five included studies were described as randomised. However, none of these studies provided any details about the method of randomisation. We therefore judged the studies to have an unclear risk of bias for the generation of the randomisation sequence. Allocation concealment was not described in any of the included studies, leading to a judgement of unclear risk of selection bias.

Blinding

It was unclear if outcome assessors were blinded in four studies (Chatwin 2003; Chatwin 2009; Mustfa 2003; Sivasothy 2001). In one study the physiotherapists rating the perceived efficacy of the interventions were blinded to participants' peak expiratory flow measurements (Senent 2011). Considering the nature of the interventions, it is unlikely that participant or clinician blinding would have been possible, and this was not discussed in any of the included studies.

The lack of blinding may represent an important risk of bias, particularly for subjective outcome variables but also for some objective outcome variables such as duration of treatment, which may have varied due to perceived, and not actual, difference in clinical response.

Incomplete outcome data

All participants completed the interventions in all studies and were included in the analyses, leading to a judgement of low risk of attrition bias.

Selective reporting

Four studies scored as "low risk" for reporting bias (Chatwin 2003; Mustfa 2003; Senent 2011; Sivasothy 2001). One study did not present any data for its primary outcome measures (physiological variables of oxygen saturation, heart rate and transcutaneous carbon dioxide tension), merely stating there was no significant difference between groups (Chatwin 2009). In this study results of VAS for comfort, breathlessness and mood were only presented as graphs, and data could not be extracted precisely (Chatwin 2009). Efforts to obtain missing data were not successful.

Other potential sources of bias

One study author disclosed a relationship with a healthcare company that manufactures ventilation equipment (Chatwin 2009), although the nature of the relationship and the relevance to this study was not clear. Another study author disclosed a financial relationship with a manufacturer of a mechanical in-exsufflator (Senent 2011). Funding sources and declarations of interest were not reported in one study (Sivasothy 2001). Considering that it was unclear whether assessors were blinded in any of the studies, this is a potential source of bias.

Effects of interventions

Four studies applied every intervention to every included subject, in random order (Chatwin 2003; Mustfa 2003; Senent 2011; Sivasothy 2001). Thus each subject received several interventions. The study reports did not present individual responses to the interventions. Meta-analysis and pooling of results was not possible owing to the repeat counting that occurred, which would cause a unit of analysis error from the un-addressed correlation between the estimated intervention effects of multiple comparisons (Higgins 2011b). The cross-over study did not report the data separately for the two study periods precluding inclusion in a meta-analysis (Chatwin 2009).

Primary outcomes  

1. Pooled mortality hazard ratio (HR) calculated using deaths throughout follow-up or pooled mortality risk ratio (RR) at a single follow-up time (at six months)

No data were available in any of the studies for this outcome.

Secondary outcomes

Acute use
1. Measures of gaseous exchange (oxygenation (PaO2) and/or expired CO2),  or pulmonary function measured by forced expiratory volume in one second (FEV1), FVC and peak expiratory flow rate (PEFR), at six months

Short-term effects of interventions on physiological variables of heart rate, transcutaneous oxygen saturation, and transcutaneous CO2 tension were primary outcomes in one study. However, the report merely stated, without providing data, that there was no difference between intervention groups in these physiological parameters (Chatwin 2009).

None of the included studies investigated the effects of treatment on FEV1, FVC or PEFR at six months.

2. Morbidity, measured by duration of hospital stay

None of the included studies reported on duration of hospital stay.

3. Quality of life measured by any validated measure at six months

No data were available in any of the studies regarding this outcome.

4. Proportion of patients intubated and invasively ventilated at six months

No data were available in any of the studies for this outcome.

5. Adverse events occurring as a consequence of the intervention

One study reported fatigue as an adverse effect of MI-E, using a 10-point VAS (Chatwin 2009). This study reported data for the group that received physiotherapy and manually assisted cough (MAC) with MI-E but not for the group that received physiotherapy and MAC without MI-E. In the latter group, fatigue was only reported as not significantly different pre- to post-intervention. In the MI-E group, fatigue VAS (mean ± SD) increased from 3.2 ± 2.2 before the intervention to 5.1 ± 2.6 after intervention (P = 0.005). The lack of comparative data makes meaningful conclusions difficult.

No serious adverse events attributed to the interventions were reported in any of the studies; however, it is unclear whether these were systematically investigated.

Chronic use

None of the trials investigated chronic use of MI-E.

Other reported outcome measures

Other short-term outcome measures were reported in the included studies; see the 'Other results' section of the Discussion. Individual study results are presented in Table 1.

Table 1. Study results
  1. VAS visual analogue scale; SD standard deviation; CI confidence interval. Paired comparisons between MI-E assisted cough and other interventions are marked across rows: * P < 0.001; # P < 0.01; ** P < 0.05

  Study interventions
Outcome measures Study and/or participant groups Data presentation Unassisted cough Manually assisted cough Insufflation- assisted cough Exsufflation assisted cough Combination of mechanical in-exsufflation and manually assisted cough Manually assisted cough after air stacking on spontaneously deep breath, using resuscitator MAC after normal ventilator- assisted inspiration

 

MAC  after ventilator- assisted inspiratory augmentation

Mechanical insufflation- exsufflation- assisted cough

Peak cough expiratory flow (PCEF) (L/min)

 

 

 

 

 

n = 22 Chatwin 2003Mean (95% CI)169 (129 to 209)#188 (146to 229)182 (147 to 217)235 (186 to 284)* - - - -297 (246, 350)#*
With scoliosis n = 4 Sivasothy 2001Median (range)288 (175 to 367)193 (185 to 287) - -362 (218 to 440) - - -231 (148 – 597)

Without scoliosis n = 8

Sivasothy 2001

Median (range)104 (43 to 188)185 (93 to 355) - -248 (110 to 343) - - -156 (61 - 247)
Bulbar ALS n = 21 Mustfa 2003Mean ± SD178 ± 61**197 ± 63188 ± 64225 ± 76 - - - -212 ± 75**

Non-bulbar ALS n = 26

Mustfa 2003

Mean ± SD217 ± 84*244 ± 83226 ± 86279 ± 87* - - - -264 ± 73*

n = 16

Senent 2011

 

Median (interquartile range)

 - - - - -284 (146 to 353)212 (99 to 595)233 (100 to 389)488 (243 to 605)

Transpulmonary pressure (TPP) (cmH2O)

 

Bulbar ALS n = 21 Mustfa 2003Mean ± SD45 ± 35*53 ± 3941 ± 3562 ± 42 - - - -57 ± 38*

Non-bulbar ALS n = 26

Mustfa 2003

Mean ± SD49 ± 51#65 ± 6048 ± 5568 ± 58 - - - -63 ± 53#

Cough expiratory volume (CEV) (l)

 

With scoliosis n = 4 (Sivasothy 2001)Median (range)0.9 (0.5 to 1.1)0.5 (0.41 to 1.01) - -0.6 (0.4 to 1.01) - - -0.7 (0.3 to 1.3)

Without scoliosis n = 8

Sivasothy 2001

Median (range)0.5 (0.3 to 0.8)0.7 (0.31 to 1.07) - -0.6 (0.4 to 2.19) - - -0.6 (0.3 to 1.61)

Peak value time (PVT) (ms)

 

With scoliosis n = 4 Sivasothy 2001Median (range)44 (40 to 50)50 (35 to 55) - -50 (45 to 120) - - -45 (30 to 60)

Without scoliosis n = 8

Sivasothy 2001

Median (range)80 (40 to 220)118 (35 to 360) - -75 (20 to 420) - - -85 (20 to 420)

Cough gastric pressure (Pgas) cm H2O

 

Bulbar ALS n = 21 Mustfa 2003Mean ± SD68 ± 5286 ± 5968 ± 6167 ± 59 - - - -57 ± 38

Non-bulbar ALS n = 26

Mustfa 2003

Mean ± SD75 ± 60101 ± 7176 ± 6373 ± 63 - - - -75 ± 60
Treatment time up to 30 min

n = 2

Chatwin 2009

 - -Not reported - - - - - -Not reported
Continued treatment time beyond 30 min

n = 6

Chatwin 2009

Median (range) -17 (0 – 35)** - - - - - -0 (0 = 26)**
Heart rate

n = 8

Chatwin 2009

 - -Data not reported - - - - - -Data not reported
Oxygen saturation

n = 8

Chatwin 2009

 - -Data not reported - - - - - -Data not reported
Transcutaneous carbon dioxide tension

n = 8

Chatwin 2009

 - -Data not reported - - - - - -Data not reported
Efficacy (VAS 10 point score)

n = 16

Senent 2011

 - - - - - -7 (5 to 8)7 (6 to 8)6 (5 to 7)8 (6 to 8)

Comfort (VAS 10 point score)

 

 

n = 8

Chatwin 2009

 - -Not reported - - - - - -Not reported

n = 16

Senent 2011

 

Median (interquartile range)

 - - - - -6 (5 to 8)8 (7 to 8)6 (5 to 7)7 (3 to 8)

n = 22

Chatwin 2003

Mean (95% CI)5.4 (4.5 to 6.3)5.9 (5.3 to 7.0)5.8 (4.8 to 6.8)6.9 (5.3 to 7.0) - - - -7.3 (6.6 to 8.0)
Secretions (VAS 10 point score)

n = 8

Chatwin 2009

Mean ± SD -Change 4.4 ± 2.5 before to 3.0 ± 1.4 after intervention (P = 0.03) - - - - - -Change 4 ± 2.2 before to 1.7 ± 0.4 after intervention (P = 0.03)
Auscultation (VAS 10 point score)

n = 8

Chatwin 2009

Mean ± SD -3.4 ± 2.0 before to 2.3 ± 2.2 after intervention (P = 0.007) - - - - - -Change 2.9 ± 1.9 before to 1.8 ± 2.0 after intervention (P = 0.02)
Breathlessness (VAS 10 point score)

n = 8

Chatwin 2009

Mean ± SD -Data not reported - - - - - -Data not reported

 

Mood (VAS 10 point score)

n = 8

Chatwin 2009

Mean ± SD -Data not reported - - - - - -Data not reported
Efficacy (VAS 10 point score)

 n = 16

Senent 2011

 Median (interquartile range) - - - - - 7 (5 to 8) 7 (6 to 8) 6 (5 to 7) 8 (6 to 8)
Fatigue (VAS 10 point score)

n = 8

Chatwin 2009

Mean ± SD -Data not reported - - - - - -3.2 ± 2.2 before to 5.1 ± 2.6 after intervention (P = 0.005)

Discussion

Summary of main results

We identified no prospective RCTs or quasi-RCTs that measured important clinically relevant short- and long-term outcomes of MI-E compared to other cough augmentation interventions, or no treatment. There is, therefore, no clear evidence for or against the use of MI-E in patients with a NMD for either chronic use or use during an acute pulmonary exacerbation.

Based on three studies (Chatwin 2003; Mustfa 2003; Sivasothy 2001), MI-E may improve PCEF compared to unassisted cough. Other measures of cough efficacy had variable response to MI-E, but were also poorly reported. By increasing cough flow, there could be improved secretion clearance, which could lead to improved clinical outcome in terms of morbidity (i.e. prevention of infection and hospitalisations), and possibly survival. The included studies did not clearly show MI-E to improve cough expiratory flow more than other airway clearance or cough augmentation techniques, and it is notable that all techniques increased cough flow to above the critical flow level required to effectively clear airway debris (Bott 2009).

It was suggested, based on one study with potential assessor bias, that MI-E may reduce treatment time when added to a standard airway clearance regimen with manually assisted cough (Chatwin 2009). MI-E appeared to be as well tolerated as other cough augmentation techniques, based on three studies which reported comfort scores (Chatwin 2003; Chatwin 2009; Senent 2011).

One study suggested that MI-E caused significant increases in fatigue (Chatwin 2009). None of the included studies reported any serious adverse events of interventions, but they were not adequately powered to determine safety.

Other results

PCEF was the most common outcome measure, reported in four studies (Chatwin 2003; Mustfa 2003; Senent 2011; Sivasothy 2001).

Chatwin 2003 found that, in adults and children with severe respiratory muscle weakness due to a NMD, MI-E assisted cough produced a higher PCEF than a voluntary unassisted cough or a cough assisted by NIV. Mustfa 2003 reported that the application of negative pressure (i.e. exsufflation) during cough augmented PCEF in patients with ALS. The application of positive (i.e. insufflation) pressure only improved PCEF further in those with non-bulbar ALS and a vital capacity less than 50%. Senent 2011 reported that PCEF can be improved using different augmentation tools, even in patients with severe bulbar symptoms. PCEF was highest when using MI-E assisted cough. Sivasothy 2001 found that the combination of MI-E and MAC, and MAC alone, improved PCEF in patients with a NMD without scoliosis, but not in those with scoliosis, although the small sample size suggests that one should interpret this subgroup analysis with caution. MI-E followed by maximal voluntary cough did not enhance cough efficacy. It is notable that all interventions increased cough flow to > 160 L/min, which is generally accepted as the critical flow level required to effectively clear secretions (Bach 2003).

Chatwin 2009, a randomised cross-over trial of eight subjects, reported that there was a reduction in treatment time after 30 min of treatment with MI-E plus standard physiotherapy and MAC compared with standard physiotherapy and MAC alone.

The most common subjective outcome measure, which was reported in three studies, was of patient comfort, using a VAS (Chatwin 2003; Chatwin 2009; Senent 2011). In one study the comfort VAS for the groups was reported only in a graph and data could not be extracted (Chatwin 2009). In another study (Chatwin 2003), participants were reported to feel most comfortable with MI-E assisted cough. Senent 2011 reported that MAC following normal ventilated inspiration was deemed most comfortable by the participants in this study (Senent 2011).

Excluded studies

Considering that RCTs provide the best quality evidence and we did not search systematically for non-randomised studies, the results of these excluded studies should be interpreted with caution. Sancho 2003 compared MI-E to tracheal suctioning in six ventilator-dependent adults with ALS and severe bulbar dysfunction and intercurrent respiratory tract infections, using a prospective cross-over study design. Deep tracheal suctioning was applied when indicated and MI-E was applied 30 min later. All participants felt that MI-E was more comfortable and effective than deep tracheal suctioning, and there were no adverse effects. The authors concluded that, for ventilated subjects with ALS, the use of MI-E can produce a greater improvement in oxygen saturation and other pulmonary variables than that obtained by suctioning alone.

Trebbia 2005 measured vital capacity and PCEF in 10 participants aged 16 to 80 years with neuromuscular disease during cough augmentation using mechanical insufflation (MI), MAC, a combination of both MI and MAC, and during unassisted coughing (baseline). Mechanical exsufflation was not used. Results suggested that both MI and MAC significantly improved PCEF and vital capacity compared to unassisted cough. Furthermore, both parameters were higher when combining MI and MAC than using either intervention alone. There was no difference in PCEF or vital capacity between MI and MAC.

Vitacca 2010) was an observational study of 39 adults aged 39 to 83 years with ALS (27 on ventilator support) who accessed different components of an on-demand consultation, MI-E and MAC, and oximetry feedback programme. Twenty-one participants used MI-E during home visits. The programme was reported to be feasible, well tolerated and cost effective.

Overall completeness and applicability of evidence

We identified five short-term studies for inclusion in this review, none of which randomly assigned participants to separate interventions. None of the included studies assessed long-term use of MI-E. The included studies were thus not able to assess the efficacy and safety of MI-E in reducing mortality and morbidity in people with NMDs.

Participant numbers were generally small, with no possibility of subgroup analyses for different age groups, conditions, or intervention methods. Very few participants in the paediatric age group were included and it is not appropriate to apply results from adult studies to paediatric practice. The pressures, times and application of MI-E amongst the different studies varied and were not consistently reported. As seen in Table 1, a variety of outcome measures were reported. All studies were conducted in Europe and therefore lack external validity and generalisability to other geographical areas and socioeconomic contexts.

Quality of the evidence

Five studies of 105 participants were included in this systematic review. Key limitations of the studies were: study design (multiple interventions randomly assigned to each participant); small sample sizes; lack of representation of different age groups and conditions; unreliable or clinically irrelevant outcome measures; and unclear to high risks of bias; specifically, poor blinding of outcome assessors, unclear methods of randomisation and allocation concealment and insufficient reporting of data. MI-E was compared with a large variety of alternative cough augmentation techniques, according to individual centre practice. None of the studies investigated clinically relevant endpoints of MI-E therapy for use in acute cases of respiratory infection or for chronic use in patients with respiratory insufficiency. The body of evidence included in this review does not allow any conclusions to be reached regarding the efficacy or safety of MI-E in people with a NMD.

Potential biases in the review process

Our search was comprehensive, including searches for unpublished studies through trial registration platforms and congress abstract reports. We did not contact MI-E manufacturing companies for data on file, and this may be a source of bias. There were no geographical or language constraints to the review. The corresponding authors of some identified studies were contacted to obtain missing results; however, we did not receive any replies to our enquiry emails. It is unlikely that obtaining these missing data would have impacted substantially on this review, as the missing outcomes were not primary or secondary outcomes prespecified in the protocol. We adhered to the published protocol in conducting this review.

We cannot rule out the possibility that some studies may have been missed, particularly if they were published in non peer-reviewed or non-accredited journals or presented at local congresses, and we cannot control for potential publication bias.

Agreements and disagreements with other studies or reviews

There were no appropriate studies or reviews with which to compare.

Authors' conclusions

Implications for practice

The results of this review do not provide sufficient evidence to guide clinical practice as they were unable to address important short- and longer-term outcomes, including adverse effects. There is currently insufficient evidence for or against the use of MI-E in people with neuromuscular disease.

Implications for research

Further research is required to establish the effects of mechanical insufflation-exsufflation in patients with a neuromuscular disease and acute respiratory infection, and for those using mechanical insufflation-exsufflation as maintenance therapy over the long term. Studies to establish the relative effectiveness of mechanical insufflation-exsufflation compared with other techniques alone and in combination are also warranted. It is recommended that adequately powered, high-quality randomised controlled trials that test the effects of a course of mechanical insufflation-exsufflation treatment rather than a single intervention be performed on participants with neuromuscular diseases. Sufficiently large sample sizes may be obtained by randomised controlled trials that involve multiple study sites. The end-points of these studies should be clinically meaningful; and objective, reliable outcome measures (including patient reported outcomes) should be fully reported, including measures of safety and efficacy. People with neuromuscular disease should be consulted when designing such clinical trials in order to ensure that measurements are appropriate to patient needs and experiences. Cost-effectiveness analyses are also warranted to compare the relative cost of therapy to efficacy and harms.

Sufficiently powered studies would enable meaningful subgroup analysis; for example, patients with and without scoliosis and patients with and without bulbar dysfunction. Studies specific to the paediatric population are urgently needed, as adult data cannot be applied to this population group, and the potential for adverse events in this group may be different. Further research is also warranted to determine the optimal pressures, insufflation and exsufflation times and frequency of mechanical insufflation-exsufflation in patients with neuromuscular disease.

Considering that peak cough expiratory flow was the most commonly reported outcome, future versions of this review should include peak cough expiratory flow as a secondary outcome measure. Studies of amyotrophic lateral sclerosis or motor neuron disease should be addressed in a separate review from hereditary neuromuscular disease owing to the different course and outcomes; and mechanical insufflation-exsufflation for acute and chronic respiratory failure should be addressed in two separate reviews with different outcome measures.

Acknowledgements

Brenda Morrow was funded in part by the Medical Research Council (MRC) of Southern Africa.

The editorial base of the Cochrane Neuromuscular Disease Group is supported by the Motor Neurone Disease Association, the Muscular Dystrophy Campaign and the UK MRC Centre for Neuromuscular Diseases.

The Cochrane Neuromuscular Group Trials Search Co-ordinator, Angela Gunn, carried out searches of the Cochrane Neuromuscular Disease Group Specialized Register, CENTRAL, MEDLINE and EMBASE.

Data and analyses

Download statistical data

This review has no analyses.

Appendices

Appendix 1. NMD Register (CRS) search strategy

aeration or insufflation or exsufflation or (cough and assist*) or (cough and augment*)

Appendix 2. CENTRAL search strategy

#1 (insufflation or exsufflation)
#2 cough NEAR/2 assist*
#3 cough NEAR/2 augment*
#4 (#1 OR #2 OR #3)
#5 MeSH descriptor Neuromuscular Diseases explode all trees
#6 MeSH descriptor Muscular Dystrophies explode all trees
#7 MeSH descriptor Muscular Atrophy explode all trees
#8 MeSH descriptor Muscular Diseases explode all trees with qualifiers: CN,ME
#9 MeSH descriptor Peripheral Nervous System Diseases explode all trees
#10 MeSH descriptor Neuromuscular Junction Diseases explode all trees
#11 MeSH descriptor Peripheral Nervous System Diseases explode all trees
#12 MeSH descriptor Motor Neuron Disease explode all trees
#13 "motor neuron disease" or "motor neurone disease" or "motoneuron disease" or "amyotrophic lateral sclerosis"
#14 "muscle weakness" or "metabolic myopath*" or "muscular dystroph*" or "neuromuscular junction"
#15 myopath* or neuropath*
#16 myopathy and (congenital or inherited or metabolic)
#17 myopathic and (congenital or inherited or metabolic)
#18 (#5 OR #6 OR #7 OR #8 OR #9 OR #10 OR #11 OR #12 OR #13 OR #14 OR #15 OR #16 OR #17)
#19 (#4 AND #18)

Appendix 3. MEDLINE (OvidSP) search strategy

Database: Ovid MEDLINE(R) <1946 to September Week 4 2013>
Search Strategy:
--------------------------------------------------------------------------------
1 randomized controlled trial.pt. (387443)
2 controlled clinical trial.pt. (89701)
3 randomized.ab. (285140)
4 placebo.ab. (156097)
5 drug therapy.fs. (1759458)
6 randomly.ab. (198174)
7 trial.ab. (300246)
8 groups.ab. (1269256)
9 or/1-8 (3282453)
10 exp animals/ not humans.sh. (4048625)
11 9 not 10 (2794973)
12 insufflation/ (1526)
13 (insufflation or exsufflation).mp. (5146)
14 (cough adj1 assist$).tw. (77)
15 (cough adj2 augment$).tw. (37)
16 or/12-15 (5223)
17 exp Neuromuscular Diseases/ (249877)
18 exp Muscular Dystrophies/ (21768)
19 muscle weakness/ (5649)
20 exp muscular atrophy/ (9614)
21 metabolic myopath$.mp. (405)
22 exp Muscular Diseases/cn [Congenital] (1811)
23 exp Muscular Diseases/me [Metabolism] (8244)
24 exp Neuromuscular Junction Diseases/ (16569)
25 exp Peripheral Nervous System Diseases/ (124547)
26 exp Motor Neuron Disease/ (19838)
27 muscular dystroph$.tw. (17597)
28 (myopath$ and (congenital or inherited or metabolic)).tw. (3235)
29 neuromuscular junction.tw. (6621)
30 neuropath$.tw. (89212)
31 exp Motor Neuron Disease/ (19838)
32 (moto$1 neuron$1 disease$1 or moto neuron$1 disease$1).mp. (6267)
33 amyotrophic lateral sclerosis.tw. (13269)
34 (myopath$ or neuropath$).tw. (106891)
35 or/17-34 (333545)
36 11 and 16 and 35 (24)
37 remove duplicates from 36 (22)

Appendix 4. EMBASE (OvidSP) search strategy

Database: Embase <1980 to 2013 Week 40>
Search Strategy:
--------------------------------------------------------------------------------
1 crossover-procedure.sh. (38578)
2 double-blind procedure.sh. (118011)
3 single-blind procedure.sh. (18318)
4 randomized controlled trial.sh. (357438)
5 (random$ or crossover$ or cross over$ or placebo$ or (doubl$ adj blind$) or allocat$).tw,ot. (1005116)
6 trial.ti. (153634)
7 or/1-6 (1142249)
8 exp animal/ or exp invertebrate/ or animal.hw. or non human/ or nonhuman/ (20083653)
9 human/ or human cell/ or human tissue/ or normal human/ (14931987)
10 8 not 9 (5184303)
11 7 not 10 (1002342)
12 limit 11 to embase (766843)
13 aeration/ (7760)
14 (insufflation or exsufflation).mp. (5936)
15 (cough adj1 assist$).mp. (166)
16 (cough adj2 augment$).mp. (46)
17 or/13-16 (11766)
18 exp neuromuscular disease/ (128421)
19 exp muscular dystrophy/ (30270)
20 exp muscle weakness/ (225530)
21 muscle atrophy/ (20918)
22 metabolic myopath$.mp. (624)
23 exp neuromuscular junction disorder/ (22941)
24 exp peripheral neuropathy/ (49105)
25 exp motor neuron disease/ (26471)
26 (myopath$ and (congential or inherited or metabolic)).mp. (3580)
27 neuromuscular junction.mp. (8301)
28 exp motor neuron disease/ (26471)
29 (moto$1 neuron$1 disease$1 or moto neuron$1 disease$1).mp. (9148)
30 amyotrophic lateral sclerosis.mp. (23713)
31 (myopath$ or neuropath$).mp. (229167)
32 or/18-31 (492516)
33 12 and 17 and 32 (26)

Appendix 5. ClinicalTrials.gov and WHO ICTRP search strategy

cough and assist*

cough and augment*

insufflation or exsufflation

Contributions of authors

Brenda Morrow: drafted the protocol; developed search strategy criteria; searched identified titles and abstracts; obtained copies of trials; selected trials for inclusion, extracted data; performed data entry, conducted analysis and interpretation; and completed final write-up of the review.

Heleen van Aswegen: searched identified titles and abstracts for trials; selected trials for inclusion; interpreted the analysis; and assisted with final write-up of the review.

Marco Zampoli: extracted data; interpreted the analysis; and assisted with final write-up of the review.

Andrew Argent: assisted with protocol development; checked data entry; interpreted the analysis; and assisted with final write-up.

Declarations of interest

The authors have no conflicts of interest to declare.

Sources of support

Internal sources

  • No sources of support supplied

External sources

  • Medical Research Council of South Africa, South Africa.

    Brenda Morrow was supported in part by a research grant from the MRC.

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Chatwin 2003

MethodsProspective randomised trial, comparing MI-E to other cough augmentation techniques
Participants

22 adults and children (aged 10 to 56 years; 6 female) with NMD and a history of recurrent chest infections or ineffective cough were included. NMD conditions were: intermediate spinal muscle atrophy (n = 10); Duchenne muscular dystrophy (n = 6); poliomyelitis (n = 3); other congenital myopathies (n = 3)

Comorbid conditions occurred in all patients: 11 had severe scoliosis with a spinal jacket and 11 had previous spinal surgery

Participants were studied during a period of clinical stability; therefore, exclusion criteria were antibiotic therapy in the preceding month, resting arterial oxygen saturation < 90% or end-tidal carbon dioxide tension of > 7kPa or both

Additional exclusion criteria were severe bulbar dysfunction and a previous history of pneumothorax

19 age-matched controls were also recruited from staff and their families.

Interventions

For all participants, unassisted cough was compared, in random order, with standard physiotherapy and assisted cough; cough after inspiration supported by noninvasive positive pressure ventilation (BiPAP); exsufflation-assisted cough with negative pressure initiated manually at the end of inspiration; insufflation given manually during the inspiratory phase; and in-exsufflation-assisted cough with delivery of negative pressure immediately prior to coughing

The CoughAssist device was used (JH Emerson Co, Cambridge, USA)

Insufflation and exsufflation pressures were titrated to patient comfort, with mask pressure for the NMD group 15 ± 3 cmH2O to -15 ± 9 cmH2O; and for the control group 17 ± 5 cmH2O and -8 ± 11 cmH2O respectively

A face mask interface was used

Outcomes

Peak cough flow

Patient Comfort Score (visual analogue scale, VAS)

Adverse events not reported

Funding sources

Jennifer Trust for Spinal Muscular Atrophy (Stratford upon Avon, UK)

Brompton Breathers Trust Fund (London, UK)

Cystic Fibrosis Trust (Bromley, UK)

British Lung Foundation (London, UK)

Association Francaise Contre Les Myopathies (Paris, France)

Declarations of interestNone reported
Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskThe method used for randomisation was not reported
Allocation concealment (selection bias)Unclear riskThe method used for allocation concealment was not reported
Blinding of participants and personnel (performance bias)
All outcomes
High risk

Given the nature of the intervention it is unlikely that participant or clinician blinding was possible

The use of blinding was not reported in the study

Blinding of outcome assessment (detection bias)
All outcomes
Unclear riskNot reported
Incomplete outcome data (attrition bias)
All outcomes
Low riskAll participants completed the study and were included in the analysis
Selective reporting (reporting bias)Low riskAll outcome measures were reported
Other biasUnclear riskOther potential biases were not discussed

Chatwin 2009

Methods

Two-day randomised cross-over trial comparing standardised chest physiotherapy with MI-E and standardised chest physiotherapy without MI-E

Participants received MI-E for one treatment session and no MI-E for a second treatment session, in a randomly assigned order, with a reverse cross-over the following day

Participants

Participants were more than three years of age with confirmed NMD and an acute respiratory tract infection

They were recruited from the adult and paediatric wards at Royal Brompton Hospital

Participants were all users of nocturnal NIV

Eight participants (6 male) with median (range) age 21.5 (4 to 44) years. Underlying conditions were: Duchenne muscular dystrophy (n = 4); spinal muscular atrophy type 2 (n = 3); and congenital myopathy (n = 1)

Comorbid conditions were not reported

Exclusion criteria were: pneumothorax, tracheostomy, severe bulbar weakness, severe uncontrolled asthma, rapidly progressive chest infection with failure to control blood gas tension with NIV, and patients being weaned off NIV after intubation

Interventions

Standardised airway clearance therapy with MI-E was compared to standardised airway clearance therapy without MI-E in all participants

Standardised airway clearance therapy included modified active cycle of breathing technique (ACBT) on NIV, with or without manual techniques of clapping and shaking and forced expiratory technique plus manually assisted cough

The MI-E treatment consisted of the same treatment as described above with the addition of in-exsufflation (in manual mode) during the cough manoeuvre and a manual assisted cough

The CoughAssist (Philips Respironics, Murraysville, Pennsylvania) was used

The in-exsufflation pressures were +20 cmH2O (range +15 to +35 cmH2O) and -20 cmH2O (range -20 to -40 cmH2O). Insufflation and exsufflation times were 2 to 4 and 4 to 5 s respectively

A facemask interface was used

Participants were treated for 30 min and then reassessed

If treatment was considered incomplete at reassessment, the session was continued and the extra treatment time recorded

Outcomes

Primary outcome measures of heart rate, transcutaneous oxygen saturation (SpO2) and transcutaneous carbon dioxide tension (PtcCO2) were continuously measured.

Treatment time up to 30 min and continued treatment time beyond 30 min were recorded.

Subjective outcome measures were auscultation score; VAS for comfort, breathlessness, mood, secretion production, and fatigue.

Funding sourcesJennifer Trust for Spinal Muscular Atrophy (Stratford upon Avon, UK)
Declarations of interestDr Chatwin disclosed a relationship with Breas Medical, Mölnlycke, Sweden
Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskThe method used for randomisation was not reported
Allocation concealment (selection bias)Unclear riskThe method used for allocation concealment was not reported
Blinding of participants and personnel (performance bias)
All outcomes
High risk

Given the nature of the intervention it is unlikely that participant or clinician blinding was possible

The use of blinding was not reported in the study

Blinding of outcome assessment (detection bias)
All outcomes
Unclear riskPerson performing auscultation and scoring blinded to allocation, but it is not clear if other outcome assessors also blinded
Incomplete outcome data (attrition bias)
All outcomes
Low riskAll participants completed the study and were included in the analysis
Selective reporting (reporting bias)High riskData for the primary physiological outcome parameters were not reported
Other biasHigh riskHigh level of assessor bias for the outcome measure of treatment time

Mustfa 2003

MethodsProspective randomised trial, comparing MI-E to other cough augmentation techniques
Participants47 participants with bulbar (n = 21; 9 women) and non bulbar (n = 26; 6 women) amyotrophic lateral sclerosis and 10 healthy volunteers (4 women) were enrolled.
Interventions

Each participant had 3 reproducible cough flows measured with:

  1. Maximal unaided coughs

  2. Manually assisted cough using abdominal pressure

  3. Manually initiated exsufflation using the mechanical in-exsufflator device (the negative pressure was gradually titrated to the maximum tolerated exsufflation), initiated just prior to coughing

  4. Insufflation with the in-exsufflator incrementally increased to the maximum tolerated pressure prior to a maximal cough;

  5. MI-E coordinated with the patients' cough efforts, using a mechanical in-exsufflator (JH Emerson Co, Cambridge, USA) (ME-I interface was a face mask)

In-exsufflation pressures and times were not reported

Outcomes

The following cough flows and pressures were measured during cough augmentation techniques: transpulmonary pressure; cough gastric pressure; PCEF; and time integrated cough volume

Adverse events were not reported

Funding sources

Muscular Dystrophy Association of America

Motor Neurone Disease Association

The King's MND Care and Research Centre

National Health Services Research and Development

Declarations of interestNone reported
Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskThe method used for randomisation was not reported
Allocation concealment (selection bias)Unclear riskThe method used for allocation concealment was not reported
Blinding of participants and personnel (performance bias)
All outcomes
High risk

Given the nature of the intervention it is unlikely that participant or clinician blinding was possible

The use of blinding was not reported in the study

Blinding of outcome assessment (detection bias)
All outcomes
Unclear riskBlinding of evaluators was not reported
Incomplete outcome data (attrition bias)
All outcomes
Low riskAll participants completed the study and were included in the analysis
Selective reporting (reporting bias)Low riskAll outcome measures were reported
Other biasUnclear riskOther sources of bias were not clearly described

Senent 2011

Methods

Prospective randomised trial, comparing MI-E to other manual and instrumental cough augmentation techniques

Randomisation was reported only for the order of the instrumental techniques, therefore the manual cough augmentation arm of the study was not included in this review

Participants

Stable patients with amyotrophic lateral sclerosis, who had been on home mechanical ventilation for > 2 months, were enrolled during scheduled routine day-hospital visits

Exclusion criteria were: occurrence of any "respiratory event" in the preceding month; presence of a tracheostomy; unassisted peak cough flow > 270 L/min

28 patients were screened; 16 participants (12 male) were included in the study; 9 bulbar and 7 non-bulbar

Age (range) of included participants was 63 (57 to 68) years

Interventions

One hour after performing three manual cough techniques, the following instrumental cough augmentation techniques were applied to each participant in random order at 10 to 15 min intervals:

  1. Expiratory abdominal thrust after air stacking on spontaneous deep breath, using an Ambu® Silicone Resuscitator (Ambu, Ballerup, Denmark)

  2. Expiratory abdominal thrust from end-inspiratory volume using bi-level pressure ventilator with normal settings

  3. Expiratory abdominal thrust from end- inspiratory volume obtained by increasing inspiratory positive airway pressure, iPAP, to +30 cmH2O

  4. MI-E assisted cough using a face mask interface. Maximum insufflation and exsufflation pressure were gradually increased to -40 cmH2O and +40 cmH2O. Four to six in- exsufflation cycles were given with a 1 to 3 s inter-cycle pause. Insufflation and exsufflation times were not reported. The CoughAssist Mechanical In-Exsufflator (Respironics, Murraysville, Pennsylvania, USA) was used

Outcomes

PCEF and efficacy and comfort VAS

Adverse events were not reported

Funding sourcesAssociation pour le Développement et l'Organisation de la Recherche (ADOREP), Paris, France; Association d'Entraide des Polio et Handicapés (ADEP)
Declarations of interest

Jesus Gonzalez-Bermejo received 2000 Euro in 2009 from Philips-Respironics for educational fees regarding cough assistance

The other authors have no conflicts of interest

Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNot reported
Allocation concealment (selection bias)Unclear riskNot reported
Blinding of participants and personnel (performance bias)
All outcomes
High risk

Given the nature of the intervention it is unlikely that participant or clinician blinding was possible

The physiotherapists were blinded to peak expiratory flow measurements, but not to allocation

Blinding of outcome assessment (detection bias)
All outcomes
Unclear riskNot reported
Incomplete outcome data (attrition bias)
All outcomes
Low riskAll participants completed the study and included in the analysis
Selective reporting (reporting bias)Low riskAll outcomes reported
Other biasHigh riskAuthor disclosed financial relationship with manufacturer of MI-E device

Sivasothy 2001

  1. a

    PCEF: peak cough expiratory flow

MethodsProspective randomised trial comparing MI-E to manually assisted coughing (MAC), and a combination of MI-E and MAC
Participants

The study enrolled 12 participants (1 female) with respiratory muscle weakness; 8 participants with chronic obstructive pulmonary disease (COPD) and 9 healthy volunteers

Participants with muscle weakness had amyotrophic lateral sclerosis (n = 7); Becker's muscular dystrophy (n = 1); previous poliomyelitis (n = 2); Duchenne muscular dystrophy (n = 1) and spinal muscular dystrophy (n = 1)

Four patients had scoliosis and 8 did not have scoliosis

Interventions

The following interventions were performed in random order on each participant:

  1. MAC consisted of manual thoraco-abdominal compression during the expulsive phase of a maximal voluntary cough. MAC hand position for those with scoliosis was on the hyperinflated hemithorax

  2. MI-E was performed using the In-Exsufflator (JH Emerson Co, Cambridge, MA, USA) set to give 20 cmH2O inspiratory and -20 cmH2O expiratory pressure. Two in-exsufflation cycles were used and after a 3rd inspiration, participants were asked to make a maximal voluntary cough without the aid of negative pressure

  3. Combination of 1 and 2 above

At least 5 min elapsed between cough manoeuvres and the best result of at least 3 attempts was analysed

Outcomes

PCEF

Cough expiratory volume

Peak value time (time from initiating positive pressure flow to the peak cough flow)

Oesophageal and gastric pressures

Funding sourcesNot reported
Declarations of interestNot reported
Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomisation method not reported
Allocation concealment (selection bias)Unclear riskNot reported
Blinding of participants and personnel (performance bias)
All outcomes
High risk

Given the nature of the intervention it is unlikely that participant or clinician blinding was possible

The use of blinding was not reported in this study

Blinding of outcome assessment (detection bias)
All outcomes
Unclear riskNot reported
Incomplete outcome data (attrition bias)
All outcomes
Low riskAll participants completed the study and included in the analysis
Selective reporting (reporting bias)Low riskAll outcome measures reported
Other biasUnclear riskOther potential sources of bias not clearly described

Characteristics of excluded studies [ordered by study ID]

StudyReason for exclusion
  1. a

    ALS: amyotrophic lateral sclerosis
    MI-E: mechanical insufflation-exsufflation
    PCEF: peak cough expiratory flow

Sancho 2003

Lack of randomisation or quasi-randomisation (Sancho 2003)

A comparison of MI-E to tracheal suctioning in six ventilator-dependent adults with ALS and severe bulbar dysfunction and intercurrent respiratory tract infections, using a prospective cross-over study design

Deep tracheal suctioning was applied when indicated and MI-E was applied 30 min later

Trebbia 2005

Mechanical insufflation with intermittent positive pressure breathing was used, without exsufflation (Trebbia 2005)

Measured vital capacity and PCEF in 10 participants aged 16 to 80 years with neuromuscular disease during cough augmentation using mechanical insufflation (MI), manually assisted cough (MAC), a combination of both MI and MAC, and during unassisted coughing (baseline)

Vitacca 2010

Lack of randomisation or quasi-randomisation (Vitacca 2010)

An observational study of 39 adult participants with amyotrophic lateral sclerosis utilising a telephone-accessed consultation, MI-E and MAC, and pulse oximetry feedback program

Characteristics of ongoing studies [ordered by study ID]

NCT01518439

Trial name or titleInstrumental and Manual Increase of Couch [sic] in Neuromuscular Patients
MethodsInterventional efficacy pilot study with randomised, single group assignment
Participants20 adult participants with neuromuscular disorders and respiratory restrictive syndrome
InterventionsComparison of Alpha 200 ® device; Alpha 200 ® + physiotherapy; CoughAssist device; CoughAssist device + physiotherapy; and physiotherapy
OutcomesPrimary outcome cough flow from the combinations of mechanical and manual cough assistance
Starting dateJanuary 2012
Contact informationFrederik Lofaso and Helene Prigent. f.lofaso@rpc.ap-hop-paris.fr
Notes 

Ancillary