Mechanical insufflation/exsufflation compared with standard of care in patients with pneumonia

Editor, Mechanical insufflation/exsufflation (MIE) was invented to help mobilise secretion from the lower respiratory tract by simulating a cough.1 Delivered by a Cough Assistor device through a face mask, MIE delivers a positive inspiratory pressure followed immediately by a negative expiratory pressure. MIE is widely used in patients with neuromuscular disease and reduced cough strength, where it prevents hospital admissions, improves survival rates, delays time to tracheostomy and has proven to be well tolerated with few adverse effects.1–3 There are no published studies evaluating MIE in patients with respiratory tract infections without neuromuscular disease. We hypothesised that MIE would be safe and well tolerated in patients hospitalised because of pneumonia. The study was conducted at the intermediate care unit (IMCU) at the Department of Infectious Diseases, Odense University Hospital, Denmark, between January 2014 and October 2015. Thirty consecutive patients admitted with pneumonia underwent 1 : 1 nonblinded randomisation to receive either the normal standard of care or MIE added to the standard of care (Fig. 1). Standard supportive treatment in patients admitted to the IMCU with pneumonia includes intermittent continuous positive airway pressure (CPAP) to improve oxygenation.Fig. 1: CONSORT diagram displaying the inclusion of patients in the study. COPD, chronic obstructive pulmonary disorder; CPAP, continuous positive airway pressure; IMCU, intermediate care unit; MIE, mechanical insufflation/exsufflation; SOC, standard of care.The primary outcome of this pilot study was the safety and tolerability of MIE. Secondary outcomes were improvement in oxygen saturation, oxygen therapy, respiratory rate, need of suction, admission time in IMCU and need for mechanical ventilation. MIE was administered through NIPPY Clearway (B&D Electromedical, Stratford-Upon-Avon, Warwickshire, United Kingdom) with standard settings of five rounds with five cycles of insufflation 2 s at 20 cmH2O, exsufflation 3 s at –20 cmH2O and oxygen flow up to 15 l min–1.4 The number of MIE treatments was prescribed by the physician in charge, and given by trained physiotherapists or nurses. Demographic, clinical and paraclinical data including comorbidities5 were registered for each patient. Statistical analyses, on an intention to treat basis were undertaken using STATA V13.0, with Fisher's exact test for categorical variables and Wilcoxon rank-sum for continuous variables, respectively. A P value less than 0.05 was considered statistically significant. The study was approved by the Danish Data Protection Agency (j.nr.2008-58-0035), Danish National Committee on Health Research Ethics (Damhaven 12, 7100 Vejle, Denmark, Project-ID S-20130102, Chairperson Birger Møller, approved 13.11.2013) and registered with ClinicalTrials.gov (identifier NCT03714321). Written informed consent was obtained from all participants. Characteristics for all 30 patients are summarised in Table 1. The median age was 67.5 years (IQR 56 to 79), 17 were men. Apart from a significantly higher prevalence of cardiovascular disease in the MIE group (P = 0.03), no significant differences in age, sex, CURB-65 score6 or comorbidities were found between the groups.Table 1: Characteristics and outcomes of hospital stays for 30 patients with pneumonia admitted to an intermediate care unitThe mean oxygen saturation at time of IMCU admittance was 94.1% (SD 3.2) in the MIE group and 92.9% (SD 3.9) in the control group (P = 0.43) (Fig. 2a), continuously higher during the first 24 h in the MIE group (mean 94.7% SD 3.0) compared with control group (mean 93.7%; SD 3.5) (P = 0.006).Fig. 2: Development over the first 24 h after admittance to an intermediate care unit in 30 patients with pneumoniae. Red: patients receiving mechanical insufflation/exsufflation. Blue: control group receiving standard of care. (a) Mean oxygen saturation (the y-axis starts at an oxygen saturation of 90%). (b) Oxygen supply. (c) Mean respiratory rate (the y-axis starts at a respiratory rate of 15 breaths/minute).All patients required oxygen therapy when entering IMCU. The MIE group had significantly lower oxygen need (mean 5.7 l/min SD 4.9) compared with the control group (mean 6.4 l/min SD 4.2) throughout the first 24 h in IMCU (P = 0.02) (Fig. 2b). Four patients no longer required oxygen therapy at time of IMCU discharge; all in the MIE group. The mean respiratory rate at IMCU admittance was 25.3 (SD7.3) breaths/minute in the MIE group, 23.3 (SD5.5) breaths/minute in the control group (P = 0.47) (Fig. 2c). Only four patients had a normal respiratory rate (12–16 breaths/minute), the rest were tachypnoeic. After 24 h, the mean respiratory rates were 22 (SD3.4) and 23.5 (SD5.6) breaths/minute in the MIE and control group, respectively (P = 0.59). Ten patients required suction (seven trachea/three pharynx) a mean 5.5 times (range 1 to 28) within the first 24 h of IMCU admittance. Of these, four were in the MIE group. After 24 h, eight patients still needed tracheal suction; only two in the MIE group. One patient (age 64 years, Charlson comorbidity score 6) did not tolerate MIE and discontinued treatment before IMCU discharge. Another two patients (age 70 and 79 years, respectively) had problems cooperating with the Cough Assistor, and therefore, received less frequent treatments than prescribed. One patient in the control group did not tolerate CPAP, which was discontinued. No adverse events related to the use of MIE were registered. The mean time of hospitalisation was 16.9 days (range 3 to 70), 61 h in the IMCU (range 2 to 163). Six patients were transferred to the ICU because of respiratory failure, three from each group. One patient from the MIE group had only received one MIE treatment prior to transfer. Three patients in the control group and two patients receiving MIE required mechanical ventilation. Three patients (10%) died during hospitalisation. Four patients were readmitted within 30 days of discharge; three were in the control group (P = 0.32). In this first study to evaluate MIE in hospitalised patients with pneumonia without neuromuscular disorders, our most important findings were the fast bettering in oxygen saturation and reduced need of oxygen therapy in patients receiving MIE. Furthermore, patients in the MIE group required less tracheal and pharyngeal suction compared with controls. Investigating safety and tolerability of MIE in patients with pneumonia was our primary aim. We chose, based on prior publications, standard settings for insufflation/exsufflation, which was generally tolerated well.1,2,7 Only one patient discontinued treatment, and no adverse events related to the use of MIE was reported. However, the standard settings of the Cough Assistor needed adjusting in two older patients. This may represent a source of bias. Unlike patients with chronic neuromuscular disorders and long-term use of MIE, our patients used it a mean 2.7 days. Choosing the right settings can be challenging in short-term use, and lower inspiratory/expiratory pressure settings should be considered in older, weaker patients to ensure treatment cooperation. In conclusion, we found MIE to be well tolerated and comparable with the current standard of care on all outcomes. Acknowledgments relating to this article Assistance with the letter: We thank Kim Oren Gradel, Cand.med.vet., PhD, at the Department of Epidemiology, Odense University Hospital, for help with the statistical analysis. The authors wish to thank the nurses and physiotherapists at the Department of Infectious Diseases, Odense University Hospital, for administrating the MIE. Financial support and sponsorship: none. Conflicts of interest: JV has received honoraria from AstraZeneca, Boehringer-Ingelheim, Chiesi and Novartis. For the remaining authors none were declared.