Study design
The Ethics Committee (Comité de Protection des Personnes) of Saint-Germain-en-Laye, France, approved the trial protocol. Participants or their legally authorized next of kin provided written informed consent before inclusion whenever possible. Otherwise, deferred written informed consent was obtained from patients. This investigator-led trial was publicly funded. This single-center, controlled trial, conducted with two parallel groups aimed to evaluate the impact of early initiation of mechanical ventilation on pneumonia occurrence in GBS patients. All authors had full and independent access to all data, and vouch for the integrity, accuracy, and completeness of the data and analysis, and for the adherence to study protocol.
The trial was registered at ClinicalTrials.gov under the number NCT00167622, before inclusion of the first patient.
Participants
Adults patients admitted to the intensive care unit with Guillain-Barré syndrome were enrolled in the study if they met all the following risk factors for endotracheal intubation, (1) time from onset to admission of less than 7 days, (2) inability to lift the head, and (3) forced vital capacity of < 60% of predicted. Patients were excluded when they had either criterion of the presence of respiratory distress signs, PaC02 greater than 6.4 kPa, PaO2 of less than 7.5 kPa, and forced vital capacity of less than 20% of predicted value or less than 15 mL/kg of body weight. Others exclusion criteria were an age below 18 years old, altered consciousness (Glasgow coma score < 8), pregnancy, hemodynamic instability, pre-existing pneumonia.
Randomization and masking
Patients were randomized 1:1 to early mechanical ventilation or to physiotherapy and oxygen whenever needed (control group). Randomization was based on computer-generated variable block sizes stratified according to presence or not of swallowing difficulties. The randomization sequence was prepared by the study statistician who did not take part in randomization. Allocation was concealed by means of opaque sealed envelopes until qualifying patients were consented and ready for mechanical ventilation. Health care professionals taking part in the intervention were aware of the treatment assignment, because they were responsible for implementing the designated respiratory management. However, assessors of the primary endpoint were fully blinded to treatment. Investigators were unaware of all data until the trial concluded.
Interventions
In all aspects of treatment, except regarding respiratory management, all patients in both groups of the trial, were treated according to the most recent guidelines for the management of Guillain-Barré syndrome, including immunotherapy [17]. Intravenous immunoglobulins were chosen in case of suspected infection, they were perfused at 0.5 g/kg daily for 5 days. Plasma exchange were realized at number of 4 amounting to 200 to 250 mL/kg with albumin 20% or crystalloids or colloids solution. All patients with suspected hospital acquired pneumonia were treated following identification of pathogens. In patients with sepsis, empiric antibiotherapy was initiated including one of piperacillin–tazobactam, cefepime, or imipenem, meropenem, owing to the intensive care unit bacterial ecology at the time the trial was conducted.
In the control arm, patients benefited from physiotherapy, including active clearance of bronchial secretion and oxygen via a face mask whenever needed. In this group of patients, mechanical ventilation with endotracheal intubation was not permitted unless acute respiratory failure occurred defined by the presence of respiratory distress signs, PaC02 greater than 6.4 kPa, arterial oxygen tension of less than 7.5 kPa, and forced vital capacity of less than 20% of predicted value or less than 15 mL/kg of body weight.
In the experimental arm, mechanical ventilation was initiated immediately after randomization either via a full-face mask or endotracheal intubation owing to the presence or absence of impaired swallowing (early mechanical ventilation—EMV). If patients underwent acute respiratory failure or new swallowing impairment, they received endotracheal intubation. They benefited daily from physiotherapy including active clearance of bronchial secretion with the same ex-insufflation maneuver as the control arm.
Investigated parameters
At baseline, we systematically recorded age, gender, Knaus disability scale [18], SAPSII [19], MRC sumscore [20], the presence of facial palsy or of swallowing impairment, spirometric data (forced vital capacity (mL and percent of predicted value), inspiratory pressure and expiratory pressure (cm of water) and arterial blood-gas.
We collected daily over the first week and weekly up to day 60 post-randomization, vital signs, core temperature (°C), biological data (liver enzymes, white blood cells count, PaO2/FiO2 ratio), bacteriological data (bloodstream culture and culture of sampling from any suspected site of infection) and immunotherapy (plasma exchange or intravenous immunoglobulin). We evaluated daily swallowing impairment by clinical deglutition test including the Volume-Viscosity Swallow Test.
Outcomes
Primary outcome was the incidence of first episode ventilator-acquired pneumonia up to ICU discharge (or 90 days, pending which occurred first).
Secondary outcomes included time to onset of pneumonia, time to and on mechanical ventilation, length of hospital stay, requirement for tracheostomy, mortality and any serious adverse events (septic shock, hepatic failure, renal failure, intravascular coagulopathy, acute respiratory distress syndrome).
Definitions
The definition for ventilator-acquired pneumonia was adapted from ATS/IDSA criteria [21] and required the new onset, i.e. > 48 h following randomization, of lung infiltrate or progressive radiographic infiltrate, a temperature higher than 38.3 °C, purulent secretions, leukocytosis or leukopenia (white blood cell count higher than 12,000/mm3 or lower than 4000/mm3), and altered oxygenation.
If pneumonia was suspected, broncho-alveolar lavage or protected brush bronchial samples were performed. Microbiological studies of the respiratory samples were considered positive if the number of colony-forming units (CFUs) for any isolated pathogen was greater than 104/ml.
The evaluation of the presence or absence of ventilator-acquired pneumonia was made by three independent adjudicators (an infectious disease physician, a radiologist and a pneumologist) who were blinded to study treatments.
Statistical analysis
The sample size was based on of the estimation that ventilator-acquired pneumonia will occur in 75% of patients with Guillain-Barré syndrome and dependent on mechanical ventilation [16]. Then, to detect a 50% absolute reduction in the incidence of pneumonias with a 2-sided significance level of 0·05 and power of 80%, we calculated that 25 subjects should be included per study arms, for a total of 50 participants.
The statistical analysis was performed according to the intent-to-treat principle, and once after all included patients reached the last time point of follow-up. Statistical data are reported as median [interquartile] for continuous variables and number (percentage) for qualitative variables.
The time to onset of pneumonia up to ICU discharge (or 90 days, pending which occurred first) was compared using the Gray’s test. Others outcomes were compared using the Wilcoxon signed-rank test or Fisher exact test as appropriate.
The open-source R software version 2.15.2 (2012-10-26), and LATEX, on a platform i386−w64−mingw32 was used for all statistical tests. Figures and tables were automatically created with Sweave.
Role of funding source
The trial was funded by Assistance Publique Hôpitaux de Paris. The funder had no role in designing the trial, collecting or analyzing the data, data interpretation, writing the manuscript, or decisions related to submission for publication.