Abstract
Rationale: The use of noninvasive ventilation (NIV) as an early weaning/extubation technique from mechanical ventilation remains controversial.
Objectives: To investigate NIV effectiveness as an early weaning/extubation technique in difficult-to-wean patients with chronic hypercapnic respiratory failure (CHRF).
Methods: In 13 intensive care units, 208 patients with CHRF intubated for acute respiratory failure (ARF) who failed a first spontaneous breathing trial were randomly assigned to three groups: conventional invasive weaning group (n = 69), extubation followed by standard oxygen therapy (n = 70), or NIV (n = 69). NIV was permitted as rescue therapy for both non-NIV groups if postextubation ARF occurred. Primary endpoint was reintubation within 7 days after extubation. Secondary endpoints were: occurrence of postextubation ARF or death within 7 days after extubation, use of rescue postextubation NIV, weaning time, and patient outcomes.
Measurements and Main Results: Reintubation rates were 30, 37, and 32% for invasive weaning, oxygen-therapy, and NIV groups, respectively (P = 0.654). Weaning failure rates, including postextubation ARF, were 54, 71, and 33%, respectively (P < 0.001). Rescue NIV success rates for invasive and oxygen-therapy groups were 45 and 58%, respectively (P = 0.386). By design, intubation duration was 1.5 days longer for the invasive group than in the two others. Apart from a longer weaning time in NIV than in invasive group (2.5 vs. 1.5 d; P = 0.033), no significant outcome difference was observed between groups.
Conclusions: No difference was found in the reintubation rate between the three weaning strategies. NIV decreases the intubation duration and may improve the weaning results in difficult-to-wean patients with CHRF by reducing the risk of postextubation ARF. The benefit of rescue NIV in these patients deserves confirmation.
Clinical trial registered with www.clinicaltrials.gov (NCT 00213499).
The use of noninvasive ventilation (NIV) as an early weaning/extubation technique from invasive mechanical ventilation remains controversial.
NIV used as an early weaning/extubation technique in difficult-to-wean patients with chronic respiratory failure did not reduce the reintubation rate within 7 days as compared with conventional weaning and early extubation with standard oxygen therapy. Nevertheless, NIV may improve the weaning results in these patients by shortening the intubation duration and reducing the risk of postextubation acute respiratory failure The potential benefit of rescue postextubation NIV in these patients needs further study
Noninvasive ventilation (NIV) has been recently developed for the management of weaning/extubation from invasive mechanical ventilation (MV) and postextubation acute respiratory failure (ARF) (1), the main goal being to shorten intubation time and to prevent or avoid reintubation and subsequent complications (2, 3). The weaning/extubation period represents an important clinical issue for clinicians and patients, and prediction of its outcome may be difficult in most weak patients. Difficult weaning requiring a progressive withdrawal from MV may occur, in fact, in 25% of intensive care unit (ICU) patients (4, 5) and in 40 to 60% of patients with chronic obstructive pulmonary disease (COPD) (6, 7). The weaning time may also account for up to 40% of the total invasive MV duration (8, 9). Moreover, reintubation may be necessary within 48 to72 hours in 5 to 25% of planned extubation, even if a spontaneous breathing trial (SBT) has been successful (3, 10). Reintubation represents per se an independent risk factor for nosocomial pneumonia, increasing ICU and hospital stay as well as mortality (3, 8). Hence, the ICU clinician has to find the optimal compromise between the risks of unduly prolonged intubation and those of too early weaning and extubation process (11, 12). Therefore, any strategy with the aim of reducing morbidity and mortality of prolonged invasive MV or reintubation appears relevant and should be developed to improve patient prognosis.
Consequently, NIV has been evaluated as an early weaning and extubation technique in difficult-to-wean patients (13–17). Despite encouraging results regarding the incidence of reintubation, complications, and patient outcome, the role of NIV in this indication remains debated (11, 18, 19). A recent metaanalysis found that a noninvasive weaning strategy could be of potential benefit as compared with conventional invasive weaning, particularly in patients with COPD (20). However, these authors acknowledged that larger controlled trials were still needed. In addition, despite the negative results of NIV to treat postextubation ARF (i.e., rescue postextubation NIV) in medical patients (21, 22), the interest of this approach probably requires further evaluation in more selected medical populations.
We conducted a prospective randomized multicenter study to investigate the effectiveness of NIV as an early weaning/extubation technique in patients with chronic hypercapnic respiratory failure (CHRF) considered difficult to wean from invasive MV. We also evaluated the role of rescue postextubation NIV when a postextubation ARF occurred in these patients.
Some of the results of this study have been previously reported in the form of an abstract (23).
The study was approved by the Local Ethics Committee of Rouen University Hospital and all patients or their families gave their signed informed consent before inclusion. The methodology used has been previously published elsewhere (24).
Patients were consecutively recruited between January 2002 and March 2006 from 13 French and Tunisian ICUs experienced in NIV for more than 5 years. Eligible patients were those with known or suspected CHRF based on their clinical history (COPD, persistent asthma, bronchiectasis, obesity-hypoventilation syndrome, chest wall deformity, sequelae of pulmonary tuberculosis), chest X-ray, arterial blood gases (ABG) in steady state, and/or admission bicarbonates level, as well as previous pulmonary function tests if available. They had to be intubated for at least 48 hours for ARF regardless of etiology. They also had to be clinically stable for at least 24 hours to undergo an SBT after they had met the following weaning criteria based on a daily screening evaluation: PaO2/FIO2 greater than or equal to 150 mm Hg, positive end expiratory pressure (PEEP) less than or equal to 5 cm H2O, no vasopressor or sedation and Glasgow Coma Score greater than or equal to 12, effective cough (11). Patients were included in the study only if they exhibited SBT intolerance, administered via a T-piece, occurring between 5 minutes and 2 hours (7). SBT intolerance was based on one of the following clinical and/or ABG criteria if available: sweats, agitation, somnolence, or consciousness impairment; respiratory rate greater than 35/min or greater than or equal to 50% increase relative to baseline; decrease in SaO2 greater than or equal to 5%; increase in heart rate or systolic blood pressure greater than or equal to 20%; PaO2 less than or equal to 60 mm Hg (8 kPa) with FIO2 of 40% and/or pH less than 7.35. For the purpose of the study, these patients, who failed a first SBT, were considered as difficult to wean from MV. In the event of SBT intolerance within 5 minutes, suggesting insufficient ventilatory autonomy, the patient was reventilated in assist-controlled mode and SBT was attempted again the next day. The noninclusion criteria were as follows: good SBT tolerance for at least 2 hours, the patient being then extubated; hemodynamic and/or clinical respiratory instability; initial intubation considered as difficult; suspected swallowing disorders; ineffective cough and/or persistent bronchial hypersecretion at the time of weaning; uncooperative patient; contraindications for the use of a face mask (skin lesions, etc.); recent gastrointestinal surgery or myocardial infarction; home NIV; tracheostomy; refusal of consent or participation in another study (Figure 1).
Figure 1. Flowchart of the trial. * For the main outcome criterion. hemod. = hemodynamic; NIV = noninvasive ventilation; resp. = respiratory; SBT = spontaneous breathing trial.
[More] [Minimize]In cases of SBT intolerance, patients were randomized into the following three groups: conventional invasive weaning group, extubation followed by standard oxygen therapy, and extubation followed by NIV. The weaning strategy was allocated based on a computer-generated randomization table using variable blocks of four, sealed and opaque envelopes according to center stratification.
Conventional invasive weaning was performed using one of the following techniques: one or more daily SBT with T-piece or pressure support ventilation (PSV) with or without PEEP, and a progressive decrease in PSV level until less than or equal to 7 cm H2O. Regardless of the technique, FIO2 was set to maintain SaO2 greater than or equal to 90%. Patients whose SBT was successful were then extubated. In contrast, patients were reventilated according to the ventilatory mode previously used, and invasive weaning was continued on the following days.
For both oxygen-therapy and NIV groups, SBT failure was followed by a reventilation period of at least 30 minutes, and extubation, performed the day of randomization after informed consent was obtained, was followed either by standard oxygen therapy to maintain SaO2 greater than or equal to 90% or immediate NIV after the technique was explained to the patient. NIV was applied via a face mask as first choice, with an ICU or a specific NIV ventilator depending on the center practice. NIV was performed in a semirecumbent position for a daily duration greater than or equal to 6 hours, as continuous initially, then intermittently with oxygen spontaneous breathing (SB) periods. The preferred recommended ventilatory mode was PSV ± PEEP or bilevel positive airway pressure, initially set to obtain an expiratory tidal volume greater than 7 ml/kg and a respiratory rate less than 30 cycles/min. The oxygen flow on SB, FIO2, and PEEP with NIV were set to maintain SaO2 greater than or equal to 90%. These initial settings, type of mask, and NIV sequences were secondarily adjusted to the patient's comfort and tolerance, presence of air leaks, and ABG controls. Criteria to discontinue NIV were as follows: daily need less than 6 hours or respiratory stability with standard oxygen therapy for at least 12 hours, thus allowing for a PaO2 greater than or equal to 64 mm Hg (8.5 kPa) with pH greater than or equal to 7.35 and PaCO2 less than or equal to 60 mm Hg (8 kPa).
The use of NIV was permitted as rescue therapy, before eventual reintubation, in both invasive and oxygen-therapy groups in cases of postextubation ARF occurrence (rescue postextubation NIV) defined by at least two of the following criteria: tachypnea greater than 30/min or bradypnea less than 12/min; hypoxemia under O2 greater than or equal to 6 L/min or FIO2 greater than or equal to 50% with a Venturi mask (i.e., SaO2 < 90% or PaO2 < 64 mm Hg [8.5 kPa] or PaO2/FIO2 ≤ 130 mmHg); hypercapnia with respiratory acidosis under nasal O2 less than or equal to 3 L/min (i.e., increase of PaCO2 ≥ 10% relative to pre-extubation value [SBT] and pH ≤ 7.35); clinical signs of ARF (i.e., cyanosis, sweats, involvement of accessory respiratory muscles, paradoxical abdominal motion, consciousness impairment). NIV application and discontinuation modalities were similar to that of the NIV weaning group.
For the three weaning groups, the reintubation decision was based either on (1) one of the following major criteria: respiratory or cardiac arrest, persistent severe hypoxemia (PaO2/FIO2 ≤ 130 mm Hg) despite NIV, hemodynamic instability with systolic blood pressure less than or equal to 85 mm Hg despite adequate vascular filling, severe cardiac arrhythmia; or (2) at least two of the following minor criteria: ineffective ventilation due to agitation and/or major air leaks under NIV; clinical signs of severe ARF with tachypnea greater than 35/min and/or pH less than 7.20; occurrence, persistence, or worsening of ARF under NIV (tachypnea, sweats, cyanosis, involvement of accessory respiratory muscles, paradoxical abdominal motion, and/or respiratory acidosis impairment); consciousness deterioration or respiratory encephalopathy score (0: normal consciousness; 1: mild flapping tremor; 2: severe flapping tremor and mild confusion; 3: severe confusion with diurnal sleepiness or mild agitation; 4: coma or major agitation) worsening (≥ 3) under NIV; bronchial hypersecretion under NIV; development of other organ failure. In all cases, the attending physician made the final decision regarding reintubation as needed.
In all three groups, weaning from invasive MV was associated with continuous noninvasive respiratory and hemodynamic monitoring as well as optimized medical treatment determined by the attending physician in light of the underlying conditions and the cause of ARF responsible for intubation.
The following characteristics were recorded on admission: demographic data, simplified acute physiological score II (SAPS II), McCabe score, underlying CHRF type, cause of ARF, home oxygen therapy, NIV failure before intubation. The following characteristics were collected on randomization: duration of prior hospitalization and invasive MV, number of prior SBTs, duration of tolerance for the last SBT, ABG data after SBT failure if available, initial weaning technique and settings.
We evaluated the weaning/extubation results according to the weaning strategy allocated. The primary endpoint of the study was defined as the need for reintubation within 7 days after extubation for all three groups. Causes and time to reintubation were recorded.
The secondary endpoints were: secondary occurrence of postextubation ARF (as defined above) or death from all causes within 7 days after extubation. To assess more globally the weaning failure, we then determined a composite criterion including postextubation ARF, reintubation, or death occurrence within 7 days after extubation. We also evaluated the time to rescue postextubation NIV and its related success rate (i.e., the probability of reintubation or death within 7 days after its initiation), duration of the weaning procedure, weaning or invasive MV complications, ICU and hospital length of stay and survival, and respiratory condition on hospital discharge. Patients were followed up until Day 28 or hospital discharge. In addition, we performed a post hoc analysis for the subgroup of patients with CHRF due to COPD regarding the weaning results and survival.
According to a two-sided 5% type 1 error and 80% power to detect a difference in the reintubation rate of at least 10% between each group based on previous data (i.e., for weaning failure rates of 20, 30, and 10% for the invasive weaning, oxygen-therapy, and NIV groups, respectively) (4, 5, 13, 14, 25), and anticipating a postrandomization secondary exclusion risk of approximately 5% (e.g., death before extubation), the sample size was 66 patients in each group, leading to a total of 208 allocated patients. An interim analysis was planned for the primary endpoint after the 99th patient was included, based on a type 1 risk of 2.94%. The trial was monitored by an independent data and safety monitoring board.
All endpoints were analyzed applying the intention-to-treat principle. Continuous data were reported as median with 1st and 3rd quartiles (Q1–Q3) and categorical data as absolute number with percentage. Comparisons between the three groups were performed with the Kruskal-Wallis test for ordinal variables, and Fisher exact, Freeman-Halton, or the likelihood ratio test for nominal variables as appropriate. Time-to-event analyses were based on the actuarial method for estimation and the log-rank test for the overall comparisons of several independent groups. A P value less than 0.05 was considered statistically significant. In addition, a center effect was controlled for using the Cochran-Mantel-Haenszel test. Analyses were performed using the SAS 9.2 statistical software (SAS Institute Inc., Cary, NC).
Of the 388 eligible patients, 180 (46%) were not included, mainly due to SBT success in 127 patients (71%) (Figure 1). The remaining 208 patients were randomized as follows: 69 in the invasive weaning group, 70 in the oxygen-therapy group, and 69 in the NIV group. Two and one patient died before reintubation in the first and third group, respectively. As a result, 67 and 68 patients in these two groups were kept in the primary endpoint analysis (Figure 1).
There was no significant difference between the characteristics of the three groups on admission and randomization (Table 1). The underlying CHRF was mainly due to COPD (69%). For the invasive group, the weaning technique was primarily SBT with T-piece (80%) then PSV (20%). A PS mode was used for 97% of patients in the NIV group, with median levels of PS and PEEP of 15 (12–18) and 4 (3–5) cm H2O, respectively.
| Parameters | Invasive Weaning Group (n = 69) | O2 Group (n = 70) | NIV Group (n = 69) | P Value |
| Characteristics on admission | ||||
| Age, yr | 70 [60–75] | 72 [65–79] | 71 [66–76] | 0.188 |
| Sex, M/F | 42 (61)/27 (39) | 55 (78)/15 (22) | 52 (75)/17 (25) | 0.051 |
| SAPS II | 40 [34–57] | 40 [34–50] | 46 [34–56] | 0.495 |
| McCabe score | 2 [1–2] | 1 [1–2] | 1 [1–2] | 0.453 |
| CHRF type | 0.215 | |||
| Obstructive | 50 (73) | 46 (66) | 48 (69) | |
| Restrictive | 7 (10) | 8 (11) | 2 (3) | |
| Mixed | 9 (13) | 14 (20) | 15 (22) | |
| Unknown | 3 (4) | 2 (3) | 4 (6) | |
| LTO (yes/no/unknown), n | 21/45/3 | 16/52/2 | 16/49/4 | 0.521 |
| ARF causes | n = 66 | n = 68 | n = 65 | 0.493 |
| Acute exacerbation | 27 (41) | 24 (35) | 33 (51) | |
| Pneumonia | 21 (32) | 22 (32) | 12 (18) | |
| ACPE | 4 (6) | 5 (7) | 4 (6) | |
| Other | 14 (21) | 17 (25) | 16 (25) | |
| NIV failure before ETMV | 15 (23) | 19 (28) | 16 (25) | 0.787 |
| Characteristics at randomization | ||||
| Duration of prior hospital stay, d | 10 [5–15] | 10 [7–17] | 8 [5–14] | 0.421 |
| Duration of prior ICU stay, d | 6 [4–12] | 7 [4–13] | 6 [4–9] | 0.669 |
| Duration of prior ETMV, d | 6 [4–11] | 7 [5–11] | 6 [4–10] | 0.721 |
| Number of SBT before randomization | 1 [1–3] | 1 [1–2] | 1 [1–2] | 0.927 |
| Duration of last SBT tolerance, min | 50 [30–80] | 45 [20–90] | 45 [17–60] | 0.362 |
| ABG for SBT intolerance | n = 46 | n = 52 | n = 40 | |
| FIO2 used, % | 40 [40–40] | 40 [40–50] | 40 [40–50] | 0.445 |
| PaO2/FIO2, mm Hg | 195 [158–221] | 171 [152–213] | 183 [145–233] | 0.523 |
| PaO2, mm Hg | 73 [60–85] | 73 [61–85] | 77 [60–90] | 0.760 |
| SaO2, % | 994 [90–97] | 94 [88–96] | 94 [91–96] | 0.503 |
| pH | 7.37 [7.32–7.43] | 7.36 [7.30–7.40] | 7.36 [7.31–7.40] | 0.436 |
| PaCO2, mm Hg | 52 [46–66] | 53 [47–64] | 53 [45–61] | 0.859 |
| HCO3−, mmol/L | 31 [29–34] | 31 [29–33] | 29 [26–33] | 0.131 |
The main weaning results are shown in Table 2 and Figure 2. The probability of reintubation was not significantly different between the three weaning strategies, even when including the occurrence of death within 7 postextubation days. Causes and time for reintubation were similar for the three groups. Regarding the primary endpoint, no significant center effect was observed (P = 0.654).
| Parameters | Invasive Weaning Group (n = 69) | O2 Group (n = 70) | NIV Group (n = 69) | P Value |
| Weaning results | ||||
| Reintubation ≤ 7 d | 20/67 (30)* | 26/70 (37) | 22/68 (32)* | 0.654 |
| Reintubation or death ≤ 7 d† | 22/69 (32) | 26/70 (37) | 23/69 (33) | 0.758 |
| Postextubation ARF or death ≤ 7 d† | 32/69 (46) | 41/70 (59) | 6/69 (9) | <0.001 |
| Postextubation ARF, reintubation, or death ≤ 7 d† | 37/69 (54) | 50/70 (71) | 23/69 (33) | <0.001 |
| Rescue NIV for postextubation ARF | 31/69 (45) | 40/70 (57) | — | 0.176 |
| Success rate (no reintubation or death ≤ 7 d) | 14/31 (45) | 23/40 (58) | — | 0.386 |
| Causes of reintubation‡ | n = 20 | n = 26 | n = 22 | |
| Consciousness deterioration or encephalopathy score ≥ 3 | 15 (75) | 16 (62) | 11 (50) | 0.239 |
| Persistent severe hypoxemia | 8 (40) | 15 (58) | 10 (45) | 0.513 |
| ARF occurrence, persistence. or worsening under NIV | 13 (65) | 16 (61.5) | 15 (68) | 0.949 |
| Hypotension/shock | 4 (20) | 3 (12) | 1 (5) | 0.323 |
| Arrhythmia | 0 | 2 (8) | 0 | 0.328 |
| Other organ failure | 2 (1) | 1(4) | 0 | 0.380 |
| Time to reintubation, d | 1.5 [0.5–3.5] | 0.5 [0.5–3.5] | 0.5 [0.5–1.5] | 0.141 |
| Time to postextubation NIV, d | 0.5 [0.5–1.5] | 0.5 [0.5–0.5] | — | 0.139 |
| Duration of intubation, d | 1.5 [0.5–3.5] | 0 | 0 | — |
| Duration of ventilatory support for weaning, d§ | 1.5 [0.5–3.5] | — | 2.5 [0.5–3.5] | 0.033 |
| Weaning results in patients with COPD | n = 50 | n = 46 | n = 48 | |
| Reintubation ≤ 7 d | 12/48 (25)* | 17/46 (37) | 13/47 (28)* | 0.420 |
| Reintubation or death ≤ 7 d† | 14/50 (28) | 17/46 (37) | 14/48 (29) | 0.516 |
| Postextubation ARF or death ≤ 7 d† | 21/50 (42) | 24/46 (52) | 4/48 (8) | <0.001 |
| Postextubation ARF, reintubation, or death ≤ 7 d† | 24/50 (48) | 32/46 (70) | 14/48 (29) | <0.001 |
Figure 2. Probability of weaning failure within the first 7 days after extubation for the overall population according to study group. ARF = acute respiratory failure; NIV = noninvasive ventilation.
[More] [Minimize]Weaning results differed significantly between groups, favoring the interventional NIV group, when postextubation ARF occurrence was considered in the weaning outcome. Thus, NIV significantly decreased the incidence of postextubation ARF as compared with both non-NIV groups (log-rank test, P < 0.001) (Table 2, Figure 2). In cases of postextubation ARF, rescue postextubation NIV was used more frequently in the oxygen-therapy (57%) than in the invasive group (45%), but this difference was not significant. The time for rescue therapy was similar between these two groups. In both non-NIV groups, the use of rescue postextubation NIV avoided reintubation or death in 37 of 71 patients overall (52%) without any difference in the success rates between the two groups. By design, intubation duration was found to be longer for the invasive group by 1.5 (0.5–3.5) days than for the two other groups. The duration of ventilatory support for weaning was significantly longer for the initial NIV group than the invasive group (P = 0.033).
Apart from the intubation and ventilatory support for weaning durations, no difference was observed between the three weaning strategies regarding invasive MV and weaning complications during the study or for other outcomes (Table 3, Figure 3). For the subgroup of patients with COPD, similar weaning and survival results were found as in the overall population (Table 2).
| Parameters | Invasive Weaning Group (n = 69) | O2 Group (n = 70) | NIV Group (n = 69) | P Value |
| ETMV or weaning complications | 35 (51) | 43 (61) | 33 (52) | 0.247 |
| Type of complications* | ||||
| Autoextubation | 5 (7) | 6 (9) | 6 (9) | 1.000 |
| Postextubation stridor | 11 (16) | 14 (20) | 6 (9) | 0.161 |
| Tube obstruction | 1 (1) | — | 1 (1) | 0.551 |
| Respiratory encephalopathy | 4 (6) | 7 (10) | 6 (9) | 0.735 |
| Bronchial hypersecretion | 7 (10) | 12 (17) | 6 (9) | 0.308 |
| Nosocomial pneumonia | 10 (14) | 17 (24) | 9 (13) | 0.184 |
| Sinusitis | 1 (1) | 2 (3) | 1 (1) | 1.000 |
| Atelectasis | 1 (1) | 3 (4) | 2 (3) | 0.873 |
| Cardiac arrhythmia | 6 (9) | 5 (7) | 5 (7) | 0.949 |
| Hemodynamic collapse | 7 (10) | 9 (13) | 12 (17) | 0.462 |
| ACPE | 2 (3) | 1 (1) | 1 (1) | 0.848 |
| Paralytic ileus | 3 (4) | 3 (4) | 1 (1) | 0.703 |
| Gastric distension | — | — | 5 (7) | — |
| Mask intolerance | — | — | 5 (7) | — |
| ICU stay, d | 7.5 [4.5–14.5] | 7.5 [4.5–17.5] | 7.5 [4.5–15.5] | 0.691 |
| Hospital stay ≤ 28 d | 18.5 [9.5–28] | 19.5 [12.5–28] | 17.5 [9.5–28] | 0.616 |
| ICU survival | 64 (93) | 61 (87) | 56 (81) | 0.101 |
| Hospital survival ≤ 28 d | 60 (87) | 61 (87) | 53 (77) | 0.154 |
| Respiratory support at discharge in last surviving patients | n = 55 | n = 54 | n = 50 | 0.512 |
| SB in ambient air | 30 (55) | 25 (46) | 20 (40) | — |
| SB with oxygen therapy | 12 (22) | 19 (35) | 22 (44) | — |
| NIV | 8 (15) | 7 (13) | 4 (8) | — |
| Tracheostomy | 3 (5) | 2 (4) | 2 (4) | — |
| Unknown | 2 (4) | 1 (2) | 2 (4) | — |
Figure 3. Probability of intensive care unit and hospital survival for the overall population according to study group. ICU = intensive care unit; NIV = noninvasive ventilation.
[More] [Minimize]In the present study, which focused on NIV as an early weaning/extubation technique in difficult-to-wean CHRF population, we found no statistically significant difference in the probability of reintubation between the three weaning strategies. However, the study demonstrates that NIV may improve weaning results in these patients by reducing the risk of postextubation ARF occurrence. Although the duration of intubation was shortened, the ventilatory support time dedicated to weaning was found to be slightly increased with NIV. Our results also suggest that rescue NIV might be useful to avoid reintubation when postextubation ARF occurs in these patients.
The clinical benefit of NIV is well recognized for the initial management of hypercapnic ARF (18, 19, 26), leading to its increasing use in France (27), Europe (28), and worldwide (8). Several randomized trials have assessed NIV as a weaning/extubation technique relative to conventional methods in patients with initial (13, 14, 16, 17) or persistent (15) weaning difficulties from invasive MV. As expected, these studies demonstrated a reduction in the invasive MV duration with NIV, but they also showed some discrepancies in terms of weaning success or failure, incidence of reintubation, total MV duration, complications, and patient outcome. These controversial results have been recently underlined in a metaanalysis involving 12 randomized studies and 530 patients (20).
The VENISE trial (VEntilation Non Invasive et SEvrage) involved a large cohort of patients with CHRF considered as potentially difficult to wean based on the first SBT failure, 69% of whom were patients with COPD. No significant difference was found between the three weaning strategies regarding the incidence of reintubation, even in the subgroup of patients with COPD. These findings are consistent with those previously published (13–17), as well as with conclusions of the aforementioned metaanalysis (20). Not surprisingly for the oxygen-therapy group, we also observed relatively high rates of reintubation in the two other groups as compared with those previously reported (4–7, 10, 11). Nevertheless, the interpretation of this finding should consider that the study involved difficult-to-wean patients with CHRF, the time between randomization and SBT was relatively short, as well as the fact we have observed the reintubation incidence within a 7-day period after extubation. Furthermore, in line with previous studies (13–17, 20) and by design, NIV allowed decrease of the intubation duration as compared with invasive weaning without increasing the risk of weaning failure in terms of reintubation. In contrast to two previous studies (13, 15), this benefit was obtained at the expense of an obviously increased ventilatory support time related to weaning with NIV. Similar results have been previously reported (14), suggesting that NIV might, in fact, allow earlier extubation but not necessarily more rapid “de-ventilation” in difficult-to-wean patients with CHRF according to their severity. In addition, our study demonstrated that NIV used as an early weaning/extubation technique may be useful to prevent or avoid the risk of postextubation ARF as compared with conventional weaning and early extubation with standard oxygen-therapy groups. In these two latter groups, we can reasonably assume that some patients would have been reintubated for postextubation ARF if NIV was not applied as rescue therapy. These positive results with NIV for reducing postextubation ARF occurrence appear, in fact, very similar to those reported in studies assessing the preventive effect of postextubation NIV in patients who successfully passed an SBT but were considered at high risk for extubation failure (29–32). Interestingly, these latter patients were also hypercapnic during or after the successful SBT.
The study also suggests a potential benefit of rescue NIV applied in cases of postextubation ARF occurrence in our selected population. Indeed, it was applied frequently in the two non-NIV groups and reduced the risk of reintubation or death in 45% of patients assigned invasive weaning and 58% of those assigned to the oxygen-therapy group. Furthermore, the reintubation rate was not different between the three weaning strategies, nor was rescue postextubation NIV time between the two non-NIV groups. Rescue NIV for the management of postextubation ARF is currently not recommended for medical patients (11, 19) based on no proven benefit to avoid reintubation and potential harmful reintubation delay on mortality (21, 22). However, these randomized studies were conducted in very heterogeneous populations, including mainly hypoxemic postextubation ARF and only 10% (21) to 12% (22) of patients with COPD. Moreover, they were performed in centers that were sometimes poorly experienced with NIV. Although our encouraging results enhance those previously reported in hypercapnic patients with CHRF (30, 33), because our patients were not randomized on postextubation ARF occurrence, prospective randomized trials are therefore still necessary to further demonstrate the benefit of rescue postextubation NIV in this particular population.
By contrast to previous studies (13, 15–17), we did not observe any benefit of NIV in terms of MV complications, length of stay, and patient outcomes. One explanation could be the frequent use of rescue postextubation NIV in both non-NIV groups, particularly in the oxygen-therapy group, and its benefit to avoid reintubation. Indeed, this might have contributed to narrowing the difference in outcome between the three groups. Nevertheless, the difference in ICU and hospital survival observed between groups, although not significant, may require further studies sufficiently powered for these outcome criteria.
The process of discontinuing invasive MV represents an important clinical issue for the ICU physician, particularly in patients with COPD. The rate of weaning success/failure may vary widely according to various studies, weaning techniques, and modes of ventilation used, as well as the definition of weaning outcome (4–9). This definition has been recently revised and weaning failure should no longer be limited as either SBT failure or the need for reintubation within 48 hours after extubation (11). In fact, the need for postextubation NIV should also be considered now as part of the weaning outcome. To our knowledge, the weaning failure, including reintubation and/or the need for rescue NIV in cases of postextubation ARF, has been applied in only one previous study (13). Finally, the recently recommended definition of weaning outcome (11) should certainly be considered in the interpretation of our results and its comparison with previous (20) and future studies.
The VENISE trial is currently the largest prospective randomized multicenter study on the use of NIV for early weaning/extubation from invasive MV. One major feature of the study is the design that included a control group that received standard oxygen therapy only. As the real need for programmed NIV for weaning/extubation may have been questioned in previous studies (13, 14), we did not use this group as another way to perform weaning but to demonstrate more objectively the real usefulness of NIV in this indication. Interestingly, additional information is also provided by the oxygen-therapy group, because a substantial number of patients who failed SBT were successfully extubated on the same day. The study also involved a selected population of patients with CHRF considered as difficult to wean, mainly due to COPD. Moreover, we followed the weaning results up to 7 days after extubation to better track the late occurrence of NIV failure (34) and consider patients with a potentially prolonged weaning (11). Nevertheless, several limitations of the present study should be acknowledged. First, it was not possible, for obvious reasons, to conduct a double-blind investigation. Despite clearly predefined evaluation criteria and interventional procedures during the trial, the impact of such a bias cannot be definitely excluded. However, most attending physicians were not directly involved in the study. Moreover, the oxygen-therapy group enabled us to better control this bias and underscored the interest of NIV in the interventional group. Second, the use of rescue NIV in cases of postextubation ARF might have contributed to reduce the difference in patient outcome between the three groups, mainly for the primary endpoint. Indeed, as stated above, reintubation would have been highly likely if rescue NIV was not allowed in the two non-NIV groups. Nevertheless, based on previously published data (30, 33) and for ethical reasons, we felt it was difficult not to offer the possibility of rescue NIV before reintubating the patient. Furthermore, we cannot formally exclude that some patients of the three weaning groups, particularly those of the interventional NIV group, might have experienced ARF worsening under NIV but benefited from NIV intensification, and hence were not reintubated. Last, the trial was conducted in centers with extensive experience with NIV. Therefore, present results could probably not apply to those centers less experienced with this technique. In our view, in agreement with other authors (1, 20, 30), a skilled team is a major determinant to efficiency and safety of NIV applied within the weaning and postextubation period.
In conclusion, NIV used as an early weaning/extubation technique in difficult-to-wean patients with CHRF did not reduce the reintubation rate within 7 days as compared with conventional weaning and early extubation with standard oxygen therapy. Nevertheless, NIV may improve the weaning results in these patients by shortening the intubation duration and reducing the risk of postextubation ARF. Our results should support the implementation of NIV in clinical practice to manage weaning difficulties in patients with CHRF (20). The potential benefit of postextubation NIV used as a rescue therapy requires confirmation by further randomized trials in this specific population.
The authors thank the medical and nursing staff of all the participating centers for their valuable cooperation in this trial, as well as Richard Medeiros, Rouen University Hospital Medical Editor, for editing the manuscript. They also thank the Mechanical Ventilation Research Group from the Société de Réanimation de Langue Française for their help in designing the study.
| 1. | Nava S, Navalesi P, Conti G. Time of noninvasive ventilation. Intensive Care Med 2006;32:361–370. |
| 2. | Girault C. Noninvasive ventilation for weaning from mechanical ventilation and postextubation acute respiratory failure. In: , Muir JF, Ambrosino N, Simonds AK editors. Noninvasive ventilation, 2nd ed. Plymouth: European Respiratory Society Journals. Eur Respir Mon. 2009;41:143–153. |
| 3. | Rothaar RC, Epstein SK. Extubation failure: magnitude of the problem, impact on outcomes, and prevention. Curr Opin Crit Care 2003;9:59–66. |
| 4. | Brochard L, Rauss A, Benito S, Conti G, Mancebo J, Rekik N, Gasparetto A, Lemaire F. Comparison of three methods of gradual withdrawal from ventilatory support during weaning from mechanical ventilation. Am J Respir Crit Care Med 1994;150:896–903. |
| 5. | Esteban A, Frutos F, Tobin MJ, Aia I, Solsona JF, Vallverdu I, Fernandez R, de la Cal MA, Benito S, Tomas R, et al.. Spanish Lung Failure Collaborative Group. A comparison of four methods of weaning patients from mechanical ventilation. N Engl J Med 1995;332:345–350. |
| 6. | Nava S, Rubini F, Zanotti E, Ambrosino N, Bruschi C, Vitacca M, Fracchia C, Rampulla C. Survival and prediction of successful ventilator weaning in COPD patients requiring mechanical ventilation for more than 21 days. Eur Respir J 1994;7:1645–1652. |
| 7. | Vallverdu I, Calaf N, Subirana M, Net A, Benito S, Mancebo J. Clinical characteristics, respiratory functional parameters, and outcome of a two-hour T-piece trial in patients weaning from mechanical ventilation. Am J Respir Crit Care Med 1998;158:1855–1862. |
| 8. | Esteban A, Alia I, Ibanez J, Benito S, Tobin MJ; Spanish Lung Failure Collaborative Group. Modes of mechanical ventilation and weaning. A national survey of Spanish hospitals. Chest 1994;106:1188–1193. |
| 9. | Esteban A, Ferguson ND, Meade MO, Frutos-Vivar F, Apezteguia C, Brochard L, Raymondos K, Nin N, Hurtado J, Tomicic V, et al..; VENTILA Group. Evolution of mechanical ventilation in response to clinical research. Am J Respir Crit Care Med 2008;177:170–177. |
| 10. | Esteban A, Alia I, Tobin MJ, Gil A, Gordo F, Vallverdu I, Blanch L, Bonet A, Vazquez A, de Pablo R, et al.. Effect of spontaneous breathing trial duration on outcome of attempts to discontinue mechanical ventilation. Am J Respir Crit Care Med 1999;159:512–518. |
| 11. | Boles JM, Bion J, Connors A, Herridge M, Marsh B, Melot C, Pearl R, Silverman H, Stanchina M, Vieillard-Baron A, et al.. Weaning from mechanical ventilation. Statement of the sixth international consensus conference on intensive care medicine. Eur Respir J 2007;29:1033–1056. |
| 12. | Heyland DK, Cook DJ, Griffith L, Keenan SP, Brun Buisson C. The attributable morbidity and mortality of ventilator-associated pneumonia in the critically ill patient. Canadian Critical Care Trials Group. Am J Respir Crit Care Med 1999;159:1249–1256. |
| 13. | Nava S, Ambrosino N, Clini E, Prato M, Orlando G, Vitacca M, Brigada P, Fracchia C, Rubini F. Noninvasive mechanical ventilation in the weaning of patients with respiratory failure due to chronic obstructive pulmonary disease. Ann Intern Med 1998;128:721–728. |
| 14. | Girault C, Daudenthun I, Chevron V, Tamion F, Leroy J, Bonmarchand G. Noninvasive ventilation as a systematic extubation and weaning technique in acute on chronic respiratory failure. A prospective randomized controlled study. Am J Respir Crit Care Med 1999;160:86–92. |
| 15. | Ferrer M, Esquinas A, Arancibia F, Bauer TT, Gonzalez G, Carrillo A, Rodriguez-Roisin R, Torres A. Noninvasive ventilation during persistent weaning failure. A randomized-controlled trial. Am J Respir Crit Care Med 2003;168:70–76. |
| 16. | Trevisan CE, Vieira SR; Research Group in Mechanical Ventilation Weaning. Noninvasive mechanical ventilation may be useful in treating patients who fail weaning from invasive mechanical ventilation: a randomized clinical trial. Crit Care 2008;12:R51. |
| 17. | Prasad SB, Chaudhry D, Khanna R. Role of noninvasive ventilation in weaning from mechanical ventilation in patients of chronic obstructive pulmonary disease: an Indian experience. Indian J Crit Care Med 2009;13:207–212. |
| 18. | International Consensus Conference in Intensive Care Medecine. Noninvasive positive pressure ventilation in acute respiratory failure. Am J Respir Crit Care Med 2001;163:283–291. |
| 19. | Consensus conference on non invasive positive pressure ventilation in acute respiratory failure (excluding newborn infants). Conférence de Consensus Commune SFAR, SPLF, SRLF 2006:13–20 [accessed December 1, 2010]. Available from http://www.srlf.org |
| 20. | Burns KE, Adhikari NK, Keenan SP, Meade M. Use of noninvasive ventilation to wean critically ill adults off invasive ventilation: meta-analysis and systematic review. BMJ. 2009;338:b1574. |
| 21. | Keenan SP, Powers C, McCormack DG, Block G. Noninvasive positive-pressure ventilation for postextubation respiratory distress. A randomized controlled trial. JAMA 2002;287:3238–3244. |
| 22. | Esteban A, Frutos-Vivar F, Ferguson ND, Arabi Y, Apezteguía C, González M, Epstein SK, Hill NS, Nava S, Soares MA, et al.. Noninvasive positive-pressure ventilation for respiratory failure after extubation. N Engl J Med 2004;350:2452–2460. |
| 23. | Girault C, Bubenheim M, Benichou J, Bonmarchand G; for the VENISE Study Group. Non invasive ventilation (NIV) and weaning from mechanical ventilation (MV) in chronic respiratory failure (CRF) patients: The VENISE study preliminary results [abstract]. Am J Respir Crit Care Med 2009;177:A2165. |
| 24. | Girault C, Chajara A, Dachraoui F, Chambaretaud V, Hellot MF, Benichou J, Bonmarchand G. VENISE: Noninvasive ventilation for the weaning of mechanical ventilation in acute-on-chronic respiratory failure patients. A prospective randomized controlled and multicenter trial. Rev Mal Respir 2003;20:940–945. |
| 25. | Esteban A, Alia I, Gordo F, Fernandez R, Solsona JF, Vallverdu I, Macias S, Allegue JM, Blanco J, Carriedo D, et al.. Extubation outcome after spontaneous breathing trials with T-tube or pressure support ventilation. Am J Respir Crit Care Med 1997;156:459–465. |
| 26. | Lightowler JV, Wedzicha JA, Elliott MW, Ram FSF. Noninvasive positive pressure ventilation to treat respiratory failure resulting from exacerbations of chronic obstructive pulmonary disease: Cochrane systematic review and meta-analysis. BMJ 2003;326:185–187. |
| 27. | Demoule A, Girou E, Richard JC, Taille S, Brochard L. Increased use of noninvasive ventilation in French intensive care units. Intensive Care Med 2006;32:1747–1755. |
| 28. | Crimi C, Noto A, Princi P, Esquinas A, Nava S. A European survey of noninvasive ventilation practices. Eur Respir J 2010;36:362–369. |
| 29. | Ferrer M, Valencia M, Nicolas JM, Bernadich O, Badia JR, Torres A. Early noninvasive ventilation averts extubation failure in patients at risk: a randomized trial. Am J Respir Crit Care Med 2006;173:164–170. |
| 30. | Ferrer M, Sellarés J, Valencia M, Carrillo A, Gonzalez G, Badia JR, Nicolas JM, Torres A. Noninvasive ventilation after extubation in hypercapnic patients with chronic respiratory disorders: randomised controlled trial. Lancet 2009;374:1082–1088. |
| 31. | Nava S, Gregoretti C, Fanfulla F, Squadrone E, Grassi M, Carlucci A, Beltrame F, Navalesi P. Noninvasive ventilation to prevent respiratory failure after extubation in high risk patients. Crit Care Med 2005;33:2465–2470. |
| 32. | El Solh AA, Aquilina A, Pineda L, Dhanvantri V, Grant B, Bouquin P. Noninvasive ventilation for prevention of postextubation respiratory failure in obese patients. Eur Respir J 2006;28:588–595. |
| 33. | Hilbert G, Gruson D, Portel L, Gbikpi-Benissan G, Cardinaud JP. Noninvasive pressure support ventilation in COPD patients with postextubation hypercapnic respiratory insufficiency. Eur Respir J 1998;11:1349–1353. |
| 34. | Moretti M, Cilione C, Tampieri A, Fracchia C, Marchioni A, Nava S. Incidence and causes of noninvasive mechanical ventilation failure after initial success. Thorax 2000;55:819–825. |
* A complete list of members may be found before the beginning of the References.
This study was funded by the Rouen University Hospital and supported by the National Clinical Research Hospital Program 01/012-HP-PHRC (French Ministry of Health).
Funding and support agencies were not involved in the study design, data collection, data analysis and interpretation, writing or approval of the manuscript.
Presented in part at the American Thoracic Society International Conference in San Diego, May 15–20, 2009.
Originally Published in Press as DOI: 10.1164/rccm.201101-0035OC on June 16, 2011
Author contributions: C.G.: idea of the study; study design and conception; obtaining funding, administrative, and material support; study supervision; literature research; data collection; data analysis and interpretation; manuscript writing; final approval. M.B.: statistical analysis, data analysis and interpretation, final approval. F.A., J.L.D., S.E., P.B., J.R., E.L., G.H., G.C., A.R., M.B., C.G., P.G.: data collection, data analysis and interpretation, final approval. J.B.: study design and conception, statistical analysis, data analysis and interpretation, final approval. G.B.: study design and conception, administrative and material support, study supervision, data analysis and interpretation, final approval.
Author disclosures
The coinvestigators of the VENISE (VEntilation Non Invasive et SEvrage) Trial Group: Tunisia: Fekri Abroug, M.D., Ph.D., Lamia Besbes, M.D. (Medical ICU, Monastir University Hospital); Souheil Elatrous, M.D. (Medical ICU, Mahdia University Hospital); Mohamed Besbes, M.D. (Medical ICU, Ariana University Hospital).
France: Jean-Luc Diehl, M.D., Ph.D., Christophe Faisy, M.D., Emmanuel Guérot, M.D. (Medical ICU, Georges Pompidou University Hospital, Paris); Guy Bonmarchand, M.D., Ph.D., Fahmi Dachraoui, M.D., Christophe Girault, M.D., Bouchra Lamia, M.D., Fabienne Tamion M.D., Ph.D. (Department of Medical Intensive Care, Rouen University Hospital); Pascal Beuret, M.D., Marie-José Carton, M.D., Mahmoud Kaaki, M.D. (Medico-surgical ICU, Roanne Hospital); Jack Richecoeur, M.D., Eléonore Klebeev, M.D. (Medico-surgical ICU, Pontoise Hospital); Erwan L'Her, M.D., Ph.D. (Medical ICU, Brest University Hospital); Gilles Hilbert, M.D., Ph.D., Hoang-Nam Bui, M.D. (Medical ICU, Bordeaux University Hospital); Gilles Capellier, M.D., Ph.D. (Medical ICU, Besançon University Hospital); Antoine Rabbat, M.D., Aurélie Lefebvre, M.D., Christine Lorut, M.D. (Respiratory and Medical ICU, Hôtel-Dieu University Hospital, Paris); Claude Guérin, M.D., Ph.D., Michel Badet, M.D. (Medical ICU, Lyon University Hospital); Philippe Guiot, M.D. (Medical ICU, Mulhouse Hospital).
Data Monitoring and Management Board: Véronique Chambaretaud, Abdesslam Chajara, Patrice Testut, Karim Lallouche (Clinical Research Assistants, Rouen University Hospital); Marie-France Hellot (Biostatistician, Rouen University Hospital); Philippe Lagoutte (Data-manager, Rouen University Hospital).
Safety Monitoring Board: Hervé Lévesque, M.D., Ph.D. (Department of Internal Medicine, Rouen University Hospital); Dominique Robert, M.D. (Medical ICU, Lyon University Hospital); Benoit Veber, M.D., Ph.D. (Surgical ICU, Rouen University Hospital).
Mechanical Ventilation Research Group from the Société de Réanimation de Langue Française (SRLF): Laurent Brochard, M.D., Ph.D. (Medical ICU, Créteil University Hospital); Alain Mercat, M.D., Ph.D (Medical ICU, Angers University Hospital); Jean-Christophe Richard, M.D., Ph.D. (Medical ICU, Rouen University Hospital); Jordi Mancebo, M.D., Ph.D. (Medical ICU, Barcelona University Hospital, Spain).
