现在的位置: 首页研究点评, 进展交流>正文
[JAMA发表述评]: 脱离机械通气的新曙光
2019年06月19日 研究点评, 进展交流 暂无评论

Editorial June 11, 2019

Revisiting, Reframing, and Casting a New Light on Liberation From Mechanical Ventilation

Timothy D. Girard, Karen E. A. Burns

JAMA. 2019;321(22):2167-2169. doi:10.1001/jama.2019.7364

Every year, more than 1 million patients throughout the world receive mechanical ventilation for acute respiratory failure. One of the most important decisions clinicians make in managing these critically ill patients is how to liberate them from invasive ventilation. Ventilator liberation poses an important dilemma for clinicians because premature or failed attempts at extubation, which typically necessitate reintubation, increase rates of ventilator-associated pneumonia, mortality, and other adverse outcomes.1 Conversely, delaying extubation also increases a patient’s risk of being oversedated and developing delirium or ventilator-associated events.2

The current standard of care for managing patients with invasive ventilation involves daily screenings to identify patients who are ready to undergo a spontaneous breathing trial (SBT) to ascertain those ready to be separated from ventilator support.3,4 Additionally, clinicians determine whether patients can be liberated from an endotracheal tube by assessing for extubation failure risk factors (eg, weak cough, heavy secretions, and depressed level of consciousness) before deciding to extubate patients who have had a successful SBT. Before the SBT became the standard assessment of ventilator liberation readiness, a series of investigations searched for the ideal predictor, but the “weaning parameters” examined were limited in their ability to predict successful extubation. Consequently, observing a still-intubated patient during a trial of unassisted breathing (ie, the SBT) remained the preferred approach. In 1995, Esteban et al5 published a randomized trial that found that using daily SBTs hastened successful liberation compared with gradually titrating mandatory ventilation rate or inspiratory pressure. Shortly thereafter, Ely et al6 showed that a respiratory therapy–driven SBT protocol resulted in earlier extubation than a physician-directed approach.

When conducting an SBT, a number of different techniques can be used. During a T-piece SBT, which was used by Esteban et al,5 a patient receives supplemental oxygen but no ventilatory assistance. Alternatively, a patient may receive pressure support ventilation (PSV), during which a small amount of positive pressure (eg, 5-8 cm H2O) assists inspiration. Additionally, the safety criteria used to determine when an SBT should be started can be either conservative or liberal, and SBTs can be conducted for variable durations. Thus, more than 2 decades after Esteban et al5 and Ely et al6 published trials ushering SBTs into routine use, new evidence to guide decision making about liberation from invasive ventilation in current practice is still warranted and needed.

In this issue of the JAMA, Subirà and colleagues7 present the results of the largest randomized clinical trial ever conducted comparing alternative SBT techniques. The authors randomized 1153 mechanically ventilated adults in 18 Spanish intensive care units (ICUs) to 2 different SBT techniques: (1) a “highly demanding” 2-hour T-piece SBT or (2) a “less demanding” 30-minute PSV SBT with 8-cm H2O inspiratory pressure and 0-cm H2O positive end-expiratory pressure. By study design, participants in both groups who successfully completed an SBT were extubated because the trial eligibility criteria ensured that enrolled participants were good candidates for extubation. In this trial, participants in whom the initial SBT was unsuccessful were ventilated at the discretion of the ICU team, and subsequent SBTs were not directed by the trial protocol.

Subirà et al7 found that participants in the 30-minute PSV SBT group were more likely than those in the 2-hour T-piece SBT group to have a successful initial SBT and therefore to be extubated. Thus, successful extubation (the primary outcome) occurred in 82.3% of patients in the PSV group and in 74.0% of participants in the T-piece group, an absolute increase of 8.2% (95% CI, 3.4%-13.0%). Additionally, although more participants in the less demanding SBT group had a successful initial SBT and were extubated, they did not experience a higher reintubation rate in the 72 hours after extubation. Moreover, those in the PSV SBT group were also significantly less likely to die in the hospital or during the 90 days after randomization—results that are difficult to explain given that the SBT techniques did not appear to affect other secondary outcomes. Nevertheless, these results are important because they provide evidence that, at least during an initial SBT, most patients can be tested for 30 minutes using 8-cm H2O PSV.

Less demanding SBT techniques were previously studied in 2 multicenter randomized trials conducted in the late 1990s.8,9 Similar to the trial by Subirà et al, these older investigations found no significant differences in reintubation rates when comparing 30-minute SBTs with 120-minute SBTs9 and comparing SBTs conducted with 7-cm H2O PSV vs T-piece SBTs.8 Given these results, it is worth considering why many clinicians continue to rely on 120-minute SBTs and many still prefer to conduct T-piece SBTs.4 Clinicians’ tendency to be risk averse and delays in knowledge transfer may explain some of the variation in ventilator liberation practices, but limitations in the evidence are also likely a contributing factor. No equivalence trials have been conducted comparing different SBT techniques, and questions remain regarding the quality and generalizability of the evidence. These new results reported by Subirà and colleagues7 are therefore an important and welcome addition to the literature. At the same time, their trial raises a number of important questions.

First, when should SBTs be started during a patient’s recovery from acute respiratory failure? Similar to previous trials,5 a high percentage of participants in the trial by Subirà et al had a successful initial SBT, which leads to the question of whether some participants were ready to undergo an SBT and to be considered for extubation at an earlier time point. That only half of the patients who experience unplanned extubation require reintubation10 suggests that recognizing the earliest time when patients are ready for extubation is a challenge for ICU clinicians.

Second, should a 30-minute PSV SBT be used regardless of a patient’s characteristics and circumstances, or should patient-related factors influence how clinicians conduct SBTs? Even though Subirà et al enrolled a diverse group of participants in multiple centers, their protocol applied only to each participant’s initial SBT. Consequently, the results of this trial cannot be generalized to patients in whom an initial SBT is unsuccessful, a group that may be more likely to have extubation failure after a successful 30-minute SBT.11Future trials of ventilator liberation techniques should continue to apply study interventions until patients achieve successful extubation (or another trial end point).

Third, does the use of related interventions—eg, computerized ventilator weaning12 or extubation to noninvasive ventilation or high-flow nasal cannula oxygen therapy after a successful (or even a failed) SBT13—modify the effects or the implementation of different SBT techniques? The availability of noninvasive modes of support, for example, may encourage clinicians to extubate high-risk patients after a less demanding SBT. It is notable that in the trial by Subirà and colleagues, the use of prophylactic noninvasive ventilation or high-flow nasal cannula oxygen therapy after extubation was more common in the 30-minute PSV SBT group than in the 2-hour T-piece SBT group (25% vs 19%; P = .01).

Fourth, does the SBT technique affect long-term outcomes? If so, how? The effects of interventions on short-term outcomes are important and generally simple to ascertain in clinical trials. Long-term outcomes, however, including long-term survival and functional status, are equally or perhaps even more important to patients.14 The trial by Subirà and colleagues7 suggests that 90-day mortality may be reduced by using 30-minute PSV SBTs rather than more demanding SBTs. If this finding is true, identifying the mechanism(s) underlying this putative benefit will not only promote knowledge transfer but may also help in the identification of other interventions that favorably influence long-term outcomes.

Fifth, how should outcomes be defined and reported in ventilator liberation trials? Although successful extubation is an outcome valued by clinicians, patients, and patients’ families,14 there is no consensus regarding how to define, measure, and report this outcome. Are patients who are extubated to noninvasive ventilation, for example, successfully extubated? At what time point should investigators report successful extubation? Which study participants should be included in the denominator (all participants randomized to an intervention or only those who are extubated)? Which is more important, the proportion of participants successfully extubated or the time to successful extubation?

In summary, mechanical ventilation research in the ICU is challenging. With their publication in this issue of JAMA, Subirà and colleagues have addressed an important knowledge gap, advancing the science of ventilator liberation and revitalizing an important area of critical care research. Their results support the use of 30 minutes of PSV during an initial SBT for most patients. Yet much work remains to be done to improve the outcomes of mechanically ventilated ICU patients, a population that both is highly vulnerable and has great potential to respond to evidence-based care.15

给我留言

您必须 [ 登录 ] 才能发表留言!

×
腾讯微博