现在的位置: 首页研究点评, 进展交流>正文
[JAMA述评]:无创通气对高流量鼻导管氧疗预防高危患者再次插管的额外价值
2019年10月15日 研究点评, 进展交流 暂无评论

Editorial October 2, 2019

Added Benefit of Noninvasive Ventilation to High-Flow Nasal Oxygen to Prevent Reintubation in Higher-Risk Patients

Irene Telias, Niall D. Ferguson

JAMA. Published online October 2, 2019. doi:10.1001/jama.2019.14609

Liberating patients from ongoing invasive mechanical ventilation is typically a 3-step process. First, clinicians must recognize that patients may no longer require mechanical ventilation—ie, when the reasons for intubation are substantially improved and a number of clinical stability criteria are met. Second, patients who meet these “readiness to wean” criteria are then assessed for their ability to breathe through the endotracheal tube with little or no assistance from the ventilator during a so-called spontaneous breathing trial (SBT). Third, for patients who pass their SBT, clinicians must judge whether those patients will be able to sustain ventilation once the endotracheal tube has been removed, maintaining airway patency and clearance of secretions.

Over more than 30 years, investigations have followed this stepwise approach, with knowledge generated around the mechanisms of weaning failure,1 the best strategies to identify patients potentially ready for extubation,2 and the modalities to test patients’ ability to breathe without assistance (SBTs).3,4 The decision to extubate, however, remains a clinical challenge, with 10% to 25% of patients requiring reintubation despite having passed an SBT, and no simple way to identify those at prohibitively high risk of reintubation.5 Failure to achieve successful extubation and requiring reintubation can be a catastrophic event and has been associated with an absolute risk of death up to 50%.6 Although residual confounding may account for some of this association, clinicians agree that it is an outcome to be avoided, while balancing this risk against the harms of unnecessarily prolonged ventilation.

A key aspect of the patient’s ability to remain successfully extubated relates to the balance between the load imposed on the patient’s respiratory muscles after extubation and the muscles’ capacity. Following extubation, work of breathing is often increased (ie, increased load) for reasons including changes in the upper airway, weaning-induced pulmonary edema, or difficulty in managing secretions.4 Moreover, respiratory muscle weakness (ie, decreased capacity) is common among patients weaning from ventilation, therefore, the possibility of overcoming an increased load is often limited.7

One technique to decrease the load and help restore balance is noninvasive ventilation (NIV), which delivers positive pressure through a tightly fitting oral-nasal mask. When applied directly after extubation in several small single-center randomized clinical trials (RCTs) to patients at higher risk of extubation failure, NIV has been shown to reduce reintubation and is strongly recommended in recent guidelines.8 However, some patients cannot tolerate NIV at all, and others require frequent breaks or interruptions. A newer technique is the use of high-flow nasal oxygen (HFNO), which delivers high flows of warm humidified gas and can decrease work by contributing to clearance of secretions, providing low levels of positive end-expiratory pressure and decreasing dead-space by carbon dioxide washout in the upper airway.9 HFNO has been shown to be more comfortable than standard oxygen and to reduce the need for reintubation in a small RCT involving 105 patients.10 These techniques were compared directly in an RCT of 604 patients at higher risk of reintubation from 3 ICUs in Spain, showing that HFNO (23%) was noninferior to NIV (19%) in preventing reintubation at 72 hours, albeit with a relatively large noninferiority margin of 10%.11 Based on these data, and with HFNO being more practical than NIV, many clinicians opted to use HFNO as standard of care to prevent reintubation in higher-risk patients.

On this background, Thille and colleagues hypothesized that these 2 therapies could be synergistic and conducted a multicenter RCT, which is published in this issue of JAMA.12 The investigators compared the use of prophylactic intermittent NIV interspersed with HFNO vs HFNO alone in patients older than 65 years or those with underlying chronic cardiac or respiratory conditions—a cohort they had previously shown to be at higher risk for reintubation (up to 30%).13 Based on their evaluation of 641 patients from 30 intensive care units (ICUs) in France, the authors found a statistically significant absolute risk reduction in the primary outcome of reintubation within 7 days of extubation of −6.4% (95% CI, −12.0% to −0.9%) among patients treated with the combination of NIV and HFNO (11.8%) vs HFNO alone (18.2%). The intervention specified use of NIV (mean positive end-expiratory pressure of 5 and pressure support of 8 cm H2O) for at least 12 hours per day during the first 48 hours after extubation (mean duration, 22 hours) and HFNO at 50 L/min between NIV sessions. The control group used continuous HFNO at 50 L/min for 48 hours (mean duration, 42 hours). Investigators allowed extended use of NIV or HFNO beyond 48 hours based on clinical conditions but discouraged use of NIV as a rescue therapy in the control group based on the evidence of worse outcomes in this context.14

This trial has several positive aspects. These include a strong physiological rationale and previous clinical data to estimate the baseline risk of the target high-risk population and the possible effect size of the intervention. Adherence to the protocol was high, with few patients not tolerating NIV (6%) and the time of use and settings of NIV and HFNO were roughly as planned in both groups. Additionally, an intention-to-treat analysis was performed and few patients were lost to follow-up. The study was multicenter and the other weaning techniques used were consistent with current practices, reinforcing external validity.

The effect of the intervention was numerically larger among patients with hypercapnia at the end of the SBT compared with those without, but as indicated by the negative test for interaction (P = .25), this degree of heterogeneity of treatment effect was not statistically significant. The suggestion that hypercapnic patients may benefit more from this therapy is consistent with previous data showing that prophylactic NIV in hypercapnic patients resulted in lower rates of respiratory failure and mortality after extubation.15The present RCT suggests that all included patients, and especially those with hypercapnia, experience an additional benefit from higher positive end-expiratory pressure levels and inspiratory support to unload the respiratory muscles. Even though the effect size may be lower in this group, nonhypercapic patients were the majority (83%), the reintubation rate was still lower in the NIV and HFNO group by 5%, and when the arterial pressure of carbon dioxide level was included as a covariate in an exploratory multiple logistic regression model, the adjusted odds ratio for reintubation still favored the use of NIV (odds ratio, 0.60 [95% CI, 0.38–0.93]). Investigators promoted the use of NIV at night, which might have contributed to the added benefit of NIV in the context of frequently underdiagnosed and newly acquired sleep disorders of breathing in the ICU.16

The authors suggest that the main limitation of their study is the use of HFNO as the control group. However, this choice seems justified by the growing evidence of its benefit in the postextubation period.11Moreover, the use of HFNO instead of standard oxygen likely biased the results toward the null. Instead, the main limitation is that clinicians, who by necessity were unblinded to treatment assignment, were then charged with determining when a patient needed reintubation, the primary end point. The authors attempted to address this by using predefined criteria for reintubation with high reported adherence (96% in the NIV and HFNO group and 97% in the HFNO alone group). However, the clinical state of patients at the time of reintubation in each group was not reported. Readers may be somewhat reassured by similar times to reintubation and mortality rates between groups among reintubated patients, which if anything, both tend to be higher in the HFNO group, pointing away from “soft calls” for reintubation in controls.

Despite being less convenient than using only HFNO for many clinicians, this study inconveniently but probably correctly shows that the combination of NIV and HFNO provides the best support for patients at higher risk of reintubation. Many clinicians may be ready to apply NIV and HFNO to those with hypercapnia at the end of the SBT based on these results and prior data. Before suggesting that all patients undergoing invasive ventilation for more than a day who are either older than 65 years or of any age with chronic cardiac or pulmonary diseases should receive NIV and HFNO for 48 hours after extubation, some caveats should be considered. This trial was conducted in ICUs with significant experience and expertise in the application of NIV and the results may not be applicable in less-experienced centers. Additional limitations include the potential restriction for mobilization and communication while undergoing NIV and the implied increase in ICU length of stay if these patients need at least 48 hours of ICU care after extubation.

How should the results of this trial then be incorporated into clinical practice? Clinicians should recognize that in this study, as in almost all RCTs, even with clear and explicit inclusion and exclusion criteria, some clinical judgment is involved with patient enrollment, either implicitly in the application of the inclusion and exclusion criteria or explicitly in the form of physician refusals to enroll eligible patients. Accordingly, to get the most of this strategy, clinicians will have to use their clinical judgment to decide which patients within the high-risk nonhypercapnic group will benefit the most based on individual clinical and physiological characteristics.

给我留言

您必须 [ 登录 ] 才能发表留言!

×
腾讯微博