Authors: Theodore J Segarra, Lee Grodin, Taylor Conrad, Ray Beyda, Kelly Maurelus, Michael C. Bond
Originally Published: Common Sense September/October 2017
Over the last decade, non-invasive ventilation (NIV), including both bilevel positive airway pressure (BPAP) and continuous positive airway pressure (CPAP) modes, has become an important tool in the management of ED patients with respiratory distress due to acute pulmonary edema (APE) and chronic obstructive pulmonary disease (COPD) exacerbations. Many studies have shown its utility in successfully reducing the need for intubation and reducing length of stay (LOS) in the ICU. Given these positive results, interest in NIV for patients with undifferentiated respiratory distress has increased but very few studies have compared the outcomes of using NIV for other causes of acute respiratory distress, such as asthma, pneumonia, malignancy, or interstitial lung disease. This review aims to discuss the current literature on the non-standard use of NIV for for other causes of respiratory distress in the emergency setting and to identify potential areas for further research.
Management of critical illness with non-invasive ventilation by an Australian HEMS. Andrew R Coggins, Erin N Cummins, Brian Burns. Emergency Medical Journal. 2016
An Australia group recently reviewed the pre-hospital use of NIV for the management of acute respiratory distress. Coggins and colleagues performed a retrospective observational study comparing outcomes in 106 patients who received NIV at some point during their pre-hospital course. One group was transported entirely on NIV (n = 58), one group failed NIV at the referral hospital and required intubation prior to transport (n = 20), and one group was able to be taken off NIV prior to transport (n = 28). All patients were transported by the Greater Sydney Area Helicopter Medical Services and transport was either by helicopter or by ambulance.
The authors determined that of the 86 patients placed on NIV and stable for transport, none required intubation during transport, and none died within 24 hours of transport. Among the 106 total patients, the median age was 63, and the most common causes of respiratory distress were pneumonia (34%), cardiogenic (27%), and COPD (26%). However, they found that patients with cardiogenic causes (heart failure and cardiogenic shock) had the highest rates of intubation at 24 hours (38%) despite low rates of early failure of NIV. In addition, nearly 20% of the non-intubated patients eventually required intubation by 24 hours. They further noted an increased trend in failed NIV when the decision to choose NIV was made by a physician in training (registrar) rather than a trained physician (consultant).
The authors concluded that the use of NIV in the pre-hospital setting is safe, but that failure of NIV does lead to increased admission and treatment times. Though limited by a small sample size and the inherent limitations of a retrospective observational design, this study was able to highlight the safe use of NIV in patients with diverse etiologies of respiratory failure. As such, it lends support to the use of NIV in patients with undifferentiated respiratory distress. Furthermore, it highlights the potential use of NIV in the pre-hospital setting as well as in the ED. Despite these positive findings, this study also demonstrates the need for additional randomized prospective studies to better compare outcomes and reduce potential confounders.
Green E, Jain, P, Bernoth, M. Noninvasive ventilation of acute exacerbations of asthma: a systematic review of the literature. Australian Critical Care. 2017 Jan 27. pii: S1036-7314(17)30017-6. doi: 10.1016/j.aucc.2017.01.003. [Epub ahead of print]
Green et al conducted a systematic review of the use of NIV for asthma. They searched EBSCOhost, MEDLINES, and PubMed using the terms “noninvasive ventilation,” “BiPAP,” “CPAP,” “wheeze,” and “asthma,” and excluded reviews, studies on topics other than asthma, those not in English, and pediatric studies. Ultimately, thirteen studies were included. While NIV use is considered safe and is common in patients with asthma, this study sought to evaluate its efficacy. However, this review highlighted the need for more research on the topic and the barriers to doing so.
Most of the included studies examined NIV use in the ED. None of the studies were blinded and most had poor randomization. Given the obvious nature of the intervention, blinding providers and patients to its use is difficult. Six of the reviewed studies do not include biometric data for greater than four hours. The authors identify that traditionally respected endpoints such as mortality and LOS are difficult to use in studying asthma. Mortality is difficult given the rarity of inpatient deaths associated with asthma. The LOS is a difficult endpoint given that each study set different parameters and have confounding results.
The authors conclude there is inadequate information to endorse the use of NIV for acute asthma exacerbations. Background information does suggest it is safe and warrants further investigation.
Confalorieri M, Potena A, et al. Acute Respiratory Failure in Patients with Severe Community-acquired Pneumonia. Am. J. Respir. Crit. Care Med. 1999;160(5):1585-1591.
This multicenter, prospective, randomized controlled trial was designed to compare the efficacy of NIV versus standard treatment of supplemental oxygen delivered by Venturi mask in patients with severe community acquired pneumonia (CAP) and concurrent acute respiratory failure. The primary endpoint was the requirement of endotracheal intubation as determined by preselected criteria at any point during treatment. Secondary endpoints included complications during hospital stay, duration of required ventilator support, length of ICU and hospital stay, in-hospital survival, and 2-month survival. Severe community acquired pneumonia was defined as per the American Thoracic Society criteria. Pre-selected values for hypoxia, hypercapnea, and tachypnea with altered respiratory mechanics were used to define acute respiratory failure. Baseline characteristics were similar between the NIV (n= 28) and the standard treatment (n= 28) groups including APACHE II scores, presence of concomitant COPD, age, blood gases, and respiratory rate. Both groups received similar medical management with antibiotics and goal SpO2 > 90%. All patients were admitted and managed in an intermediate respiratory ICU and were only transferred to the full ICU if they required intubation or invasive monitoring.
Reasons for intubation included worsening hypoxemia, worsening hypercarbia, severe hemodynamic instability, and inability to tolerate secretions. Only 21% (n= 6) of the NIV group versus 61% (n= 17) of the standard treatment group met criteria for intubation (p = 0.007). Of note, only 14 patients in the standard treatment group were ultimately intubated as 3 patients in this group with concurrent COPD were given NIV after meeting criteria for intubation and subsequently improved. Duration of ICU stay was significantly less in the NIV group (1.8 ± 0.7 days) compared to standard treatment (6 ± 2 days), p 0.04. Patients randomized to the NIV group also noted a rapid and sustained decrease in respiratory rate within 24 hours compared to the standard treatment group. Importantly, time to intubation was similar in both groups (mean of 44 hours with wide variability). Survival during hospital stay, 2-month mortality, and required intensity of nursing was similar between the two groups. The authors performed a post-hoc analysis comparing the subset of patients with COPD. The significant difference in a lower number of patients meeting criteria for mechanical ventilation was only seen in the COPD population. Additionally, a significant reduction in 2-month mortality was noted in the NIV group for patients with COPD.
For patients with severe CAP and respiratory failure who don’t require immediate intubation, this study demonstrates that NIV decreases the intubation rate compared to standard treatment. An important feature of this study is that both groups noted similar APACHE II scores as well as time to intubation in those individuals requiring it. Both of these confounders have been implicated in skewing results of other studies regarding the use of NIV in severe CAP. Ultimately, the usefulness of NIV is likely in its ability to improve hypoxia and hypercarbia. This study is limited by the relatively small number of patients enrolled as well as the realization from the post-hoc analysis that the main effect was seen by the efficacy of NIV in COPD patients which is supported in other research. It is also important to note that the centers where the research was performed all had patients admitted initially to a more intensely monitored setting when compared to a regular floor ward, specifically to their respiratory intermediate ICU (likely comparable to some stepdown units although arguably different). Thus, it is difficult to generalize this to patients admitted to the hospital wards with lower levels of care. Nonetheless, NIV is useful as an early intervention in severe CAP with acute respiratory failure, especially in patients with COPD. It is important to note that NIV does not replace appropriate invasive monitoring and interventions required in patients with worsening respiratory failure.
Mosier JM, Sakles JC, et al. Failed noninvasive positive-pressure ventilation is associated with an increased risk of intubation-related complications. Annals of Intensive Care. (2015) 5:4; 1-9.
In this single center, retrospective, cohort study of medical ICU patients, the authors studied the potential detrimental effects of delayed intubation should a trial of NIV for acute respiratory distress fail. They compared patients intubated following failed NIV to patients intubated primarily without an initial trial of NIV. Specifically, they looked at the composite primary outcome of desaturations, hypotension, or aspiration events during intubations. There were no significant differences in age or gender between the NIV-trialed (n=125) and non-NIV-trialed group (n=110). However, there were differences in what the authors termed “difficult airway characteristics.” Specifically, in the NIV-trialed group there was less blood in the airway (5.6% vs 22.7%, p=0.001), less airway edema (4.0% vs 13.6%, p=0.01), and fewer desaturation events to less than 88% (20.8% vs 39.1%, p=0.003). There was a higher frequency of short necks in the NIV-trialed group (37.6% vs 20.0%, p=0.004). The severity of illness assessment scores (e.g. APACHE II, SAPS II, and APACHE IV) were also significantly lower in the NIV-trialed group. Of the patients in the failed NIV group, there was a higher incidence of intubation for hypoxemic respiratory failure (64% compared to 45.5%, p=0.006), and 49% of these were intubated for pneumonia or acute respiratory distress syndrome (ARDS).
Most of the patients in this study had NIV initiated in the ICU (77.7%) as compared to the in ED (11.6%) or on the medical ward (10.7%). Causes for NIV were acute hypoxemia (55.9%), acute hypercapnia (25.4%), and increased work of breathing (8.5%). The main reasons listed for intubation in the study differed between the two groups; the NIV-trialed group was intubated less often for airway protection (6.4% vs 26.4%), patient control (0.8% vs 1.8%), hemodynamic instability (0.8% vs 5.5%), and severe metabolic acidosis (0.8% vs 3.6%), but was intubated more often for respiratory failure (91.2% vs 62.7%), with all of these differences being statistically significant (p=0.001). There was no statistical difference between the two groups in the rates of hypotension, desaturation, or aspiration. While there was no statistical difference in the unadjusted odds of the composite complication, there was increased odds of the composite outcome in a propensity-adjusted multivariate regression analysis with an odds ratio of 2.20 (95% CI 1.14-4.25), when adjusted for the presence of pneumonia or ARDS. Finally, the unadjusted odds of death was 1.79, when a composite complication occurred (95% CI 1.03-3.12).
This study suggests that in patients intubated after a trial of NIV, there may be an increased risk of the composite outcome of hypotension, desaturation, or aspiration, and if one of these should occur there may be an increased unadjusted risk of death in the ICU. Several possible explanations may account for these findings. The increase in the rate of intubation complications could account for the increased mortality associated with delayed intubation following NIV-trial failure. Respiratory failure, either hypercapnic, hypoxemic, or mixed, is caused by multiple different etiologies, and while patients with COPD or CHF exacerbations may benefit from NIV, those with other causes of respiratory failure may not.
Furthermore, in patients with non-COPD or non-CHF causes of respiratory failure, delaying primary intubation may lead to worsening respiratory failure and reserve, as well as make intubating conditions more difficult. This study has several limitations including its retrospective design, and the authors address these, encouraging more research into this topic. While propensity scoring was utilized in this study in an attempt to address variables that may have influenced the decision to pursue a trial of NIV, not all variables and clinical factors may have been considered in the model. In addition, the study population was of patients that were all intubated in the ICU, and not all NIV-trialed patients, as data on NIV success rates was not available to the authors. While this study has its limitations, it does suggest a possible association between failure of NIV in acute respiratory failure and complications during subsequent intubation. The benefits of NIV in COPD and CHF exacerbations have led to the adoption of this mode of ventilatory support in a broad spectrum of acute respiratory failure patients. While this may benefit some patients with respiratory failure, some may be harmed. In these subsets, earlier intubation may offer a more favorable risk benefit ratio.
Overall, these studies highlight the potential for expanding the use of NIV for the treatment of acute respiratory distress due to causes other than just APE and COPD exacerbations. Although the authors in these studies were able to show some promising results regarding newer applications of NIV, the overall strength of the current evidence is still significantly lacking, and the data from different studies has been conflicting. Furthermore, we must keep in mind the potential for harm in choosing to use NIV in the wrong settings or for the wrong patients. The most important point derived from these studies is that the field would benefit greatly from additional prospective, randomized trials comparing the risks and benefits of applying NIV more broadly in patients with undifferentiated respiratory distress.