Special Feature

Three Current Controversies in Mechanical Ventilation

By David J. Pierson, MD, Editor, Professor, Pulmonary and Critical Care Medicine, Harborview Medical Center, University of Washington, Seattle, is Editor for Critical Care Alert.

The literature on the technical and clinical aspects of mechanical ventilation for patients with acute respiratory failure continues to expand, with nearly 50,000 citations appearing in PubMed for the topic "mechanical ventilation," and more than a thousand articles in English reporting the results of clinical trials under this heading published during the last 5 years. In spite of this avalanche of publications—or maybe because of it—many aspects of ventilator management remain unsettled and controversial.

To address several clinically important areas in which clinicians commonly disagree, last fall the American Respiratory Care Foundation convened a group of recognized experts on ventilator management and critical care, charging them with identifying the issues involved in each area and performing evidence-based reviews of current knowledge, in an attempt to bring clarity and up-to-date practice to the bedside in the ICU. Articles reviewing the issues addressed, along with transcripts of the vigorous discussions that accompanied them, appear in the April and May issues of Respiratory Care.1,2 Most of the topics discussed at the conference dealt with mechanical ventilation, and in this essay I summarize the controversies addressed during 3 of these discussions that pertain to most ventilated patients.

Should All Patients with Acute Respiratory Failure Get A Trial of Noninvasive Ventilation Before Intubation?

During the last decade noninvasive positive-pressure ventilation (NPPV) has become the standard of care for acute respiratory failure complicating COPD. Numerous well-designed randomized controlled trials have shown that the use of NPPV, as compared to usual management, reduces the need for intubation, decreases complications, shortens ICU and hospital stays, costs less, and saves lives. It is often the case in medicine that a therapy shown to be effective in one clinical setting tends to "metastasize," and to be applied to patients with other diagnoses and in other clinical settings than those in which its benefits have been demonstrated. This has definitely been the case with NPPV. Although NPPV is still not used when it should be by many clinicians, paradoxically it has become the default initial approach for anyone in respiratory distress in the hands of others.

Consider the following patients with acute respiratory failure:

  • An elderly patient with advanced COPD presents with 3 days of increasing dyspnea, a PaCO2 of 85 mm Hg, and a pH of 7.24.
  • A patient with cardiomyopathy and recurrent episodes of congestive heart failure presents with acute shortness of breath and typical findings of pulmonary edema on chest X-ray.
  • A young asthmatic is seen in the emergency department with severe wheezing, obvious signs of hyperinflation, a heart rate of 140 beats/min, and a pulse oximetry saturation of 88% while breathing nasal oxygen at 2 L/min.
  • Three months following liver transplantation, a patient presents with fever, dyspnea, bilateral infiltrates on chest X-ray, and a PaO2 of 80 mm Hg on 100% oxygen by mask.
  • A middle-aged motorcyclist collides with a truck and presents with multiple rib fractures, bilateral lung contusions, and a PaO2 of 80 mm Hg on 100% oxygen by face mask.
  • A patient with amyotrophic lateral sclerosis and a history of recurrent aspiration presents with respiratory distress, signs of lobar pneumonia, and acute-on-chronic respiratory acidosis.
  • A patient develops increasing respiratory distress and hypoxemia 12 hours following extubation after coronary artery bypass grafting.

These hypothetical patients represent the spectrum, with respect to the advisability of NPPV as an initial approach to ventilatory support, from "clearly indicated" to "clearly contraindicated" as supported by the evidence. In their article, Hess and Fessler3 review this evidence, which is now extensive for several of the clinical circumstances illustrated above.

The evidence best supports the use of NPPV in patients who fit the first two descriptions above: exacerbation of COPD and acute cardiogenic pulmonary edema (although in the latter setting continuous positive airway pressure may be as effective as NPPV). In acute severe asthma, fewer patients have been studied and the evidence is less clear-cut, since in the great majority of instances either the attack will improve without the need for ventilatory assistance or the patient will require intubation because of altered mental status or other contraindications to NPPV.

Several series have reported avoidance of intubation and improved outcomes in patients with acute respiratory failure complicating solid-organ transplantation or other forms of immunocompromise when NPPV is used. Such series have generally excluded patients with the factors shown in Table 1 as relative or absolute contraindications to NPPV. The same is true for acute hypoxemic (as opposed to hypercapnic) respiratory failure in other types of patients. Patients with depressed mental status or impaired bulbar function are more likely than others to aspirate when NPPV is used, and the latter is generally regarded as inadvisable in these settings.

NPPV has been shown in several studies to shorten the period of invasive mechanical ventilation and to decrease the need for reintubation when used as an adjunct to weaning in selected patients. In these studies the patients have been extubated directly to NPPV as part of a deliberate step-down protocol. However, failed extubation, the situation in which a patient who has been extubated in the usual sequence subsequently develops recurrent respiratory failure, is a distinctly different setting. In extubation failure, NPPV has been shown not only to be ineffective in forestalling reintubation but also associated with worse outcomes. When a patient has been extubated following mechanical ventilation for acute respiratory failure and subsequently develops respiratory distress, hypoxemia, and/or acute respiratory acidosis, that patient should be reintubated without a trial of NPPV.

In general, the more severely ill the patient is, and the more additional active medical issues (co-morbidities) are present, the less likely NPPV is to be successful. This has recently been confirmed by Confalonieri and associates,4 who derived a predictive equation for success vs failure of NPPV in acute respiratory failure complicating COPD, using a database of more than 1000 patients, and validated it in a second, prospective patient series. These authors used their predictive equation to produce color-coded charts for both initial assessment and patient status after 2 hours of NPPV, to indicate the relative likelihood of failure based on the patient's arterial pH, APACHE II score, Glasgow Coma Scale score, and other clinical data.4 These charts are also reproduced in the article by Hess and Fessler.3

The bottom line is that NPPV is life-saving in several clinical settings and should be more widely applied as the standard of care in those settings, but that it is also not a panacea and can be ineffective or frankly harmful if applied inappropriately or in the wrong patients. The pro-con article by Hess and Fessler nicely summarizes the current evidence supporting this important modality and its practical implications for the ICU.

Should Tidal Volume Initially Be 6 mL/kg for All Patients with Respiratory Failure?

The landmark Acute Respiratory Distress Syndrome (ARDS) Network tidal volume study5 showed that limiting delivered tidal volume to 6 mL/kg predicted body weight, as compared to 12 mL/kg, substantially reduced mortality and improved other outcomes in patients with acute lung injury (ALI) or ARDS. Similar benefits have been found in other studies using various forms of low-tidal-volume, lung-protective ventilation. In addition, compelling findings in animal studies and highly suggestive data in patients indicate that excessive lung stretch involving large tidal volumes can cause ALI/ARDS.6 In light of the benefits of low-tidal-volume ventilation with respect to ALI/ARDS, there has been considerable discussion about whether low tidal volumes should be used in managing all patients who require mechanical ventilation—not just those who have or are at risk for developing ALI/ARDS. This issue was debated by Steinberg and Kacmarek7 at the controversies conference.

During the conference, consensus (if not unanimity) was reached on most of the debated topics. However, whether all ventilated patients should receive 6 mL/kg was one for which this was not the case, with the proponents of both positions holding them vigorously and the assembled experts being more evenly split than was the case with most other questions.8

Table 2 summarizes the main points for and against the routine use of low tidal volumes. The "con" position revolves around two main issues: that low-tidal-volume ventilation may be harmful, and that the target tidal volume of 6 mL/kg predicted body weight is not the appropriate target, especially in ALI/ARDS.

In the ARDS Network and other studies, arterial oxygenation has not been as good, particularly in the first few days of ventilatory support, in the patients receiving low tidal volumes as in those on higher volumes. Low-tidal-volume ventilation may cause hypercapnia, or worse hypercapnia, as compared to the use of larger volumes, and more sedation (or even paralysis) may be required in order for patients to tolerate it, although published studies so far do not confirm this. And atelectasis may be more common when low tidal volumes are used, although the clinical importance of this observation is uncertain. In general, participants in the conference did not regard the potential adverse effects of low-tidal-volume ventilation—either in patients with ALI/ARDS or in ventilated patients in general—as of sufficient importance to outweigh its potential benefits.

The biggest controversy has to do with how best to gauge the risk of VILI and the excess mortality associated with high-tidal-volume, high-pressure ventilation. The "plateau pressure" faction contends that limiting trans-pulmonary pressure is the essential element, and that end-inspiratory plateau pressures less than 30 or 35 cm H2O are clinically safe regardless of the tidal volume delivered. The "tidal volume" faction points to data from the ARDS Network study9 showing that mortality in that trial was correlated with both tidal volume and plateau pressure, observed mortality decreasing progressively even at pressures below 30 cm H2O. The arguments here were mainly in the context of how best to manage ALI/ARDS, and data from patients without ALI or ARDS or substantial risk for developing these conditions are largely nonexistent.

The bottom line is that although no one doubts that lung-protective ventilation saves lives in ALI/ARDS, the experts disagree on whether limiting tidal volume or end-inspiratory plateau pressure (or both) is the crucial management element. Clearly, ventilating ALI/ARDS patients with tidal volumes substantially over 6 mL/kg and plateau pressures exceeding 30-35 cm H2O when these could readily be reduced by ventilator adjustment is contrary to present evidence. Using a tidal volume of 6 mL/kg (predicted body weight) in all ventilated patients to prevent VILI seems unnecessary in many instances and may cause practical problems with patient tolerance, although this practice is gaining in acceptance and becoming more widespread.

Should All Ventilated Patients Be Monitored with Capnography?

Capnometry (digital display of data) and capnography (graphical display of data) can be either time-based or volume-based. The technology for expired CO2 monitoring has improved substantially over the last 15-20 years, and current apparatus accurately and rapidly provides end-tidal partial pressure (PetCO2) and volume as well as calculation of dead-space ventilation. Manufacturers of end-tidal CO2 monitoring devices point out that arterial blood gases (ABGs) are invasive and expensive, and that continuous direct monitoring of gas exchange via ABGs is not a practical option in today's ICU. Capnography is touted by its proponents for titrating ventilator settings, detecting airway mishaps, monitoring the course of a patient's critical illness, guiding weaning, and diagnosing such events as acute pulmonary embolism and the onset of ARDS.10

Capnography is the gold standard for confirmation of endotracheal intubation, several studies having demonstrated its clinical superiority over auscultation, the self-inflating bulb, and trachea light. For more than 20 years it has also been a standard of care for continuous patient monitoring in the operating room. However, whether capnography accurately indicates what is going on in a mechanically ventilated patient in the ICU remains hotly contended. The issues involved, and the available evidence, are discussed at length by Cheifetz and Myers.11

Table 3 lists the arguments for and against the use of capnography for the routine monitoring of ventilated patients. The claimed advantages and values of continuous capnographic monitoring in the ICU are mainly extrapolated from data generated in the operating room, and few studies have examined the clinical accuracy of capnography in critically ill patients. What data are available from this setting tend to emphasize its differences from the controlled anesthesia environment.

One study of mechanically ventilated patients with severe head trauma12 found that the gradient between arterial and end-tidal PCO2 values (P(a-et)CO2) before and after endotracheal suctioning varied between -5.5 mm Hg and +19.7 mm Hg, and that the two values had least agreement in patients with atelectasis, pneumonia, or a chest tube. On the basis of their measurements these authors concluded that PetCO2 was less valid as a surrogate for PaCO2 in patients who were spontaneously breathing, were on assist-control ventilation, had PEEP > 5 cm H2O, or had worse oxygenation according to PaO2/FIO2 ratio, or any combination of these factors.12 In another study of simultaneous PetCO2 and PaCO2 values in patients with severe trauma, only 40% of the changes showed a linear relationship, and changes in PetCO2 falsely predicted the changes in PaCO2 in 27% of instances.13

The bottom line here is that capnography is no substitute for either an arterial PCO2 or a skilled clinician at the bedside, particularly in an unstable patient with underlying pulmonary dysfunction and multiple comorbidities. The more physiologically normal the patient, the more accurately capnography reflects lung function and gas exchange, but, it could be argued, the less that patient needs the monitor. When polled at the conclusion of the discussion, only a small minority of the conference participants agreed with the starting premise that all ventilated patients should be monitored with capnography from intubation to extubation.8


  1. Respiratory Controversies in the Critical Care Setting. Part 1. Respir Care. 2007(Apr);52(4):406-488.
  2. Respiratory Controversies in the Critical Care Setting. Part 2. Respir Care. 2007 (May);53(5):568-644.
  3. Hess DR, Fessler HE. Respir Care. 2007 May;52(5):568-581.
  4. Confalonieri M, et al. Eur Respir J. 2005 Feb;25(2):348-355.
  5. [No authors listed] Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. The Acute Respiratory Distress Syndrome Network. N Engl J Med. 2000;342(18):1301-1308.
  6. Tremblay LN, Slutsky AS. Intensive Care Med. 2006;32(1):24-33.
  7. Steinberg KP, Kacmarek RM. Respir Care. 2007;52(5):556-567.
  8. Cheifetz IM, Macintyre NR. Respir Care. 2007 May;52(5):636-644.
  9. Hager DN, et al. Am J Respir Crit Care Med. 2005;172(10):1241-1245.
  10. Thompson JE, Jaffe MB. Respir Care. 2005;50(1):100-108.
  11. Cheifetz IM, Myers TR. Respir Care. 2007;52(4):423-38.
  12. Kerr ME, et al. Crit Care Med. 1996;24(5):785-790.
  13. Russell GB, Graybeal JM. J Trauma. 1994;36(3):317-322.